When it comes to analyzing thin layers or multilayer coatings, glow discharge optical emission spectrometry (GD-OES) is one of the most powerful techniques for obtaining in-depth elemental profiles. However, until recently, a major challenge remained: how to accurately convert the sputtering time into an actual depth measurement, given that the erosion rate varies depending on the layers traversed? This is where the DIP (Differential Interferometric Profiling) accessory, developed and patented (WO15166189A1 & WO18119410A1) by HORIBA's Innovation team, comes in. It is integrated into the latest generation of GD devices and can be retrofitted to earlier devices. DIP brings unprecedented accuracy to GD-OES analyses, improves the reproducibility of results, and paves the way for a more detailed understanding of multilayer structures.
The DIP principle is based on differential interferometry, an optical method that enables continuous, real-time measurement of the depth of the crater formed by the discharge. By combining a reference laser beam with a beam directed at the eroded area, the system detects minute variations in light phase caused by material erosion.
The result is a direct, nanometric, and instantaneous measurement of depth, without relying on assumptions about the density or composition of the material.
The optical method used is as follows: a red laser (wavelength of 635 ± 5 nm) is split into two beams of orthogonal polarizations. The first beam is directed towards a reference area on the surface of the substrate that is unaffected by erosion. The second beam illuminates the center of the crater as it forms.
After passing through these two zones, the beams are recombined and analyzed by four photodiodes (S1–S4). These detectors measure variations in the phase and amplitude of the light signals, which directly reflect changes in the depth of the crater.
For reflective materials, the depth measurement is obtained directly using the relationship:
where D is the depth of the crater, λ is the wavelength of the laser, and Φ is the measured optical phase shift. Each phase variation therefore corresponds to an actual variation in the eroded depth.
A key advantage of the DIP accessory is its stability. The optical reference is taken directly from the surface of the sample itself, rather than from an external element, which minimizes thermal and mechanical drifts. Thanks to this design, measurements remain reliable even in the presence of plasma-induced temperature variations, ensuring nanometric accuracy throughout the analysis.
The DIP accessory therefore serves several purposes and offers a number of advantages. First, it replaces the estimated time-depth conversion, which is based on assumed erosion density and speed, with a direct and continuous measurement of the actual depth. DIP thus improves the quantification of elementary profiles and the accuracy of layer thickness measurements.
DIP is ideal for all plasma coatings (PVD/CVD) as well as for many types of opaque samples, including metal and ceramic coatings.
Some figures concerning the DIP accessory. Measurement accuracy is better than ±5% over a range from 100 nm to several tens of µm. It is in excellent agreement with measurements taken by an external profilometer: the average relative error is 2–3%. However, DIP provides a point measurement linked to the positioning of the laser; it does not provide information on the shape of the crater (concave/convex) or on roughness.
Here are some examples of applications:
· TiN on WC: measured depth 20 µm by DIP vs. 20.6 µm by profilometer (3% difference).
· TiAlN on WC: 9.5 µm (DIP) vs. 9.8 µm (profilometer).
Finally, the DIP accessory allows local reflectivity to be measured, which is useful for detecting buried layers (e.g., TiN under TiAlN) but also in simple cases for measuring the thickness of transparent optical layers.
In summary, the DIP accessory offers numerous benefits for GD-OES analyses, starting with direct time/depth conversion without assumptions about density. It thus provides greater accuracy in depth concentration profiles by significantly reducing the cumulative uncertainties of GD-OES quantification.
