Identifying the Geographic Origin of Roasted Coffee

In a recent study, a team of researchers from the University of Bergen have developed a spectroscopy-based method that can identify the geographic origin of roasted coffee in minutes without destroying the sample.¹ Because destructive sampling was the primary method of conducting this type of analysis in the past, the study adds to the building body of work that is exploring the use of new, nondestructive technologies that can analyze food products such as coffee.

Sleshi Fentie Tadesse will present the findings of this study during a poster session at McCormick Place Convention Center, In Chicago, Illinois, on Wednesday, August 26, from 7 to 9 p.m. EDT. Tadesse will show how the study combined two infrared (IR) spectroscopy techniques (in this case, attenuated total reflectance Fourier-transform infrared (ATR-FTIR) and near-infrared (NIR) spectroscopy) with statistical modeling to classify roasted, ground coffee by origin, variety, and roast level.¹

Why is knowing the geographical origin of food products important?

Understanding the geographical origin of coffee is important because it helps the market accurately price the coffee and determine consumer trust. Currently, existing verification methods, which include chromatography and sensory panels, are accurate but slow, require destroying the sample, and depend on subjective human judgment.¹ Because of these three factors, using chromatography and sensory panels are not practical methods for routine quality control.¹

What did the researchers do in their study?

The researchers used coffee samples sourced from various countries known for their coffee quality. The countries selected were Ethiopia, Brazil, Colombia, Costa Rica, Kenya, Mexico, Peru, and Rwanda. Once the samples were obtained, the team scanned samples across the mid-infrared (MIR) and NIR ranges before processing the resulting spectral data to remove noise and scattering artifacts.¹ Statistical modeling revealed that samples clustered distinctly by origin and variety, indicating that chemical differences between growing regions leave a measurable signature.¹

Which classification approach performed better in this study?

As part of the poster presentation, Tadesse will explain that the study tested two classification approaches, soft independent modeling of class analogy (SIMCA) and partial least squares discriminant analysis (PLS-DA). SIMCA correctly distinguished coffee origins with 100% specificity and better than 92% accuracy, though it performed less reliably when sorting samples by roast level.¹ Meanwhile, PLS-DA achieved perfect classification across all metrics tested, with strong statistical confidence.¹ The ATR-FTIR data outperformed NIR data in predictive accuracy, which is a difference the researchers attribute to ATR-FTIR's greater sensitivity to specific molecular structures.

What are the implications of this study?

The coffee industry is currently lacking a quality-control tool that can be deployed at scale. Because the method Tadesse will present is non-destructive, it could also be integrated into existing inspection workflows without disrupting supply chains.¹ As a result, the combined technique can be a candidate for routine authentication and quality assurance applications going forward, pending further validation on roasting-degree discrimination.¹

What should attendees know about the upcoming ACS 2026 Fall Meeting?

The ACS 2026 Fall Meeting will be taking place from August 23rd to the 27th in Chicago, Illinois. It is expected to draw over 10,000 attendees for technical sessions, professional development courses, and an exposition.² The meeting's theme is "Chemists Making History," marking ACS's 150th anniversary under the banner #ACS150. The technical program is expected to be extensive, with approximately 5,255 in-person oral presentations and 4,206 poster presentations set to take place during the conference.² Meanwhile, there will also be a special anniversary symposium featuring Nobel laureates Carolyn Bertozzi, Stefan Hell, Jack W. Szostak, and Stanley Whittingham.²

References

1. Tadesse, S. F. Characterization and Classification of Roasted Ground Coffee Using ATR-FTIR and NIR Spectroscopy Combined with Multivariate Analysis. Presented at the American Chemical Society Fall 2026 Meeting, Chicago Illinois, August 26, 2026. Available at: https://acs.digitellinc.com/live/37/page/1374?search=Sleshi%20Fentie%20Tadesse&window-sxlznr
2. Wetzel, W. Why Spectroscopists Should Attend the ACS Fall 2026 Conference. Spectroscopy Online, 2026. https://www.spectroscopyonline.com/view/why-spectroscopists-should-attend-the-acs-fall-2026-conference (accessed July 23, 2026).