Using Principal Component Analysis for Photomask CD signature investigations G. R. Cantrell, Christian Bürgel, Axel Feicke, Martin Sczyrba and Clemens Utzny Advanced Mask Technology Center GmbH & Co. KG, Rähnitzer Allee 9, 01109 Dresden, Germany ABSTRACT Reticle critical dimension (CD) errors must be minimized in order for photomask manufacturers to meet tight CD uniformity (CDU) requirements. Determining the source of reticle CD errors and reducing or eliminating their CDU contributions are some of the most relevant tasks facing process engineers. The AMTC has applied principal component analysis (PCA) to reticle resist CD measurements in order to examine variations in the data. PCA provided the major components of resist CD variation which were rescaled into reticle CD signatures. The dominant component of CD signature variation is very similar in shape and magnitude between two different chemically amplified resist (CAR) processes, most likely indicating the variation source is a common process or tool. CD variational signatures from PCA were used as a basis for launching investigations into potential reticle CD error sources. PCA was further applied to resist CD measurements from alternate process tools to assist efforts in judging the effectiveness of resist CD signature matching. Keywords: critical dimension (CD) errors, CD uniformity (CDU), principal component analysis (PCA), chemically amplified resist (CAR), stability 1. INTRODUCTION Reticle critical dimension uniformity (CDU) continues to be one of the most challenging specifications for photomask manufacturers. Each new wafer technology node has reticle CDU specifications tighter than its predecessor, requiring the elimination or at least the minimization of systematic reticle CD errors, such as center to edge, side to side effects, and localized CD deviations, often referred to as hot/cold spots. For some CARs, the optimization of resist bake and/or develop processes have proven effective in lowering reticle CDU. 1,2,3,4 Other methods such as feedback correction strategies have also been employed when the source of reticle CD error is unknown. 5 Despite these improvement techniques sub 45nm wafer technology nodes demand mask manufacturers fulfill 2nm 3σ CDU specifications, 6 which will require novel approaches for photomasks manufacturers to elucidate the source of CD errors as well as eliminate or minimize their impacts. To this end, the AMTC examined resist CDU signatures using PCA, a statistical technique for revealing patterns in large data sets. 7,8,9 While PCA examines only variations in a data set the resulting components of this variation can be converted into reticle CD signature variations giving process engineers evidence of potential CD error sources. PCA enabled the identification of systematic resist CD signature variations on two different CAR systems which were not evident by other examination methods. The shape and magnitude of the dominant CD signature variations were very similar between two different CAR processes suggesting a common process or tool as the potential error source. PCA was similarly applied to determine the effectiveness of resist CD signature matching using alternate processes tools such as resist develop and post exposure bake (PEB). This report presents AMTC’s application of PCA to photomask resist CD signatures. 2. EXPERIMENTAL The AMTC processes test reticles at regular intervals in order to gauge line performance and stability. Commercially available 193nm phase shift blanks with widely used positive and negative CARs (pCAR, nCAR) are written with 50kV e-beam (PG) lithography and processed identically to customer reticles, except without the use of feedback  variational  line  require  to