From bean to bar: the detectives break down quality control of chocolate

Chapter 4
📁  Case overview: The detectives have been working so hard lately, Shallot Holmes was happy to give them a sweet break. In this treat of a post, the detectives solve a case on how to speed up quality control of chocolate, specifically through NIR analysis of fat and moisture content of ingredients, intermediate and final products.

Shallot Holmes walks happily into the office. He greets the other four detectives. The next file case is in his hands. Hurriedly, the private investigators drop everything they are doing and head over to the roundtable eager to hear about their next challenge.

Once seated comfortably, Shallot Holmes places the file on the table and addresses his colleagues. He happily explains their next client comes right on time following their excellent NIR workshop. Then he offers each of the other four detectives a chocolate bar. They exchange surprised glances, but unwrap the chocolates and dig in.

As they enjoy their treats, Shallot Holmes explains that their next customer has come to them for some valuable advice. They are chocolate manufacturers who are looking for faster methods of performing quality control of their products that can match the data of established traditional methods.

Lieutenant Cornlumbo clears his throat and offers to take the lead in this case. The others look somewhat taken aback, but Lieutenant Cornlumbo laughs that just because he is a corn does not mean he does not have as much of a sweet tooth as Miss Mapple.

Lieutenant Cornlumbo starts by explaining that chocolate manufacturers typically use analytical methods to measure quality control parameters such as moisture, fat, protein and sugar content in their incoming, in-process and finished products. These parameters affect taste, texture, shelf-life and cost of chocolate treats.
He then states that:

The fastest and least destructive method to control products during chocolate production would be NIR analysis.

Lieutenant Cornlumbo walks over to the whiteboard and proceeds to outline various quality control steps during the production of chocolate.

Cocoa beans

As the main ingredient in chocolate, the quality of cocoa beans is largely governed by the dynamic environmental conditions of the region where they were grown. Typically, manufacturers quantify the fat content in the beans and further intermediate products to ensure consistency in the final products. Technicians also analyze moisture content to monitor the roasting process.

Miss Mapple jumps in to explain that the reference method for measurement of fat content in chocolate is the Weibull-Stoldt method. This method involves traditional acid hydrolysis, followed by Soxhlet extraction in ether. For moisture content measurement, adds Miss Mapple, the primary reference method is Karl Fischer titration. Lieutenant Cornlumbo agrees but points out that both methods could be quite cumbersome, as they involve sample preparation, chemical reagents, skilled technicians, and extended analysis times.

Instead, he again suggests using NIR analysis. For this method, a technician with minimal training would have to perform four super simple, super speedy steps:

1. Pour beans in a sample cup.
2. Place the cup on NIR system.
3. Press a button.
4. Simultaneously measure both fat and moisture content in less than 30 seconds.

With the advantage of being non-destructive and incredibly quick, the NIR method can be used to make decisions regarding cocoa bean processing, including if roasting is complete.

Here, Nancy Beef decides to jump in and refresh the detectives’ minds by quickly explaining how the NIR method works. The measurements are based on the interaction of light with the sample, in this case cocoa beans or chocolate. The carbon-hydrogen functional groups are representative of the sample’s fat content, whereas hydrogen-oxygen functional groups are representative of the sample’s water content. These bonds are detected by NIR spectroscopy. By using a calibration model, the sample spectra can be related back to their composition of fat and moisture. These calibration models are based on samples of known compositions, where primary reference methods, including Weibull-Stoldt and Karl Fischer are used to generate and validate the relationship between the spectra and the parameter of interest.

Intermediates: cocoa mass, cocoa butter and cocoa powder

Lieutenant Cornlumbo moves on to the next step in chocolate production, the intermediate products. He is happy that the client can also use NIR systems for quality control in intermediate products. For example, the following analysis is possible with NIR spectroscopy:

• Moisture and fat in nibs and cocoa mass
• Free fatty acids and iodine in cocoa butter
• Moisture and fat in cocoa powder

These measurements can then be used to optimize chocolate production efficiency and profitability through:

• Maximizing the cocoa butter yield from cocoa liquor,
• Ensuring the standard of identity specifications are satisfied without excess addition of expensive ingredients, such as cocoa butter
• Determining the fat content on which the products should be sold

Final confectionary products

Lieutenant Cornlumbo comes to the logical conclusion of discussing how NIR spectroscopy can be used for quality control of final products, including milk chocolate and dark chocolate. He states that the method can be readily used to measure parameters, such as moisture, fat, lactose, sucrose and theobromine. Again, with calibration models, NIR can be used to quickly perform quality control on these chocolate treats before they heat the shelves.

Shallot Holmes is satisfied with this information and decides to wrap it up before the team starts getting cavities. He is certain that the client will be interested in trying out NIR analysis and might even come with future cases for them to test out the method with experimental data.

But he turns out to be wrong about one thing. As they head out for an afternoon break, instead of going for coffee or tea, most of the private investigators order hot chocolate instead! The team’s appetite for chocolate, just like for food analytics cases, seems hard to satisfy indeed 🍫🍫🍫.