A Methodology for Efficient Dynamic Spatial Sampling and Reconstruction of Wafer Profiles

McLoone, Seán; Johnston, Adrian; Susto, Gian Antonio

doi:10.1109/tase.2017.2786213

Cited by 9 publications

(9 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this paper, we propose a novel approach to dynamic sampling, called Induced Start Dynamic Sampling (ISDS), that provides a natural extension of greedy selection methods to dynamic sampling, and is able to achievable comparable reconstruction accuracy to the static sampling gold standard. In so doing, it consistently outperform SDS [27], the current state-of-the-art approach for wafer spatial dynamic sampling. To demonstrate the efficacy of the ISDS strategy, it is implemented with the four greedy selection algorithms described above, namely, FSCA, OPFS, ITFS and FP, and the resulting dynamic sampling algorithms benchmarked against other dynamic methods proposed in the literature using both simulated and real industrial case studies.…”

mentioning

confidence: 81%

“…This can be achieved by using the measured sites as regressors and estimating prediction models for each of the unmeasured sites using the historical data. While linear and nonlinear regression approaches can be employed, linear models have proven to be sufficient in practice [27], and with much lower complexity are the preferred option. Given a set of input values v i ∈ R p×1 , and a target output variable…”

Section: B Linear Regression For Profile Reconstructionmentioning

confidence: 99%

“…(ii) Simulated case study [RBF] [27] -In this case study the wafer profiles have been artificially generated as a linear combination of Gaussian Radial Basis Functions (RBF) defined on the unit radius disk centered at the origin with additive Gaussian measurement noise. The model for the wafer height at coordinates x, y is given by…”

Section: Dataset Descriptionmentioning

confidence: 99%

“…A dynamic extension of FSCA was proposed in [27], referred to as Sequential Dynamic Sampling (SDS), where after the static selection, each unselected feature is associated with a cluster induced by the selected ones. Then at each process run, one feature is selected sequentially from each cluster guaranteeing a complete span of the available sites after a limited number of process runs.…”

mentioning

confidence: 99%

“…This approach presents some substantial limitations: on the one hand, at each iteration the group of measured sites is selected based on a similarity measure (e.g. correlation [27]) that generated the clusters, but the resulting reconstruction error is not considered. On the other hand, the number of possible combinations of measured sites is typically very large (proportional to the product of the cluster sizes) and there are no guarantees that the selected combination at each iteration is optimal in terms of reconstruction capabilities.…”

mentioning

confidence: 99%

See 4 more Smart Citations

Induced Start Dynamic Sampling for Wafer Metrology Optimization

Susto

Maggipinto

Zocco

et al. 2020

IEEE Trans. Automat. Sci. Eng.

Self Cite

View full text Add to dashboard Cite

Metrology, which plays an important role in ensuring production quality in modern manufacturing industries, incurs substantial costs, both in terms of the infrastructure required, and the time needed to perform measurements. In particular, in the semiconductor manufacturing industry, measuring fundamental quantities on different sites of a wafer surface is associated with increased production time. To increase metrology efficiency, a typical strategy is to limit the number of sites measured and to exploit statistical models (soft sensing) to reconstruct the wafer profile. Moreover, for quality reasons, spatial dynamic sampling strategies may be employed to ensure that all regions of a wafer surface are checked periodically during production. In this work, we propose a new sampling strategy, called Induced Start Dynamic Sampling (ISDS), that adapts greedy feature selection algorithms to the spatial dynamic sampling problem such that the number of measured sites at each process run is minimized while achieving good wafer profile reconstruction accuracy and process visibility. The superiority of the proposed strategy with respect to the state-of-the-art is demonstrated using both simulated data and an industrial chemical vapour deposition case study. Note To Practitioners-In this work we tackle a practical metrology problem encountered in semiconductor manufacturing, namely, the design of a dynamic wafer measurement plan to monitor the accuracy of a process across the whole wafer surface. The measurement plan is called 'dynamic', since, the measurement locations, which are drawn from a candidate set that provides coverage of the whole wafer, change at each process iteration. Our methodology addresses the challenge of finding an optimized trade-off between the number of measurements performed on each wafer and the reconstruction accuracy that can be achieved for the unmeasured areas on the wafer, while at the same time, for quality assurance purposes, ensuring that all locations on a wafer are visited in a finite number of process runs. The major benefit of the methodology is that it can significantly reduce the number of sites that need to be measured on each wafer enabling greater throughput on metrology tools without sacrificing process monitoring and anomaly detection capability.

show abstract

mentioning

confidence: 81%

Section: B Linear Regression For Profile Reconstructionmentioning

confidence: 99%

Section: Dataset Descriptionmentioning

confidence: 99%

mentioning

confidence: 99%

mentioning

confidence: 99%

See 3 more Smart Citations