Specific Process Knowledge/Characterization/Stylus Profiler Measurement Uncertainty
Stylus Profiler Measurement Accuracy
The accuracy of a height measurement with the profiler depends on the measurement settings, the sample, the step or trench shape and height, the instrument calibration, and the resolution.
Optimal Measurement Settings
Both the force setting and the scan speed are important: Too high force may compress a soft material like Al, Au or some polymers, while too low force may lead to the stylus "jumping" over features, especially if the scan speed is high. Too low scan speed may result in drift of the measurement and noise from vibrations while too high scan speed may mean that the stylus tip does not have time to reach the bottom of the features you are measuring and can also give rise to increased noise levels (see the Dektak XTA manual on LabManager, Figure 3 for details).
A sharp vertical step is easiest to measure. If the step is gradual or the surface is rough, it can be difficult to determine where to measure and how the scan should be leveled. For very small steps, underlying scan noise can also make it hard to level properly and will convolute with the shape of the step. See further below under Total Uncertainty for steps < 1 µm.
Calibration Standard Uncertainty
Nanolab staff periodically check most of the the stylus profilers' measurement accuracy with a standard step height of 923 nm for the smaller ranges and 23.058 µm (previously almost 25 µm) for the larger ranges, so that the middle ranges are checked with both standards. The 95 % confidence intervals for the standards are 18 nm for the 917 nm standard and 0.068 µm for the 23.058 µm standard. If the control measurement is beyond the limit set in our Quality Control procedure, the instrument is calibrated and the users informed (see LabManager for details, for instance the Dektak XTA control instruction and control measurement data)
This means the 95 % confidence interval of a 1 µm step measured with the smallest measurement range is at least the 1.8 % error of the standard step while the 95 % confidence interval of a 25 µm step measured with a larger range is at least the 0.3 % error of the standard step. Steps between 1 and 25 µm measured with the intermediate ranges will presumably have an intermediate error just due to the intrinsic uncertainty on the standard step height. See Figure 1.
Total Uncertainty for steps > 1 µm
Apart from the error due to the standard step height's intrinsic uncertainty, there is a contribution to the overall uncertainty from the deviation of the value that the profiler measures from the theoretical height of the standard step. This can be called "QC error" since we accept some variation in our quality control routine. In practice for the 917 nm standard this range is about ± 4 nm for the Dektak XTA and ± 8 nm for the P17, while for the large standard it ranges from about 0.06 µm on the Dektak XTA to about 0.15 µm on the P17. You can see this deviation in the QC data in LabManager for a given instrument.
There is also a tiny contribution to the total error from the instrument's limited resolution and finally of course there is random noise in any measurement. For many repeated measurements of the same line on the standard step height (a rigid, well defined vertical step) we have found the random error from the Dektak XTA is on the order of ± 5 nm for the 917 nm standard and 0.05 µm for a 24.925 µm standard. The random error with the P17 profiler is even smaller (about 0.001 µm).
To estimate the overall accuracy of the profiler's measurements you can convolute these various sources of error. The error sources are shown graphically in Figure 2 for measuring the small standard step with the Dektak XTA's smallest range. You can see an uncertainty budget for the Dektak XTA measurements here (made by Rebecca Ettlinger in 2020): Uncertainty budget Dektak rev.xlsx. It is based on the assumption that all the error sources are independent and can therefore be added by the sum of squares method.
The resulting error calculation for the Dektak XTA of a 1 µm very well defined standard step is about 2 % (as the uncertainty on the calibration standard dominates), while for a very well defined step of 25 µm the cumulative error is about 0.7-1 %. However, in real devices the random error will often be much larger than for our standard samples and so the real confidence interval will be larger.
To calculate the accuracy of your particular measurement, you should repeat the measurement several times and estimate the standard deviation. If the scatter is quite small you can try to calculate the total error on your measurement using the sum of squares method to combine the intrinsic step height error, the QC error and the random scatter error. If the scatter of your measurements is large that will be the dominant source of error in your measurement and you can safely ignore the other contributions.
Total Uncertainty for steps < 1 µm
Apart from the uncertainty described above, the underlying "shape" of the scan due to the instrument itself will influence the measurement accuracy for steps <1 µm. Although the stylus profilers use an optically flat surface as a reference, the basic underlying scan is never truly completely flat. We have in June 2025 found both random and repeatable scan bowing / noise in the range of 10-60 nm across a 2 cm (2000 µm) scan. This is enough to influence the measurement accuracy of steps <1 µm and especially <500 nm.
We therefore cannot recommend using the stylus profilers for measuring steps <100 nm and one must expect relatively large uncertainty on steps in the range of hundreds of nm. See further discussion below.
If measuring steps <100 nm it is important to check a measurement of a "flat" area nearby and to play with positioning the measured step in different locations along the scan length (to show variation in the underlying scan noise right under the step).
The figures on the right show representative "flat" scans made with the P17 and DektakXT. The shape is influenced a lot by the position of the leveling cursors and a bow-like shape as the one seen in the P17 scan can easily be found in DektakXT scans of flat surfaces as well. More "flat" scans for the Dektak and P17 can be seen here: File:Dektak XT and P17 scan flatness comparison summer 2025.pptx.
How much flatness can we expect of a "flat" scan?
KLA, the manufacturer of the P17 profiler, provides some information on the flatness one can expect from the scans, some of which is freely available in this document. Here we see that we can expect a variation of up to 40 nm across a flat scan of 3 cm. Thus the variation of up to 40 nm across 2 cm that we have seen in practice is not far beyond the presumably best-case scenario stated by the manufacturer (our tool could use some lubrication of the scan axes at the time of measurement).
For the P17 the underlying scan noise or bow is extremely reproducible for many scans in the same position whether or not the vacuum hold is turned on. Therefore it seems that the non-flatness of the scans derive from some underlying structure on the optically flat surface or the scan rails.
For the Dektaks, we have in some cases seen that the scan noise is reproducible not only for scans in the same location, but even for scans in different locations on the stage, which must be due to some kind of dirt on the "feet" of the stage as they move across the optical flat. However in other cases, as in all the scans shown in the document above, the bowing of the scan varies even though the scan coordinates and sample are the same. The latter might be because the stage positioning is a little less accurate than for the P17 or because there is more environmental noise influencing the measurement.
How does the underlying scan noise affect the measurement accuracy?
The step determination will usually be better than the noise of the underlying scan - we don't expect an error in the measured step sizes as big as 30-60 nm. This comes down to removing some of the noise due to averaging parts of the scan during leveling and measuring. Also, usually it is not necessary (or recommended) to measure such steps across more than 1 cm. The closer together the measurement points, the "flatter" the underlying scan will be.
For example, in the DektakXT we regularly measure shadow masked metal films made by e-beam evaporation that are around 100 nm thick. We have found good agreement between the DektakXT measurements of a 500 µm wide step and XRR measurements of films of the same thickness, so we see no reason not to trust these measurements to within around +/- 5 nm. We level the scan with the leveling cursors close to the step before and after it, so only the scan noise in an ~700 µm range is important. Additionally, since we average the thickness across approx. 400 µm at the top of the step compared to the ~40 µm leveling intervals nearby (where the height is also averaged), the variation comes out much smaller than what is seen across a full 2000 µm scan.
In practice we have found the step heights around 100 nm can be measured with higher accuracy in the DektakXT than the P17: We tend to get a measurement with the DektakXT within a few nm of the XRR measurement while the measurement with the P17 has been off by around 10-15 nm. We don't yet have a good explanation for this - nor do we have enough data to prove it's a general trend.
It may be even better to just level with cursors near each other on what one knows is a flat surface and then measure just before and after the step very near the leveling cursors to get the best height estimate.