Two Reviews of SciGlass 6.5
Reviewed by:
Arun K. Varshneya
Professor of Glass Science & Engineering
New York State College of Ceramics
Alfred University
Alfred, NY 14802
I recall writing a review for SciGlass 4.0 back in 1997 (Ceram. Bull. 76(5) 82-83(1997)). Back then, I was impressed with its property-composition search capabilities. I remember asking the database which glass compositions would have a Tg greater than a specified value and then limiting the fields to exclude certain oxides. It would take just a few seconds to compile that information rather than spending hours in a library (and getting frustrated because your library cut back on its holdings of periodicals as a result of cash-crunch.)
It shouldn’t take much to guess that I continue to be impressed with the latest member of its family. So, what is new? Then the database only had about 40,000 glasses. Now it contains a whopping 267,000 or more compositions. Those days, it was limited to oxide glasses. Now it has 21,000 chalcogenides (my favorite!), and 13,000 halides, in addition. (I wish it had glassy metals too…) Bunch more properties have been expanded. There are over 12,000 glasses compiled just for their chemical durability data, for instance. Back then, the database simply wasn’t sufficient to make a serious test of property predictions using models. This is now possible simply because of the enormity of the collection. If you don’t like some particular author’s model, you can discard that and choose another one. Of course, you can also drop outliers (or someone’s measurements if you didn’t trust their work habits). After exporting to Excel spreadsheet, you can add your own data, fit to different mathematical functional forms (including one of your own) and see if the data conform to acceptable statistical tolerances. This is important – since, one of the questions most serious researchers do ask is, “Do their data make sense?” You can plot these into colorful graphs using a variety of units on the axes and export them to PowerPoint for a formal presentation.
Yes, there is a considerable amount of information on areas of glass formation, including data on critical cooling rates. The 3,600 ternaries carry the additional feature that their apex compositions can be selected using the zoom option. You can draw iso-property lines through the datafield. And there are nearly 7,000 optical spectra that are compiled in the SciGlass 6.5. There are patents and trademarks related to glass technology.
All of the database elements are presented with citation with full search capability. The authors of SciGlass seem to have put up their best effort in presenting them.
One of the more interesting use of SciGlass 6.5 I found was the fact that, whereas much of the data in the literature has been scattered all over the place in tabular or graphical using all sorts of units, it is all now in one single place – in SciGlass 6.5 CD. Don’t go too far.
While writing the second edition to my book, “Fundamentals of Inorganic Glasses”, it occurred to me that I could perhaps present a better book to the student myself by teaming with a wekll-known authority on glass, the one and only Oleg Mazurin, who is the primary author of SciGlass. It has been a bit too late, unfortunately, to change figures and the associated discussion, however, a “student version” of SciGlass 6.5 CD will be included with the second edition of my book to get the student started in the right direction early in his/her career.
Surely, the database and its future family members will be of great use to a glass engineering science student, a university professor of glass science who teaches and directs research, a technical working in an industrial manufacturing environment and a senior researcher in an industrial laboratory. My congratulations to Drs. Oleg Mazurin, Alexander Priven, and the SciGlass team for their painstaking efforts in bringing a compilation of glass properties literature – all in one handy SciGlass 6.5 CD.
Reviewed by:
James E. Shelby
McMahon Professor of Ceramic Engineering
New York State College of Ceramics
Alfred University
Alfred, NY 14802
The SciGlass 6.5 database is an invaluable tool for anyone interested in glass research and development, ranging from students to academics to industrial scientists and engineers. SciGlass is obviously of great benefit to those involved directly in glass research, where a knowledge of glass properties and trends in properties, as well as glass formation regions are needed on a regular basis. This database is, however, equally useful to the glass technologist, or for the engineer who is not a specialist in glass technology, but must use glass in the manufacture of products which rely on glass as an essential component of a more complex system. Access to this database greatly simplifies the search for glasses having specified properties, or for the published data for glasses of specified compositions. In particular, the extensive number of commercial glasses covered by this version of the database is invaluable to those searching for a readily available glass to perform a specific function, where well defined properties are essential to the final product. Furthermore, the ability to calculate properties within or beyond the compositional boundaries of the published data is particularly useful for development of new glasses having properties which may not be readily available, or where compositional restriction forbid use of the traditional glasses with the needed properties. The property predictive power of the database, i.e. the ability to specify a property and have the program provide potential compositions which will satisfy those needs is a powerful tool for the glass technologist.
The scope of coverage of SciGlass 6.5 is amazing. Over 265,000 compositions are represented, ranging from oxides to halides and oxyhalides, to chalcogenides. Data can be found for compositions ranging from simple, one component glasses such as vitreous silica, and two component glasses such as alkali silicates, borates, or germanates, to complex, many component commercial and non-commercial glasses. The list of components covered by the database has been greatly expanded from earlier versions, where only oxides were included. The ability to switch between mol%, wt%, and atomic% for composition specification is also very useful.
In the past, obtaining data for less common compositions or for more complex systems required a tedious literature research, which often missed many of the articles published in less common journals, in conference proceedings, or in languages other than that of the searcher. In particular, the massive literature on glass available only in Russian has often been ignored, or at least not well utilized, by the English speaking world. Literature outside the field of experience of the searcher, e.g. the geochemical literature, was usually missed. The patent literature is also frequently ignored due to difficulties in finding items of specific interest. This version of SciGlass includes a large number of commercial, trademarked, and other patented glasses, simplifying searches for material of interest for such materials. Inclusion of non-oxide glasses further increases the utility of the database. Since these glasses are particularly useful in many new, high technology areas (ultraviolet and infrared transmitting, semi-conducting, and photonic applications), inclusion of these materials radically increases the number of those who can benefit from use of this database.
Inclusion of glass formation regions in this database is particularly useful for those working with compositions which might be less commonly studied. The first step in any systematic study of the properties or structure of glasses of a specified compositional type must begin with the formation of those glasses. A prior knowledge of previous work indicating those compositions which will form a glass greatly simplifies planning of such studies. The ternary plotting feature of SciGlass, in particular, is extremely useful for anyone approaching a new compositional system, where they have no prior experience. In the same way, the ability to plot ternary property data from a number of studies is of great aid in understanding the compositional trends occurring within a system. Since many published studies only consider one or two compositional series in a ternary system, the ability to see the results of all such studies on a single diagram greatly improves understanding of compositional trends in 3-component systems. The prediction of regions of composition in these systems where the glasses have properties lying between specified limits is also of great use.
A number of features of SciGlass are particularly useful in my own areas of research and teaching. Since my own work deals primarily with the relations between the properties and compositions of glasses and melts, I find the ability to query the SciGlass database invaluable in planning studies of new binary and
ternary systems. Since I often know authors who may have published in the area of interest, I appreciate the ability to access a list of all publications of a given author contained in the database. The added feature which allows specifying a particular journal or a year or period of years in which the publication occurred further improves the utility of this feature.
The plotting capabilities of the program are very useful in determining the relative reliability of the data from different sources, especially when there are several published sets of data for a particular property of a series of glasses. As Dr. Mazurin has recently noted, some fraction of the published work on the properties of glasses is of questionable quality. Plots of all of the literature values in a system readily reveal sources of suspect data. In fact, I heartily agree with the suggestion that all papers published which involve systems where others have previously published should make an effort to establish the reliability of the new data by comparison with that from past studies. If there is disagreement, the authors of the new publication have an obligation to present reasons why they believe that their data are superior to that previously published. If their data does agree with that from the literature, it will only strengthen their arguments regarding the validity of their sample compositions and the extension of their results to new compositions. In particular, papers which only present spectroscopic results frequently fail to establish the reliability of their compositions. Since any experienced glass researcher knows that production of new compositions often can result in erroneous compositions due to preferential volatilization of components or reactions with crucible materials or the furnace atmosphere, it is important that the authors of any study confirm the validity of their work by the simple act of comparison to history. There is no doubt that SciGlass radically reduces the effort needed to make such a comparison.
As a teacher, I have also found SciGlass to be a valuable teaching tool. Students are able to find more complete information regarding subjects for term papers and for background for their research projects. I have used this program to find data for homework assignments and for examples used in the classroom. Plots illustrating the scatter found in real, as opposed to theoretical, discussions of property trends are particularly useful in discussions of property/structure relations.
The tables option of this program is especially useful for presenting material which is difficult to specify otherwise. For example, chemical durability of glasses cannot be represented by a single property listing in the way that, e.g. density or refractive index, can be. There are many ways of studying chemical durability, each with its own advantages and disadvantages, and many of which are specific for some application. The tables listings not only provide access to such information, but provide sources with details of the various approaches to such studies.
The inclusion of optical spectra in SciGlass is another useful table function. It is often very difficult to find spectra for many glasses. As an example, I recently need a typical visible region spectrum of erbium in vitreous silica for a course on optical glasses. After searching through several books in my office, I was unable to find such a spectrum. Two minutes using SciGlass produced exactly what I needed. In the future, this program will be my first choice for finding this kind of information.
I sure that this review has neglected many of the features of SciGlass. In summary, this program is an invaluable tool for anyone needing information about glasses or their melts. I can no longer imagine carrying out the tedious, painful searches of the literature for property data, glass formation regions, and other information which was necessary before the advent of this essential research and development tool.