Figure 1: Photograph of beans moving from unroasted, to medium roast, to dark roast
Roasting coffee beans causes meaningful changes in their microstructure, such as the increase in total pore volume, porosity, and density due to the rupture of bonds in the internal structure of the coffee beans during roasting. These changes affect the stability as well the grinding and brewing performances, and ultimately the taste.
From Harvesting through Roasting
The harvesting of coffee cherries begins the process of developing coffee beans for use. After harvesting, the samples are brought down to 11% moisture and milled to remove the cherry husk from the bean. After sorting for size and quality, these green coffee beans are then exported worldwide for further processing1. While many are familiar with the flavor and aroma of roasted coffee beans, green beans are less commonly encountered in public. These beans, while similar in shape to the beans you will find on the shelf, have a vastly different flavor and texture than what would be expected from a finished coffee bean. The roasting process converting these green coffee beans to a finished product is critical for flavor development.
During roasting, coffee beans are rapidly heated in specialized roasters, which causes the loss of moisture from the beans and the development of flavors through the Maillard reaction. Once beans reach around 400F, beans begin to change in color and the fragrant oil caffeol begins to release through the process of pyrolysis1. The longer beans are left in the roaster, the more intense flavors will be developed.
Traditional examination of the physical properties of coffee beans relies on light microscopy, electron microscopy, and mercury porosimetry2. Micro-CT is highly suited for food and beverage product development and quality control due to its non-destructive quantitative view within samples at atmospheric temperature and pressure with no complicated sample preparation required. In this work, we examined replicate samples at three time points during the roasting process from a single batch of Guatemalan coffee beans. We examined the unroasted green beans, a medium roast set of beans, and a dark roast set of beans to explore the physical changes occurring during the roasting process.
X-Ray Microscopy Imaging of Coffee Beans
For our article this month, we chose to focus on imaging four samples extracted at three time points from a single lot of Guatemalan coffee during the roasting process using the SkyScan 1272 high-resolution desktop micro-CT. As can be seen from the view in Figure 1 above, obvious physical changes are observed based on the surface appearance of the beans. However, micro-CT examination allows for the qualitative and quantitative comparison of internal features and structures without compromising sample integrity.
Figure 2: Planar 2D and rendered 3D views of a green coffee bean
As shown in Figure 2, green coffee beans generally have a low porosity and a high water content. We can also see a region of lower density within the bean near one of the ends. This low density band within the bean likely represents the previous connection path between the bean and the stem from the tree on which the coffee cherry originally grew. The natural shape of the bean is a curled c-type shape with a large open void present through the central axis of the bean.
Figure 3: Planar 2D and rendered 3D views of a medium roast coffee bean
When comparing the medium roast beans to the unroasted, drastic physical changes are observed within the sample (Figure 3). Most notably, a sharp increase in porosity is observed throughout the structure and larger voids are starting to develop.
Figure 4: Planar 2D and rendered 3D views of a dark roast coffee bean
When comparing the dark roast beans to the medium roasted beans, the changes are less obvious (Figure 4). The overall porosity appears relatively similar from a qualitative view, though the cracks that began to form in the medium roast sample are now larger and more pronounced.
Figure 5: Planar 2D and rendered 3D high resolution views of a dark roast coffee bean
While the images we have examined so far were all captured at standard resolution on the SkyScan 1272, there is room for improvement in resolving the pores by switching to the highest resolution imaging mode (Figure 5). The increased resolution allows for a finer representation of the small pores that developed during the roasting process.
Figure 6: Quantitative pore maps (red) overlaid on green (top) and dark roast (bottom) imaging data
Using our CTAn software, the pores within each bean sample were measured quantitatively during a volumetric analysis and the resulting pores were mapped in red and overlaid upon the original image data (Figure 6). These views are particularly helpful in examining the drastic increase in porosity arising from the roasting process within the beans.
Relevant morphometric parameters for each sample calculated within CTAn were compiled and presented in Table 1.
|Bean surface area||570.58||72.54||6537.49||1117.90||7545.89||1961.91||mm2|
|Bean surface / volume ratio||4.08||0.40||48.80||3.95||43.66||7.61||1/mm|
|Average pore diameter||186.30||57.28||81.95||6.29||105.05||20.82||um|
|Total porosity (percent)||8.07||2.09||28.38||1.76||26.30||3.38||%|
Table 1: Morphometric parameters from the quantitative analysis of coffee beans
Visually we observe the green beans to have low porosity, so it is reasonable to believe the average of 8 percent porosity we observe for this sample is based primarily on the large open channel present within the folded structure of the bean. The medium and dark roasted beans both show an approximate four-fold increase in porosity. Interestingly, the porosity does not change with statistical significance between the medium and dark roast timepoints. As the porosity of the beans increases during the roasting process, so does the surface area to volume ratio, as the volume remains relatively unchanged while the presence of pores increases the surface area. As with porosity, no significant differences were observed between the medium and dark roast time points.
The SkyScan 1272 allowed us to non-destructively compare samples of Guatemalan coffee beans sampled through the roasting process. This examination detailed the nearly four-fold increase in porosity arising from the roasting process. This comparison highlighted several key differences internal to the sample and not visible upon external examination. We hope you found this Image of the Month informative and encourage you to subscribe to our newsletter and social media channels in preparation for the continuation of our image of the month series next month.
|Sample||Standard Resolution||Maximum Resolution|
|Pixel Size (µm)||8||4|
|Scan Time (HH:MM:SS)||01:35:37 x 5 connected scans||05:02:48|
These scans were completed on our desktop SkyScan 1272 system at the Micro Photonics Imaging Laboratory in Allentown, PA. Reconstructions were completed using NRecon and visualization of 2D and 3D results were completed using DataViewer, CTVox, and CTAn.
Coffee samples were kindly collected in collaboration with Liberty Beans Coffee Company (https://libertybeanscoffee.com/) located in Cherry Hill, NJ. Details on the batch are listed below.
Asociacion de Productores de Café Diferenciados y Especiales de Guatemala (ASPROCDEGUA)
Farm: About 394 organic-certified members of ASPROCDEGUA
Altitude: 1600-2200 masl
Region: Cuilco, Colotenango, Santa Barbara, San Sebastian, Huehuetenango, Sipacapa, San Antonio Huista, Union Cantinil, San Pedro Necta, Todos Santos, Concepcion Huista, San Marcos
Varietals: Caturra, Catuai, Pache
General Cup: Clean, sweet, smooth and citric with praline pecan flavor
Process: Fully Washed
Would you like your work to be featured in our monthly newsletter? If so, please contact us by calling Seth Hogg at 610-366-7103 or e-mailing email@example.com.
- Karen J. Cloete, Žiga Šmit, Roya Minnis-Ndimba, Primož Vavpetič, Anton du Plessis, Stephan G. le Roux, Primož Pelicon, Physico-elemental analysis of roasted organic coffee beans from Ethiopia, Colombia, Honduras, and Mexico using X-ray micro-computed tomography and external beam particle induced X-ray emission, Food Chemistry: X, Volume 2, 2019, 100032, ISSN 2590-1575, https://doi.org/10.1016/j.fochx.2019.100032.