You are here

Raman spectroscopy for MEA-triazine QA/QC testing

A customer approached OndaVia interested in measuring their triazine-based scavenger solutions. The OndaVia team analyzed the sample, determining that the triazine content was 58%. The customer expressed initial surprise at this result: their production approach targeted a 70% triazine solution!

We very commonly see high concentration triazine samples that report far below what is expected. In fact, I have measured multiple 70% solutions in the mid-50% range, which started me wondering: is there something unusual with high concentration samples that affects our Raman measurements? Or is there something in the production chemistry that makes producing high-percent triazine difficult?

MEA-triazine is produced by mixing stoichiometrically equal amounts of monoethanolamine and formaldehyde. Three molecules of MEA plus three molecules of formaldehyde produced one MEA-triazine molecule and three molecules of water. The molecular weights of the materials involved means the reaction should start with approximately 33% formaldehyde and 67% monoethanolamine. The reaction products are 80% by weight of MEA-triazine with 20% water. In order to produce an 80% triazine solution, producers need to start with nearly pure formaldehyde and monoethanolamine.

In this sample, the spectrum (right) looked unusual compared to a reference standard. We built our calibration curve starting from an analytical grade MEA-triazine—a 97% sample of triazine solid. Compared to the reference standards, the peaks at 870-cm-1 that correspond to bonds in monoethanolamine were high relative to the primary peak we attribute to triazine (920-cm-1). In other words, it appeared that the sample contained far more MEA relative to triazine than one would expect.

The customer produced this material by mixing approximately 26% by weight of 91% paraformaldehyde with 74% by weight of a 85% monoethanolamine solution. The resulting solution (unreacted) contains 63% monoethanolamine. If all the monoethanolamine reacted to form triazine, we would expect a solution of more than 70% triazine.

But this reaction is formaldehyde-limited: all of the formaldehyde reacts with MEA, resulting in a 58% triazine solution with 14% excess MEA. Our initial measurement that determined the triazine content at 58% was quite accurate. And the spectral regions that indicated a high MEA concentration are also correct—MEA was in excess. A little excess MEA is common to avoid having an remaining formaldehyde. For this reaction, if the conditions are adjusted to 31.4% by weight of the paraformaldehyde material then the triazine content will approach 70%.

These results could have multiple negative consequences. If you are shipping a 70% solution that is really 60%, at a minimum, your product is out-of-specification. More likely, your customers are going to find that the product works poorly and are going to switch vendors. If they use an OndaVia Analysis System to determine the triazine content, they are going to reject your shipments immediately upon delivery. The value difference for 70% versus 60% triazine is approximately $300 per 1000-L tote.

How do you avoid this error? One common method for triazine analysis is via a titration to determine a total amine value. This approach is fundamentally flawed for triazines. First, MEA-triazine rapidly decomposes in acidic solutions. Below pH 7, the decomposition is almost immediate into formaldehyde and monoethanolamine. When a titration is performed, the pH is decreased until an endpoint is reached—but this process breaks the triazine and simply measures the amount of monoethanolamine present!

I demonstrate this in the figure to the left. I started with a 60% triazine standard and added HCl to a final concentration of 50-mM. The MEA peak increases sharply while the triazine peak decreases. Adding additional HCl to 100-mM further decreases the triazine signal. The half-life of triazine at pH 9.5 is about 1000 seconds. At pH 8.5, the half-life decreases to 1 second.

The weight of three monoethanolamine molecules is 84% of the triazine molecular weight. In the above example, the user would be misled by an amine value that would read 63%, the sum of the excess MEA plus the MEA released from the triazine molecule.

Why is this important? Monoethanolamine is an H2S scavenger so any MEA present could perform the intended function. In fact, monoethanolamine scavenges H2S 3:3 while triazine scavenges 2:3. But monoethanolamine is a reversible scavenger. As it gets warm, it will release the H2S. And in an alkaline solution—like in high concentration triazine—monoethanolamine will not scavenge. You are paying for excess MEA that is either temporarily capturing H2S or is not capturing at all.

As a further example of the power of Raman spectroscopy, we took the sample with excess MEA and added extra formaldehyde. The resulting spectrum is labeled “adjusted” in the figure above. The addition of formalin (37% formaldehyde in water) to the solution cooled on ice resulted in an small increase in the triazine signal and a sharp drop in the MEA signal. The reaction chemistry indicates the triazine content should increase to 61% triazine while the excess MEA will be completely consumed. The resulting signal should be similar to a 60% triazine analytical standard. The measurement determined a triazine content of 62%, a decrease in the MEA signal, and a spectrum similar to a standard triazine solution—just as expected.

If you manufacture MEA-triazine, you need better tools for analysis. A single 80,000-L reaction vessel pays for our analysis tool. If you use MEA-triazine, you better carefully monitor what your supplier provides. A test report that claims an MEA-triazine content based on TAV is worthless. Proxy measurements like density are easily fooled by solvents or excess amines. Only a direct measurement tool like Raman spectroscopy can provide reliable triazine content values.