Scaled adaptive lithography

Abstract

An adaptive method of providing electrical interconnections for a plurality of feed-through lines each having a respective end extending to at least one substrate surface includes generating an artwork representation for the electrical interconnections using specified feed-through line end positions on the at least one substrate surface. The at least one substrate surface may include a surface of a stack of substrates with at least two substrates having feed-through line ends facing a common direction. Actual positions of the at least two of the feed-through line ends are determined, and a scale factor is estimated using the determined actual positions. Actual positions of others of the feed-through line ends are estimated using the scale factor, and the artwork representation is modified to properly include electrical interconnections to ones of the feed-through line ends which are not in their specified positions. The artwork representation can be converted into a turn point polygon (TPP) format for effective data handling and modification of the artwork representation, and the TPP format can then be can be converted to a rectangular representation. A laser can be controlled in accordance with the rectangular representation of the modified artwork.

Description

SPEECH INFORMATION

A portion of the disclosure of this patent document contains material which is subject to copyright protection. The copyright owner has no objection to the facsimile reproduction by anyone of the patent document or the patent disclosure, as it appears in the Patent and Trademark Office patent file or records, but otherwise reserves all copyrights whatsoever.

BACKGROUND OF THE INVENTION

Co-fired packages are used as substrates for multi-chip modules and can include materials such as alumina or aluminum nitride with buried conductors which offer a variety of physical and electrical advantages. Such packages can be used in two dimensional high density interconnect (HDI) processes to reduce the number of metal interconnect layers, to fabricate hermetically sealed modules, and to provide vertical feed-through connections. As disclosed in commonly assigned and concurrently filed Saia et al., "Stacking of Three Dimensional High Density Interconnect Modules with Metal Edge Contacts," U.S. application Ser. No. 08/453,109, still pending, which is herein incorporated by reference, co-fired packages can be used in three dimensional HDI processes to provide edge feed-through connections.

When co-fired substrates are used for two dimensional HDI, an overlay pattern must be aligned with respective ends of the buried feed-through conductor paths (hereinafter referred to as feed-through lines). If the substrate end pitch of a feed-through line is greater than about 20 mils, it would be possible to align a fixed grid pattern of cover pads over the ends. For smaller end pitches, the tolerance of via placement due to movement in the substrate material during firing can hinder the overlay of a fixed grid array of cover pads. Likewise, an end pitch of at least 18 mils is preferred for most three dimensional HDI designs.

Eichelberger et al., U.S. Pat. No. 4,835,704, is a patent in which adaptive lithography was developed to align pattern transfer mask artwork to die positioned on a substrate by compensating for translational and rotational errors in chip placement. In U.S. Pat. 4,835,704, the chip and HDI artwork are both known, fixed arrays.

SUMMARY OF THE INVENTION

It would be desirable to have an adaptive lithography method for situations in which one of two patterns being aligned is not fixed.

The present invention includes a method of adapting a fixed grid lithographic artwork pattern to a variable expanded or contracted artwork array by introducing a scale factor in addition to translational and rotational factors.

The invention can be used in one embodiment, for example, for precision aligning a fixed pattern transfer artwork array to a variable expanded or contracted pattern array that exists on co-fired substrates with thick film feed-through conductor paths. The added alignment accuracy will permit the use of co-fired substrates with a feed-through path via pitch of about 10 mils or greater.

Briefly, according to an embodiment of the present invention, an adaptive method of providing electrical interconnections for a plurality of feed-through lines each having a respective end extending to at least one substrate surface comprises generating an artwork representation for the electrical interconnections using specified feed-through line end positions on at least one substrate surface; determining actual positions of at least two of the feed-through line ends; estimating a scale factor using the actual positions of at least two of the feed-through line ends; estimating actual positions of the other feed-through line ends using the scale factor; modifying the artwork representation for the electrical interconnections for the feed-through lines to properly include electrical interconnections to ones of the feed-through line ends which are not in their specified positions; and forming the electrical interconnections for the feed-through lines using the modified artwork.

A substrate surface may comprise a surface of a stack of substrates with at least two substrates having feed-through line ends facing a common direction. The specified feed-through line end positions can be situated in a row and at least two of the feed-through line ends for which actual positions are determined can be situated at opposite ends of the row, for example. The artwork representation can be converted into a turn point polygon (TPP) format for effective data handling and modification of the artwork representation. A laser can be controlled in accordance with the TPP representation of the modified artwork. In one embodiment, the TPP format is first converted to a rectangular representation which is then used to control the laser.

BRIEF DESCRIPTION OF THE DRAWINGS

The features of the invention believed to be novel are set forth with particularity in the appended claims. The invention itself, however, both as to organization and method of operation, together with further objects and advantages thereof, may best be understood by reference to the following description taken in conjunction with the accompanying drawings, where like numerals represent like components, in which:

FIG. 1 is a sectional side view of stacked and interconnected two dimensional electronic modules.

FIG. 2 is a sectional view along line 2--2 of FIG. 1.

FIG. 3 is a side view of a stack of two dimensional electronic modules illustrating positions of electrically conductive feed-through lines.

FIG. 4 is an illustration of a system for determining the positions of selected ends of the feed-through lines and estimating the positions of the other feed-through lines.

FIG. 5 is a flowchart of steps in one embodiment of the present invention.

FIG. 6 is a top view of a representative conductor path modification expected from use of the present invention.

DETAILED DESCRIPTION OF A PREFERRED EMBODIMENT OF THE INVENTION

FIG. 1 is a sectional side view of a plurality of stacked and interconnected two dimensional electronic modules. As described in aforementioned Saia et al., "Stacking of Three Dimensional High Density Interconnect Modules with Metal Edge Contacts," a method for fabricating a stack of circuit modules includes providing a plurality of substrates 10 having module interconnection layers 36 (defined below). Each of the substrates has a circuit chip (not shown) with chip pads positioned therein and an electrically conductive feed-through line 12 having one end extending to a first surface 11b of the substrate and having another end extending to a second surface 11a of the substrate with the first surface being perpendicular to the second surface.

Each of the module interconnection layers is situated over a respective first surface of each of the substrates and may include, for example, a module dielectric layer having module vias therein and a module pattern of electrical conductors extending through the module vias to couple the selected chip pads to the feed-through lines (not shown). Base metallization layers 14 can be applied on the first surfaces of the substrates prior to the module interconnection layers, if desired. The plurality of substrates are stacked. In one embodiment, an adhesive 46 is used between adjacent substrates. A side interconnection layer is applied over a side surface including the second surfaces 11a of the substrates. The side interconnection layer includes a side dielectric layer 38 having side vias 48 therein aligned with predetermined ones of the feed-through lines and a side pattern of electrical conductors 42 extending through the side vias. A passivation layer 44 can be applied over the side pattern of electrical conductors.

FIG. 2 is a sectional view along line 2--2 of FIG. 1. Each substrate 10 can include one or more electrically conductive feed-through lines 12 extending to the substrate edge. Furthermore, if desired, multiple edges (not shown) of each substrate can have feed-through lines exposed.

FIG. 3 is a side view of a stack of two dimensional electronic modules comprising substrates 10a, 10b, 10c, and 10d and respective module interconnection layers 16a, 16b, 16c, and 16d illustrating positions of electrically conductive feed-through lines 12a, 12b, 12c, and 12d.

Precise alignment and positioning of the feed-through lines is difficult to achieve. One problem, which is especially difficult to overcome for co-fired ceramic substrates, is that the feed-through lines tend to move during substrate fabrication. Another challenge arises when the distances d.sub.1, d.sub.2, and d.sub.3 are varied due to differing module interconnection layer thicknesses or adhesive layer (shown in FIG. 1) thicknesses. Therefore, it can be difficult to accurately predict the positions of feed-through lines.

As an illustrative example, feed-through lines 12a in substrate 10a represent the positions of an aligned row, feed-through lines 12b and 12d in substrates 10b and 10d represent respective rows of feed-through lines which have been compressed with the feed-through lines being closer together, and feed-through lines 12c in substrate 10c represent a row of feed-through lines which has been expanded. Similarly, d.sub.3 represents a distance between adjacent rows of feed-through lines which is greater than either of d.sub.1 or d.sub.2.

As can be seen from FIG. 3, individual feed-through lines within a single row can be shifted in addition to the potential of the adaptive region 40 of the row itself to be translated and/or rotated. A similar situation can be detected with respect to an adaptive region 41 around a column.

FIG. 4 is an illustration of a system for detecting the positions of selected ends of the feed-through lines and estimating the positions of other ones of the feed-through lines. FIG. 4 additionally illustrates a laser 18 (with associated alignment controls 20 and lens 24) which can be used in a later discussed step of forming a pattern of electrical conductors for interconnecting selected feed-through lines. As described in aforementioned Eichelberger et al., U.S. Pat. No. 4,835,704, a table 30 can be used to support a workpiece 28 for scanning with a CID (charge inducted device) or CCD (charge coupled device) camera 26. A computer 32 exchanges data with control electronics 34 which can be used to control the table positioning.

In one embodiment of the present invention, the workpiece comprises a stack such as shown in FIG. 3 with the feed-through lines 12a, 12b, 12c, and 12d facing the camera. Preferably the stack is surrounded by a support fixture 29 including a material such as titanium.

FIG. 5 is a flowchart of steps in one embodiment of the present invention. At step 49, ideal case artwork can be generated using a computer aided design system or an automatic routing system. The term "artwork" refers to a pattern in the form of a database (referred to as a "first" database) which describes the ideal positions of the feed-through lines and the interconnect structure for providing interconnects between electronic components and/or feed-through lines. When the ideal artwork is modified, that modified artwork can be used to drive a laser beam under computer control to directly write interconnection patterns in metallization and/or photoresist that have been modified to adjust for the measured and estimated positions of the feed-through lines, for example.

Any computer aided design (CAD) system can interface to an adaptive lithography system with changes in the software which convert to the stripped level of data coding being necessary. In the embodiment of FIG. 5, for example, at step 52, ideal artwork data is converted into turn point polygon (TPP) format (which is stored in a "second" database). Other techniques that may be used include bit maps in binary formats, for example.

At step 49 in FIG. 5, a CAD system can output a drawing interchange file (DXF) which is an ASCII output file of many commercial CAD packages. TPP format describes a closed convex polygon by a listing of respective (x, y) coordinates of all the vertices. The CAD to TPP data conversion can be accomplished using software programmed in a language such as C to convert DXF format to TPP format. For example, Appendix A is a source code listing of a computer program which can be used to convert DXF format to TPP format.

Instead of measuring each individual feed-through line end position, in the present invention the positions of a smaller number of feed-through line ends can be obtained at step 50 by the process discussed with respect to FIG. 4, and the positions of the remainder of the feed-through line ends can then be estimated at step 56 using a technique including the calculation of scale correction factors. In one embodiment what is detected is the positions of the extents (the two feed-through line ends which are farthest apart in a respective substrate, such as 12e and 12f in substrate 10b of FIG. 3). The larger the number of adaptive regions and measured feed-through line end positions, the greater the accuracy of estimations of the other feed-through line end positions and the greater the complexity of the adaptive lithography technique.

For a respective one of the substrates in an adaptive region 40 (shown in FIG. 3), for example, reference points (x.sub.1, y.sub.1) and (x.sub.2, y.sub.2) for feed-through lines 12e and 12f can be measured.

Scale factors for the group of feed-through line end positions between the reference points can be calculated at step 56, for example, using the following equations which compare the measured positions to the ideal (specified) positions of the original artwork:

% x scale=(measured x2-measured x1)/(specified x2-specified x1) % y scale=(measured y2-measured y1)/(specified y2-specified y1) These calculations provide respective (x, y) scale factors for an adaptive region. A database can be generated which describes the actual locations of each of the feed-through line ends. This database can be compared to the ideal case positions listed in the first database, and translation and rotation errors can be calculated for the feed-through line end positions of the adaptive region. The scale, translational errors, and rotational errors can be used to generate a "third" database.

After the data has been converted to TPP format and the scale factors have been calculated, at step 54 the scale transformation is performed in conjunction with translation and rotation with respect to a fixed reference point in the adaptive region.

An adaptive region bounds the adaptive lithography algorithms to specified areas and is described in rectangular format by two (x, y) coordinates of diagonal vertices. HDI metal interconnection patterns and via openings which reside within or pass through the adaptive region can be scaled using the calculated scale factors in conjunction with translational and rotational transformations. The basic formulas for translation, rotation, and scale transformation are as follows:

TRANSLATION

x'=x+Tx y'=y+Ty

ROTATION

x'=xR+(x-xR)cos .theta.-(y-yR)sin .theta. y'=yR+(y-yR)cos .theta.-(x-xR)sin .theta.

SCALE

x'=xF+(x-xF)Sx y'=yF+(y-yF)Sy, wherein the translational distance is (Tx, Ty) is the distance that an adaptive region has moved, the rotation of the adaptive region point is (xR, yR), and the scale factors for individual feed-through lines within an adaptive region are (Sx, Sy) with a fixed point of (xF, yF).

The three transformations can be combined as follows:

xx=(((x-xref�ic!*scale.sub.-- x�ic!+xref�ic!)+((y-yref�ic!)*rotate�ic!)+xoffset�ic!; yy=(((y-yref�ic!*scale).sub.-- y�ic!+yref�ic!-((x-xref�ic!)*rotate�ic!)+yoffset�ic!, wherein (xx, yy) is the transformed (x, y) vertex of TPP; (x, y) is the (x, y) vertex of TPP; (scale.sub.-- x, scale.sub.-- y) is the (x, y) scale factors; (x.sub.-- ref, y.sub.-- ref) is the fixed reference point; rotate�ic! is the rotation units in radians; and (xoffset, yoffset) is the linear offset from the fixed reference point.

The vertices of a designated TPP structure inside the adaptive region are transformed and output to a "fourth" database. New vertices are added to TPP structures which cross an adaptive region to provide the fixed points for interconnect stretching and bending.

At step 58, the adapted TPP data (the fourth database) is fractured and converted to rectangular format. For example, Appendix B is a source code listing of a computer program which can be used to convert adapted TPP data to rectangular format.

At step 60, the rectangular data is then converted to laser direct write (RLL) format for laser lithography processing. For example, Appendix C is a source code listing of a computer program which can be used to convert rectangular data to RLL format. At step 62, the interconnection steps discussed with respect to FIG. 1 of applying a side dielectric layer and pattern of electrical conductors are performed using the data in the laser direct write format to form vias 48 and fabricate the pattern of electrical conductors 42. The pattern of electrical conductors can be scaled, translated, and rotated in accordance with the RLL format. In an alternative embodiment, software could be designed to convert TPP format directly to RLL format.

The method was reduced to practice using co-fired ceramic packages including 80 feed-through lines having pitches of 18 mils along two opposite edges. The positioning of the feed-through lines on the finished substrates relative to the planned position can be referenced in several ways. The simplest method is to measure the overall length of a row of feed-through lines in an adaptive region. This method does not take into account the rotational error in the placement of the feed-through lines, but the method will provide an indication of the ability to overlay a fixed pitch of cover pads. In more complicated applications, it is possible to obtain feed-through line position information at selected points in a curve or other appropriate configuration using the above discussed information with appropriate modifications to the formulas and software.

A pattern with 14 mil cover pads was applied on the 18 mil feed-through lines in several patterning tests. The process of aforementioned U.S. Pat. No. 4,835,704 was used and, on the substrates with larger placement errors, the lines were not completely covered by the pads. When the scaled turn point polygon adaptive lithography method of the present invention was used, the artwork expanded and contracted as necessary, providing complete line coverage.

FIG. 6 is a top view of a representative conductor path modification expected from use of the present invention. Feed-through line end positions 64 and 66 represent compressed and expanded positions respectively with respect to the fixed artwork portions 68 of a pattern of electrical conductors. Portions 70 of the pattern of electrical conductors link the fixed artwork to adaptive regions 40a and 40b.

While only certain preferred features of the invention have been illustrated and described herein, many modifications and changes will occur to those skilled in the art. It is, therefore, to be understood that the appended claims are intended to cover all such modifications and changes as fall within the true spirit of the invention. ##SPC1##

Claims

1. An adaptive method of providing electrical interconnections for a plurality of feed-through lines, each of the feed-through lines having a respective end extending to at least one substrate surface, the method comprising:

generating an artwork representation for the electrical interconnections using specified feed-through line end positions on the at least one substrate surface;

determining actual positions of at least two of the feed-through line ends;

estimating a scale factor using the actual positions of the at least two of the feed-through line ends;

performing calculations to estimate actual positions of others of the feed-through line ends using the scale factor and the determined actual positions of the at least two of the feed-through line ends;

modifying the artwork representation to properly include electrical interconnections to ones of the feed-through line ends which are not in their specified positions; and

forming the electrical interconnections for the feed-through lines using the modified artwork.

2. The method of claim 1 wherein the at least one substrate surface comprises a surface of a stack of substrates with at least two substrates having feed-through line ends facing a common direction.

3. The method of claim 2 wherein the specified positions of the feed-through line ends are stored in a first database and the step of generating the artwork representation comprises the steps of:

providing a circuit layout of a pattern of electrical conductors including the specified feed-through line end positions; and

storing a representation of the circuit layout in a second database describing an electrical interconnection structure for providing the electrical interconnections for the feed-through line ends in their specified positions.

4. The method of claim 3 wherein the step of storing the representation of the circuit is performed by storing a turn point polygon representation of the circuit layout in the second database.

5. The method of claim 4, wherein the step of modifying the artwork representation comprises:

comparing the positions of the at least two of the feed-through line ends with respective specified positions; and

after performing calculations to estimate the actual positions of the others of the feed-through line ends, calculating a translation and rotation for each of the feed-through line ends based on the comparing step.

6. The method of claim 5 wherein the step of modifying the artwork representation further includes generating a third database describing the actual locations of each of the feed-through line ends with respect to the at least one substrate surface and wherein said step of comparing compares entries in the first and third databases.

7. The method of claim 6 wherein the step of modifying the artwork representation further comprises producing a fourth database by altering the second database according to the scale factor estimation and the translation and rotation calculations for each of the feed-through line ends to produce a fourth database storing a turn point polygon representation of the modified artwork.

8. The method of claim 7 wherein the step of forming the electrical interconnections comprises

converting the turn point polygon representation of the modified artwork in the fourth database to a rectangular representation; and

controlling a laser in accordance with the rectangular representation of the modified artwork.

9. The method of claim 1 wherein the specified feed-through line end positions are situated in a row and the at least two of the feed-through line ends for which actual positions are determined are situated at opposite ends of the row.

10. An adaptive system for providing electrical interconnections for a plurality of feed-through lines, each of the feed-through lines having a respective end extending to at least one substrate surface, the system comprising:

means for generating an artwork representation for the electrical interconnections using specified feed-through line end positions on the at least one substrate surface;

means for determining actual positions of at least two of the feed-through line ends;

means for estimating a scale factor using the actual positions of the at least two of the feed-through line ends;

means for performing calculations to estimate actual positions of others of the feed-through line ends using the scale factor and the determined actual positions of the at least two of the feed-through line ends;

means for modifying the artwork representation to properly include electrical interconnections to ones of the feed-through line ends which are not in their specified positions; and

means for forming the electrical interconnections for the feed-through lines using the modified artwork.

11. The system of claim 10 wherein the at least one substrate surface comprises a surface of a stack of substrates with at least two substrates having feed-through line ends facing a common direction.

12. The system of claim 11 wherein the means for generating the artwork representation for the electrical interconnections using specified feed-through line end positions on the at least one substrate surface include means for storing the specified positions of the feed-through line ends in a first database, and

wherein the means for generating an artwork representation for the electrical interconnections using specified feed-through line end positions on the at least one substrate surface include:

means for providing a circuit layout of a pattern of electrical conductors including the specified feed-through line end positions; and

means for storing a representation of the circuit layout in a second database describing an electrical interconnection structure for providing the electrical interconnections for the feed-through line ends in their specified positions.

13. The system of claim 12 wherein the means for storing the representation of the circuit layout in the second database comprise means for storing a turn point polygon representation of the circuit layout in the second database.

14. The system of claim 10 wherein the specified feed-through line end positions are situated in a row and the means for determining actual positions of the at least two of the feed-through line ends include means for determining actual positions of two of the feed-through line ends which are situated at opposite ends of the row.