Incorporating CAChe Component into the Aspirin Lab
Ping Y. Furlan
University of Pittsburgh at Titusville
Titusville, PA 16354
Lisa Bell-Loncella
University of Pittsburgh at Johnstown
Johnstown, PA 15904
One of our students’ favorite General Chemistry II Lab exercises is to make aspirin from
oil of wintergreen, a naturally occurring substance [1]. The conversion involves a twostep sequential synthesis: oil of wintergreen to salicylic acid to aspirin:
O
CH3
C
O
O
O
C
OH
OH
C
OH
OH
O
C
CH3
O
Oil of Wintergreen
Salicylic Acid
Aspirin
In this lab, students are introduced to FTIR spectroscopy which is used to characterize the
product aspirin and its precursors. The technique is based on the fact that bonded atoms
vibrate much like balls bound by spring. These vibrations require energy input in the IR
region (100-10,000 cm-1). Different combinations of bonded atoms absorb different
amounts of energy showing absorptions at different wavenumbers or frequencies in IR
spectra. (Wavenumber, cm-1, often called frequency, is directly proportional to the energy
absorbed.) The absorptions for hydroxyl (–OH) and carbonyl (C=O) functional groups,
for example, occur at around 3000 cm-1 and 1700 cm-1, respectively. Additionally, a
group of atoms experiencing a change in its environment generally shows a shift in its IR
absorption. Spectral differences are therefore observed among the three compounds,
namely, oil of wintergreen, salicylic acid, and aspirin, at around 1700 cm-1 due to
different C=O absorptions, and at around 3000 cm-1 due to different -OH absorptions. As
a result, IR spectroscopy can serve as fingerprints for identifying these compounds.
However, we find that many introductory students experience difficulty grasping the
correlation between the structural groups and their vibration energies in the IR region.
CAChe molecular modeling software from Fujitsu may prove to be an effective tool in
teaching concepts of IR spectroscopy. As the IR instrument becomes more and more
available, many students in first year chemistry are exposed to IR spectroscopy,
especially involving characterizing aspirin related compounds [1-3]. The most
remarkable feature of the CAChe software is its ability to provide a direct and visual
correspondence between the IR peaks and the functional groups. The program runs fast
and is user friendly. It also provides exposure to computational chemistry at the
introductory level.
In the Aspirin lab, we are interested in predicting the IR absorptions due to the carbonyl,
C=O, and hydroxyl, OH, functional groups at around 1700 cm-1 and 3000 cm-1. The
CAChe 6.1 Software package provides nine different quantum-mechanical methods for
IR spectrum predictions: six MOPAC (Molecular Orbital Package) methods and three
DFT (Density Function Theory) methods. The MOPAC is a semi empirical computing
method. It optimizes the geometry and determines the IR transitions in a single
experiment. The DFT method, on the other hand, is based on mathematical
approximations. When using DFT methods, the geometry must be optimized first before
computing the IR transitions.
To compare these methods, eight small molecules (ethanol, 1-propanol, 1-dodecanol,
phenol, acetone, 2-butanone, acetic acid, and ethyl acetate) containing the groups of
interest along with oil of wintergreen, salicylic acid, and aspirin were used. Their IR
spectra were compared with those predicted using different methods. It was found that
different methods yielded quite different values and the relative errors ranged from 125% for the C=O group and 1-40% for the OH group. Within a given method, the
calculation errors were observed to be systematic and the predicted wavenumbers were
noted to be fairly fixed within a small range for a given functional group. Although the
variation was small, the methods demonstrated sensitivity to the group’s surroundings,
i.e., it was able to differentiate the absorptions of the same type of group surrounded by
different environments.
MOPAC PM5 seemed consistently to give better results although in many cases the
predicted IR profiles were different from the observed ones. For both functional groups,
the calculation errors using MOPAC PM5 method were under 10%. The MOPAC AM1
usually gave predictions fairly well for the OH absorptions in the 3000 cm -1 region and
the profiles around the region were closer to the experimental ones. The standard
procedure and the MOPAC PM3 yielded similar values, and the MOPAC MNDOd and
PM3/Cl always gave too high values for both groups.
The three DFT methods in general gave good predictions especially for C=O groups.
However, the C=O peaks at around 1700 cm-1 often overlapped with C-H vibrations in
the predicted spectra. Moreover, DFT methods require performing two experiments: the
geometry optimization and the IR absorption computation. Each step was found to take a
considerably long time to run, particularly when the molecule involved a benzene ring or
became larger. Among the three DFT methods, the B88-LYP IR Spectrum gave the best
prediction.
It was believed that it would be valuable for the introductory students to run through
different methods with small molecules and then select the better one to use for
predicting their sample IR spectra. This experience would give them a chance to see that
the calculations are based on different models which are not exact replicates of the
reality. It would also show them that some models are better at predicting certain
properties than others, and these results are capable of shining light into our
understanding of the world of chemistry.
In the first part of the lab, students will use CAChe to learn how different functional
groups absorb IR energy at different frequencies. They will then be given a chance to
compare different methods including MOPAC AM1, PM3, PM5 and B88-LYP DFT to
see which results best match the experimental absorptions for the groups under study.
Acetic acid, a small molecule containing both carbonyl and hydroxyl groups, can be used
for this investigation. The typical results are shown in Table 1.
O
H3C
C
OH
Acetic Acid
Table1. Comparison of Experimental and Calculated IR Absorptions using Various
Methods for Acetic Acid
Acetic
Acid
C=O
OH
Experimental,
cm-1
1717
3056
MOPAC AM1
cm-1
Error
22%
2088
11%
3431
MOPAC PM3
cm-1
Error
15%
1980
24%
3851
MOPAC PM5
cm-1
Error
6%
1827
1%
3097
B88-LYP DFT
cm-1
Error
2%
1747
14%
3534
In this practice, students will first use the CAChe software to draw the acetic acid
molecule and use the “comprehensive” command under “beautify” to optimize the 3-D
structure. They will then save the file four times including the ending (AM1, PM3, PM5,
LYP) on the name to signify the computation method being used. After computing each
“IR transitions”, students can display both the molecule and the IR spectrum on the
screen using the windows “tile” function. When students select a particular absorption on
the spectrum, the atomic motions responsible for that absorption will be shown (See
Figure1). Students should be able to quickly identify that C=O vibration occurs at 17002000 cm-1 and OH at 3000-4000 cm-1. They should also realize that calculations using
MOPAC methods are very quick, usually done within a few seconds. On the other hand,
the DFT method is much slower, taking about 3-5 minutes to finish even for this small
molecule. From a simple error analysis, students can see that MOPAC PM5 method
gives the predictions that best agree with the experimental ones within a time frame of a
few seconds.
Figure1 shows the 3-D structure (right) and the predicted IR spectrum (left) of acetic acid. The highlighted
IR peak around 3400 cm-1corresponds to the indicated vibration of the OH group.
Since students are expected to have performed an earlier activity on molecular structures
and on the use of the CAChe software to draw molecules and optimize structures, this
part of the lab should be done within 15-20 minutes.
The second part of the lab is for students to build the three molecules involved in the
aspirin synthesis lab, namely oil of wintergreen, salicylic acid, and aspirin. Students may
build and beautify one molecule and save it three times using the name indicating the
molecule being studied. They then modify and beautify the other two structures. After
this, students will predict the IR transitions of each molecule using the MOPAC PM5 IR
Spectrum procedure followed by analyzing and recording the IR peaks. They again
should quickly comprehend that different OH groups and different C=O groups would
absorb IR energy at slightly different frequencies. They should anticipate these peaks
from the actual IR spectra of their samples. For instance, for aspirin they should expect
two C=O peaks at around 1700 cm-1; one occurs at a slightly higher frequency due to the
ester type C=O vibration and the other at a slightly lower frequency due to the acid type
C=O vibration (See Figures 2a and 2b). The program runs fast, and the calculations
should be done within 15-20 minutes.
Figure 2a shows the predicted IR peak due to
the ester type C=O vibration that occurs at
1845 cm-1.
Figure 2b shows the predicted IR peak due to
the acid type C=O vibration that occurs at
1819 cm-1.
The last part of the lab is for students to obtain the actual IR spectra of the starting
material (oil of wintergreen) and the two products they made (salicylic acid and aspirin).
They will then match the predicted C=O and OH peaks with the observed ones to
understand the origins of these peaks and to see if they have successfully converted their
compounds. The typical experimental and calculated IR absorptions using MOPAC PM5
are listed in Table2.
Table2. Experimental and Calculated IR Absorptions using MOPAC PM5 for the Listed
Molecules
Experimental,
cm-1
Calculated,
cm-1
Error
Oil of Wintergreen
C=O
OH
(ester)
1681
3188
C=O
(acid)
1669
Salicylic Acid
OH
OH
(acid)
3238
3017
1826
3109
1824
3108
9%
2%
9%
4%
C=O
(ester)
1754
Aspirin
C=O
(acid)
1693
OH
(acid)
2971
3091
1845
1819
3087
2%
5%
7%
4%
In conclusion, CAChe software can be incorporated into the Aspirin Synthesis and
Characterization Lab easily and smoothly. The added component can be finished less
than 30-40 minutes. The exercise will help the introductory students to understand how
different structural groups absorb IR energy at different wavenumbers, how a subtle
change in the group environment can result in a shift in its IR absorption, and how IR can
be effectively used as fingerprints for compound identification. Additionally, it will
provide the students an opportunity to be exposed to molecular modeling and
computational chemistry at the introductory level.
References
1. OlmsteadIII, J “Synthesis of Aspirin: A General Chemistry Experiment”, J.
Chem. Educ. 1998, 75, 1261.
2. Byrd, Houston; O’Donnell, Stephen E. “A General Chemistry Laboratory Theme:
Spectroscopic Analysis of Aspirin”, J. Chem. Educ. 2003, 80, 174.
3. Mirafzal, G. A.; Summer, J. M. “Microwave Irradiation Reactions: Synthesis of
Analgesic Drugs”, J. Chem. Educ. 2000, 77, 356.