In a linear chemical process, as opposed to a convergent one, the cost of an expensive starting material for the first transformation may become predominant as the number of linear steps rises.
A consequence is that the first reaction’s overall yield from a sophisticated starting material can easily become more important than the entire combined costs of reagents, catalysts, processing chemicals, and solvents for all the other steps.
Since the overall yield for a reaction step is the product of the assay yield (the fraction of product in the reactor at quench ) and the isolation yield (the fraction of the assay material actually obtained separated pure ) expressed as a percentage, these two need to be optimized separately to raise the overall yield efficiently. Increasingly sophisticated (and more expensive) reagents, equipment and processing chemicals in the service of this first transformation usually lead to improved profits so long as they significantly improve either the assay yield or the isolation yield. Another consequence of the rapidly increasing value of the main input starting material for each subsequent step is that throughput becomes less important because the moles of starting material being processed are progressively being reduced while the capacities of the general-purpose reactors that are being used in the campaign remain the same. That is to say, for example, it might take 6 full load repeats of the first step of a sequence followed by three full-load repeats of the second step but the only one run of each of the third and fourth steps to produce a required amount of final product, because your reactor size is not big enough to do the first or second steps just once. So for the early steps in a linear process improving assay yield and/or isolation yield using inexpensive technology is key.
Steps towards the end of the linear process chain batches are fewer in number and smaller in size. High assay and isolation yields remove key, but now this is where polymer-bound reagents, solid-state synthesis, and isolation by forming and then decomposing derivatives; technologies that may be too expensive to use in the early steps of a sequence or which temporarily increase the maximum stirrable volume of a batch, can lead to lower overall costs by using even more expensive means to raise these yields.
I do not usually think of it in this way but what the calculation shows is that using an expensive reagent should be placed as late in a process as possible because there will be fewer moles of intermediate to transform the later in the process the reagent is used. Just to illustrate, consider the synthesis of a molecule consisting of two functional groups that are not interacting but sit at the two ends of an alkyl chain. Each of the two functional groups is created by a chemical transformation from a different functional group in the starting material. The reagents and conditions required for the transformations, one at each end, do not interfere with each other. They could be conducted, either the left end first then the right end, or vice-versa. If the left end transformation uses much more expensive processing than the right end, the left end transformation should be done second, because it will that way use fewer molar equivalents of the processing chemicals since the overall yield in the first transformation is unlikely to be 100%.
A decision about positioning chiral resolution in a process chain is frequently required. The rule of thumb is to perform the resolution as early as possible in the synthesis because this decreases the amount of useless mass that is being carried through the synthetic stages wasting processing chemicals. When the resolving agent is very expensive, however, a confounding factor is at play. If the resolution is delayed more matter must be carried through more steps as we have noted, but when the resolution is done later in the sequence fewer moles of resolving agent will be used. So if throughput is not a problem and the processing chemicals that will be used going from the early resolvable intermediate to the later resolvable intermediate are relatively inexpensive, it could be worth holding off the resolution.
I do not usually think of it in this way but what the calculation shows is that using an expensive reagent should be placed as late in a process as possible because there will be fewer moles of intermediate to transform the later in the process the reagent is used. Just to illustrate, consider the synthesis of a molecule consisting of two functional groups that are not interacting but sit at the two ends of an alkyl chain. Each of the two functional groups is created by a chemical transformation from a different functional group in the starting material. The reagents and conditions required for the transformations, one at each end, do not interfere with each other. They could be conducted, either the left end first then the right end, or vice-versa. If the left end transformation uses much more expensive processing than the right end, the left end transformation should be done second, because it will that way use fewer molar equivalents of the processing chemicals since the overall yield in the first transformation is unlikely to be 100%.
A decision about positioning chiral resolution in a process chain is frequently required. The rule of thumb is to perform the resolution as early as possible in the synthesis because this decreases the amount of useless mass that is being carried through the synthetic stages wasting processing chemicals. When the resolving agent is very expensive, however, a confounding factor is at play. If the resolution is delayed more matter must be carried through more steps as we have noted, but when the resolution is done later in the sequence fewer moles of resolving agent will be used. So if throughput is not a problem and the processing chemicals that will be used going from the early resolvable intermediate to the later resolvable intermediate are relatively inexpensive, it could be worth holding off the resolution.
No comments:
Post a Comment