Semiconductor device manufacturing: process – Formation of electrically isolated lateral semiconductive... – Having substrate registration feature
Reexamination Certificate
2000-05-26
2001-10-30
Dang, Trung (Department: 2823)
Semiconductor device manufacturing: process
Formation of electrically isolated lateral semiconductive...
Having substrate registration feature
C438S018000, C438S725000, C257S797000
Reexamination Certificate
active
06309944
ABSTRACT:
BACKGROUND OF THE INVENTION
(1) Field of the Invention
The invention relates to the fabrication of integrated circuit devices, and more particularly, to a method to improve overlay alignment accuracy for wafer stepper tools.
(2) Description of the Prior Art
During the manufacturing of semiconductor devices, many of the processing steps that are applied have as objective to create individual device features in addition to creating device interconnect patterns. For each of these steps, specific geometries are created in layers of materials that are used to form the semiconductor device, a most common method that is applied for the creation of these various geometries uses methods of photolithography whereby a mask contains patterns that are to be created. Since many of the layers that are created as part of a semiconductor device interact and are mutually dependent, it is of key importance that the successively overlying layers are within strict limits of alignment so that the relative positioning of each overlying layer is precisely controlled and adjusted with respect to one or more underlying layers that have previously been exposed. For the purpose of alignment of successive and overlying patterns, a number of different methods are typically used, these methods are mostly dependent on alignment markers that have been provided for this purpose in the surface of the substrate that is used to create the semiconductor device.
Alignment errors can occur in a number of manners. For instance, and hereby referring to a lower field and an upper field whereby the lower field is the field to which the upper field must be aligned, the upper field may be rotated with respect to the lower field, the upper field may be shifted in a X-direction or a Y-direction or in both X and Y directions with respect to the lower field, the upper field may be divided into a multiplicity of fields (a field grid) whereby a subset of this multiplicity of fields is either magnified or reduced in size relative to the lower field or individual sections of the upper field are shifted in unequal amounts with respect to the lower field, the individual sections of an upper field may be skewed or rotated around an axis with respect to the lower field, the individual sections of the upper field (the field grid) may have shifted in an outward direction either uniformly or non-uniformly with respect to the lower field, and the like.
In order to control and evaluate alignment accuracy, the typical parameter that is used for this purpose is to measure the intra-field error that is incurred when exposing two overlying fields. This method makes use of sets of (square or rectangular shaped) alignment marks whereby one set of alignment marks is centered within the geometric boundaries of a second set of alignment marks. These alignment marks can be provided to individual fields within the field grid (thereby surrounding each field on all sides) or these alignment marks can be omitted between adjacent fields of a field grid. The number and placement of the alignment marks is open to interpretation and specific design requirements, key to this method however is that the individual alignment marks for the upper field fit within and are surrounded by the individual alignment marks of the lower field. It is from this easy to realize that a measurement can readily be made that measures the distance between the sides of the two sets of alignment marks, a measurement that is directly indicative of the accuracy of alignment between the two layers. A measurement of equal value of the distance between the sides of the upper level and the lower level alignment marks in both the X and the Y direction along all four sides of the alignment marks indicates that the two sets of alignment marks are in perfect alignment and, with that, the exposure for the upper level is in alignment with the exposure for the lower level. Where these measurements differ in value, elementary observations and deductions will lead to conclusions of the extent and the direction of misalignment that occurs between the upper and the lower level. This measured misalignment can directly be converted into a readjustment of the stepper tool to the point where the misalignment is eliminated and again the upper and the lower level are in perfect alignment. This method of evaluating alignment between two overlying layers leads not only to measuring misalignment between an upper and a lower layer in the X and Y direction but further serves to identify any problems of rotation of the upper layer with respect to the lower layer. It must be realized that the alignment marks are not limited to relatively small sections on the surface of the substrate so that any error of alignment will be highlighted and is therefore readily identifiable due to the extended distance between the various alignment marks.
The conventional method of measuring and correcting alignment errors requires the steps of measuring the overlays of the alignment marks as previously indicated, calculating the required adjustments based on the relative field size of the lower and the upper field, updating or interfacing with (if required) an existing data base to further influence future alignment procedures and finally affecting the correction in alignment, if any.
Various field sizes of the field grid can have an effect on the process of mark alignment. Overlying fields are not necessarily of uniform size or of uniform ratio between upper and lower field size. Furthermore, fields of one level may overlay a field of the next level causing further confusion as to how exactly to coordinate the corrections, if any, between the various overlying fields. It must thereby be understood that errors may occur not only between overlying levels (the level-to-level error) but also between fields within one level (within-level between-fields). In making multiple exposures it may therefore be required to correct alignment problems between overlying levels and between fields within a given level. This latter problem may be aggravated for levels that contain fields that differ considerably in size from the size of the fields that are contained in the preceding level. Multiple, smaller size fields in an overlying layer are in this respect particularly prone to cause problems. For all measurements and corrections that are applied in order to gain acceptable inter-level alignment, there is a tolerance of misalignment whereby the alignment is considered acceptable if the alignment between layers falls within the tolerance for the step of alignment measurement. It is clear that this may still lead to considerable misalignment between levels of patterns that are separated by a large number of interposed levels. This can occur in the case where for example all or a large number of tolerances are skewed in the same direction thereby accumulating a build up of a misalignment error between widely separated levels that belong to one construction this is unacceptable.
To further highlight the effect that different field size can have on alignment errors, the example of two tools that have different field and exposure sizes will be highlighted. The first tool has a smaller field size whereby one chip within the surface of the substrate is exposed per field size using one reticle. A second tool has a larger field size whereby four chips on the substrate surface are exposed per field using one reticle. If errors in alignment occur between the one field of the first tool and the four fields of the second tool, the alignment that will be affected between the two tools and based of the measured alignment error may result in alignments for the second tool that cannot compensate as desired for all four fields of the second tool. Not all four fields of the second tool will be correctly adjusted if for instance a rotational error has been measured between the two tools since this rotational error adjustment affects the four fields within the second tool in different amounts of adjustment. Conventional methods address this problem by limiting eit
Chao Kuo-Hung
Huang Chu-Wen
Kuo Cheng-Chen
Sheng Han-Ming
Ackerman Stephen B.
Dang Trung
Kebede Brook
Saile George O.
Taiwan Semiconductor Manufacturing Company
LandOfFree
Overlay matching method which eliminates alignment induced... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Overlay matching method which eliminates alignment induced..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Overlay matching method which eliminates alignment induced... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2597314