Dear Dr. Pranjala Tiwari,

XPS analyses are able to provide qualitative / quantitative information regarding the elements present in the extreme surface area (2-5 nm). This interval depends upon the kinetic energy of the electrons and the material being analyzed In your case, knowing the composition of the different layers that make up your sample, performing a "weak" and controlled etching, you could go back - by consulting the suitable scientific bibliography - to an "approximate" thickness. You could also try Angle-resolved XPS (ARXPS); it is a technique that varies with the emission at the electrons are collected, thereby enabling electron detection from different depths. ARXPS provides information on the thickness and composition of ultra-thin films. Such measurements are non-destructive, unlike sputter profiling.

For further information, you can see at: https://xpssimplified.com/angle_resolve_xps.php

Best regards, Pierluigi Traverso

Dear Dr. Pranjala Tiwari,

XPS analyses are able to provide qualitative / quantitative information regarding the elements present in the extreme surface area (2-5 nm). This interval depends upon the kinetic energy of the electrons and the material being analyzed In your case, knowing the composition of the different layers that make up your sample, performing a "weak" and controlled etching, you could go back - by consulting the suitable scientific bibliography - to an "approximate" thickness. You could also try Angle-resolved XPS (ARXPS); it is a technique that varies with the emission at the electrons are collected, thereby enabling electron detection from different depths. ARXPS provides information on the thickness and composition of ultra-thin films. Such measurements are non-destructive, unlike sputter profiling.

For further information, you can see at: https://xpssimplified.com/angle_resolve_xps.php

Best regards, Pierluigi Traverso

Yes that possible but you should try to measure it with UV Vis or atomic force microscope

Hi Pranjala,

Pierluigi Traverso is perfectly right, the only thing that I would add is that you should be very careful when evaluating the thickness of a multicomposite just based on the sputtering time. You could have preferential sputtering of some elements with respect to others according to the probe you use, so same thicknesses of different elements could be sputtered in different times - hope this makes sense.

You can use the sputtering time for a "rough" estimation of the thickness, but it would be better to back it up with AR-XPS or, if you're lucky enough to have one, with cross-sectional TEM. There's a slightly fancier option, called energy-resolved XPS, but for that you need to have access to synchrotron radiation facilities.

Best of luck with your project!

Iolanda

Dear Dr. Pranjala Tiwari,

I am in complete agreement with what was said by Dr. Iolanda Di Bernardo. The evaluation of "abrasion" due to sputtering is particularly complex, if a comparison standard was available (sample with layers of the same or similar chemical composition of those to be measured and with known thickness) a useful "calibration" of the sputtering system could be made .

Best regards, Pierluigi Traverso

I would recommend to go through the works of H.-H. Strehblow and coworkers (Hoppe, Collisi, Druska, Borthen, Jabs) about the inner structure of passive films on metals - they have developed adequate model for angle-resolved data analysis, and I remember a work on Ni-Cr allows in J. Electrochem. Soc. (DOI: doi: 10.1149/1.1837577 ), where the details of the evaluation procedure are nicely described! Good luck, Dirk

This is a classic example of the gap between textbooks and practical experiments. The textbooks often make thickness analysis by XPS appear straightforward, whereas in practicality only a relatively small number of works are able to perform this correctly and accurately, and only for the most ideal films. This is true both for sputtering and for AR-XPS, and it requires many assumptions to be made. Some are listed below.

Conversely, XRR works very well for such thicknesses (as long as the layer is continuous). It takes us 10 min measurement + 10 min alignment to get publish-grade data. If it's only a couple of samples - you're welcome to contact me and we'll be happy to help with the measurements for free ( [email protected] )

Just a few of the assumptions required for using sputtering in XPS to calculate thicknesses:

- You need to assume the sputtering rate is uniform for the different layers

- You need to assume the photoelectron attentuation is uniform across the layers and the different elements (or you can model it, using additional assumptions)

- You need to assume the crater roughness during the entire experiment remains small enough to be neglected

- When you sputter compounds, you need to assume an identical sputtering rate for all the elements in the compound. This is an often-overlooked assumption which is sometimes (e.g. oxides) grossly incorrect.

AR-XPS also has its quirks, most notably the roughness has to be much smaller than the other parameters that you're looking at (or you can model it using further assumptions). Also, unless you're using a synchrotron source, the range of thicknesses where this works is very narrow.

From your question, I am assuming that you are looking for input using XPS and not using other techniques that might be more definitive. Knowing the identity and order (from surface to deepest) of the layers in your multilayer film is helpful, but not necessary. You may also consider the following:

1. Throughout the sputter profile, the thickness of each layer removed is equal to the time for that layer multiplied by the sputter rate for that film. Lists of sputter rates (Å/min) or sputter yields (Å/incident ion) for elemental samples are available (typically from sputter target providers). Therefore,

Total thickness of multilayer = (time1) x (sputter rate 1) + time2 x rate2 + time3 x rate3 + etc.

If the layers are elemental, this is relatively straightforward. If layers are compounds this is much more complicated and corrections for preferential sputtering should ideally be included. Preferential sputtering becomes more problematic, however, with thicker rather than thinner layers.

Ideally this requires knowing the current density of your sputter beam, but if your system has a Faraday cup or other current monitor, this is easy to get. It also assumes your layers have bulk property sputter rates (big assumption).

Another caveat is choosing from your profile where to take the time for each layer – e.g. at half height of signal, or something else. You can increase your depth resolution or sensitivity (make sharper profiles) by using lower energy and/or lower current density sputtering. When possible, use the highest binding energy peak (of sufficient intensity) for each element because these are the lowest kinetic energy, and thus signify shallower origin.

2. Another alternative is to use the QUASES program, in which case all you need is a single survey spectrum (at largest pass energy) – best signal to noise – at normal take-off angle. You will need to learn how to use this software, but it is very powerful. It uses the inelastic background to the elastic elemental XPS peaks to fit to depth and thickness, etc. You are fitting your data to a model. There are work arounds with this even if you have elemental peak overlaps.

Hope this is helpful to you.

Good luck with your research.

Good intervention

Absolutely right, but depend on the composite working on.

I would go with X-ray reflectivity measurements. XPS is not as straightforward as explained in an earlier reply.

In theory you can. The sputter rate of individual elements and compounds differ from each other. So if you know the sputter rate of your material with given amount of argon and voltages, you can in principle get the amount of material that is sputtered. Given the geometry of your sample, you can get the thickness of the sputtered material.

However, there are some serious considerations involved:

• The material you are sputtering is extremely uniform. XPS is an extremely surface sensitive technique, if there is some imperfection in uniformity, it will reflect in your data.
• Sputter rates are calculated for extremely pure substances. You must keep in mind that there are some adsorbed oxygen and adventitious carbon in your sample surface (unless you experiment in-situ).

There are some ways you can avoid some of these flaws. One of them is to use smallest possible samples. This ensures better uniformity in sputter etching. Another technique is very much used for depth profiling is the Angle Resolved- XPS. However, to correlate your angle to the information depth in your samples you need to carry out some simulations or some theoretical work.

In either case, a direct correlation with thickness is not possible. You have to correlate the results with some known quantity to calibrate your thickness.

XRR also has substantial disadvantages - you need to have a contrast (i.e. different indices of refraction) between the different layers, otherwise there is no reflection at the inner interfaces. See the example with a TiO2 layer on a float glass substrate. the thickness increases from scan "0" for a pure glass substrate to about 5 nm for the last curve "8" ...

And, with XRR it is hard to determine light elements quantitatively - since e.g. O, N, C have a very similar refractive index for hard X-rays far above the respective K absorption edges...

Best regards, Dirk

Hi Dirk,

This is absolutely correct (I believe that the electron density, or average Z is the right parameter, because we're talking about refractive indices for x-rays). At least in my field (thin oxides), I didn't get to encounter many examples where there was no contrast across the interface (and you can always replace the substrate to circumvent this).

Lior

If you know the sputtering rate (nm/s), then you can calculate the estimated depth in nm. Then plot oxygen profile with respect to estimated depth, half of the maximum concentration of oxygen locates the estimated depth of substrate/film interface.