[curiouscat]

In one of our production-scale reactions there's a side-reaction that leads to a high-BP by-product mixture (let's call it "crap"). Since this leads to a loss of selectivity ($$$) we keep track of how much crap we form in each batch. The "crap" is an uncharacterized gooey mixture of possibly tens of different high-MW, high-BP, possibly polymeric products. They stay dissolved in the liquid product.

The conventional estimation protocol is to take a fairly large quantity (say, 1000 gm) of the product (liq) & subject it to a high vacuum (10 mmHg) distillation. The residue remaining at the end ("crap") is weighed & thus we estimate our crap %. A good batch may have 2% "crap" (w/w) & a really bad batch 15%.

Question: Would there be a better, faster, possibly instrumental process to quantitate "crap"? The conventional vacuum distillation is messy, with vacuum pumps, glassware, heating mantles etc. and needs approx, an hour.


Would something like a GC with an internal standard work at estimating the non-volatiles in the sample? Or an HPLC? Would I need to select a solvent that dissolves all the "crap"?

The volatile component of the product mix is anyways analyzed via a GC for the product distribution.

Thoughts? Ideas?

[kriggy]

IMO its hard to tell if you dont know what exactly is the crap. If its high BP and MW then GC is probably not a good way to go. HPLC might work but you would need a some kind of standard to quantify the amounts also, there is probably no need to separate each single compound from the crap. Does it absorb in UV/VIS area? Maybe you can analyze it by just UV/VIS? Make a calibration curve and then just analyze the reaction mixture at wavelenght where has the crap absorption maximum. The problem with this method might be that you would need to make a calibration curve at least every day if you want to have reliable results, that means having a crap stored somewhere and of course, the composition of crap might be slightly different from batch to batch with obvious consequences

[marquis]

Try thermo gravimetric analysis (TGA).  There are a lot of instrument variations.  You should be able to find one close to what you need.  Maybe an off the shelf unit with the vendor making a few modifications?

Good luck

[curiouscat]

Thanks! A TGA sounds really promising.

Because, what we really want to measure here is "Net mass fraction of all components with a BP above x C"