Come to see our talks and posters at SPIE Advanced Lithography.
Improved sub-surface AFM using photothermal actuation
One of the key challenges for metrology and process control at front-end semiconductor fabrication is the ability to see through optically opaque layers. By the advent of 3D structures such as FinFETs and 3D NANDs, and ongoing miniaturization of the circuits, the conventional optical and e-beam based imaging techniques are facing physical limitations in inspecting the new IC components. One promising technology to overcome some of these challenges is the sub-surface Atomic Force Microscopy (AFM).
Recent developments in AFM have enabled its application for measuring the sub-surface features based on the effect of their mechanical properties locally present at the surface. Presence of a subsurface structure changes the effective elasticity and damping characteristic of the surface, and such a change can be measured using an AFM. Researchers have reported many different techniques for measuring these changes in mechanical properties of the surface. A common description of all these techniques is that a microcantilever with a nano-scale sharp tip is brought into contact with the surface of the sample and is scanned over the surface, while its contact resonance frequency is monitored. A shift in contact resonance frequency of the cantilever indicates a change in local mechanical properties, which is indicative for a sub-surface feature. This contact resonance frequency is normally measured indirectly using a amplitude modulation scheme.
One main challenge regarding all these techniques is that the frequency response function (FRF) of the cantilever in contact is commonly very vague. As shown in Fig. 1, there exist normally more than one peak at the amplitude signal, with highly ambiguous phase behavior.
In this paper, we use a different type of excitation which provides a much more clean FRF and therefore enables frequency modulation. In the proposed method, the AFM cantilever is exposed to a 405 nm laser beam with an AC modulated power, while its deflection is measured using a 830nm laser. Since the cantilever has a bilayer structure containing gold, it absorbs some of the 405nm laser power as heat and the changes of its temperature cause stress between the layers which bends the cantilever. As shown in Fig.1 with this technique, a very clean amplitude and phase signal can be captured.
Since the proposed techniques provides a more clean FRF, one can use a Phase Lock Loop (PLL) to quantitatively measure the resonance frequency of the cantilever and perform a frequency modulated experiment which was as elusive as impossible with acoustic excitation techniques. Fig.2 clearly depicts the advantage of using the proposed method over state of the art sub-surface techniques. Here, the resonance frequency histogram and cross-section of a subsurface image of a photoresist sample with sub-surface trenches is shown. As it can be seen, using the PTA in combination with frequency modulation, the sub-surface features can be resolved with much better SNR.
Quantitative tomography with Subsurface Scanning Ultrasound Resonance Force Microscopy
Extracting quantitative numbers about dimensions and material properties of buried structures is continuing to be an important but difficult task in metrology. Examples of questions needing this capability include critical dimensions of fins such as the profile (bottom width, top width, height) or the presence and extend of voids. In recent years TNO has shown the concept of using AFM in combination with ultrasound to image buried structures based on their (visco-)elasticity in a technique called Subsurface Scanning Ultrasound Resonance Force Microscopy (SSURFM). We have successfully imaged structures of less than 15nm wide as well as structures buried up to a micrometer deep. However, extracting quantitative results from this data is not trivial as the induced stress field in the sample depends on many parameters in often a non-linear way: experimental parameters such as applied force and tip size and shape, and geometry and material properties of the buried structures themselves. Therefore measurements based on this technique have a point spread function which varies in a complicated way with the sample properties that need to be measured. However, as we show, a solid understanding of the physics and mechanics involved and modeling of the expected structures and their response to externally applied stress, enable to obtain quantitative measurements. We specifically show our progress on characterizing a sample containing fins from a 7nm manufacturing test run.