Monday 13 March 2017

On the mammoth table - JAP 2005, JAP 2013 - next step VPHA?



The "mammoth table" of two-reactant ALD processes that many ALD researchers will have seen, was created for Puurunen's JAP 2005 review (open access here), and updated for the Miikkulainen et al. JAP 2013 review (open access here).


An idea has come up that it might be possible to update the table again within the Virtual Project on the History of ALD (VPHA). No promises yet---just an idea.


Screen capture of the first rows of the updated "mammoth table" spanning over 17 pages. Reprinted from V. Miikkulainen, M. Leskelä, M. Ritala, R. L. Puurunen, J. Appl. Phys. 113 (2013) 021301, http://dx.doi.org/10.1063/1.4757907, with the permission of AIP Publishing.


With the view of possibly updating the mammoth table, this post describes how the mammoth table was originally created and how it has been once updated. Whether or not an update will be made within VPHA, the background information shared here should be useful regarding an update that eventually at some point in some form most likely will be made. A previous post on the unique article codes used in VPHA is related to this post and the mammoth table: the unique coding was already started for the JAP 2005 review and re-used in the Miikkulainen et al. JAP 2013 review as well as in VPHA's ALD-history-evolving-file. If/when an update to the mammoth table will be made within VPHA, the unique article codes will be used again.


ALD surface chemistry review and the mammoth table - Puurunen, JAP 2005

The mammoth table aims to summarize all two-reactant ALD processes developed by a given timeline. The table was first created for the JAP 2005 review.


When writing the JAP 2005 review, I had a vision to try to write the kind of review article which I hoped would have been available when I started with ALD in 1998.


The core of the review was to review in detail the surface chemistry of the Me3Al/H2O process, for which in my own research new significant results had been found. Before discussing the details of this one process, I needed to give a general description of the surface chemistry concepts related to ALD, as suitable description did not exist in scientific literature earlier. At the same time, I took the opportunity to discuss some historical problematic terminology.


In addition to these, I thought that it was generally useful to give an overview of the existing ALD processes. This was the easiest made in a tabular form. I also decided to illustrate with chemical drawings how the ligands of the metal reactants look like, as especially for non-chemists (and sometimes for chemists, too), the ligands used can appear complex. To further compactly visualise two-reactant ALD processes made, I created a version of the periodic table which overviews the different materials made - see the image below.


The original periodic table summary of two-reactant ALD processes, from the JAP 2005 review. Reprinted from R. L. Puurunen, J. Appl. Phys. 97, 121301 (2005), http://dx.doi.org/10.1063/1.1940727, with the permission of AIP Publishing.


My overview was, of course, not the first tabular summary of ALD processes. There were summary tables which existed before and of which I was aware, and these served as inspiration for the mammoth. The first one that I was aware of, was in the Leskelä & Ritala review [J. Phys. IV France  9, Pr8/837 (1999)]. An updated version, this time with references, was published in the Ritala & Leskelä 2002 book chapter. The latter one I found very useful, because it contained the references. Both these tables were created from the viewpoint of and for the use of ALD process developers, however. The material was ordered on the basis of the reactant used; the material which was made, was secondary. My primary interest was to see, which processes have been developed for a given material.  


To create the mammoth table, I first went through the material of the table of the Ritala & Leskelä 2002 book chapter and re-ordered the information. As I recall, some references were left out in this process (e.g., I did not include conference presentations). Second, to fill possible gaps and to update for references published after creating the 2002 book chapter, I went again through all ALD literature through an ISI Web of Science search. The following key words were used (listed in the JAP 2005 review, reference no. 5): atomic layer chemical vapo(u)r deposition, atomic layer deposition, atomic layer epitaxy, molecular lamination, molecular layer epitaxy, molecular layering, molecular stratification.


The following criteria were used in creating the mammoth table:
  • The ISI Web of Science search was made from the beginning of the database. I cannot recall the year with 100% certainty; my gut feeling is that the database started from 1964. The search was made until February 2005 (date specified in the text preceding Table III in the the JAP 2005 review). This means that all works published in 2004 or earlier have been included, and some which have been published after that.  
  • As already mentioned, only references were included, which have been published in refereed scientific journals and can thus be found through searching the scientific databases such as ISI Web of Science. Conference presentations were on purpose not included.  
  • References were included in the table with the principle of being inclusive, not exclusive. Thus, the works were included if the authors themselves had reported the work as ALD. In a more critical review by a more critical author, some of the references might have been left out because of not sufficiently following the principles of ALD but having a significant part of non-ALD growth, such as CVD through thermal decomposition of the metal reactant.
  • The principle in building the table was that the article should report some physicochemical characterization of the ALD process, relevant to anyone studying the surface chemistry of ALD. Minimum for this would be to report the thickness measured for a given number of cycles at a given temperature, allowing one to calculate the growth per cycle (GPC). If in the paper the ALD material was just used for some function (e.g. electrical) and no physicochemical characterization was made for the material, the reference was not included.
  • The table is for two-reactant processes only. Processes, where more than two reactants are used, are not listed. Works, where the main emphasis is on nanolaminates or ternary or quaternary compounds, are included only if results are reported also for the binary compounds grown in two-reactant processes. Special cases of ternary compounds are included, if both reactants are metal compounds and thus a two-reactant process results in a ternary compound.
  • A positive choice of including a paper in the list was often for clear cases made on the basis of the abstract of the article. For less clear cases, a copy of the article was obtained and a decision of whether to include the reference in a table was made after a more detailed check.  


In general, the choices made in building the table aimed at (1) covering literature significant to as large as possible a community, while still (2) minimising the effort to be done by setting clear boundaries to what is done and what remains outside the scope of the work.


To build the table, I had to create an order. Many small decisions had to be made, and someone else building the table would undoubtedly have arrived at a table that looks slightly different. Ordering of the processes was made on the following basis.


  1. Atomic number of the element Z.
  2. Material made, on the order of roughly speaking most often grown by ALD: oxides, nitrides, phosphides, fluorides, tellurides, element itself, and mixed materials.
  3. Metal reactant, also on the order of roughly speaking how often the class of metal reactants is used: halides, organometallic compounds, metal-organic compounds, on the order of increasing size (mass) of the molecule.  
  4. Non-metal reactant, on the order from simple to more complex.


Updating the mammoth table - Miikkulainen et al., JAP 2013

In just few years, I got the feeling that it would be good to update the mammoth table. The chance came when with a joint Tekes project "ALEBOND" with Prof. Markku Leskelä from the University of Helsinki. We investigated among other things, whether the roughness originating from the crystallinity of ALD films can be utilized for an anti-stiction function in MEMS. There was no source available at that time of the experimental studies of the crystallinity of ALD films, and an overview was needed. Simultaneously with creating the tables of crystallinity of ALD films, it was natural to update the mammoth.


For JAP 2013, it was Ville Miikkulainen, then a postdoctoral researcher at University of Helsinki (PhD from the University of Eastern Finland, Prof. Tapani Pakkanen's group, where quantum chemical studies of ALD were started with Tuomo Suntola in the 1980s) who made the literature search. For quite some time, we worked on the review the the three of us - Puurunen, Miikkulainen, Leskelä. At a later stage, Prof. Mikko Ritala was invited to join the review author list. It was Prof. Leskelä who proposed the final elegant and appropriate solution for the author order---Miikkulainen first and myself last.


Notations from the update:
  • All the same key words were used in the search as for the JAP 2005 review and the search was made with ISI Web of Science again. In addition to the older terms, "atomic layer deposited", a term which have been used more often recently, was used.
    • The first round of literature search was not made with the full keyword list. A second round was needed later to fill the possible gaps. Lesson learned: think carefully of key words before starting.  
  • The ISI Web of Science search was made to cover literature until the end of year 2010. In practice, this means that the last search was made during the first months of 2011.
  • We continued to build upon the same BibTeX file as used for the JAP 2005 review.
  • Auto-export of BibTeX references from ISI Web of Science was used. While this accelerates the updating of the reference database, it was noticed later in the manuscript preparation that this way, the special characters of e.g. many Finnish (and many other) names are lost, and typographical errors introduced in addition to these.
    • The very last stage in the preparation of the JAP 2013 review, then, was correcting for these errors and this too two workdays. Lesson learned: best to check the names while importing the references and make the necessary corrections.
  • The same concept of periodic table as created for the JAP 2005, was used.
  • The ordering of the material in the mammoth table was made mainly with the same principles as for the JAP 2005 review. One small change made that was made was that if an elemental film has been made, that comes first in the order.
The updated periodic table summary of two-reactant ALD processes, from the JAP 2013 review. Reprinted from V. Miikkulainen, M. Leskelä, M. Ritala, R. L. Puurunen, J. Appl. Phys. 113 (2013) 021301, http://dx.doi.org/10.1063/1.4757907, with the permission of AIP Publishing.


JAP 2013 and the Drozd SnSe process

While the purpose of the JAP 2013 review, as of the JAP 2005 review had been, was to cover all two-reactant ALD processes published by the date of making the literature review, after the publication, we noticed that (at least) one reference remained missing.

The missing work is by Drozd et al. and reports a material made by ALD not made in other studies, SnSe. This work did not come up in the literature search as it did not contain any of the used key words which were used in the search. The abbreviation ALD was there in the article, but we did not use "ALD" in the search. The next image points where the process is missing from the periodic table of two-reactant ALD processes.


Periodic table of materials made with two-reactant ALD processes from the JAP 2013 review, with the missing material by Drozd et al. pointed to.


A lesson from this to all researchers is to provide sufficient key words with their articles. Abbreviations in general and "ALD" in particular is not a good word to use as a search criteria, as it as multiple meanings.

Mammoth table and VPHA - could we update it again?

The latest version of the mammoth shows works until the end of year 2010, and it is partly outdated. The two articles, in which the mammoth table has been published and thereafter updated, are among the most popular reviews on ALD ever published. An update would undoubtedly in the form of an article certainly be useful for the ALD research community.


As explained above, the material for the mammoth has been collected so that updating the earlier work should be relatively simple as the update can be directly built on existing material. I have once done the hard work of going through the literature, and once it has been done by Ville Miikkulainen. Quite some time ago, I have thought and we have agreed with Ville to be willing to proceed in an open science manner and to allow others to build upon our work, would someone appear who would like to make an update of the material. Years have passed, and nobody has shown up who would be both willing to make and actually making the update. (There have been discussions in the background of various kinds of updates in the past years.)


Personally, I am at the moment gaining momentum towards updating the mammoth table again. This is because I have just started as an associate professor of Catalysis Science and Technology at Aalto University, and for my own research, having easy-to-access up-to-date information would be beneficial. Just like for everyone else.


While I will not have time/motivation to do the "hard work" of going through the literature again by myself, I would be willing coordinate a volunteer-based effort to do this. Perhaps we could even make this update within the framework of the collaboration in the Virtual Project on the History of ALD---I believe that it is largely the same people who want to know of the old ALD literature as well as of new literature. I have a vision that we will still write a compact end review from VPHA, to complement the narrow-goal review already published in JVSTA. Making an update of the two-reactant ALD processes could perfectly fit to be a a part of that review article.


Thinking of the possible update to come, it is good to make some planning and be open for new ideas. Below are some actions that would be needed and questions related to them. Comments welcome regarding the actions needed.
  • Search by ISI Web of Knowledge should be made for years 2011 to 2016. Make the search in the first months of 2017, to be sure that all references listed for 2016 get included.
  • Use the same key words as for the JAP 2005 review, plus some additional ones if judged necessary. The 2005 key words were: atomic layer chemical vapo(u)r deposition, atomic layer deposition, atomic layer epitaxy, molecular lamination, molecular layer epitaxy, molecular layering, molecular stratification.
    • "Atomic layer deposited" was used for the JAP 2013 review and should be included now, too.
    • Are there key words which should be added?
  • Should the material be ordered in the same way as for the earlier reviews, or would there be suggestions for improvement?


One last thing - why "mammoth"?

Where does the name "mammoth" actually come from? While I am not sure who has actually used first, I remember to have heard it from Prof. Leskelä from the University of Helsinki, probably during the ALEBOND project collaboration.

In Finnish, the name mammoth, "mammutti", is descriptive of something that is really large. Unbelievably large. Maybe even too large.  


References mentioned/linked



Related posts in the ALD History Blog



Peer-reviewed scientific articles created within VPHA

  • R. L. Puurunen, "A Short History of Atomic Layer Deposition: Tuomo Suntola's Atomic Layer Epitaxy", Chem. Vap. Deposition 20 (2014) 332-344. Published with open access. http://dx.doi.org/10.1002/cvde.201402012.
  • A. A. Malygin, V. E. Drozd, A. A. Malkov, V. M. Smirnov, "From V. B. Aleskovskii's “Framework” Hypothesis to the Method of Molecular Layering/Atomic Layer Deposition", Chem. Vap. Deposition 21 (2015) 216-240. http://dx.doi.org/10.1002/cvde.201502013
  • Focused ALD history review, 2017, 62 co-authors, alphabetical order: Esko Ahvenniemi, Andrew R. Akbashev, Saima Ali, Mikhael Bechelany, Maria Berdova, Stefan Boyadjiev, David C. Cameron, Rong Chen, Mikhail Chubarov, Veronique Cremers, Anjana Devi, Viktor Drozd, Liliya Elnikova, Gloria Gottardi, Kestutis Grigoras, Dennis M. Hausmann, Cheol Seong Hwang, Shih-Hui Jen, Tanja Kallio, Jaana Kanervo, Ivan Khmelnitskiy, Do Han Kim, Lev Klibanov, Yury Koshtyal, A. Outi I. Krause, Jakob Kuhs, Irina Kärkkänen, Marja-Leena Kääriäinen, Tommi Kääriäinen, Luca Lamagna, Adam A. Łapicki, Markku Leskelä, Harri Lipsanen, Jussi Lyytinen, Anatoly Malkov, Anatoly Malygin, Abdelkader Mennad, Christian Militzer, Jyrki Molarius, Małgorzata Norek, Çağla Özgit-Akgün, Mikhail Panov, Henrik Pedersen, Fabien Piallat, Georgi Popov, Riikka L. Puurunen, Geert Rampelberg, Robin H. A. Ras, Erwan Rauwel, Fred Roozeboom, Timo Sajavaara, Hossein Salami, Hele Savin, Nathanaelle Schneider, Thomas E. Seidel, Jonas Sundqvist, Dmitry B. Suyatin, Tobias Törndahl, J. Ruud van Ommen, Claudia Wiemer, Oili M. E. Ylivaara, Oksana Yurkevich, "Review Article: Recommended reading list of early publications on atomic layer deposition—Outcome of the “Virtual Project on the History of ALD”", Journal of Vacuum Science and Technology A, 35 (2017) 010801. Open Access as published. http://dx.doi.org/10.1116/1.4971389


********************* 
Virtual Project on the History of ALD (VPHA) - in atmosphere of Openness, Respect, and Trust

No comments:

Post a Comment