CHEM 184/284 (Chemical Literature) - Huber - Winter 2023: Lecture 15

Searching Reactions

CASREACT -- the CA database for organic chemical reaction searching

It is important to be aware that when you use the Reaction search feature in SciFinder-n, you are searching a different database than when you search for substances. Substance searching uses the REGISTRY file of over 175 million substances (approximately 100 million of which are simple organic and inorganic substances), covering the literature from 1907 to present, plus selected earlier references.

CASREACT, on the other hand, is more focused. It covers selected journal articles in organic chemistry (1840-present) and organic chemistry patents (1982-present). It indexes EVERY single-step and multi-step reaction described in the papers, including reactants, reagents and products, with yield information where available. Recently, reaction data from the Wiley reference works, Organic Reactions, Organic Syntheses and Enccyclopedia of Reagents for Organic Synthesis were added to the file. At present  Dec. 2020), over 132 million reactions are included in CASREACT, with over 150,000 new reactions added per week.

Therefore, when searching for organic reaction information, if a truly comprehensive search is needed, both Reaction Searching and the other two approaches mentioned above should be employed. Reaction Searching alone is very detailed for organic reactions, but covers selected journals and patents only. Substance Searching is less powerful, but will extend your search to a wider range of source documents. Research Topic searching may pick up some additional reactions from the 1907-1966 literature.

Finally, remember that the CASREACT database does not index reactions by "name reactions", so you can't use "Diels-Alder" or "Heck reaction" in a reaction search as such. However...you can do a reaction search, Get References and then refine the references for the "name reaction" you're looking for or you can do a Research Topic search for the desired name reaction, then Get Reactions, and Refine the resulting set of reactions by a further Substance or Reaction search.

Searching Reactions by Keyword/Substance Name

• SciFinder-n provides a text search window allowing you to search the reactions database by keyword or chemical name. 
• Note, however, that the keyword searching only applies to reaction notes, which can be highly variable. In particular, the reactions database DOES NOT index name reactions as such. So, for instance, if you want to find Diels-Alder reactions, or aldol condensations, you should either use the reaction drawing tool to create the appropriate reaction diagram, or, do a References search on your reaction name, and then Get Reactions from that answer set, and filter the resultant Reactions answer set to find what you need.
• Searching substance names is more successful. Below see part of the answer set for searching for "iron pentacarbonyl" in Reactions

• Starting from the top,  the Reactions header describes the search.
• Below that, on the feft, is a button to Get References for the reactions in the answer set.. On the right-hand side re the Download, E-mail and Save and Alert icons, which work in much the same way as their counterparts in References and Substances answer sets.
• Below that, n the left is the Filer Behavior header, followed by the total number of results in the answer set, 
• To the right, is a drop-down menu of Group by options:
  • By Scheme - groups together reactions with the same reactants and products.
  • By Document - Groups reactions by the document in which they apper. Only one reaction is displyed for each. To see all reactions in a given document, click View Related Reactions.
  • By Transformation - Groups reactions by the transformation which occurs. To see individual reactions within a group, click on View Related Reactions. See below for an example of transformation grouping

• At the right-hand side, there id a drop-down menu to choose how reactions in the answer set are sorted.   The options are: Relevance, Publication date (ascending or descending), Yield (highest to lowest), and Number of Steps (ascending or descending). Note that there are no sort options when the reactions are Grouped by Transformation.
• Next to that is the drop-down menu for choice of view. The default is Expanded; the other alternative ic Collapsed, a less detailed view.
• On the left, the Filter options for Reactions answer sets. Note that you may Filter by or Exclude any of the options:
  • Substance Role - Includes Product, Reactant, Reagent, Catalyst and Solvent, as relevant.
  • Yield - listed in 10% ranges from highest to lowest.
  • Number of Steps - from fewest to most.
  • Non-participating functional Groups - This highly useful filter identifies functional groups which exist in both the reactants and products. So, for example, if you are trying to carry out a reaction in which one amine group in a reactant is converted to an amide, but a different amine group is left unreacted, you can use this filter under Amine to find potentially useful reactions.
  • Reaction Mapping -  Allows you to select reactions that have atom mapping data available.
  • Experimental Protocols - Reaction records in SciFinder-n may include one or two types of very detailed reaction procedure infomration. Experimental Procedures are copied directly from the original paper. They are only available where the publisher has given permission. So, there are lots of these from articles in ACS journals, not so much from Elsevier journals. Synthetic Methods  are detailed procedures extracted by CAS indexers from the original documents  There will be examples of each below. 
  • Reaction Type - Can be Full, Product Only or Failed. Note that records of failed reactions can be very useful, as they may suggest procedures to avoid, or procedures that you might be able to creatively fix!
  • Stereochemistry - Identifies the presence or absence of stereochemical centers in the reaction diagrams.
  • Reagent - tabulates the reagents (if any) used in the reactions.
  • Catalyst - substances used as catalysts in the reactions
  • Solvent - solvents used in the reactions. This can be very useful for identifying "green" alternatives for synthetic procedures. Note that SciFinder-n only allows filtering by specific solvents, not by classes of solvents.
  • Commercial Availability - Identifies reactions in which the product is commercially available, or in which all the reactants are commercially available.
  • Reaction Notes - terms in the reaction notes. These can include reaction types like photochemical or stereoselective, or reaction conditions like "high pressure" It also includes the term Safety, indicating that there is some hazard warning associated with the reaction. Note that the lack of a Safety term is not a guarantee that there are not hazards associated with the reaction. Another note term which is sometimes used is "Green Chemistry" which may be useful in filtering your results.
  • Search Within Results - Allows you to search by structure drawing for substances or reactions within your answer set.
  • Source Reference Filters - You can filter the  by the following document information: Document Type, Language, Publication Year  Publication Name or CA Section. If you wish to filter by other document characteristics, such as Author, use the Get References option to create a References answer set, then filter that set.
• Brief Record Information (when grouped by Scheme)
  • Scheme Number and number of Reactions in the scheme.  At the right: Number of Steps and Yield If you click on the three dots ar the right, you can copy the reaction diagram to the structure drawing tool for editing.
  • Structure Diagrams of Reactants and Products - Each substance has a link to Supplier information, where available. If you click on the structure diagram, a pop-up Quick View of the substance appears (see Lecture 13 for more on Quick View).
  • Reaction Summary - Gives number of steps and yield. , brief description of the reaction, reaction conditions, such as temperature and solvent(s) for each step, brief bibliographic record of the source document, and links to Reaction Detail, Experimental Procedure (where available) and Full Text. of the source document (including PatentPak, where applicable.)

Reaction Detail Example

• Reaction Detail Content includes:
  • Next to the Reaction Detail header, the scheme number and the number of the reaction in the scheme. If you had grouped by Document, it would give the Document number and the number of the reaction in the document.
  • At the far right, there are navigation arrows to go to previous or next reactions within the scheme or document.
  • Below that are the Download, E-mail and Save options for this reaction
  • Next is the reaction diagram for this reaction. As in the brief record, each substance links to a Quick View of its substance record, and has links to Supplier information, if available. At the right is the Number of steps and Yield for this reaction.
  • Below that is a listing (where relevant) of the Stage, Reagents, Catalysts, Solvents and Conditions for this step. Note the link to Alternative Steps.Clicking on this will open a new tab/window with an answer set of one or more aternative steps for this reaction.
  • At the right is the information about the source document, including authors and their company-organization. Patent sources will include inventors and assignee.
  • The CAS Reaction Number is an internal ID number. Unlike CAS Registry Numbers, they are not, at present, searchable in SciFinder-n.
  • Notes - can include reaction type details, such as photochemical. This is where safety notes will appear. However, safety details do no appear here. See the experimental protocols, or full text of the source document for safety information.
  • Experimental Protocols - Here you will find tabs for MethodsNow and/or Expericmental Procedures, if available. The image above shows the MethodsNow protocol for this reaction.
    • Products - Gives the CAS RN or common name of the product(s) and the percent yield.
    • Reactants - Gives the CAS RN or common name of the reactant(s)
    • Catalysts - Gives the CAS RN or common name of the catalyst(s), if any.
    • Solvents - Gives the CAS RN or common name of the solvent(s) if any
    • Note that, for each of the above, the name or RN is hotlinked to the Quick View of the substance record.
    • Procedure - Detailed step-by-step procedure for the reaction, including quantities, reaction conditions and purification steps if any.
    • Transformation - Describes what the molecular transformation is in the reaction.
    • Scale - The quantity category by weight of product produced in this procedure, such as milligram, gram, kilogram, etc.
    • Characterization Data - What will be available will vary, but may includ spectral data (NMR, IR, mass, UV-Vis, etc.), physical properties (melting point, boiling point, refractice index, etc.) and state (solid, liquid, gas; possibly crystalline structure and/or color.)
    • CAS Method Number is another internal ID number which is not, at present, searchable in SciFinder-n.
  • Below is the Experimental Procedure tab for the same reaction.
    • This is the procedure copied directly from the Experimental section of the source document. It is precisely as detailed and systematic as the author chose to make it, following the editorial practices of the journal.

 

Searching by Reaction Drawing

As seen above, it is easy to work from a substance record to reactions in which that substance is a product. But you can also achieve more specificity by drawing both reacants and prodcuts in a single query. Let's say I am looking for reactions which will transform aryl amines to nitro=aryls, while preserving an unreacted amine group elsewhere in the molecule.  I start be drawing the basic reactant and product molecules. Note that since I am modifying the amine group and the nitro group, I cannot use the shortcuts for these groups, but mus draw them out fully.

Now I use the Reaction Arrow tool to specify which is the reactant and which the product. Click and hold the cursor near the reactant, and drag toward the product. With the arrow drawn, note how the reaction roles are now listed below each structure. If you wanted to change a role, you could use the Reaction Role tool, click on the structure and reassign.

Now I wish to specify that the amino nitrogen in the reactant becomes the nitro nitrogen in the product, To do that, I use the Map Atoms tool. Select the tool, then click first on the reactant nitrogen, then on the product nitrogen.  Note that each is now labeled with a 1. Map Atoms can be used repeadedly for multiple atoms if necessary. If my reaction had bonds being formed or broken without atoms being added or lost, as in a Dield-Alder reaction, I could use the Map Bonds tool to indicate which bonds were being changed.

Now I'm ready to search.  I click OK, to transfer the reaction diagram to the query box, the click the Search icon and get my initial reaction search results. Note the option to Start Retrosynthetic Analysis.

Below, I have selected the Substructure answers, as I wish to take the larger answer set and Filter the reults to get some very specifoc reaction information.

First, I want to get some detailed instructions right away, so I finter to reactions with Synthetic Methods procedures. .I also want high yield reactions, so I then select Yield 90-100%. As you see below, I'm down to 515 reactions.

I'm not quite done. I also have an amine group elsewhere in my reactant that I don't want oxidized. So I take a look at the Non-participating Functional Groups filter, look at the whole set, and search for "amine". If you check the numbers associated with each group, you will see that "Amine" includes Primary amine", "Secondary amine" and "Tertiary amine", but selecting all those groups does not hurt my results.

I select all the terms that mention "amine" and click Apply. and now I'm down to 14 reactions.

Now, if I wish, I can take a look at some of the other filters, like Catalyst ,Solvent and Reaction Notes to see if I wish to refine further. I might want to avoid expensive palladium catalysts, so I can Exclude them or I might want a reaction I can carry out in water to avoid polluting solvents, or I might find the notion of a microwave stimulated reaction interesting.

Here's the reaction detail for one of the reactions we found:

Reaction Similarity Searching

Reaction similarity searches are done in the same way as other reaction searches. Simiarity is defined in three categories:

• Broad - Atoms at the reaction centers must match, but any other variation is allowed.
• Medium - Reaction centers, plus adjacent atoms and bonds must match
• Narrow - Reaction center, plus atoms and bonds up to two atoms away must match.

Below is an example from a reaction similarity search:

 

Retrosynthesis 

This feature of SciFinder-n allows you to create a synthesis plan for either known or hitherto unknown molecules with tremendous control over the synthesis steps. For known substances, you may access the Retrosynthesis tool  from the substance record, or, for known or predicted substances, you may draw the structure with a version of usual structure drawing tool. Note that some of the structure drawing options are unavailable, because you are drawing an exact structure, not a substructure, similarity or reaction search.

For a known molecule, all you need to do is find the substance using any of the substance search methods, then click on the structure diagram to pull up the Quick View. Below is the Quick View for the drug, glipizide.

 

Note the fourth option on the left, Start Retrosynthetic Analysis. You can click on that to begin the process. The first time you use the system, you'll see the pop-up message below:

• Plan Otions
  • Select Synthesis Depth - Lets you select how far back you want the synthesis to go. Depending on the cost of some of your reactants, if may be moreeconomical to sythesize them rather than buy them. On the other hand, it may be more economical of your time to purchase as much as you can.
  • Break and Protect Bonds - This lets you speicy where you want, or don't want to put fragments together in making the compound.
  • Set Rule Supporting Predicted Reactions - SciFinder-n uses an artificial intelligence reaction planning software developed by ChemPlanner to craft its predicted reactions. It relies on the examples of similar reactions from the literature. You can choose, in effect, how many predictive reactions have to exist fora reaction step to be considered.
• Steps - By clicking on the Steps tab, you can view the Yield, Evidence (that is the literature references which the step is derived from) and Alternative Steps for that poitn in the reaction.
• Scoring - By clicking on the Scoring Tab, you can vary what desirec characteristics your syntheis plan will emphasize, including:
  • Complexity Reduction - How important is the simplicty of each step?
  • Convergence - How many precursors do you want in the plan?
  • Evidence - How many literature exmples do you want suppoting each step in the plan?
  • Cost - How important is controlling the cost of the synthesis?
  • Yield - How imporant is having high yield in each step of the synthesis?
  • Atom Efficiency - How much to you want to control waste byproducts?

After clicking Create Retrosynthesis Plan, here's the opening screen of a retrosynthetic plan for glipizide, showing only experimentally verified reaction steps.

 

Here are some of the significant features to note:

• A three stage synthesis is shown, with the ultimate product on the left, and the stages working backwards from left to right.
• Currently the plan is displaying only experimentally verified stips. To see predicted possibilities, turn Predicted ON on the left. Predicted plans may not appear immediately. When they become available, you may view them from your Search History.
• Note on the left that SciFinder-n has taken the reported yields of each reaction step and calculated an estimated yield for the reaction as shown. It has also calculated, based on yields, and the prices of commercially purchased reactants, an overall price per 100 grams produced..
• Note that the structures for each of the substances in the plan is hotlinked to its supplier information (if any) and to the Quick View of the substance record.
• You can change steps at any point by clicking on the appropriate icon.
• If you click on the Plan Options link on the left, you can tweak the rules for the retrosynthesis to fit your particular needs. See below:

To create a retrosynthesis plan for a hypothetical molecule, open the Substance drawing window, draw your molecule, click OK, and then switch the query option to Reactions. You should see a screen like this:

 

T

You can click on Create Retrosynthesis Plan directly, or go to Set Plan Options first.  Note that in the absence of an experimental synthesis for your molecule, the predicted syntheses may take a relatively long time to generate. However, you can log out, and the system will continue working. When the predictive retrosynthetic plan is complete, you'll get an e-mail. Log back into SciFinder-n, go into your Search History at the lower part of your home page and click on Open Plan.

Here's the result on a hypothetical trifluoromethyl derivative of glipizide.

• © 2023 Charles F. Huber

This work by Charles F. Huber is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License.

Screenshots are © 2023 American Chemical Society, and are used here for fair use educational purposes only.
Based on a work at guides.library.ucsb.edu