In situ spectroscopic ellipsometry and ALD
– How ellipsometry became a popular technique to monitor ALD film growth

Serendipity is often cited as a key factor in scientific innovations. I can fully confirm this as will also be obvious from the subtitle of this blog site: “Blog about strategy, serendipity and vision in nanoscience”. From my point of view, the fact that in situ ellipsometry has become a popular technique to monitor ALD film growth has for a large part been a matter of serendipity as I will explain below.

Before I started working on ALD, my research focused especially on plasma deposition (or plasma-enhanced chemical vapor deposition, PECVD) of thin films. Monitoring film growth during the deposition process was an important aspect of our work. It allowed us to study the dynamic processes during film growth and to speculate about surface processes taking place. For this reason, we built our own ellipsometers. These were single wavelength ellipsometers using a HeNe laser (632 nm). We started with a rotating analyzer version and subsequently we built a rotating compensator ellipsometer. Early 2000, we got the opportunity to buy a spectroscopic ellipsometer from J.A. Woollam Co., one of the type M-2000. At first, the wavelength range was 250 – 1000 nm (1.2 – 5.0 eV) but after a while we had the wavelength range extended to the near-infrared (250 – 1700 nm, 0.7 eV – 5.0 eV). So this ellipsometer was available in our lab and at that time it was mostly used to study deposition processes of silicon-based thin films, by PECVD and also by hot-wire CVD [1,2].

In 2002, I decided to start working on ALD after visiting the AVS 2^nd International Conference on ALD in Seoul, Korea. Through a proposal submitted to the Dutch Technology Foundation STW, we received money to build an ALD reactor, more specifically a remote plasma ALD reactor as I decided that plasma ALD would be the focus of our work. After receiving some advice by Dr. Suvi Haukka from ASM Microchemistry, Stephan Heil – my very first PhD student on ALD – and I decided that the first material we were going to develop by ALD was TiN. The process that we employed was based on TiCl₄ and H₂-N₂ plasma. However, when preparing for these experiments – our first ALD experiments ever – we realized that we had no direct method to see whether we were growing a film or not. We were unexperienced in ALD and we realized that it might not be trivial at all to have growth in our initial experiments with this new reactor. It might take a while before we would find the “sweet spot” (at that time we didn’t know that the process wouldn’t be too complicated after all). We preferred not to depend solely on ex situ experiments such as Rutherford Backscattering Spectrometry or the like as this would make the process optimization very slow and expensive. It would have meant that we would have to wait a couple of days every time we had made a new sample.

At some point I thought it might be possible to use spectroscopic ellipsometry on the TiN films. I thought “Okay, the TiN films will be metallic but if they are very thin, they should still be transparent for light of the ellipsometer”. At that time, we were still somewhat naïve regarding optical measurements on metallic films (we thought “metallic = non-transparent”). We had no clue that one can do lots of interesting optical measurements on such films. Only when reading the papers of Patsalas and Logothetidis about spectroscopic ellipsometry on PVD TiN  [3,4], we realized that ellipsometry would be a powerful tool for studying ALD growth of TiN. We immediately fitted ellipsometry ports on our new ALD system (see figure below) and there we went!

The first TiN ALD experiments we did were very successful. We got film growth, we had self-limiting reactions and our results were in line with previous results on ALD of TiN. Moreover, the in situ spectroscopic ellipsometry measurements turned out to be very exciting. We were able to measure the thickness of the films and we “discovered” that in situ spectroscopic ellipsometry measurements are a very powerful method to quickly check for saturation. Interestingly, with in situ spectroscopic ellipsometry we were also able to monitor initial film growth and thereby check for nucleation delays. These nucleation delays were an important topic at the time (and they still are, especially due to the hot topic of area-selective ALD (see our ASD blog)). And last, but not least, with spectroscopic ellipsometry we could determine the dielectric function of the TiN and subsequently – with Drude’s theory – we could extract information on the electrical properties of the films: we had a way to determine the resistivity of the films in situ, which we could compare to ex situ measurements with a four-point probe. We presented this work during the AVS 4^th International Conference on ALD in Helsinki in 2004 (see slide below) and it was written down in our first ALD paper that we published in 2005 [5]. Together with my PhD students Stephan Heil and Erik Langereis, who played a very important role in the ellipsometry work, we had made our first steps in the world of ALD! Soon after that, several other publications followed with in situ spectroscopic ellipsometry in a central role, for TiN [5,6] but also for TaN [7]. And obviously, many more papers have followed afterwards. I guess nearly all of our publications about ALD contain in situ spectroscopic ellipsometry results.

Slide as presented during the AVS 4^th International Conference on ALD (ALD 2004) by Stephan Heil, Erik Langereis, Richard van de Sanden and Erwin Kessels. This was our first conference in which we presented ALD results.

One other crucial step was taken in 2005/2006. In 2005, I was approached by Oxford Instruments as they were interested in launching a tool for plasma ALD. Although they had ample experience in the field of plasma etching and deposition, they were new to the field of ALD and they approached three people to serve as consultants: Dr. Steve Rossnagel (U.S.A.), Prof. Hyeongtag Jeon (Korea) and I (Europe). It is a story by itself (something for another blog post) but the end result was a new tool for remote plasma and thermal ALD: the Oxford Instruments FlexAL. In the discussions, we had with them they also asked which lab would be well suited as a beta-site for their first tool. I immediately saw a great opportunity and informed them that we, the Eindhoven University of Technology, would be very much interested. This is how it happened (after some negotiations) that their first tool was installed in our NanoLab@TU/e clean room (see press release below). When building the tool, they asked me whether I had certain wishes and I told them that it would be great if they could fit ellipsometry ports on the tool. At first they were quite hesitant but in the end they agreed. And I guess they have never regretted it. It formed a great unique feature on their tools and many of the systems they have sold have been equipped with ellipsometry ports. They even started a collaboration with J.A. Woollam Co. (see the article in their newsletter below) such that customers could buy an ellipsometer already fitted to the tool. So it has also been a good deal for J.A. Woollam Co., as they have confirmed at several instances.

Press release by Oxford Instruments to announce the installation of a beta version of their FlexAL thermal and plasma ALD tool in the clean room of the Eindhoven University of Technology (15 March 2006). The J.A. Woollam Co. M-2000 ellipsometer fitted to the tool is visible in the photograph.

Newsletter by J.A. Woollam Co (issue 10, 2009) describing the collaboration between J.A. Woollam Co. and Oxford Instruments. The CompleteEASE®  software is integrated within the Oxford Instruments operating software.

To give an example from our own lab: we currently have 7 ALD tools (3 FlexALs, 1 OpAL, 3 home-built systems) and all of them have ellipsometry ports on which one of the 5 ellipsometers (all J.A. Woollam Co. M-2000) can be fitted. And although the Oxford Instruments FlexAL reactor was the first commercial reactor with in situ ellipsometry, several other ALD equipment manufacturers have followed in the mean time. As far as I know, in addition to Oxford Instruments, Veeco Cambridge Nanotech, Picosun, Kurt J. Lesker and Sentech have sold ALD tools with the option to fit ellipsometry (click on the links to see the related websites)

As alluded to, we have published many papers on ALD in which in situ spectroscopic ellipsometry plays a major role. Of course this holds also for many research groups (see e.g., the CoCoon group at the Ghent University). One of the important publications is the topical review paper that was written by Erik Langereis in 2009: In situ spectroscopic ellipsometry as a versatile tool for studying atomic layer deposition [8]. This paper addresses the use of in situ spectroscopic ellipsometry for metal oxides and metal nitrides (the materials covered in this review are: Al₂O₃, HfO₂, Er₂O₃, TiO₂, Ta₂O₅, TiN and TaN_x). In addition to some basics about ellipsometry and the fitting of dielectric functions, it covered aspects such as:

• ALD thickness and growth-per-cycle
• ALD saturation curves
• Initial film growth
• ALD half-cycle reactions
• Electron-impurity scattering in ALD films
• Size effects in conductive ALD films
• Distinction of material phases
• Amorphous to crystalline phase transitions during ALD

Moreover, quite recently we extended this work with a paper about in situ spectroscopic ellipsometry during ALD of metals (Pt, Pd, Ru) by Noemi Leick et al. [9].

Although we did a lot of work on in situ ellipsometry, which was for a large part a case of serendipity, we were certainly not the first ones reporting on in situ ellipsometry during ALD. After we published our first papers, I stumbled upon an article from the George group. In a paper about ALD of SiO₂ [10] from Klaus et al., they described already the application of in situ spectroscopic ellipsometry (probing 44 wavelengths) for the monitoring of film growth (see figure below). Lucky enough we found this paper before publishing our review paper such that the following sentence could be added to the review: “The concept of in situ spectroscopic ellipsometry for thickness monitoring during an ALD process was already reported in the late nineties by Klaus et al.; however, the application of in situ spectroscopic ellipsometry did not settle in ALD research at that time.” Note that Klaus et al. also used in situ spectroscopic ellipsometry to study ALD of tungsten [11] and tungsten nitride [12].

Figure in the paper by Prof. Steven George and coworkers (Atomic layer controlled growth of SiO₂ films using binary reaction sequence chemistry, by Klaus et al., Appl. Phys. Lett. 70, 1092 (1997)) [10]. The figure shows the thickness of SiO₂ as a function of the number of ALD cycles.

But to my great surprise, there was even earlier work. Last year during the 14^th International Baltic conference on ALD [http://bald2016.ru/] in St. Petersburg Russia, Prof. Victor Drozd of the St. Petersburg State University reported on the history of molecular layering (ML; the name ALD had in Russia during the time of its invention in the 70s). On one of the slides he showed a scheme of a ML-ALD device with in situ ellipsometer and quartz microbalance. This scheme came from a paper published in 1980! Please read more about the ellipsometry and the interesting story about the early ML-ALD work in the paper “Progress in device from molecular layering to atomic layer deposition worldwide technology” by Prof. Drozd [13].

Figure shown by Prof. Victor Drozd during his presentation the 14^th International Baltic conference on ALD about the early days of Molecular Layering in Russia in the 70s [13].

Clearly in situ ellipsometry is a powerful tool for monitoring ALD growth and obtaining a quick insight into the ALD process characteristics and material properties. Prof. Drozd and Prof. George and their co-workers implemented in situ ellipsometry already at a very early stage however the real potential of the method as a tool for studying ALD was only realized later. Not being very modest here, I consider the use of in situ spectroscopy ellipsometry in the field of ALD as one of our success stories. From discussions with colleagues in the ellipsometry industry, it is estimated that well over 50 ALD chambers already include in situ ellipsometry. That ALD is an interesting market has also been realized by this ellipsometry industry. In addition to J.A. Woollam Co. and Sentech, there have recently also been announcements by the company Film Sense, one related to the installation of an ellipsometer on a Beneq ALD reactor and one related to the installation of an ellipsometer on a Kurt J. Lesker ALD reactor. Interestingly, J.A. Woollam has announced a new ellipsometer which is slightly more basic than their M-2000 systems (and therefore lower priced) but which should be perfectly suited for most ALD studies. This ellipsometer was showcased at the recent 17^th International Conference on ALD and is pictured below.

Figure showing the new, compact iSE spectroscopic ellipsometer from J.A. Woollam on a mock chamber.

References

[1] Substrate temperature dependence of the roughness evolution of HWCVD a-Si:H studied by real-time spectroscopic ellipsometry, W.M.M. Kessels, J.P.M. Hoefnagels, E. Langereis, and M.C.M. van de Sanden, Thin Solid Films 501, 88 (2006).

[2] Real-time study of a-Si:H/c-Si heterointerface formation and epitaxial Si growth by spectroscopic ellipsometry, infrared spectroscopy, and second-harmonic generation, J. J. H. Gielis, P. J. van den Oever, B. Hoex, M.C.M. van de Sanden, and W.M.M. Kessels, Phys. Rev. B 77, 205329 (2008).

[3] Optical, electronic, and transport properties of nanocrystalline titanium nitride thin films, P. Patsalas and S. Logothetidis, J. Appl. Phys. 90, 4725 (2001).

[4] Interface properties and structural evolution of TiN/Si and TiN/GaN heterostructures, P. Patsalas and S. Logothetidis, J. Appl. Phys. 93, 989 (2003).

[5] Plasma-assisted atomic layer deposition of TiN monitored by in situ spectroscopic ellipsometry, S.B.S. Heil, E. Langereis, A. Kemmeren, F. Roozeboom, M.C.M. van de Sanden, and W.M.M. Kessels, J. Vac. Sci. Technol. A 23, L5 (2005).

[6] In situ spectroscopic ellipsometry study on the growth of ultrathin TiN films by plasma-assisted atomic layer deposition, E. Langereis, S.B.S. Heil, M.C.M. van de Sanden, and W.M.M. Kessels, J. Appl. Phys. 100, 023534 (2006).

[7] Synthesis and in situ characterization of low-resistivity TaN_x films by remote plasma atomic layer deposition, E. Langereis, H.C.M. Knoops, and A.J.M. Mackus, F. Roozeboom, M.C.M. van de Sanden and W.M.M. Kessels, J. Appl. Phys. 102, 083517 (2007).

[8] In situ spectroscopic ellipsometry as a versatile tool for studying atomic layer deposition, E. Langereis, S.B.S. Heil, H.C.M. Knoops, W. Keuning, M.C.M. van de Sanden, and W.M.M. Kessels, J. Phys. D: Appl. Phys. 42, 073001 (2009).

[9] In situ spectroscopic ellipsometry during atomic layer deposition of Pt, Ru and Pd, N. Leick, J.W. Weber, A.J.M. Mackus, M.J. Weber, M.C.M. van de Sanden, and W.M.M.  Kessels, J. Phys. D: Appl. Phys. 49, 115504 (2016).

[10] Atomic layer controlled growth of SiO2 films using binary reaction sequence chemistry, J.W. Klaus, A.W. Ott, J.M. Johnson, S.M. George, Appl. Phys. Lett. 70, 1092 (1997).

[11] Atomic layer deposition of tungsten using sequential surface chemistry with a sacrificial stripping reaction, J.W. Klaus, S.J. Ferro, S.M. George Thin Solid Films 360, 145 (2000)

[12] Atomic Layer Deposition of Tungsten Nitride Films Using Sequential Surface Reactions, J. W. Klaus, S. J. Ferro, and S. M. George, J. Electrochem. Soc. 147, 1175 (2000)

[13] Progress in device from molecular layering to atomic layer deposition worldwide technology, V. Drozd, IEEE proceedings of 14th International Baltic Conference on Atomic Layer Deposition (BALD), St. Petersburg Russia, pg. 2-4 (2017)