Systems and methods for analyzing defects in CVD films
Embodiments of the present technology may include semiconductor processing methods that include depositing a film of semiconductor material on a substrate in a substrate processing chamber. The deposited film may be sampled for defects at greater than or about two non-contiguous regions of the substrate with scanning electron microscopy. The defects that are detected and characterized may include those of a size less than or about 10 nm. The methods may further include calculating a total number of defects in the deposited film based on the sampling for defects in the greater than or about two non-contiguous regions of the substrate. At least one deposition parameter may be adjusted as a result of the calculation. The adjustment to the at least one deposition parameter may reduce the total number of defects in a deposition of the film of semiconductor material.
1. A semiconductor processing method comprising:
causing a film of semiconductor material to be deposited on a substrate in a substrate processing chamber;
receiving samples for defects in the deposited film at greater than or about two non-contiguous regions of the substrate with scanning electron microscopy, wherein the defects are characterized by a size less than or about 10 nm;
calculating a total number of defects in the deposited film based on the sampling for defects in the greater than or about two non-contiguous regions of the substrate;
identifying, using a trained neural network, a deviation in the deposited film from a reference based at least in part on the defects in the deposited film; and
receiving, from the trained neural network, a corrective action for the deposited film.
2. The semiconductor processing method of claim 1 , further comprising identifying, from the trained neural network, one or more deposition parameters of the deposited film based on the deviation.
3. The semiconductor processing method of claim 1 , further comprising identifying, from the trained neural network, one or more characteristics of the substrate based on the deviation.
4. The semiconductor processing method of claim 3 , wherein the one or more characteristics of the substrate comprise a stress in the substrate.
5. The semiconductor processing method of claim 3 , wherein the one or more characteristics of the substrate comprise a bow in the substrate.
6. The semiconductor processing method of claim 3 , wherein the one or more characteristics of the substrate comprise a thickness of the deposited film.
7. The semiconductor processing method of claim 1 , wherein the corrective action comprises causing the substrate to be rejected and removed from further production processes.
8. The semiconductor processing method of claim 1 , wherein the deviation comprises a trend indicating progressive deviation from a reference library of substrate data.
9. The semiconductor processing method of claim 8 , wherein the corrective action comprises identifying a series of wafers for removal from further production processes based on the trend.
10. The semiconductor processing method of claim 8 , wherein the corrective action comprises augmenting or adjusting a recipe for a deposition process that generated the deposited film.
11. The semiconductor processing method of claim 1 , wherein the corrective action comprises indicating that a processing chamber that generated the deposited film should be cleaned.
12. The semiconductor processing method of claim 1 , wherein the corrective action comprises indicating that a showerhead in a processing chamber that generated the deposited film should be replaced.
13. The semiconductor processing method of claim 1 , wherein the corrective action comprises adjusting at least one deposition parameter chosen from deposition temperature, deposition pressure, an amount of spacing between the substrate and a showerhead in the substrate processing chamber, and a flow rate ratio of two or more deposition precursors, wherein the adjustment to the at least one deposition parameter reduces the total number of defects in a deposition of the film of semiconductor material.
14. A semiconductor processing system comprising:
a deposition chamber to deposit a film of semiconductor material on a substrate in the deposition chamber;
a defect sampling unit comprising a scanning electron microscope, wherein the defect sampling unit is operable to receive the substrate with the deposited film of semiconductor material and sample the substrate for defects in the deposited film at greater than or about two non-contiguous regions of the substrate; and
a control unit in electronic communication with the deposition chamber and programmed to perform operations comprising:
calculating a total number of defects in the deposited film based on the sampling for defects in the greater than or about two non-contiguous regions of the substrate;
identifying, using a trained neural network, a deviation in the deposited film from a reference based at least in part on the defects in the deposited film; and
receiving, from the trained neural network, a corrective action for the deposited film.
15. The semiconductor processing system of claim 14 , wherein the corrective action reduces the total number of defects in a subsequently-deposited film of the semiconductor material by greater than or about 80%.
16. The semiconductor processing system of claim 14 , wherein the sampled film of semiconductor material is characterized by a calculated total number of defects greater than or about 10,000, and the corrective action reduces the calculated total number of defects in a subsequently-deposited film of the semiconductor material to less than or about 2000.
17. The semiconductor processing system of claim 14 , wherein the corrective action reduces an average surface roughness in a subsequently-deposited film of the semiconductor material by greater than or about 50%.
18. The semiconductor processing system of claim 14 , wherein the non-contiguous regions that are sampled for defects in the deposition film are characterized by a field of view of about 5 μm.
19. The semiconductor processing system of claim 14 , wherein the deposited film of semiconductor material is characterized by a thickness of less than or about 50 Å.
20. One or more non-transitory computer-readable media storing instructions that, when executed by one or more processors, cause the one or more processors to perform operations comprising:
causing a film of semiconductor material to be deposited on a substrate in a substrate processing chamber;
receiving samples for defects in the deposited film at greater than or about two non-contiguous regions of the substrate with scanning electron microscopy, wherein the defects are characterized by a size less than or about 10 nm;
calculating a total number of defects in the deposited film based on the sampling for defects in the greater than or about two non-contiguous regions of the substrate;
identifying, using a trained neural network, a deviation in the deposited film from a reference based at least in part on the defects in the deposited film; and
receiving, from the trained neural network, a corrective action for the deposited film.