Reaxys provides information about millions of chemical structures, reactions, and properties. It is an important tool for researchers who are synthesizing and characterizing chemical compounds, or searching for data on the physical, chemical, spectral, or toxicological properties of chemical substances.
This feature parses your terms and attempts to find the most relevant hits across the substances, reactions, and literature segments of the Reaxys database. It works fine for some very straightforward searches, such as "preparation of 12078-17-0" where the algorithm can make a relational match between a concept and an unambiguous substance identifier. But it is still being developed and doesn't always get it right. When you use it for non-chemical concept searches it'll drop you into the Literature search, or into external substance files like PubChem. For more precise searching of the core database of substances and reactions, use the Query Builder feature.
One could write a whole book on the MarvinJS editor tool. Consider visiting and viewing various tutorials and help materials at the Reaxys support site, or send a direct feedback comment from the Reaxys interface.
Yes, but don't give up if it doesn't get a hit. Beilstein substance records created prior to 1994 ususally contain RNs as chemical identifiers. Some substance records added since 1994 do not. Many older Gmelin inorganic substance records do not contain CAS RNs. CAS RNs should not be confused with Reaxys registry numbers, which are accession numbers within this database alone. A single substance record in Reaxys may contain several CAS RNs because the two systems use different registration rules. CAS assigns RNs to structures more narrowly than Reaxys.
Yes, but this is the least reliable kind of search. Reaxys indexes various synonyms for compounds, but don't rely solely on name string searching. Some names are shown only in German, and some substance records lack names altogether. CAS Registry Numbers and exact structures are better starting points. You can also generate a structure from a name, which you can then edit for further searching.
The Molecular Formula search is especially useful for finding hard-to-draw inorganic and organometallic/coordination compounds, and for building queries with variable element counts, shortcuts (Me, Et, Ph), and periodic groups. Select Query Builder, then the MF menu option, then click on the periodic table button to pull up a tabular tool where you can select elements and groups to include.
TIP: To specify element counts/ranges you have to type in a query with the correct syntax - the interface doesn't have an easy element count feature. Example: To search for substances with 1-3 P atoms and 1-4 Al atoms, type P[1-3]Al[1-4] in the MF search box. If you want to allow any other atoms as well, append ?? to the query. The order of the elements in the query doesn't matter, but it may be case sensitive to avoid ambiguity.
Yes, but it can get complicated. Start with the Query Builder tab, and select Properties fields from the menu on the right. Enter values and select the numeric operator. You can also use the "Find any" checkbox to limit to substances that contain a particular field without having to enter values for that field. Due to the huge number of property values contained in the database, this technique is most useful with multiple property fields or in conjunction with a substructure search. If you enter a range of values for a property it's best to keep that range narrow to avoid large results sets. However, even if you specify a narrow range you'll still get all substances where your ranges falls within a much larger reported range, which limits the usefulness of this searching. Remember also that a substance record might contain many values for a given property in a table, based on different conditions, and from different sources.
For example, you want to search for substances that have a reported melting point between 82-84 C, a UV absorption maximum between 312-315 nm, and for which some IR data have been reported. Your query box would look like this:
(When you enter multiple parameters in a single field form, such as a solvent with MP, they are automatically combined using the PROXIMITY operator to improve relevance.)
When you view results, the fields you searched in will be highlighted for examination.
The "Doc Index" tab allows you to search across the bibliographic database that includes the source articles and patents for Reaxys chemical information (structures, reactions, properties), as well as about 16,000 other periodical sources that were added from various Elsevier bibliographic files (Scopus, Embase, Compendex, etc.) more recently. These additional sources are not necessarily chemical or even scientific in nature - keyword searches may turn up almost anything.
The "Document basic index" field searches for words found in titles, abstracts, and indexing terms. For more complex bibliographic queries, such as authors, DOI, patent number, date, etc., open the Documents form from under the Forms menu.
Reaxys is not a comprehensive, integrated literature index in the way that SciFinder is. If you need to do a general topical/keyword literature search on chemical topics, it's better to use SciFinder.
No, you have to look up the actual compound. Some handbooks, such as the Aldrich Catalog, still list Beilstein volume/page references (e.g. "Beil. 13, III, 2311"), but Reaxys does not include these at all. Some inorganic substance data tables do provide the corresponding Gmelin Handbook reference alongside the source journal reference, but the Gmelin reference itself is not searchable.
Current database statistics can be found on the About Reaxys page.
Machine Indexing: Beginning in 2016, Reaxys expanded computer-analysis of chemical data from the full text of up to 15,000 journals that are covered by various Elsevier indexing products (Scopus, Embase, Compendex, etc.). However, relatively few of these journals contain chemical information.
Reaxys itself covers chemical compounds with defined structures and/or molecular formulas. In general, polymers, biological molecules (peptides, proteins, nucleic acids, enzymes, etc.), engineered materials (ceramics, alloys, etc.), and multicomponent mixtures are not included in the core substance-based segment of Reaxys.
Generally not. Some data derived from the old print Handbooks were originally assessed for consistency, but more recent data are excerpted directly and uncritically from the literature. (Data taken from patents should be viewed with particular skepticism.) All excerpted data are experimental, however - not derived or calculated.
Reaxys contains plentiful spectral data (peaks, solvents, etc.) on chemical compounds as reported in the primary literature, but no spectral graphs. You can use the form-based search (Query Builder in New Reaxys) to specify spectroscopic data with various parameters.
The Reaxys database does not duplicate the Handbooks. Pertinent data and references were extracted from the printed works to create the retrospective portion of the electronic database.
The printed Beilstein Handbook entries, while in highly abbreviated German, provided some textual descriptions of synthetic chemistry that are not reflected in the online version. Beilstein Supplement V (Heterocyclics, 1960-79) has only been partially converted. Conversely, there is much information later added to the Beilstein database that was never included in the printed Handbook. Consulting the print is rarely necessary.
Numeric property data from the print Gmelin Handbook up to 1975 can be found in Reaxys. However, Handbook data from 1976 to 1998 (when it ceased) have not been added. Print Gmelin remains useful because it was more textual and narrative in presentation and provided useful tables and graphs that are not included in Reaxys. It is a good idea to follow up a database search with a look in the corresponding Gmelin volumes.
Because the print and online formats are complementary, the printed Handbooks remain available in the Chemistry Library's collection.
Use both if you can, especially when trying to determine if a specific molecule has been reported. SciFinder and Reaxys are complementary resources, and although both are ultimately based on the primary literature, they are quite different in the ways they scan and index the literature, the ways they register compounds and index reactions, and in the time periods they cover.
Reaxys is valuable for obtaining preparation/reaction methods and physico-chemical data for pure compounds, and especially for its thorough coverage of pre-1960 organic literature. Since its modern journal coverage is more selective than SciFinder's, search results can be more manageable.
SciFinder excels with its wider and deeper coverage of literature and patents from ca. 1965 forward; its comprehensive registration of all types of chemical compounds, polymers, mixtures, and advanced materials; and its more thorough coverage of organic reactions after 1985. Results can be voluminous because of the sheer number of references, so it's important to do focused searches and use the refining and analysis tools.
This work is licensed under a Creative Commons Attribution-NonCommercial 2.0 Generic License.