Logo PTB

Entwicklung und Kalibrierung eines neuen EUV Referenz-Fotomasken-Normals mittels CD-AFM


In collaboration with the company Advanced Mask Technology Center (AMTC) in Dresden, a new EUV photomask standard has been developed recently. The EUV photomask was manufactured by the Advanced Mask Technology Center (AMTC) using state-of-the-art lithography processes to pattern the absorber of a regular EUV blank, i.e. a quartz substrate with EUV multilayer-mirror and an absorber stack including a backside metallic coating. Different dimensional parameters such as the critical dimension (CD, the structure width), CD uniformity (CDU), pitch, height, sidewall angle, line edge/width roughness (LER/WR) etc. of the developed photomask standards were calibrated accurately and traceably using the CD-AFM and TEM techniques.

A home-built CD-AFM at PTB has been applied for the calibration of this reference sample. To achieve better CD measurement performance, a special probing strategy ─ referred as vector approach probing (VAP) ─ was developed. Using the VAP method, the structure is probed in a point-wise manner by an AFM tip approaching along a 3D vector path, which is typically given by the normal vector of the surface. The tip sample interaction curve is recorded for calculating the measurement result at each point. Due to the VAP method the probe wear is rather small, too. It has been shown that the wear is less than 1 nm over more than 500 repeated measurements of a feature. Also, the reproducibility of repeated feature characterizations is better than 1 nm. A typical CD-AFM measurement result is demonstrated in figure 1. The traceability path of the AFM measurements is realized in two aspects: the scaling factor of the AFM scanner was calibrated by a set of step height and lateral standards certified by metrological AFMs, while the effective tip width was ultimately traceable to the lattice parameter of silicon via High Resolution Transmission Electron Microscopy (HR-TEM) approach. This calibration approach is non-destructive, and currently offers an expanded measurement uncertainty of U = 2.5 nm (k=2) for the CD calibration.

To verify the calibration performance, a photomask which had been calibrated by the CD-AFM at PTB was cleaved and measured by a HR-TEM at the university Duisburg-Essen. The scaling factor of the HR-TEM measurements were determined by the pitch of the structures, which is also traceable to the metrological AFM of the PTB. A typical STEM image and its data evaluation is illustrated in figure 2. For all six groups of structures investigated (with nominal CD values of 120 nm, 140 nm, 160 nm, 180 nm, 200 nm and 320 nm), the deviation is as small as 1.1 nm between the CD-AFM and HR-TEM results. It is well below the measurement uncertainty and thus confirms the calibration performance of the developed method. It should be mentioned that the HR-TEM approach is firstly very time-consuming due to its sample preparation and secondly destructive due to the part of the sample being cut out by the Focused-Ion-Beam (FIB) technique for the investigation by HR-TEM and complicated measurement procedures needed. Therefore, it cannot be directly applied for calibration purpose.

The calibrated photomask standard is currently being applied at AMTC to transfer the standard to photomask production. In industry, the CD was measured by means of a commercial AFM (Bruker 3D-AFM) that provides the capability to serve as a “golden tool” for the whole CD tool fleet of the AMTC fab (mainly SEMs), thus playing an important role for quality assurance in state-of-art EUV photomask manufacturing.

AFM image measured on the DD pattern

Fig.1 AFM image measured on the DD pattern with nominal CD of 80 nm of a photomask standard, shown as the raw data without tip correction in (a); and the evaluation of a cross-sectional profile for middle CD and sidewall angle as shown in (b).

STEM image on a pair of line feature

Fig.2 A STEM image on a pair of line feature (nominal CD of 120 nm) of the EUV photomask. The pitch value (P = 239.92 nm) of the feature pair calibrated traceably to the metrological AFM was applied to determine the scaling factor of the STEM image. The CD of the left line feature is thus determined as 103.8 nm, illustrated as an example in the figure.