Get the pdf download to your inbox:

No spam, a download link will be sent directly to you.

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.

Stay on top!

Get helpful articles and special offers once a month.

Using PiFM to Characterize Area Selective Deposition (ASD) of Thin Films

Key Takeaway:

  • Photo-induced Force Microscopy (PiFM) measurement confirms what AFM topography cannot do alone: the unintended deposition of thin films outside of the selected deposition regions.

Detecting Cobalt Films on ASD Test Chips

Visualizing the Area-Selective Deposition (ASD) of monolayers in ALD and CVD processes is a challenging task, especially in cases where the deposited layer shows no topographic features to distinguish it from the surrounding region. Chipmetrics solves this challenge with their ASD test chip consisting of sub-100 nm alternating material lines embedded into a planar silicon surface, enabling process teams to evaluate area-selective depositions with standard metrology tools (1).

Figure 1: ASD test-chip with alternating SiO2 and Si3N4 lines on a Silicon substrate (left). AFM topography (right) shows that Cobalt deposition results in a 6 nm height increase on the Silicon region compared to the SiO2/Si3N4 region1.

Photo-induced Force Microscopy (PiFM) by itself can also solve the same challenge via direct visualization of the nanoscale chemical differences of ALD/CVD layers based on the material’s infrared absorption signature. Thus, PiFM can be used to independently characterize the ASD test chips in order to demonstrate and verify the chip’s area-selective deposition functionality. Here, we perform PiFM measurements using Vista 75 AFM on a Cobalt deposited ASD chip and compare it with an uncoated chip. Specifically, we would like to see whether a Cobalt nano-layer can be detected over the alternating material lines region and thus verify its area-selective deposition.

AFM topography images (Figure 1) on the ASD test chip clearly shows that Cobalt deposition is non-uniform. The surrounding Silicon substrate region experiences a 6 nm higher Co deposition compared to the alternating SiO2/Si3N4 region. While these results indicate a selective deposition of Cobalt, it does not confirm if any Cobalt has deposited over the SiO2/Si3N4 lines.

Presence of Cobalt could in principle be detected directly with PiFM by looking for the Co-O stretch band between 500-700 cm-1, which requires a suitable tunable laser which can access that wavenumber range (e.g. OPO/DFG). However, it should still be possible to indirectly detect the presence of a Cobalt layer from the Photo-induced Force (PiF) signal intensity from SiO2 (Si-O stretch at 1100 cm-1) and Si3N4 (Si-N stretch at 950 cm-1). Presence of a Co layer would shield the PiF signal arising from the underlying material’s absorption bands, and thus by comparing the PiF intensities of a coated sample with an uncoated one, one can deduce over which regions Co was deposited. Such an indirect detection method is suitable for cases when the available tunable laser (e.g. MIRcat QCL, 770 – 1500 cm-1) can only access wavenumebers corresponding to secondary material components of interest, not the primary one (Cobalt in this case).

Figure 2: Topography and PiFM images at 950 cm-1 (Si3N4) and 1100 cm-1 (SiO2) wavenumbers measured on an uncoated (top) and coated (bottom) ASD test chip. An overlaid PiFM image combining the Si3N4 (red) and SiO2 (green) PiFM images is shown on the right for each case

Figure 2 shows how this indirect method was used to detect the presence of a Cobalt layer over the alternating SiO2/Si3N4 region. PiFM images on the uncoated ASD chip shows a strong signal contrast for both Si-O stretch (1100 cm-1) and the Si-N stretch (950 cm-1) absorption bands coming from the SiO2 and Si3N4 regions, respectively. However, on the Co coated chips, the PiFM images show a smaller intensity for both absorption bands, which indicates that a thin Cobalt layer must be present over the alternating line region, shielding the PiF signal.

We sincerely thank Mr. Thomas Werner (Chipmetrix GmbH) and Mathias Franz (Fraunhofer ENAS) for preparing and providing the ASD test chip as well as the schematics used in Figure 1.

Detecting Inhibitor SAM with PiFM

In another example of selective ALD, PiFM is used to determine whether an inhibitor SAM has selectively coated only the copper (Cu) lines without infiltrating the interlayer dielectric (SiO2 in this case) region. The cartoon shows the intended final configuration of the sample surface. The AFM topography and PiFM measurements are performed in the SiO2 region as indicated in the cartoon. AFM topography distinctly shows the interface between SiO2 and Cu regions by an abrupt change in height (on the right edge of the topography image). Based on the AFM topography alone, there is no clear indication if the inhibitor SAM has infiltrated the SiO2 region or not.  

Figure 3: Cartoon of alternating material lines displays intended result of selective ALD and the measurement range for AFM topography and PiFM measurements. PiFM images at various wavenumbers are correlated with AFM topography to confirm the presence of inhibitor SAM molecules in SiO2 region.

Collecting PiFM images at a few wavenumbers confirms that the inhibitor SAM has infiltrated the SiO2 region. PiFM image at WN1 directly highlights the inhibitor SAM and displays its presence in the SiO2 region, especially near the interface.  Examination of the area in the yellow ellipse shows that the topography associated with the inhibitor SAM is taller than the background, a likely indication of small clusters of inhibitor molecules.  As further confirmation, the PiFM image at WN3, which should highlight the ALD layer, shows lower signal at those locations where inhibitor clusters are observed in the PiFM image at WN1; the presence of inhibitor molecules in the SiO2 region prevented the deposition of ALD even in the SiO2 region. The combined PiFM image shows the precise alignment of the inhibitor SAM (green) and the lack of the ALD layer (purple).  These clusters are all less than 2nm in height which indicates they contain only a small number of inhibitor molecules. 

PiFM Detects Thin Film ASD

In both cases examined here PiFM provided the necessary information to confirm the deposition of thin films in unintended regions. AFM topography alone of the selected regions could not confirm or deny deposition. On Chipmetrics test chips an indirect PiFM method was used to confirm the presence of a Cobalt layer over the alternating SiO2/Si3N4 region by showing a decrease in Si-O stretch (1100 cm-1) and the Si-N stretch (950 cm-1) absorption bands signal. On the Cu/ SiO2 sample, a direct PiFM method was used to confirm the presence of an inhibitor SAM on the SiO2 region by imaging at wavenumbers that directly highlight the inhibitor SAM and ALD material. PiFM’s sensitivity and sub-5 nm spatial resolution position it as a powerful metrology tool for ASD processes.  

References

  1. Chipmetrics. ASD-2 Test Chips with Customizable Metal-Dielectric Material Stacks, 2026. https://chipmetrics.com/chipmetrics-announces-new-asd-2-test-chip-with-customizable-metal-dielectric-material-stacks/.
  2. Franz, M.; et al. Low-Temperature ALD of Metallic Cobalt Using the CoCOhept Precursor: Simulation-Assisted Process Development for Deposition on Temperature Sensitive 3D-Structures. J. Vac. Sci. Technol. A 2025, 43 (2), 022412. https://doi.org/10.1116/6.0004248

Interested in a niche application?

Ask us, we may have already studied it.