CHAPTER 1.1
RETROSYNTHETIC ANALYSIS

Every organic synthesis problem actually begins at the end of the story, a target molecule (TM). The goal is to design a reasonable synthesis that affords the TM as the major product. In the interest of saving both time and money, an ideal synthesis will employ readily available starting materials and will be as efficient as possible. The planning of a synthesis involves imagining the possible reactions that could give the desired product; this process is called doing a retrosynthesis or performing a retrosynthetic analysis of a TM. A special arrow is used to denote a retrosynthetic step. The ? arrow leading away from the TM represents the question "What starting materials could I use to make this product?" and points to an answer to that question. The analysis begins by identifying a functional group present on the target molecule and recalling the various reactions that are known to give products containing that functional group (or pattern of FGs). The process is continued by analyzing the functional groups in the proposed starting material and doing another retrosynthetic step, continuing to work backward toward simple, commercially available starting materials. Once the retrosynthetic analysis is complete, then the forward multistep synthesis can be developed, beginning with the proposed starting materials and treating them with the necessary reagents to eventually transform them into the desired TM.

Retrosynthesis and Synthesis of a Target Molecule (TM)

A retrosynthesis involves working backward from the given target molecule (work done in our minds and on paper), while the synthesis is the forward path leading to the target molecule (experimental work done in the lab). Performing a retrosynthetic analysis is challenging since it requires not only knowledge of an enormous set of known organic reactions but also the ability to imagine the experimental conditions necessary to produce a desired product. This challenge becomes more manageable by developing a systematic approach to synthesis problems.*

When evaluating a given target molecule, it is important to consider how the functional groups present in the TM can be formed. There are two possibilities for creating a given functional group: by conversion from a different functional group (called a functional group interconversion or FGI), or as a result of a carbon─carbon bond-forming reaction (requiring a retrosynthetic "disconnection"). In order to synthesize a target molecule (or transform a given starting material into a desired product), a combination of FGIs and carbon─carbon bond-forming reactions will typically be required. While the key to the "synthesis" of complex organic molecules is the formation of new carbon─carbon bonds, the synthetic chemist must also be fully capable of swapping one functional group for another.

1.1.1 RETROSYNTHESIS BY FUNCTIONAL GROUP INTERCONVERSION (FGI)

Each functional group has a characteristic reactivity; for example, it might be electron-rich, electron-deficient, acidic, or basic. In order to synthesize organic compounds, we must construct the desired carbon framework while locating the required functional groups in the appropriate positions. This necessitates that the chemist is familiar not only with the reactivities of each functional group but also with the possible interconversions between functional groups. Such functional group interconversions (FGIs) enable the chemist to move along a synthetic pathway toward a desired target.

Examples of Functional Group Interconversions (FGI)

Let us consider a carboxylic acid target molecule (RCO2H). There are many ways to generate a carboxylic acid functional group, so there are many possible syntheses to consider (often, there may be more than one good solution to a given synthesis problem!). One reaction that gives a carboxylic acid product is the hydrolysis of a carboxylic acid derivative, such as a nitrile. Therefore, a possible retrosynthesis of a carboxylic acid TM (What starting materials are needed?) is to consider an FGI and imagine a nitrile starting material. In other words, if we had a nitrile in our hands, we could convert it to a carboxylic acid, leading to a synthesis of the target molecule.

Retrosynthesis of a TM via FGI

Synthesis of the TM

Choice of Reagents

There is almost always more than one reagent that can be used to achieve any given transformation. In fact, a quick look at a book such as Comprehensive Organic Transformations by Richard Larock* reveals that there may be dozens of possibilities. Why have so many methods been developed over the years for organic reactions? Because not every molecule-or every chemist-has the same needs. The most obvious reason any "one size fits all" approach fails is that complex synthetic targets contain a wide variety of functional groups. The molecule as a whole must tolerate the reaction conditions used, and side reactions with other functional groups must be kept to a minimum. For example, chromic acid oxidation (Na2Cr2O7, H2SO4) of a 2° alcohol to give a ketone would not be useful if the starting material contains any functional groups that are sensitive to acidic conditions. In such a case, the Swern oxidation might be preferred (DMSO, ClCOCOCl, Et3N). New reagents, catalysts, and methods are continuously being developed, with goals of having better selectivity, better tolerance for certain functional groups, being "greener" with less waste or lower toxicity, requiring fewer steps, being more efficient and/or less expensive, and so on.

The focus of this book is on the strategies of organic synthesis; it is not intended to be comprehensive in the treatment of modern reagents.┼ Instead, reagents used are those that are typically found in undergraduate organic chemistry textbooks. Hopefully, these reagents will be familiar to the reader, although they would not necessarily be the ones selected when the synthesis moves from paper to the laboratory. Furthermore, experimental details╬ have largely been omitted from this book. For example, osmium tetroxide oxidation of an alkene is given simply as "OsO4." In reality, this expensive and toxic reagent is used in catalytic amounts in conjunction with some other oxidizing agent (e.g., NMO), so the precise reagents and experimental reaction conditions are much more complex than what is presented herein.

1.1.2 RETROSYNTHESIS BY MAKING A DISCONNECTION

Rather than being created via an FGI, a functional group (or pattern of FGs) may be created as a result of a reaction that also forms a carbon─carbon sigma bond. In that case, the retrosynthesis involves the disconnection of that bond. In a typical carbon─carbon bond-forming reaction, one of the starting material carbons must have been a nucleophile (Nu:, electron-rich), and the other must have been an electrophile (E+, electron-deficient). While this is certainly not the only way to make a carbon─carbon bond (e.g., the organometallic coupling reactions explored in Chapter 8 offer an alternate approach), the pairing of appropriate nucleophiles and electrophiles serves as an important foundation to the logic of organic synthesis, and such strategies will solve a wide variety of synthetic problems. Therefore, the disconnection of the carbon─carbon bond is made heterolytically to give an anion (nucleophile) and a cation (electrophile). These imaginary fragments, called "synthons," are then converted into reasonable starting materials. By being familiar with common nucleophiles and electrophiles, we can make logical disconnections. The example below shows the logical disconnection of an ether TM, affording recognizable alkyl halide E+ and alkoxide Nu: starting materials.

A Logical Disconnection of a TM

Disconnecting that same carbon-oxygen bond in the other direction (with both electrons going to the carbon) would be an illogical disconnection, since it leads to an electrophilic oxygen synthon for which there is no reasonable equivalent reagent.

An Illogical Disconnection of a TM

Let us consider once again a carboxylic acid TM. We have seen that a carboxylic acid can be prepared by an FGI if the carbon chain is already in place, but it is also possible to create new carbon-carbon bonds in a carboxylic acid synthesis. For example, the reaction of a Grignard reagent with carbon dioxide generates a carboxylic acid functional group, so this presents a possible disconnection for the target molecule's retrosynthesis. The logical disconnection is the one that moves the electrons away from the carbonyl, giving reasonable synthons and recognizable starting materials (RMgBr Nu: and CO2 E+).

Retrosynthesis via Disconnection of a TM

Synthesis of the TM

What Makes a Good Synthesis?

The fact that multiple retrosynthetic strategies usually exist means that there will often be more than one possible synthesis of a desired target molecule. How can we determine which synthesis is best? This depends on many factors, but there are some general rules that can help us devise a good plan to synthesize the simple target molecules found in this book.

1. Start with reasonable starting materials and reagents. A good synthesis begins with commercially available starting materials....