The most important reaction in industry is the one in which you transform chemicals into money.  It’s about adding value to feedstocks in some way.  A chemical is valued because of some property.  For instance, heptane might be valued because of it’s hydrophobicity, it’s inertness, it’s moderately high boiling point, the solubility (or lack therein) of some material in it, or all of these attributes.  

Heptane is a useful example because it is often used as a substitute for hexane. It has a higher boiling point than does hexane, which raises an interesting point.  The art of synthetic chemistry is in managing reactivity.  In R&D work, when faced with sluggish reactivity, we might be tempted to find more reactive components.  For instance, if sodium tert-butoxide isn’t basic enough, try n-BuLi.  If that isn’t basic enough, try t-BuLi. This series of bases from NaOtBu to n-BuLi to t-BuLi increases in basicity, but it also increases in cost on a $/mol basis.  The hazards also increase.  

But another way to increase reactivity is to increase the reaction temperature.  It is probably the easiest and cheapest way to do it, in fact. Of course, petroleum chemists have known this for quite some time.  Hydrocarbons that are normally inert in the ordinary range of temerpatures, say -78 C to 200 C, become reactive to HF or H2SO4 or zeolites at 300 to 400 C.

A reaction that is sluggish in refluxing hexane may perk up in refluxing heptane, xylenes, or mineral oil.  Most people seem to have an aversion to running a reaction at elevated pressure. This is unfortunate and may be due in some small way to lab culture.  If monies haven’t been provided for a Parr reactor in the past, then there is an “activation barrier” to trying reactions at elevated pressure.  Also, high pressure processing in scale-up is hampered by the requirement for bigger pots & pans with higher pressure ratings.  The practical limit for high pressure in a common metal reactor vessel might be 70 or 90 psi.  general purpose production reactors have mechanical limitations that bench chemists may not have considered.  The agitator shaft has a mechanical seal that is prone to leakage.  The pot will have numerous ports with valves that can can be weak points. General purpose reactors have heating/cooling jackets on them that can leak. All reactors have pressure relief devices called rupture disks that are set to predetermined relief pressures.  Glass lined reactors may have pressure limitations due to the brittle glass that lines the interior surface of a metal pot.

It turns out that in the chemical processing industry, high pressure capability is a capacity that relatively few company’s have.  High pressure capacity is niche work and is nice to have.  Most of us have to manipulate reactivity by other means. 

One of the banes of life for a scientist in fabulous industry is having to deal with regulatory compliance.  And, in my opinion, one of the thorniest to contend with is TSCA– Toxic Substances Control Act. Now, for those people who make the same thing day-in and day-out, TSCA is practically invisible. In this mode, your product is either on the list and therefore approved for manufacture, or management has applied for and received some exemption from the EPA.  But for those intrepid characters who are in the business of making new stuff or just lots of different stuff on a regular basis, the question of TSCA compliance is an ongoing minefield concern. 

TSCA is promulgated by the EPA.  Basically, TSCA regulates what isn’t already covered by food, drug, agrochemical, cosmetic, and nuclear material regulations.  TSCA covers chemicals and formulations used in R&D and in general manufacturing.  The TSCA inventory is maintained by Chemical Abstracts Service. With certain exceptions, what is on the TSCA list can be manufactured freely and in any quantity.  The TSCA inventory has a group of listings for public viewing and a confidential group of listings. The balance of chemicals in the universe are those that are not on the TSCA inventory. These are problematic for manufacturers.

One important complication for chemicals that are on the inventory is the SNUR– Significant New Use Rule.  Even though a chemical may be on the list, certain uses may be restricted. So if you plan on manufacturing a product that is on the TSCA inventory, you really should look for SNUR’s.

A chemical product that is not on the public or confidential TSCA inventory cannot be sold for commercial use in the USA. Perversely, you can manufacture for export only.  Products that are not on the inventory can be sold at any scale for R&D use only, however. 

Let’s say that something is on the confidential inventory.  Unless you know this, you would conclude that a chemical is not on the inventory. Well, guess what? You can’t just call the EPA to find out if a chemical is on the confidential inventory. You have to submit an application as if you were going to file for real. If it is confidential, then the EPA will notify you on the normal application timeline.

In order to manufacture something for commercial use that is not on the TSCA inventory, you either have to get it listed by filing a PMN (Premanufacturing Notice) or you file for an LVE (Low Volume Exemption).  Also, any raw materials and isolated intermediates in the process have to be listed. If not, you have to file for those as well. So, initiating the manufacture of new chemicals is complicated by the requirement of performing numerous filings.

LVE’s have a 30 day evaluation period. If you screw up the application, you have to resubmit it and the clock restarts at zero again. The EPA folks look at the chemical process and all of the chemicals and evaluate the potential for harmful exposure to people and the environment.  They use numerous modeling programs to estimate toxicity and potential environmental insult.

In parts of the physical world like the lab or a production area, it is possible to have a physical disaster like a spill, fire, or explosion. In the regulatory world, you have administrative disasters.  And these administrative- or compliance- calamities can be just as costly and career threatening as an actual disaster in the plant. Fortunately, in an administrative disaster the body parts lying around are just metaphors.

[Note: I am not a regulatory specialist.  I acknowledge that I am a mere laboratory wretch and therefore deeply marbled with imperfections and inhomogeneities.  As god dog is my witness, I am prostate prostrate in supplication before those with superior understanding of this topic. I welcome- nay, beg- corrections, comments, and lashings from those with superluminal understanding of this most sacred codex.]

Process development is one of the jobs I do.  Take an existing process and find ways to make a compound faster, better, and cheaper. The matter of condensing multiple steps into fewer steps is called “telescoping”.  One of the most desired outcomes of process development is to find a way to execute a reaction with fewer labor hours and maybe even higher yield.

My comments are in the context of specialty chemical manufacture. In this domain of industrial activity, it is not unusual for a specialty chemical to be campaigned for production on demand (POD).  That is to say, instead of building an inventory and letting it sit for some time period, it might be more desirable to make material when an order comes in.  This is a valid strategy for products that have a poor shelf life or for compounds whose demand is sporadic. 

But, there are economic arguments for and against POD. On the negative side, the lack of inventory can cause customers to go elsewhere for orders that have to ship immediately. Not every customer can wait until the next hole in the production schedule for a shipment.  Also, unless one has confidence in projected demand patterns and has made a successful business case to management for excess production, POD esentially dooms one to a perpetual cycle of smaller scale production runs with the concommittant smaller economies of scale. 

On the vendor side, getting an accurate picture of demand can be very difficult. The reason is that the manufacturer of a specialty chemical is not often connected to the “final” end use of the product, so timely and accurate market data might be considered proprietary information that the direct customer is not willing to share.

On the positive side, POD assures that the dollars invested in inventory are kept to a minimum.  Management has to be watchful of inventory levels.  It is possible to accumulate large dollar investments in inventory.  Having a million dollars of slow moving inventory is equivalent to having a milllion dollars of working capital sitting on pallets that you can’t use for other applications.  But for POD to work well, the plant must have some excess capacity. And one of the reasons we have sales people is to fill up that excess capacity. So, POD may not be a strategy that works all of the time.

A fair question might be the following- why should an opportunity for process development even exist on an current process? In other words, why wasn’t it done to begin with?  Fair question.  There are a few answers. 1) In the race to get a product to market on schedule, there usually isn’t time to explore all of parameter space. Often, to meet obligations that our friends in the sales force have made, the development timeline can accomodate only a certain amount of R&D activity before something has to go to the pilot plant for scaleup.  2) The reality is that any given R&D group is likely to chose certain favored synthetic approaches from their particular tool bag.  The solution to a scaleup problem is not automatically a global solution to the problem.  A great many syntheses have alternative approaches that may find favor in a particular group. Especially if the literature search was truncated in some way.

In science it is always good to reevaluate your fundamental assumptions, and in manufacturing it is the same.  No process is perfect and every one can be tweaked in some way to optimize the economics.  Some companies have special staff to do just this thing.

Many of us have joked that it is possible to make anything in a single step if only you had the right starting materials.  True enough.  But manufacturing as a profit generating activity requires that value be added to raw materials to produce profitable finished goods. This forces manufacturers to vertically integrate a process to some extent so as to allow for sufficient added value in the finished good. In other words, the more art you can apply to the manufacture of a product, the greater the chance that several of the steps may be highly profitable. 

One way to think about high $ per kg boutique products is as follows.  A product that requires considerable art (skill) is likely to be one that has a mfg cost driven by labor costs.  Products whose costs are driven by labor are products whose costs can be driven down more readily than those driven by raw material costs. A labor intensive product stands a better chance of cost improvements than does a raw material cost intensive product.  The reason? Improved throughput in units per hour already cuts unit labor costs.  You get the picture.