We synthetikkers live in the gilded age of NMR. This analytical method is so fast and so rich in quantitative and structural details that we may forget what it’s like to produce materials that aren’t amenable to NMR rigged for liquid samples and H, C, F, B, Si, and P.

I’ve been busy making metal oxides and various complexes for sale that lend themselves to a very short list of analytical methods. When you make compounds for sale you have a responsibility to provide an unambiguous assay of purity for the lot.  Compounds that are poorly soluble, paramagnetic, or lack NMR active nuclei can be problematic for NMR assay in a production setting. Yeah yeah, I know- get a solid state NMR. Well, we don’t have one and it ain’t gonna happen in my lifetime. Meanwhile, I have 200 g of new product that needs to get certed and into inventory.

Lately I have been taking cues from catalog company web-sites and exploring other methodologies. Complexometric titrations for metals assay, AA, gravimetric AgX for halides, Karl Fischer for water, Loss On Drying (LOD) for volatiles (water, solvents), combustion analysis (C, H, & N), Glow Discharge MS (the Big Hammer) for refractory metal oxides, XRD for anything that could be in the xtal database, melting points, TGA, and I’m turning back to FTIR. 

I haven’t been using FTIR in a quantitative way, just looking for a “Conforms to Structure” result. But nonetheless, in the preparation of new compounds for the product list it is a life saver. I can convince myself that the desired ligands are there and use other methods to try to quantitate their wt %.

I always feel better if we can come up with 3 methods that corroborate the composition. You don’t always have to come up with methods that are on the specification either. It is reasonable to report results on a Certificate of Analysis that are “Report Only” and show general conformance rather than some percentage quantity.  Examples might be appearance, color, or even an NMR spectrum.

I have only recently begun to use XRD and am a mere novice in its intracies. I have sent solid solutions where components that I knew to be there were not detectable. I have also sent samples that came back with % compositions of several xtal phases. For characterization of production lots, it has utility in the detection of certain components. Amorphous phases and random, solid solutions are a blind spot for the method. On the other hand, there is ca one half million compounds in the database so it can detect xtal phases down to ~ 1%.

I have learned an expensive lesson in regard to ICP MS. The method is quite blind or unreliable with certain elements. Sulfur and halides in particular. A sample can be loaded with sulfur (often as sulfate) and the assay will come back with a wildly low value. An ICPMS assay of rare earth metal oxides can support a claim for 99.99 % total rare earth oxides. A GDMS of the same sample may show that it is 99.8x % in metals and even lower if you include halides, sulfur, and phosphorus. 

To be fair to purveyors of ICP MS, it is quite sensitive but standards at the lower limit of detection may not be available. Sub ppm numbers without an explanation of conditions and error are to be taken with skepticism.  Everything looks like a dogs lunch once you get down to the sub ppm level.