Exploring Reaxys


Reaxys is a web-based platform designed to facilitate the discovery and use of core and very broadly related chemistry information. Our aim is to explore the capabilities of Reaxys as a research tool and as an environment for chemistry teaching and learning.

Students in the Cheminformatics OLCC have access to Reaxys. Please contact your Facilitator for registration account information.


Please subscribe to this page if you wish to participate in discussions related to Reaxys. You can look at the left pane and see who will receive an email when you make a comment. You only need to subscribe to comments. Do not subscribe to updates.



Reaxys — a chemistry research tool

With an intuitive user interface and easy-to-use search tools, Reaxys delivers chemistry and chemistry-related information in the form of literature and patents, excerpted properties of substances (chemical, physical, spectral, bioactivity), and details on reactions and experimental procedures. Functions to help process results allow evaluation and selection of answers to a query.

We invite you to explore Reaxys and participate in discussions below: post a comment, write a question, share an answer. We hope to trigger thoughts and curiosity as you learn about cheminformatics and the tools available to find chemistry information.


Getting started

Once you have login information from your Facilitator, begin exploring Reaxys at the following link:


You will find a description of Reaxys in the Reaxys Fact Sheet and a quick overview of how to use Reaxys in the Quick Start Guide. The Reaxys Support Center is also a good place to find some search workflows, including the following:

Structure Drawing Workflow
Synthesis Planner Workflow
Substance Searching Workflow
Literature Searching Workflow
Reaction Searching Workflow


Materials and Resources

As topics arise in the discussion, we will develop and post activities to help you explore the topics further in Reaxys. These resources, which you can download from the top of this page, give you:

 An introduction to finding properties for substances in Reaxys (Reaxys_PropertiesSearch_200117)

 An introduction to performing text searches in Reaxys (Reaxys_TextSearch_230117)

 An overview of main approaches to searching substances in Reaxys (Reaxys_SubstanceSearch_270117)

 A description of the content and use of Substance Records in Reaxys (Reaxys_SubstanceRecords_030217)

And if you have a particular question that you would like to research in more detail, let your Facilitator know. We welcome the opportunity to help you.

No votes yet
Join the conversation.

Comments 23

OLCC S13 | Sun, 01/22/2017 - 09:51

I had never heard of Reaxys until I started this class. Our library does have SciFinder Scholar, which seems to be very similar, and I was wondering if you could compare the two. When would we want to use Reaxys, when would we want to use SciFinder? Is there a good article comparing the two?


Damon Ridley's picture
Damon Ridley | Mon, 01/23/2017 - 04:27


Great questions and thanks for posting them!

To answer your last question first, perhaps you can have a look at DOI: 10.1021/acs.jchemed.5b00601 which is a relatively recent article on the topic you raise. After I saw this publication Neelam and I had a number of discussions on the content of the paper and my quick summary is that she and her colleagues did a great job although there were a few things in some of the text that I worried about.

Can I compare the two? Well, yes, although I have not been involved with SciFinder since early 2013 and while I have kept up to date with news about SciFinder on various websites, I have not used it for almost 4 years. Nevertheless, my understanding is that the principles outlined in my last SciFinder book are still mostly valid. [Information Retrieval: SciFinder, 2nd Edition, Damon Ridley, ISBN: 978-0-470-71247-4.]

When would we want to use Reaxys, when would we want to use SciFinder? My first comment is that chemists and chemistry (very broadly defined, meaning going beyond traditional chemistry into fields such as biomedical, materials, environmental sciences and so forth) are extremely lucky to have two excellent information resources to tap into and in my opinion all universities with major science programs should have both. However, I know that funding and available time (to look through multiple systems) both are in short supply, so indeed scholars have to consider at one time or the other, which to use and when.

My second comment is that the products are continually being upgraded, not only in content but also in search, display and post-processing functionality. One consequence of continual upgrades is that any comparison needs to be kept up to date, and here I add that Reaxys made huge changes during 2013 and 2014 so any review of Reaxys before the end of 2014 is essentially not valid and in my opinion should be dismissed. Further, Reaxys introduced a super new interface in October 2016, and SciFinder has announced its new interface will be available soon.

In your comment you suggest both products seem “to be very similar”, and in the sense that both have document, substance and reaction records you are correct. Indeed it is the combination of these three types of records that clearly stand Reaxys and SciFinder apart from everything else (including web search interfaces such as Google/Google Scholar). However, there are substantial differences in the content, the organization, and the structure of data, and if you work through the document posted in the Reaxys Module in your course: Reaxys_PropertiesSearch_170117.docx, you see one major example.

To discuss such differences in more detail, however, is beyond the scope of any reply to a general question relating to comparison of the two products, but if you or others involved in the course would like further details, about specific areas, then please enter them in separate and new comments - and do it as quickly as possible since that will give us time not only for a more comprehensive discussion, but also to allow Anja and me to (possibly) create new teaching Reaxys Modules or even some (research-based) Special Projects for your course.


OLCC S50 | Tue, 01/24/2017 - 11:32

I have been going through the examples this morning on page 3 of the handout on property searches. I have found that text searches for name are not as robust or precise as structure searches. The only one that I initally had trouble with is "What can you find on the cyclovotammetry of tetraphenylporphyrin?" I tried to do a property search on voltammetry but that doesn't seem to be an indexed property. Next, I just did a document search, and limited it to abstracts contain voltammetry. This appeared to give me some useful data. Is there a better/more efficient way to do this search? Has anyone else had success with this search in a different way?
Thank you,

Anja Brunner | Tue, 01/24/2017 - 15:50

Hello Chris,

Did you try doing the search by entering cyclovoltammetry of tetraphenyilporphyrin into the Search Reaxys box of Quick Search (the main landing page of Reaxys)? Reaxys recognizes cyclovoltammetry as a property and will pull Records of relevant Substances that contain cyclovoltammetry information. This will be the first hit set listed by Reaxys. Each record will show the Hit Data, which you can explore by expanding it. Look at the Hit Data in the first record. You will see that Cyclovoltammetry is one of the indexed terms in the Description subfield of the Property Field Electrochemical Characteristics. I suspect that when you say that you did a property search on voltammetry, you searched that term under the fields available in Query builder. Is that right? Try looking for Electrochemical. The field Electrochemical Characteristics will be one of the Properties you can select and drag into the Query builder workspace. There, select show fields and click on the Lookup icon of Description. Cyclovoltammetry is one of the terms you can search.

I hope that helps! Otherwise, please keep asking questions.

Damon Ridley's picture
Damon Ridley | Wed, 01/25/2017 - 01:27


I tried to reply to your message last night (currently I am in Germany), but my reply did not go through the system.

Nevermind, since I see Anja (my colleague) has replied. A couple of extra notes:

1. Structures are precise entities, and provided you build the structure (and choose the right search options to meet your needs- the options are in the panel on the right of the structure drawing screen) you should get "robust" results (meaning very good precision and comprehension in answers).

2. Text data you search comes from authors in titles and abstracts (and authors do not enter terms systematically), and indexers in Index Terms (these are systematic terms). There clearly are a lot of variables here, and this mainly explains why text searches are not so "robust or precise".

3. Skilled searchers (e.g., information professionals) place a lot of importance on indexing, which can be found mainly in two ways: a. doing a search and looking for index terms in answers; b. consulting an index term list.

3a. This is exactly what Anja did and what she explained in her post; you get the index terms (in Substance Records) immediately in this case.
3b. Later on in the "properties module" we see how to find the index terms (Query builder => Electrochemical characteristics => click the Lookup icon in the Description Field).

I hope this helps.


Damon Ridley's picture
Damon Ridley | Wed, 01/25/2017 - 02:06

Hi Everyone,

In my reply to Chris I referred to terms 'entered' by authors and by indexers in database records.

As authors, it helps to remember that the terms we use in titles and abstracts will make up a large part of the text that is searched in databases. We want fellow scientists to find our papers, right? - so it follows that we should carefully consider the terms we use. I am not saying that we should be de facto indexers, but we definitely need to consider terms that have "meaning" within the context of being part of a searchable database.

Compare abstracts that contain:
"Compounds (1) to (5) had reduction potential between -1 and -1.2V ......" and
"Cyclovoltammetric studies of the zinc tetraphenylporphyrin derivatives prepared showed half-wave potential ....."

The first abstract has few searchable terms [and (1) to (5) means nothing as an entry in a database, right], whereas at least the second abstract has a few searchable terms.

Think about it when you write your next title and abstract!


Anja Brunner | Wed, 01/25/2017 - 06:52

Research shows that the vast majority of information searches performed by scientists are text-based. So, Damon and I have posted (at the top of this module page) a new resource introducing text searching in Reaxys (Reaxys_TextSearch_230117). In it, we comment on general issues to consider for text searches, how systems build algorithms to interpret text-based queries, and we've worked through some of the ways in which Reaxys handles text searches.

Reaxys contains trillions of searchable words, not just in Document Records but also in Substance and Reaction Records. Creating a search engine that accesses and effectively unveils this content embedded in different types of records is challenging. However, with a little understanding of the background and a little practice, we can quickly see the opportunities available in Reaxys.

Please do take a look at the resource and try some of the activities it describes in Reaxys. And we look forward to any comments or questions on text searches in Reaxys!


Anja Brunner | Fri, 01/27/2017 - 05:05

With the complexity of naming chemical substances and the different ways in which substances are systematically indexed by database producers, searching information on a substance or group of substances can pose a challenge.

Reaxys houses a very extensive database of substances and associated information, and as you will see, that information comes from different sources and is organized into several types of records. The system also offers several opportunities to retrieve that information. Learning about some of those options is the objective of another resource we have posted (at the top of this module page).

Try some of the searches! And please do not hesitate to post comments and ask questions!

Robert Belford's picture
Robert Belford | Fri, 01/27/2017 - 11:17

A student posted a question to hypothes.is, but it did not go to everyone who is subscribing to this article, so I am pasting it here

I understand from the paragraph which databases are used for sodium acetylide. On the other hand, i don't understand which databases treat sodium acetate as 2 fragments. Any help?

Please note, this deals with the article by Cooke and Ridley, which is at the bottom of this page. Here is a google short URL to the hypothes.is link, and in case it gets depreciated, I am pasting the long one at the bottom of this comment https://goo.gl/fS7PJz . If you join the hypothes.is group called OLCCSendMail, these two links will take you right to the highlighted text the student is talking about.


Olcc S14 | Fri, 01/27/2017 - 12:14

I am trying to find out articles related with adsorption of phosphate from water using wood nanocomposites. I have tried using truncation as adsor*phosphate**water**wood* but the Reaxys does not read it. So can anybody tell me how to find articles in Reaxys related to my subject with truncation or any other methods?

Damon Ridley's picture
Damon Ridley | Fri, 01/27/2017 - 14:23


Yes you certainly need to put spaces between the terms entered. You may also wish to try: absor* phosphate* *water* *wood* which you will find gives a few different hits. Also remember that when you put in 4 concepts like this you are requiring that ALL FOUR be in the database record. (You are not searching the full text and it is possible that fewer concepts are in the database record although all concepts could be in the full text.) So you may wish to try in Search Reaxys: adsorb* *phosphate* *wood* (note that I have left-truncated phosphate in case there are records such as organophosphates (if you also want to include them)). Further: absorb* *phosphate* *wood* ?

Remember that you are requiring a potentially large number of terms to be searched and Reaxys may take a little while (30 seconds?) to find all the answers.

Remember also that you are trying to enter search terms that anticipate what was written by an author (in titles and abstracts) and entered by an indexer in Index Terms. In your search you don't have one or two "really good terms" that will always be used (and on which you can focus) - so you have to try them all (or perhaps three of them as I have indicated).

You may think this is all a bit cumbersome since in something like Google you can enter: adsorption of phosphates by wood whereupon you will find some interesting hits, but then you wonder what you may have missed. The learning point here is that you may have to search a few different sources.

I consider this a "tough" search question, but it seems to me that you can pick up around 200 document records in Reaxys (if you combine the various searches and particularly if you leave out the 'water' concept). Realistically I'd be pretty happy with this at least as a start.


olcc s16 | Fri, 01/27/2017 - 14:36

I'm trying to figure out if Reaxys treating 2-hydroxypyridine and 2-pyridone as one substance or not. We know that 2-hydroxypyridine and 2-pyridone are tautomer. When I perform quick text search for each, Reaxys retrieved strictly one either 2-hydroxypyridine or 2-pyridone. However, if I perform structure search with a check on "Tautomer", I received result of both 2-hydroxypyridine and 2-pyridone as tautomer of each other. So, how can I tell if Reaxys treated these substances as one or two? Because seems like it would depend on my searching method.


Damon Ridley's picture
Damon Ridley | Sat, 01/28/2017 - 00:58

Thanks for your question.

In cases like this I'd definitely go down the structure search route since structure searches are precise and comprehensive - provided you enter the right query.

So from the Landing Page, I'd click Create Structure or Reaction Drawing, then in Create structure template from name enter say: 2-hydroxypyridine. I'd search this structure As drawn and I'd definitely include Tautomers. I'd consider the defaults but at this stage I'd probably leave them as they are.

After you Transfer and Search you find each substance has its own Substance Record. I'd browse some of the other answers just for interest, but there are not many documents for them.

The question now is: Which tautomer is indexed? The answer is that it depends on what was in the original document. If 2-hydroxypyridine was the focus then that would be the indexed substance; similarly if 2-pyridone was the focus - it would be the indexed substance. If the original document discussed both, then while I am not certain of the indexing policy, I think both would be indexed.

I hope this helps.


Damon Ridley's picture
Damon Ridley | Sat, 01/28/2017 - 02:27

People have asked about the indexing of salts in different databases.

There are two parts of the problem:
- there are different "classes" of salts;
- databases may (or may not) have different indexing policies for the different classes.

We can have salts of simple alkali/alkaline earth metals with non carbon-containing or carbon-containing acids, or of simple halides with non carbon-containing or carbon containing bases, or salts of carbon-containing acids with carbon-containing bases. We also talk about hard and soft acids and bases, but the chemical difference here usually relates to the extent of ionisation of the "bond" between the different atoms. Then, when is something a salt or a coordination compound? Calcium hydroxide is normally considered a salt, so is calcium acetate - but is calcium oxalate a salt or a coordination compound? Then, what happens with "hydrates" - are they hydrates, salts or coordination compounds? (Incidentally, I once made a lithium salt of a phosphinic acid, and it gave a strong molecular ion in the mass spectrum - possible only because of a very "covalent" Li - O bond.)

Given all that, the representation of substances in computer databases needs to have clear cut rules and the different databases may have different policies. Hence, in different databases, lithium acetylide may have a different representation from lithium acetate. Further, take ferrocene. When it was first made chemists thought of different ways to represent its structure. These structures were published in papers and databases entered the structures accodingly. However, years later it was realised the actual structure was quite different - that being the case should all the previous structures be deleted, or changed? Are databases always going to make "retrospective corrections"?

The issue here is not necessarily with computers, or the policies of people who create databases, but perhaps more with the (fascinating) complexities of chemical substances.


OLCC S19 | Tue, 01/31/2017 - 23:05

I wanted to look up a Diels Alder rxn mechanism, so I input the name into the quick search and got a gazillion hits. Then I grabbed a recent lab handout and attempted to have structure editor build 4-cyclohexene-cis-1,2-dicarboxylic anhydride by inputting the name. That didn't work at all, so I used the structure editor to draw the compound.
I then searched for rxns and got 28 hits on position/steroisomer, 130 near it, and boatload after that. I clicked the hyperlink for the 28, and found what I thought looked like the closest to what I wanted to see, but when I went to view...it's behind a paywall. Can't view it. I've already had serious trouble merely accessing Reaxys because of other issues which I thought we had ironed out by this morning. Apparantly they are not. Any suggestions?
By the way, I ran into the same thing when I tried to access the paper which is at the bottom of this module through both SciDirect and Reaxys.Thanks for your help.

Damon Ridley's picture
Damon Ridley | Tue, 01/31/2017 - 23:23


If you are interested in the mechanism of a single reaction then can you tell me exactly what that reaction is? Butadiene + Maleic anhydride?

Otherwise, what you have done makes sense in general. I simply need to explain what is happening in your searches. I shall do that in a couple of hours. First I'd like to know exactly what you want though.


OLCC S19 | Tue, 01/31/2017 - 23:41

Yessir. It is butadiene sulfone and maleic anhydride in xylene.

Damon Ridley's picture
Damon Ridley | Wed, 02/01/2017 - 01:10

Many points here, including:

- Quick search: diels alder mechanism gives choices with 20K (actual reactions indexed with Diels Alder - this normally is great, but not in your case because you want the mechanism) and 1136 documents with the terms in titles, abstracts and indexing. Way forward from the latter may be to View Results => Filters => Document Type => Reviews which will give 69 that would be helpful.

- Create structure template from name is great if you have simple names, but matching even half-way complex names is challenging (the system would have to do complicated part name searches).

- Draw the product alone is a sound approach but you would need to sort through the answers (as you have done).

- We have given you access to Reaxys, but not to full text documents through ScienceDirect (sorry); maybe your Uni has a ScienceDirect subscription and you could find the article through that.

- For a specific reaction, it is best to draw the full reaction and search As drawn. When you do this you get 5 documents (although a Similarity search gives 12). Reaction Records are searchable in Reaxys by structure diagrams but also by text terms. It so happens, though, that 'mechanism' is not systematically indexed in association with Reaction Records. So you basically have to rely on a text search in Document Records (which is what you did when you got the 1136 documents above). However, remember that you are searching text in titles, abstracts, and index terms here - and also remember that most Diels Alder reactions are done for synthetic reasons (so mechanisms would not be discussed, since the general mechanism is known (see below)).

- However, the reaction you want is a classic Diels Alder reaction which would follow the normal [4+2]cycloaddition mechanism (HOMO/LUMO) which you would find in most organic text books (or say in Wikipedia: dies alder).


- what you did was great and sound as far as the mechanics of the search goes (what you click and where you click)
- the issue really relates to understanding the database content (what is indexed, and where); this ALWAYS is the difficult part (the mechanics of searching is trivial, right?)
- you get to understand database content, but trying searches (as you did).

Many other comments could be made, but I hope this explains most of the issues. Please don't hesitate to get back to me if you need further help (on this or on other searches).


OLCC S19 | Wed, 02/01/2017 - 02:00

Thanks. That's why I chose this reaction, so that I could find familiar points. I'll go back later - much later, as it is 0200 here - and try drawing out the reaction like you said and searching that.

Damon Ridley's picture
Damon Ridley | Thu, 04/13/2017 - 17:30

Time zones! (Your first post that I answered was before 6:00 my time.) Actually, Reaxys does have "mechanism" as an index term in Reaction Records, but it is in the Subject Studied Field. Query builder => Search properties: subject => Drag Subject Studied to the working space. To see terms in the field, click Lookup icon at the very right of the field and in this case you see 16 terms, one of which is: mechanism. Check the box next to it, then Transfer. (Lookup is a very handy tool to help select terms in any field in Query builder.) If you now Search (Reactions) you get around 75,000 reactions that have been specifically indexed as mechanism. To narrow these, click back to Query builder, click Structure icon at the top of the screen, and then build the structure of your cyclohexene anhydride adduct. Click the arrow icon (bottom of panel on left), then click to the left of your structure and drag towards the structure. You then have drawn an arrow that indicates to Reaxys you want the structure to be a product in a reaction. Click Substructure on the right panel, and then Transfer query. You now have a structure query and Subject Studied: mechanism query. Click: Search (Substances). You will get 38 reactions from around 80 documents. If instead you search reaction: structure of butadienesulfone + maleic anhydride => cyclohexene As drawn you don't get any hits. You can also try other structures if you wish, although if you search a very general structure (such as cyclohexene, substructure, as product - i.e., you have drawn the reaction arrow as above) you will find the search takes several minutes to complete. In such cases you can try Similar rather than Substructure as your search type. All this is possible because you can search text terms in Reaction Records in Reaxys. I hope this helps. Damon

Robert Belford's picture
Robert Belford | Thu, 04/13/2017 - 15:40
on this tutorial,<a href="http://supportcontent.elsevier.com/RightNow%20Next%20Gen/Reaxys/New_RX_Substance_Searching_Workflow_12-16-2016.pdf">http://supportcontent.elsevier.com/RightNow%20Next%20Gen/Reaxys/New_RX_Substance_Searching_Workflow_12-16-2016.pdf</a> step 8.e has you set the substituents to 6, so the atoms are labeled (s6). What is the value of the number? Why 6? I may discover later, but I am not sure what this parameter does. Thanks, Bob

Damon Ridley's picture
Damon Ridley | Thu, 04/13/2017 - 17:51
Bob Good question! (And I have often wondered why myself.) I think the answer is 'historical' and traces back to the fact that the default structure search in Reaxys (and its predecessors) is 'As drawn' which mostly means that positions are blocked unless you 'open' then. This is the reverse for many other products where the default structure search is 'by substructure' and then you need to 'close' ("lock") positions where you don't want further substitution. When you want to 'open' a position in Reaxys you can 'open' up to various extents and the node then is shown annotated (s#) where # can be 1,2,3 etc to indicate the number of further substituents allowed. The annotation used to 'fully open' is: (s6). Personally, I don't use this to open a specific node in MarvinJS. Instead, I right click on the node, go to the Advanced Tab and then adjust the Substitution box to what I want. Note that if you want to allow all substitutions or just substitutions on heteroatoms then MarvinJS has a box at the top right of the Structure Editor to do this. Damon