A Measurement Mystery: Perfecting Sample Preparation for Precise Analysis

Chapter 60
📂 Case overview: The detectives help a client who is using the latest analysis equipment but getting inconsistent results and inexplicably large standard deviations. But as with every good mystery, things aren’t as simple as they first appear.

The detectives have reconvened and are enjoying the cherry clafoutis Nancy Beef made with the flour given to her from the case she solved on Free Fatty Acid Determination. “This is divine, Nancy,” says Miss Mapple. “C’est délicieux!” declares Eggcule Poirot. Then, seemingly out of nowhere, Cornlumbo shouts at Eggcule, “You’ve got more cherries in your slice; that’s unfair!” Miss Mapple looks horrified and tells Cornlumbo, “Well, you’re surely going to have more than one slice, so I’m sure you’ll get your fair share of cherries.”

If the methods of analysis are fine, why don’t the results match reference values?

Shallot Holmes enters the office, and his eyes immediately light up as he sees what Nancy has bought for everyone. “Oh, thank you, Nancy; you see, sometimes it really pays off solving our client's cases - and talking about cases; I have a new mystery for you all to solve… A client is having issues with inconsistent results. To get any degree of accuracy, they must perform numerous analyses, wasting time and resources, which the client is very much against. They even invested in NIR to speed up their processes, but again, the results are inconsistent. It really is a mystery because, as far as I can tell, the client is using the latest equipment, and I could not fault their methods of analysis when I spoke with them on the phone. It’s going to take more clues to get to the bottom of this case, so who wants to visit the processing facility and laboratories to see if they can solve the measurement mystery?”

Why are consistent results essential for NIR calibration?

Feeling a little embarrassed about his outburst regarding the lack of cherries in his clafoutis, Cornlumbo says that he would like to take a shot at solving the case.

“The client’s results are inconsistent; they are getting high standard deviations. To get any degree of accuracy, they have to perform numerous analyses to gain results that correspond to reference values.”

Cornlumbo, with an extra slice of clafoutis in his lunchbox, heads to see the client and is immediately impressed when he arrives at the food processing facility. The client tells Cornlumbo about the Fat and Protein analysis they perform for declaration purposes. Having previously sent their samples for 3^rd party analysis, the client wanted to bring everything in-house to streamline their process. They had invested in the latest equipment for fat extraction and analysis. The client's plan was to perform an analysis of their samples according to reference methods and use the results of the analysis to calibrate their NIR equipment; once calibrated, any subsequent analysis could be rapidly performed on-line, ensuring the full optimization of their production facility.

What can cause high standard deviations?

The client's frustrations began when the results from the first reference materials tested with the new extraction equipment didn’t match reference values. They tried another analysis, and the results were closer but still not accurate enough. It took eight results to bring the mean value within the target amount. The inconsistencies infuriated the client, who knew that the results needed to be as accurate as possible if they were to be used to calibrate the NIR equipment. “So, Mr Cornlumbo, we have invested all this money, and our process is now less accurate and taking even longer than before,” says the client. Cornlumbo is a little nervous as it appears that the client's equipment and setup seem impeccable. To buy some time, he asks if he can look around the laboratory and facility for clues.

On looking around the laboratory, Cornlumbo cannot find anything out of the ordinary. The client is performing the fat analysis according to the Weibull-Stoldt method, which involves a hydrolysis step to release encased fats. Cornlumbo could not think of a better method to ensure the accuracy of results. Unable to find any clues, Cornlumbo made his way to the processing facility to see the in-line NIR setup. Again, everything was impeccable, with sensors in the correct places and set up perfectly to perform an accurate analysis. Cornlumbo starts to panic; he has never failed to solve a case before and doesn’t know how he could face the other detectives if he failed. As he nervously paces up and down, trying to solve the case, his belly begins to rumble, as it often does when he gets nervous, or tired, or stressed, or relaxed, or happy, or sad, or well… let’s just say he is hungry again! He then remembers the cherry clafoutis in his lunchbox.

“The client is performing the fat analysis according to the Weibull-Stoldt method, which involves a hydrolysis step to release encased fats.”

Cornlumbo takes the clafoutis from its wrapper and begins to guzzle it down, and as he does so, he realizes that this slice has loads of cherries in it. ‘This is much better than the first slice I had,’ he thinks to himself when suddenly, it occurred to him. The cherry distribution was inhomogeneous; if he had only been given the first slice, he would have thought the clafoutis had a very meager amount of cherries. But had he only eaten the second slice, he would have thought it was the world's best super-duper-cherry-packed clafoutis ever made. That’s it! Cornlumbo thought. There isn’t a problem with the analysis but with the sample preparation. Cornlumbo rushes back to the laboratory and asks to see how the samples are prepared.

Why is sample preparation so important?

In the corner, a lab technician is blending samples in a mixer. The mixer is whizzing and whirring, and you can feel the heat coming from it. “Do you always blend samples for that long?” asks Cornlumbo. The technician tells him they never used to blend for so long, but they realized the particle distribution was uneven. It was like Cornlumbo had imagined, but the technician had corrected the issue by blending for longer. So, what was now causing the issue? Cornlumbo took a look at the sample after the vigorous blending and could tell immediately that they had been damaged due to heat and decomposition caused by the long mixing times. Cornlumbo took a look at the mixer, and it was by no means up to the standards of the other equipment in the laboratory; it was clearly dated and was not performing properly. He suggested investing in a high-quality mixer that would achieve consistent particle sizes.

“Small particle size does not necessarily correspond to well-homogenized samples. Well-homogenized samples are obtained when samples are treated with more mixing intervals and less mixing time per interval”

Cornlumbo reports his findings to the client and explains that the power of their new analytical methods could never be fully appreciated if the preparation of the sample is not sufficient. He explains the relevance of mixing, grinding, and homogenization processes and their influence on the accuracy of the sample analysis. “It’s all about the aliquot size,” says Cornlumbo. The client looks confused, so Cornlumbo continues, “In the context of chemical analysis, the aliquot refers to a sub-portion or fraction taken from a larger sample solution. It is often only a few grams or milligrams of homogenized sample; therefore, if the sample is not sufficiently homogenized, the aliquot will not be representative of the whole cake… I mean, sample,” says Cornlumbo. He explains that small particle size does not necessarily correspond to well-homogenized samples. Well-homogenized samples are obtained when samples are treated with more mixing intervals and less mixing time per interval. “For accurate results with as few measurements as possible, proper sample preparation is essential,” says Cornlumbo. The client is so happy and grateful that they have removed the bottleneck in their process, and Cornlumbo immediately rushes back to the office to tell the detectives the good news.

“The aliquot refers to a sub-portion or fraction taken from a larger sample solution. It is often only a few grams or milligrams; therefore, if the sample is not sufficiently homogenized, the aliquot will not be representative of the whole sample”

On entering the office, Cornlumbo immediately runs up to Nancy and shouts, “Nancy! Thank you so much! You’re the best, and your cherry clafoutis truly saved the day”. Miss Mapple looks slightly confused. “Well, I can see that you finally managed to find your manners while you’ve been gone; well done, Cornlumbo”. Everyone begins to congratulate Cornlumbo, who always seems to get hungry when being congratulated, so he looks around the room to see if there is any cake left.