Compensation for substrate doping in edge reconstruction for in-situ electromagnetic inductive monitoring

Patent metadata

Field	Value
Publication number	US-12447578-B2
Application number	US-201817280163-A
Country	US
Kind code	B2
Filing date	Sep 26, 2018
Priority date	Sep 26, 2018
Publication date	Oct 21, 2025
Grant date	Oct 21, 2025

Abstract

A method of compensating for a contribution of conductivity of the semiconductor wafer to a measured trace by an in-situ electromagnetic induction monitoring system includes storing or generating a modified reference trace. The modified reference trace represents measurements of a bare doped reference semiconductor wafer by an in-situ electromagnetic induction monitoring system as modified by a neutral network. The substrate is monitored with an in-situ electromagnetic induction monitoring system to generate a measured trace that depends on a thickness of the conductive layer, and at least a portion of the measured trace is applied to a neural network to generate a modified measured trace. An adjusted trace is generated, including subtracting the modified reference trace from the modified measured trace.

First claim

What is claimed is:

1. A method of polishing a substrate, comprising:
storing or generating a modified reference trace representing a sequence of measurements taken by a sensor, of an in-situ electromagnetic induction monitoring system, moving across a bare, n-type or p-type doped reference semiconductor wafer as modified by a neural network, wherein the neural network is configured to compensate for distortions caused by a partial overlap between an edge of a semiconductor wafer and a magnetic field generated by the in-situ electromagnetic induction monitoring system, wherein the neural network comprises an input layer, an output layer, and one or more hidden layers;
bringing a substrate comprising a conductive layer disposed over the semiconductor wafer into contact with a polishing pad;
generating relative motion between the substrate and the polishing pad;
monitoring the substrate with the in-situ electromagnetic induction monitoring system as the conductive layer is polished to generate a measured trace including measured signal values at multiple different locations on the semiconductor wafer, wherein the measured signal values depend that depends on a thickness of the conductive layer, wherein the measured trace is a sequence of measurements as the sensor moves across the substrate;
applying at least a portion of the measured trace including the measured signal values at multiple different locations on the semiconductor wafer as input to the neural network configured to compensate for the distortions caused by the partial overlap between the edge of the semiconductor wafer and the magnetic field generated by the in-situ electromagnetic induction monitoring system to generate a modified measured trace;
generating an adjusted trace to at least partially compensate for a contribution of conductivity of the semiconductor wafer to the measured trace, including subtracting the modified reference trace from the modified measured trace; and
at least one of halting polishing or modifying a polishing parameter based on the adjusted trace.

2. The method of claim 1 , comprising scanning the bare, doped reference semiconductor wafer with the sensor of the in-situ electromagnetic induction monitoring system to generate a preliminary reference trace having raw signal values, converting raw signal values in the preliminary reference trace to thickness values to generate an initial reference trace, and applying at least a portion of the initial reference trace to the neural network to generate the modified reference trace.

3. The method of claim 1 , wherein generating the modified reference trace includes scanning across the bare, doped reference semiconductor wafer with the sensor of the in-situ electromagnetic induction monitoring system.

4. The method of claim 1 , wherein generating the adjusted trace comprises scaling a difference between the modified reference trace and the modified measured trace,
wherein the difference is calculated by subtraction of the modified reference trace from the modified measured trace.

5. The method of claim 4 , wherein the adjusted trace A(x) is calculated such that
A ( x )=( T ( x )− S ( x )− b )/ k

where T(x) is the modified measured trace, S(x) is the modified reference trace, and b and k are constants.

6. The method of claim 5 , comprising selecting b and k according to a configuration of the sensor of the in-situ electromagnetic induction monitoring system.

7. The method of claim 1 , wherein at least the portion of the measured trace applied to the neural network includes a portion corresponding to an edge region of the substrate.

8. The method of claim 7 , wherein at least the portion of the measured trace applied to the neural network does not include a portion corresponding to a central region of the substrate.

9. The method of claim 1 , comprising training the neural network with a plurality of training traces representing measurements of one or more training substrates comprising a conductive layer on an undoped semiconductor wafer,
wherein the one or more training substrates with different training traces have different thicknesses of the conductive layer and different edge profiles.

10. The method of claim 1 , comprising receiving user input selecting the modified reference trace from a plurality of modified reference traces.

11. The method of claim 1 , wherein each layer of the neural network comprises one or more neural network nodes, and wherein
each neural network node of the one or more neural network nodes receives one or more node input values from one of the inputs to the neural network or from an output of one or more nodes of a preceding neural network layer.

12. The method of claim 11 , wherein the neural network comprises at least one of (A) a fully-connected, feedforward layer, (B) a non-fully-connected feedforward layer, and (C) a non-feedforward layer.

13. The method of claim 11 , wherein a number of neural network nodes of the output layer is smaller than a number of neural network nodes of the input layer.

14. The method of claim 11 , wherein a number of neural network nodes of the input layer is equal to a number of the measured signal values at the multiple different locations on the semiconductor wafer.

15. A computer program product, tangibly embodied in a non-transitory computer readable medium, comprising instructions to cause one or more computers to:
store or generate a modified reference trace representing a sequence of measurements taken by a sensor, of an in-situ electromagnetic induction monitoring system, moving across a bare, n-type or p-type doped reference semiconductor wafer as modified by a neural network, wherein the neural network is configured to compensate for distortions caused by a partial overlap between an edge of a semiconductor wafer and a magnetic field generated by the in-situ electromagnetic induction monitoring system, wherein the neural network comprises an input layer, an output layer, and one or more hidden layers;
receive from the in-situ electromagnetic induction monitoring system, as a conductive layer disposed over the semiconductor wafer of a substrate is polished, a measured trace including measured signal values at multiple different locations on the semiconductor wafer, wherein the measured signal values depend on a thickness of the conductive layer, wherein the measured trace is a sequence of measurements as the sensor moves across the substrate;
process at least a portion of the measured trace including the measured signal values at multiple different locations on the semiconductor wafer as input through the neural network configured to compensate for the distortions caused by the partial overlap between the edge of the semiconductor wafer and the magnetic field generated by the in-situ electromagnetic induction monitoring system to generate a modified measured trace;
generate an adjusted trace to at least partially compensate for a contribution of conductivity of the semiconductor wafer to the measured trace, wherein the instructions to generate the adjusting trace include subtracting the modified reference trace from the modified measured trace; and
at least one of halt polishing or modify a polishing parameter based on the adjusted trace.

16. The computer program product of claim 15 , wherein the instructions to generate the adjusted trace comprise scaling a difference between the modified reference trace and the modified measured trace, wherein the difference is calculated by subtraction of the modified reference trace from the modified measured trace.

1

Assignees

• Applied Materials Inc

Inventors

• Xu Kun
• Gage David Maxwell
• Lee Harry Q
• Ivanov Denis Anatolyevich
• Iravani Hassan G
• Bennett Doyle E
• Shrestha Kiran Lall

Classifications

• H10P90/123

  by grinding or lapping · CPC title

• H10P74/23

  characterised by multiple measurements, corrections, marking or sorting processes · CPC title

• H10P74/207

  Electrical properties, e.g. testing or measuring of resistance, deep levels or capacitance-voltage characteristics · CPC title

• H10P74/203

  Structural properties, e.g. testing or measuring thicknesses, line widths, warpage, bond strengths or physical defects · CPC title

• H10P52/402

  of semiconductor materials · CPC title