STiR coffee and tea magazine

38 STiR coffee and tea | 2023 Issue 5 (October / November) Measuring the roast's sucrose content, for example, predicts the process since the sugar level declines as the roast advances. Color too is indicative. Spectroscopy based methods are somewhat less useful in predicting when the roast will start to crack and finish cracking, however. Sensory attributes Sensory analysis of specialty coffee is done by trained personnel under the strict pro- tocols of a cupping. The cuppers score and classify a coffee based on its physical attributes and the quality of the beverage. The whole process is time consuming and often lacks reproducibility, Belchior said. Different water or different equipment can affect results. Spectroscopic methods can be useful to confirm the coffee classification, however, when inconsistencies occur. One method predicts the sensory characteristics of coffee samples by comparing a sample's spectroscopic profile to results evaluated using the SCA protocol. Similarly, cor- relation with known profiles can be used to identify different species, varieties, produc- tion regions, drying conditions, post-harvest processes, storage time, roasts, and blends. In one study, researchers analyzed 24 samples of ground and roasted coffee in capsules. They used spectroscopy to create models to predict aroma and flavor based on analysis of carbohydrates, proteins, and carboxylic acid. Acidity was predicted by measuring carboxylic and well as chlorogenic acid. Results were compared to assess- ment of sensory characteristics by a panel of coffee tasters. Defects and adulterants Spectroscopic methods can be applied to detect defective beans as well as adulter- ants such as corn, barley, coffee husks, wood, rice, and others. "In the case of adulterants, and in the case of the pres- ence of other things that are not coffee, we need to tell people that they are buying something that is not coffee, that is coffee with something else," Belchior said. Defective beans are those that are sour, broken, black, moldy, immature, or damaged by insects. All of these de- fects reduce the quality of the beverage. But once the coffee beans are roasted and ground, it's not possible to detect the presence of defective beans without spectroscopy-based methods. Identifying the geographic origin of a sample is possible with spectroscopy. This can be done to detect fraud, as when a seller claims that a coffee originates from a certain region, but actually does not. Spectroscopy also enables scientists to discriminate between arabica, liberica, and robusta beans, as well as hybrids. Some techniques can detect the mois- ture content of ground and roasted cof- fee beans or unroasted coffee. Others can predict the caffeine content in green coffee beans. Limitations Spectroscopy-based methods must be conducted by skilled professionals who understand the techniques and the ana- lytic system. Some techniques can score coffees as precisely as Q graders do. Spectroscopy does not quantify the chemical composition. Instead, it makes a prediction based on differences be- tween samples based on the presence of certain groups of compounds. So, it is not as accurate as chromatography. "But we can detect, confirm, and differentiate coffee by a certain chemical composition. And the chemical composition is impor- tant for us to understand and to know, because it differentiates the quality of the beverage as well," Belchior said. On the way Belchior is certain that analysis of coffees using spectroscopy will become part of regular practice in assessment of quality, authenticity, origin, and roast profile in the near future. "With the machinery, the samples, the results, the chemo-metrics, and the confirmation of different infor- mation that is available in the coffee in- dustry, we are able to receive more reli- able results." Spectroscopic analysis can detect mold and adulterants like grain, wood, and coffee husks even after the beans are ground and blended.

• Cover