Carbohydrate, Protein,
Photosynthetic Pigments, & Caffeine Assays Show Variation: Green vs Roasted
Coffea arabica
The
Cells: Chris Justema, Abbie Mincks, Ashley Young, and Matt Lincoln
LBS 145: Cell and
Molecular Biology
Section M1
James Hardie and
Rebecca DeGraaf
Abstract
Final Draft Responsibility: Matt Lincoln
Are
there differences between un-roasted and roasted coffee and how much caffeine
is lost when coffee beans are decaffeinated?
To find out we tested un-roasted and medium roasted caffeinated coffee
beans, and unroasted and medium roasted decaffeinated coffee beans for
carbohydrate levels, protein levels, and the presence of caffeine. All solutions in Benedict’s test, a test
which identifies reducing sugars, failed to form a red precipitate to indicate
the presence of simple sugars. Selivanoff’s test formed a red solution in
unroasted coffee bean solutions after 45 seconds indicating ketoses and after
two minutes for roasted solutions indicating aldoses. Barfoed’s test showed the
10% decaffeinated and caffeinated medium roasted solutions tested positive for
monosaccharides. All other solutions tested negative for monosaccharide
content. Bial’s test showed all our solutions contained furanose rings. Our
solutions all turned yellow/green or green/olive in color indicating
pentose-furanose rings. No solutions turned a bluish-black during the iodine
test indicating no presence of starch.
For the
Discussion
Final Draft Responsibility: Chris Justema
The
goal of this study on coffee beans was first to identify any changes that occur
in the macromolecules and photosynthetic pigments of the coffee bean after it
has been changed from a green unroasted bean to a medium roasted coffee
bean. The other main objective of this
study was to determine the difference in caffeine levels between caffeinated
and decaffeinated medium roasted and unroasted coffee beans. Based on previous research, it was predicted
that carbohydrates and proteins are broken down by the roasting process
(Parliament, 2000). In addition, it was
logical to predict that photosynthetic pigments will be broken down due to the
heat of roasting, because other components of the coffee beans such as
carbohydrates and proteins are predicted to break down as stated above. Finally for the independent part of this
study, it can be expected that the caffeinated coffee beans will have more
caffeine than the decaffeinated coffee beans.
It is the intent of this part of the experiment to investigate the
difference between caffeine levels in the caffeinated and decaffeinated coffee
beans. Overall, it was hypothesized
that the caffeinated and decaffeinated medium roasted coffee beans would show
through carbohydrate, protein, and pigment tests that they have been chemically
broken down by the roasting process. It
can also be hypothesized that the caffeinated coffee bean solutions will
contain more caffeine than the decaffeinated coffee bean solutions.
Carbohydrates were studied in each type of coffee
bean using Benedict’s, Barfoed’s, Selivanoff’s, Bial’s, and Iodine tests. Before Barfoed’s test was performed it was
predicted that both the caffeinated and decaffeinated medium roasted bean and
the green unroasted coffee bean would test positive for monosaccharides. This is because about 30% of all
polysaccharides hydrolyze to monosaccharides in medium roasted coffee beans
(Trugo, 1985). Our experiment results
differed from our predicted results in that only the 10% solutions of
decaffeinated and caffeinated medium roasted coffee beans produced a red
precipitate indicating the presence of monosaccharides. The 5% solutions of decaffeinated and
caffeinated medium roasted coffee beans produced no change, which indicates
disaccharides and polysaccharides.
There could be many reasons why no monosaccharides were present in the
5% coffee solutions. The main reason is
that the 5% coffee solutions could have been too dilute, therefore the
concentration of monosaccharides would have been much lower than pure coffee
and harder to detect. In this
experiment highly diluted concentrations had to be used in order to be able to
distinguish a color change in solution or a colored precipitate. If higher concentrations were used then the
medium roasted coffee solutions were far too dark to observe these changes.
The 5% and 10%
solutions of decaffeinated and caffeinated unroasted (green) coffee beans did
not produce a red precipitate in Barfoed’s solution, which would indicate
disaccharides and polysaccharides. They
instead formed a green precipitate that would indicate an unknown substance, or
this unknown precipitate could have formed due to the fact that the unroasted
coffee bean solutions are green. In
addition, an unknown side reaction of Barfoed’s solution with other chemicals
that make up the coffee beans could have formed the green precipitate. It was first thought that the solution was
green because the unroasted coffee bean solution was green, but to be sure, the
solution was discarded and it was observed that the precipitate itself was
green. With regard to the hypothesis
for carbohydrates, it was found that the 10% medium roasted coffee showed
monosaccharides, and therefore some polysaccharides were broken down by
roasting which agrees with the original hypothesis. The positive control was glucose, which formed an orange/red
precipitate, and the negative control was distilled water and it produced no
change.
Benedict’s test is
used to test for reducing sugars such as glucose and fructose. Some of the main simple sugars that are lost
during roasting include sucrose, hexoses, and pentoses (Parliment, 2000). Sucrose is the main free sugar found in
unroasted coffee beans (Trugo, 1985).
Even though sucrose is lost during the roasting process, aldehyde and
ketone groups are reportedly present in greater concentration in the coffee
bean (Dart and Nursten, 1985). This is
most likely because sucrose breaks down into its monosaccharide constituents of
glucose and fructose, which are aldehydes and ketones. Therefore, it was predicted that Benedict’s
test would show a positive result for the caffeinated and decaffeinated medium
roasted coffee beans, while the unroasted coffee beans would show a negative
result because they are mainly composed of sucrose which is not a reducing
sugar (Trugo, 1985). In direct
opposition to the research that was studied, our results of Benedicts test
showed that the 5% and 10% solutions of decaffeinated and caffeinated medium
roasted coffee beans and unroasted coffee beans showed no positive result for
the test. This would mean that none of
the solutions contained free aldehyde or ketone groups.
These results for
Benedict’s test may have occurred because the percentage of coffee compared to
distilled water was so small that Benedicts test was not able to pick up the
traces of free aldehyde and ketone groups.
In addition, even though research shows that aldehyde and ketone
concentrations should increase, these might not be free aldehydes and ketones
(Dart and Nursten, 1985). Therefore,
Benedict’s test would also show a negative result. A red precipitate did form with our positive control, glucose. The negative control, distilled water,
produced no change. Therefore the
Benedict’s solution was working correctly.
Selivanoff’s test allowed us to determine whether
ketoses or aldoses are predominant in each of the four types of coffee
beans. It was predicted that either
aldoses or ketoses would test positive for the 5% and 10% medium roasted coffee
beans, because aldehyde and ketone concentrations increase with roasting (Dart
and Nursten, 1985). The coffee solutions used in this experiment are not pure
solutions of aldehydes or ketones. Therefore Selivanoff’s test should detect the
one of higher concentration because there are many aldehydes and ketones mixed
together. Before doing the experiment
it cannot be determined which will be higher in concentration, because every
coffee bean is different.
The results yielded from Selivanoff’s showed a
negative presence of ketose and aldose in the 5% and 10% solutions of
caffeinated and decaffeinated medium roasted coffee beans because the solution
did not turn red even after 2 minutes. This indicated that aldose was present. The 5% and 10% solutions of caffeinated and
decaffeinated green coffee beans turned red within 45 seconds, which showed a
positive presence for monosaccharide ketoses.
Fructose, a positive control, turned red in just under one minute
showing a positive presence of monosaccharide ketoses. Glucose, another positive control, turned
red in just over two minutes showing a presence of aldoses in the
solution. Distilled water was used as
the negative control and did not change color.
Selivanoff’s test
refuted the hypothesis that the breakdown of polysaccharides to monosaccharides
should cause the medium roasted coffee solutions to test positive for ketoses
or aldoses. There are small amounts of
simple reducing sugars present along with many polysaccharides in green coffee
beans (Trugo, 1985). The green coffee
beans do not have so many types of monosaccharides so maybe it was easier for
Selivanoff’s test to pick out the monosaccharide ketoses amongst all the
polysaccharides that it should not detect.
In the medium roasted coffee solutions there should be a mixture of many
types monosaccharide ketoses and aldoses so Selivanoff’s test did not test
positive for either one, because the test is best utilized when there is a pure
solution of only ketoses or only aldoses.
Bial’s test was
performed to test for furanose, five membered, rings. In unroasted coffee beans the furan concentration in is quite
small, but becomes larger as the coffee beans are roasted (Dart and Nursten,
1985). The reactions that create the
furans are very complicated, and that made it so it cannot be predicted what
type of furan would be present.
Although, it can be predicted that furanose rings should test positive
in the caffeinated and decaffeinated medium roasted bean solutions (Dart and
Nursten, 1985). After doing research on
unroasted coffee beans, it was found that certain types contain furans while
other do not (Dart and Nursten, 1985).
Based on this a prediction for furanose rings in unroasted coffee beans
could go either way, for our study it was predicted that the unroasted coffee
bean solutions would test positive for furanose rings.
Bial’s test
yielded results that showed a positive presence for furanose. All of the coffee bean solutions that were
tested yielded positive results for the presence of pentose furanose
rings. These results agree with the
hypotheses for this test that all types of beans would contain some type of
furanose ring. This is because furans
are found in unroasted coffee beans and their concentration increases as the
coffee beans are roasted (Dart and Nursten, 1985). The positive controls, fructose and xylose, tested positive for
furans, whereas the negative control, distilled water, tested negative.
The Iodine test
was used to test for starch, which is another carbohydrate structure in coffee
beans. Some starch should have been
degraded during the roasting process, and some of it is caramelized when
roasted (Sivetz and Foote, 1963). Since
the roasting process does not degrade all of the starch in the coffee beans, it
can be predicted that starch will found in both the unroasted and roasted
coffee bean solutions.
The results that
we found did not support the prediction for this test. The 5% and 10% solutions of decaffeinated
and caffeinated unroasted (green) coffee beans and 5% and 10% solutions of
decaffeinated and caffeinated medium roasted coffee beans all produced negative
results for starch. The results we
gathered contradicts our research possibly because the 5% and 10%
concentrations of coffee solution were used were too dilute to show the
presence of starch during this test.
Distilled water, our negative control, tested negative, whereas our
positive control of 1% starch solution tested positive, which was expected.
Similar to
carbohydrates in the coffee beans, it was predicted that the roasting process
would decompose protein. When the
coffee bean undergoes roasting amino acid nitrogen in released from the coffee
bean, decreasing the amount of protein present in the bean (Macrae, 1985). The protein concentration of each of the
coffee beans was tested using the Bradford Assay to verify that the unroasted
coffee beans had a higher protein concentration than that of the medium roasted
coffee beans. The results obtained from
our experiment were analyzed using the Bradford Assay Standard Curve. We discovered that the 5% and 10% solutions
of decaffeinated and caffeinated unroasted coffee beans have a higher
concentration of protein than 5% and 10% decaffeinated and caffeinated medium
roasted coffee beans. The 5%
caffeinated unroasted coffee solution contained an average of 0.678 μg/μl, while the 5%
caffeinated medium roasted coffee solution had an average protein concentration
of 0.095 μg/μl. This
is verification of the hypothesis that proteins are degraded by roasting
(Macrae, 1985).
Another general
trend of our findings for the Bradford Assay was that caffeinated unroasted
coffee beans had a higher concentration of protein compared to decaffeinated
unroasted coffee beans. This was
evident by the 10% caffeinated unroasted coffee solution which had an average
protein concentration of 0.848
μg/μl, whereas the 10% decaffeinated unroasted coffee solution had an
average protein concentration of 0.340 μg/μl. This trend was not followed by the
decaffeinated medium roasted coffee beans, which had a higher concentration of
protein; compared to caffeinated medium roasted coffee beans. The 10% caffeinated medium solution was
calculated to have an average protein concentration of 0.124 μg/μl, while the 10% medium decaffeinated coffee
solution had an average protein concentration of 0.244 μg/μl. The difference between these concentrations
was not very large, and the numbers were very close to zero; therefore error in
the procedure, or the fact that the solutions are very diluted could have
caused these coffee solutions to not follow the same trend as the caffeinated
and decaffeinated unroasted coffee bean solutions. Even though research has shown that caffeine is a non-protein
nitrogen compound, it seems as though it did affect the protein concentrations
in the coffee beans (Clifford, 1985).
The results from this experiment showed that the presence or absence of
caffeine in the different types of coffee beans did seem to have an affect on
protein concentration, even though this was not previously expected.
In addition to
testing for changes in macromolecules due to roasting, the photosynthetic
pigments in the coffee beans were also studied using paper chromatography. It seems logical that the coffee beans
themselves should contain the same pigments as the coffee plant. The leaves of a coffee plant has the
pigments chlorophyll a, chlorophyll b, xanthophyll, and carotene according to a
study done on coffee plants regarding chilling and its affect on photosynthesis
of the coffee plant (Goncalves de Oliveira et al., 2002). Due to the inability to find resources on
pigment degradation due to roasting, it could be predicted that because many
other components of the coffee bean break down when roasted, then the
photosynthetic pigments will also break down.
This can be determined visually by the color of the unroasted coffee
bean, which is green, and a roasted coffee bean, which is dark brown. It was predicted that the unroasted coffee
beans would contain the pigments chlorophyll a, chlorophyll b, xanthophyll, and
carotene, and the roasted coffee beans would contain none at all.
The
predictions made before this experiment were then proven wrong by the actual
results. After paper chromatography was
performed on the 5% and 10% solutions of caffeinated and decaffeinated medium
roasted and unroasted coffee beans, no pigments were found in any of the
samples. There are many factors that
can attribute to these results. One of
the factors involved is the concentration of the solutions. They are 5% and 10% concentrations and not
100% pure coffee solutions; therefore they are not going to contain the same
amount of components. Pure coffee
solutions were not used because the solutions needed to be consistent
throughout the experiment, and 5% and 10% solutions were used for the
carbohydrate tests. Had the solutions
not been diluted the precipitates from the carbohydrate tests would not have
been easy to see because the medium roasted coffee solutions were dark even
after dilution. In this case the
pigments may have been diluted out of the solutions. Another factor is the processing of the coffee beans. Both the unroasted and roasted are sun-dried
before being sold (Smith, 1985). This
could have degraded any pigments that could have been found in the unroasted
coffee beans.
For
the final independent part of this experiment, caffeine was extracted from each
type of coffee bean solution. It is
obvious that the caffeinated coffee beans are going to have much more caffeine
than the decaffeinated coffee beans.
Unlike the components previously discussed, caffeine is not broken down
by the roasting process (Parliament, 2000).
Even after roasting caffeine (a nitrogen compound) survives because it
does not even begin to melt until it reaches temperatures of 236°C but does
sublime at 178°C (Trugo, 1985). Through sublimation some caffeine is lost but
the actual amount of caffeine lost is not excessive because the roasting
process does not surpass 218°C (Parliament, 2000). Therefore the caffeine does not melt. Based on this, the goal of our study was to find out how much
caffeine is lost not through roasting but through the decaffeination
process. For this reason we compared
caffeinated to decaffeinated coffee in the last part of the experiment.
The
results of the caffeine extraction assay were quite unusual and overall
inconclusive. In two cases the
decaffeinated coffee bean solutions had higher caffeine levels than the
solution with the same roast of bean that was caffeinated. In addition, when the caffeinated coffee
bean solutions did show a higher amount of caffeine than the decaffeinated
coffee beans solutions, the difference in the amount of caffeine was not
great. In one instance the 10%
caffeinated medium coffee bean solution had 0.47g of caffeine and the 10%
decaffeinated medium coffee bean solution had 0.44g of caffeine. The difference in these values should be
much larger. There could have been a
problem with the chemicals used or the procedure itself because the negative
control, water, had a higher amount of caffeine than the positive control,
caffeinated Coke-Cola®. Results of this
assay were very inconclusive.
When coffee beans
are decaffeinated, they must be at least 97% caffeine free to meet standards in
the
In the entire
experiment it was found that roasting does indeed breakdown sugars, but not
starch because there was none present.
Protein is also degraded by roasting, and interestingly enough it was
found that caffeine seemed to have an effect on increasing the protein
concentration (Parliament, 2000). After
testing for photosynthetic pigments our team found that the sun-drying part of
the processing that the unroasted coffee beans go through must have degraded
any pigments that may have been present (Smith, 1985). Therefore, there were no original pigments
found to compare to the paper chromatography result of the medium roasted
coffee. Finally, the caffeine
extraction assay failed to support our hypothesis in some cases because of a
faulty procedure and/or chemicals. In
the future, studies could be done to further identify specific aldehydes and
ketones that are produced from polysaccharides during roasting. To test for photosynthetic pigments in
future studies, coffee scientists could grow coffee plants in order to take the
unroasted coffee beans from the plant before they are dried. This may produce a better result for paper
chromatography and other pigment tests such as absorption spectrum
analysis. In addition, a different and
more precise assay to test for amounts of caffeine could be used along with a
melting point procedure to check the purity of the caffeine found.