DOI: 10.1595/147106708X362735 Platinum Group Metals Patent Analysis and Mapping A REVIEW OF PATENTING TRENDS AND IDENTIFICATION OF EMERGING TECHNOLOGIES
By Richard Seymour Johnson Matthey Technology Centre, Blounts Court, Sonning Common, Reading RG4 9NH, U.K.; E-mail: [email protected]
The patent literature contains a wealth of detailed information about existing and new uses for the platinum group metals (pgms). While excellent searching tools have existed for many years for identifying patents relating to specific topics, it is only relatively recently that it has been feasible to map the complete archive of patent literature to identify important trends relating to potential new applications. This paper summarises the results of such an exercise for the pgms carried out in early 2008 and shows that one such ‘hot spot’ relates to organic light emitting diodes (OLEDs).
Previous articles in this Journal have described Matthey Technology Centre is Aureka® (a product the importance of patents as a key source of techni- available from Thomson Reuters) (3), which cal and commercial intelligence (1, 2). The use of includes patent data sets from the Patent patent mapping to visualise large sets of patent data Cooperation Treaty (PCT) and European Patent and to identify trends contained within that data has offices, plus a range of national patent collections also been demonstrated (2). The present paper fur- including those of the U.S.A., Japan, the U.K., ther develops these themes by examining the patent France and Germany. With the exception of Japan, literature on pgms published since 1983, in particu- these collections contain full-text patent docu- lar that on the minor metals iridium and ruthenium. ments, available either as PDF or HTML files. In the case of Japanese patents, a text version of the Searching – What and Where? English-language title, abstract and other front I will begin by thinking about search strategy. In page details is available, together with a PDF file of this case, the initial objective is to create a large set the full specification in Japanese. of patents relating to the pgms, which will later be It must be borne in mind that using the French analysed and refined. The choice of keywords is and German collections would require us to search therefore straightforward: platinum, palladium, in French or German respectively, and of course rhodium, iridium, osmium and ruthenium. In the the results obtained would also be in French or patent literature it is unlikely that the names of German. The patent collections of other countries, these metals would be used in other contexts. for example China and India, are not currently However, this might be a difficult problem if we available in Aureka. However at this stage we are were searching the news or business press, where looking for the big picture. The detail can follow the names of the pgms are associated with many later if necessary, for example by adding Chinese brand names – for example there would probably patent documents retrieved from other patent be many hits on topics such as platinum credit databases. cards or iridium satellite communication systems, We also need to think about where in the and strategies for removing such material would patent document we might wish to search for need to be found. information on pgms. This is an important ques- Perhaps a more important question to ask is tion and to understand the various possibilities which patent collections to use to search for these and their implications we first need to think about words? The software package used at the Johnson the structure of a typical patent:
Platinum Metals Rev., 2008, 52, (4), 231–240 231 – Title: often deliberately rather vague and non- opportunities for suppliers of catalysts, the cus- specific. tomer being the owner of the patent. – Abstract: a short summary of the invention, in Table I illustrates the wide variation in the num- perhaps 100 to 200 words. ber of retrieved patents obtained according to – Claims: the claims of a patent govern its legal where in the patent the search is performed. The effect, that is, the areas of technology that are to table clearly shows that choosing which part of the be monopolised. Generally it can be said that a patent document to search is critical. If we search feature is not protected unless that feature is in the patent title, abstract and claims then we claimed or covered by the general language in retrieve over five times as many patents as exactly the claims. So these are key – get the claims the same search restricted to just title and abstract. wrong and your invention may be seriously If we search in the full text of the patent then we compromised. retrieve five times as many again. Then, depending on the particular country, there may also be sections on: Table I – Background: provides details on the context of Searches on the Term ‘Platinum’ Conducted in the invention, current technology, and why the U.S. Granted Patent Collection, for Patents existing solutions may be inadequate. Published between 1st January 2001 and 31st December 2007 – Description: a detailed description of the inven- tion and possible variants thereof. Criteria Number of ‘hits’ – Examples: worked examples, covering aspects such as how the invention is made. Scientists ‘Platinum’ in the patent 1611 sometimes wrongly concentrate on the exam- title or abstract ples just as they would read the experimental ‘Platinum’ in the patent 8878 sections of scientific papers. title, abstract or claims Now let us suppose we are searching for patents ‘Platinum’ in the patent 44,541 in which a new pgm chemical or material is dis- full-text closed, or in which the use of a pgm is a key part of ‘Platinum’ in the patent 35,663 the invention. In this case restricting the search to full-text but not title, terms in the title or abstract, and possibly also the abstract or claims claims, will be adequate. Clearly if the word ‘plat- inum’ appears in any of these sections then it is Table II shows the top fifteen assignees for likely to be a very important part of the invention. each set of results in Table I. It shows that we But what about the case when the name of the might expect to obtain quite different results for pgm appears somewhere in the rest of the patent, the various searches, even though the keyword is but not in the title, abstract or claims? Can these the same in each case. Apart from Micron patents safely be ignored? An example of such Technology Inc, which heads up each list, there are patents might be the use of a standard palladium some very significant differences. Engelhard (now on carbon hydrogenation catalyst in a multi-stage BASF Catalysts) comes in at number five in the organic synthesis route. The novelty is in the end- ‘title or abstract’ search but does not appear in the product, not the catalyst used, and therefore the ‘title, abstract or claims’ or full-text searches. On term ‘palladium’ is unlikely to occur in the title, the other hand, the Semiconductor Energy abstract or claims. However it may well come up in Laboratory, while it does not appear in the ‘title or the examples. While we can probably ignore such abstract’ search, and only reaches number twelve patents for the purpose of identifying key new in the ‘title, abstract or claims’ search, comes in at application areas, important information may number two in the full-text search. nevertheless be obtained from them. For example, Pfizer is another good example. Like they may provide valuable intelligence on sales Semiconductor Energy Laboratory, this company
Platinum Metals Rev., 2008, 52, (4) 232 Table II Search Results by Top Fifteen Assignees for Patents in the U.S. Granted Patent Collection, Published between 1st January 2001 and 31st December 2007
Rank ‘Platinum’ in patent title ‘Platinum’ in patent title, abstract ‘Platinum’ in patent full-text or abstract (1611 patents) or claims (8878 patents) but not title, abstract or claims (35,663 patents) 1 Micron Technology Inc Micron Technology Inc Micron Technology Inc 2 General Electric General Electric Semiconductor Energy Laboratory 3 Shin-Etsu Chemical Co IBM Fuji Photo Film Co Ltd 4 UOP LLC Samsung Electronics Co Ltd Eastman Kodak 5 Engelhard Corporation1 Advanced Micro Devices Inc Canon KK 6 Dow Corning Matsushita Electric Industrial Matsushita Electric Industrial Co Ltd Co Ltd 7 Matsushita Electric Industrial Shin-Etsu Chemical Co General Electric Co Ltd 8 Texas Instruments Inc Intel Corp 3M Innovative Properties Co 9 Dow Corning Toray Silicone Infineon Technologies AG IBM 10 IBM Hitachi Ltd NGK Insulators Ltd 11 Advanced Cardiovascular Institut Francais du Petrole Seiko Epson Systems 12 Samsung Electronics Co Ltd Semiconductor Energy Laboratory Medtronic Inc 13 Honeywell International Inc UOP LLC Pfizer 14 Infineon Technologies AG Texas Instruments Inc Sony Corp 15 BASF Hewlett-Packard Development Co Hitachi Ltd
1Now BASF Catalysts only appears in the top assignees from the full-text patent family members filed in different geograph- search. One would expect Pfizer’s main interest in ical regions, to leave one patent per invention. The platinum to be as a user of catalysts in pharmaceu- final document list contains just over 13,540 tical manufacturing, rather than as a developer of patents. new pgm-based technologies. Manual inspection of Figure 1 shows a basic breakdown of these selected Pfizer patents confirms this to be the case. patents by metal and by five-year timeslices. Overall growth in pgm patents in the period from The Results List and Initial 1983 to 2007 is about 6% per annum. In the last Analysis seven years, this growth rate has been nearly 13%. For the remainder of this paper we will be con- However, growth in the number of patents by sidering the results of searches based on the names individual metal is not completely uniform. There of the pgms in the patent title or abstract. We have has been somewhat higher growth for platinum, undertaken the search in the U.S., European and ruthenium and iridium patents, and slightly lower PCT patent collections, for patent applications or growth for palladium and rhodium, as shown in granted patents published in the period from 1st Table III. January 1983 to 31st December 2007. The search A comparison between the pgm patent picture results have then been ‘deduplicated’ to exclude and that for a number of other metals (gold, silver,
Platinum Metals Rev., 2008, 52, (4) 233 Fig. 1 PGM 7000 patents by metal – 6000 number published in 5000 Osmium five-year time Iridium periods 4000 Ruthenium 3000 Rhodium 2000 Palladium 1000 Platinum Number of patents published Number of patents
0 1983–1987 1988–1992 1993–1997 1998–2002 2003–2007
Time period nickel and cobalt) is shown in Figure 2. The num- surrounded by sea. The visualisation is helpful ber of patents on all these metals has increased. because ThemeScape groups together similar doc- However the rate of increase for pgms and gold is uments and labels these groups according to considerably higher than that for nickel, cobalt and frequently used key terms found within those silver. This is illustrated in Table IV by looking at groups. The more documents contained within the earliest (1983–1987) and latest (2003–2007) each group, the higher the ‘mountain’ appears. The time periods. automatic labelling sometimes produces meaning- Figure 3 shows the importance of these metals ful headings (e.g. silicone, rubber, polysiloxane), in selected technology areas. The analysis was but sometimes these are less obviously meaningful based on selected International Patent (compounds, preparation, reaction). Where neces- Classification (IPC) codes (4). sary these can be edited following an inspection of documents contained within the groups. Patent Mapping The grey dots represent sample documents – in The Aureka ThemeScapeTM tool (3) was used to this set of 13,540 documents only a small propor- create a visualisation of the pgm document list tion are shown in this view, but more (or all) described above. The results are shown in Figure 4. documents will be shown when specific areas are The resulting map looks like a mountainous island magnified. Clicking on specific dots will display the
Table III PGM Patents by Metal: Number and Proportion of Patents Containing Specific PGMs Published in Early (1983–1987) and Recent (2003–2007) Time Periods
Metal Patents containing specific pgms Patents containing specific pgms vs. total pgm patents, 1983–1987 vs. total pgm patents, 2003–2007 Number Proportion, % Number Proportion, % Platinum 840 46 2265 52 Palladium 659 36 1452 33 Rhodium 408 22 689 16 Ruthenium 349 19 1024 23 Iridium 184 10 613 14 Osmium 56 3 189 4 Total1 18331 – 43921 –
1This is not the arithmetic sum as more than one pgm can appear in any given document
Platinum Metals Rev., 2008, 52, (4) 234 Fig. 2 20000 Comparison of 18000 pgms with selected other 16000 Cobalt metals – 14000 number of Nickel 12000 patents published in 10000 Silver five-year time 8000 periods Gold 6000 4000 PGM 2000 Number of patents published Number of patents 0 1983–1987 1988–1992 1993–1997 1998–2002 2003–2007
Time period
Heterogeneous catalysts
Semiconductors or Electronics
Homogeneous catalysts PGM Gold Fuel cells and batteries Silver Nickel Coated products
Technology area Technology Cobalt Alloys
Medical
0 500 1000 1500 2000 Number of patents Fig. 3 Relative importance of various metals in selected technology areas, by IPC code, for the time period between 1st January 2003 and 31st December 2007
Table IV PGM Patents vs. Other Metals: Number and Proportion of Patents Containing Specific Metals Published in Early (1983–1987) and Recent (2003–2007) Time Periods
Metal Patents containing specific metals Patents containing specific metals vs. total metal patents, 1983–1987 vs. total metal patents, 2003–2007 Number Proportion, % Number Proportion, % PGMs 1833 28 4392 31 Gold 646 10 2152 15 Silver 2009 31 4062 28 Nickel 2200 34 4626 32 Cobalt 1279 20 2441 17 Total1 65361 – 14,2931 –
1This is not the arithmetic sum as more than one pgm can appear in any given document
Platinum Metals Rev., 2008, 52, (4) 235 Fig. 4 PGM patent map covering granted patents or patent applications published between 1st January 1983 and 31st December 2007 original document. The contour lines enclosing The ‘exhaust, engine, oxide’ area contains many particular areas can be used to select groups of red, green and white dots, indicating that both met- documents for inspection or further analysis. als may be used in emission control applications. In Figure 5, we have further processed the basic Figure 6 is a similar image showing the minor map shown in Figure 4 in two ways. Firstly, to cre- metals rhodium, ruthenium, iridium and osmium. ate a timeslice covering documents published only Of particular interest here are the two boxed in the period January 2003 to December 2007; sec- areas, the first just left of centre, the second cen- ondly, to show patents on platinum as red dots and tre right. These contain a cluster of mainly light patents on palladium as green dots. Where specif- blue dots (ruthenium) and dark blue dots (iridi- ic documents cover both platinum and palladium, um), respectively. Comparison of the number of these are shown as white dots. dots with the same areas in Figure 7, covering the The reason for this exercise is to show the rela- 1993 to 1997 timeslice, shows a marked increase tive importance of particular metals in specific in the numbers of iridium and ruthenium patents technology areas. For example, the ‘silicone, rub- published in 2003–2007. These are examples of ber, organopolysiloxane’ and ‘fuel cell, fuel, emerging technologies. Magnification of one of electrode’ areas are dominated by red dots, indicat- these areas (see Figure 8) shows that this area ing that platinum is the preferred metal in these includes many patents on organic light emitting applications. The ‘plating, deposited, substrate’ diodes (OLEDs), which is an important potential region is dominated by green dots, confirming the new application for iridium-based fluorescent or importance of palladium in electronic applications. phosphorescent dopants. OLEDs (see also
Platinum Metals Rev., 2008, 52, (4) 236 Fig. 5 PGM patent map timeslice 2003–2007, showing occurrences of platinum patents (red dots), palladium patents (green dots), and patents covering both platinum and palladium (white dots)
References (5, 6)) are solid-state devices com- Grubbs’ catalyst) and the application of rutheni- posed of thin films of organic molecules that um in silane production are other emerging create light with the application of an electric cur- technology areas. rent. Compared with conventional light-emitting diodes (LEDs) or liquid crystal displays (LCDs), The Non-Patent Literature OLEDs provide brighter, crisper displays which While patents are an extremely important require less power. It has been discovered that in source of technical and commercial intelligence, some iridium complexes, strong spin-orbit cou- there is also a huge amount of non-patent litera- pling leads to singlet-triplet mixing, ideal for ture covering the pgms. This is illustrated in Figure highly efficient electrophosphorescence required 9, which compares the size of the patent literature for future OLEDs. Companies with pgm patents on ruthenium with that of the non-patent scientif- in the OLED area currently include DuPont ic literature on the uses of this metal in (U.S.A.), Samsung (Korea), LG Electronics chemistry-related areas. The top ten uses for (Korea), Idemitsu Kosan (Japan) and Konica ruthenium in the non-patent literature, based on Minolta (Japan). controlled index terms used in the Chemical Further analysis of the map shows that ruthe- Abstracts database, are shown in Table V. nium-based interconnects and electrodes, Specialised software tools such as STN® iridium-based capacitor materials, new magnetic AnaVistTM (7) are now available to assist with the materials containing iridium or ruthenium, ruthe- analysis of non-patent (as well as patent) literature, nium-based metathesis catalysts (for example similar to that described above for the patent data.
Platinum Metals Rev., 2008, 52, (4) 237 Fig. 6 PGM patent map timeslice 2003–2007, showing occurrences of rhodium patents (yellow dots), iridium patents (dark blue dots), ruthenium patents (light blue dots), osmium patents (purple dots), and patents covering two or more minor metals (white dots) time of writing, there are in the region of 13,540 Table V inventions, covering the period from January Top Ten Uses for Ruthenium from the Non-Patent 1983, in which the use of one or more pgms is a Chemical Literature, 2003–2007 key part of the inventive step. There are many oth- Publication index term Proportion of total, ers in which pgms may be used, for example as 2003–2007, % part of a complex organic synthesis route. Growth Fuel cells 11.2 of this literature is expected to continue to Oxidation catalysts 10.7 increase at a rate slightly higher than that of cer- Hydrogenation catalysts 9.4 tain base metals. Nanoparticles 5.2 Oxidation, electrochemical 4.4 Patent mapping tools can be used to identify Carbon black, uses 4.1 key areas of development and ‘hot spots’ of Fluoropolymers, uses 3.7 activity which may lead to future volume applica- Magnetisation 3.7 tions. ‘Hot spot’ areas for the minor metals Spin valves 3.7 Vapour deposition process 3.4 ruthenium and iridium currently include iridium in organic light emitting diodes (OLEDs), ruthe- nium-based interconnects and electrodes, Conclusions iridium-based capacitor materials, new magnetic The patent literature is an extensive and materials containing iridium and/or ruthenium, detailed source of information on existing and and the application of ruthenium in silane potential new applications for the pgms. At the production.
Platinum Metals Rev., 2008, 52, (4) 238 Fig. 7 PGM patent map timeslice 1993–1997, showing occurrences of rhodium patents (yellow dots), iridium patents (dark blue dots), ruthenium patents (light blue dots), osmium patents (purple dots), and patents covering two or more minor metals (white dots)
Fig. 8 Magnified view of centre-right box shown in Figure 6 for timeslice 2003–2007, showing occurrences of iridium patents (dark blue dots), ruthenium patents (light blue dots), and patents covering two or more minor metals (white dots)
Iridium OLED Electroluminescent
Platinum Metals Rev., 2008, 52, (4) 239 Fig. 9 Trends in the 3500 non-patent literature – number of publications 3000 in patent and non-patent 2500 literature for ruthenium Patents 2000 1500 Non-patents 1000 (uses) 500 Number of publications 0 1983–1987 1988–1992 1993–1997 1998–2002 2003–2007
Time period
References 1 I. Wishart, Platinum Metals Rev., 2005, 49, (2), 98 http://www.wipo.int/classifications/ 2 R. Seymour, Platinum Metals Rev., 2006, 50, (1), 27 5 J. A. G. Williams, Platinum Metals Rev., 2007, 51, (2), 3 Thomson Reuters, Scientific: Products: Aureka®: 85 http://scientific.thomsonreuters.com/products/aureka/ 6 R. J. Potter, Platinum Metals Rev., 2008, 52, (3), 155 4 World Intellectual Property Organization, IP 7 CAS, STN® AnaVistTM: Services, WIPO International Classifications: http://www.cas.org/products/anavist/
The Author Richard Seymour is the Head of Technology Forecasting and Information at the Johnson Matthey Technology Centre, U.K. He is interested in the use of information in the areas of competitive intelligence and commercial development.
Platinum Metals Rev., 2008, 52, (4) 240