PNNL

An innovative approach to using Raman spectrometry to address Hanford Site nuclear waste testing challenges achieved cover status in two major scientific journals.

Pacific Northwest National Laboratory (PNNL) researchers are exploring methods to safely process and characterize nuclear waste to support Hanford Site operations. One of the biggest Superfund cleanup sites in the United States, Hanford’s 177 underground tanks contain approximately 56 million gallons of mixed radioactive and chemical waste.

Each tank’s waste varies in chemical complexity. Workers must be able to understand the chemical makeup quickly and safely as part of the vitrification process, where the waste is mixed with glass-forming chemicals and melted into large glass logs for safe long-term storage. The accurate analysis of the waste composition is vital to adding the correct glass-forming chemicals, making sure the glass is processable and meets durability requirements. Workers currently use the traditional “grab sample” collection and analysis to analyze the tank waste chemistry—a process that is time-consuming and expensive.

Raman to the Rescue

A PNNL-led team of chemists and engineers worked in the Radiochemical Processing Laboratory—a Hazard Category II non-reactor nuclear research facility—to create a new technology and unique approach. First, the Raman spectrometer-turbidity probe: Imagine a high-powered laser being submerged into different solutions ranging in turbidity—in simple words, cloudiness—and being able to record both the solution’s chemical makeup and the solution’s cloudiness. The probe gives the “fingerprints” of the chemical species, and the algorithm translates those fingerprints to measured concentrations. Second, the team developed an approach of leveraging multiple Raman wavelengths to gain robust and reliable measurements of key chemicals in the solutions.

The Raman probe and approach “is absolutely game-changing because you can understand the chemistry in real time without interrupting the process by collecting the sample, sending it off to an analytical laboratory, and then waiting for analysis,” PNNL analytical chemist Amanda Lines said. “You’re cutting your response times from potentially days down to seconds. You’re cutting your costs tremendously, and you’re making yourself a lot more agile in terms of decision-making in response to process conditions.”

This work was described in two journal cover articles published in spring 2022.

Combined Raman and Turbidity Probe

In the Analytical Chemistry article, “Combined Raman and Turbidity Probe for Real-Time Analysis of Variable Turbidity Streams,” Lines and Samuel Bryan, PNNL Lab Fellow and chemist, partnered with Job Bellow and Christina Gasbarro from Spectra Solutions to create the Raman spectrometer and turbidity probe. This project was sponsored by the Department of Energy (DOE) Office of Science, through the Small Business Innovative Research Program.

Research showed a combined spectrometer-turbidity probe could analyze different chemical solutions in real time and in situ—also known as on site. The Raman spectroscopy identified the fingerprints of the chemicals present, while the turbidimetry piece of the probe measured the cloudiness of the solution. Combining these two data streams allows for highly accurate and robust chemical composition analysis. 

The second piece of the process is building algorithms, known as chemometric modeling. “It is the science of understanding chemical information from data analytics,” Lines said. “We build algorithms that take the really complex data from those sensors and then output a meaningful stream of comprehensive information.” These outputs provide easily understood chemical information to operators looking to control the process as it runs. 

Leveraging Multiple Raman Systems

The ACS ES&T Water article, “Leveraging Multiple Raman Excitation Wavelength Systems for Process Monitoring of Nuclear Waste Streams,” highlighted work by the PNNL team of Heather Felmy, chemist; Hope Lackey, a PhD intern in analytical chemistry from Washington State University; Adan Schafer Medina, chemical engineer; Michael J. Minette, project manager; Bryan; and Lines. 

Lines explained that using multiple Raman spectrometers improves the accuracy of the characterization of chemical composition because different wavelengths provide different data. For example, longer wavelengths—such as red—reduce interference in a signal but lose sensitivity for certain analytes. However, when you use the shorter end of the spectrum—such as blue—there is higher sensitivity to lower-concentration chemicals but also more background interferences.

“It’s a balancing game that you have to play,” said Lines. “Combining three Raman spectrometers lets us interrogate the system from multiple angles, generating complementary data sets, so we can then sift through it with those advanced algorithms and more accurately provide a comprehensive analysis of the waste sample process as a whole.” 

The testing to improve the Raman detection sensitivity was sponsored by the DOE Office of Environmental Management, through the Office of River Protection. These tools can ultimately enable more efficient and informed processing and immobilization of legacy wastes.