Chemical Education Journal (CEJ), Vol. 18 / Registration No. 18-101 / Received October 13, 2015.
URL = http://www.edu.utsunomiya-u.ac.jp/chem/cejrnlE.html 


A Simple Case Study for the Introduction of the HMBC and the EI-MS Techniques to Second and Third Year Undergraduate Students.

Esther H. S. WOO1, Mackenzie J. FIELD1, Mathew L. SUTHERLAND1, Chloe A. N. GERAK2* and Nabyl MERBOUH1*

1Simon Fraser University, Department of Chemistry, 8888 University Drive, Burnaby, BC, V5A 1S6, Canada.

E-mail: nmerbouhsfu.ca

2University of British Columbia, Department of Biochemistry and Molecular Biology, 2350 Health Sciences Mall, Vancouver, BC, V6T 1Z3, Canada.

E-mail: cagerakalumni.ubc.ca


Abstract

The structure elucidation and spectral assignments of two constitutional isomers of phenylbutyric acid proved to be more challenging than anticipated for second year undergraduate students. When given 3-phenylbutyric acid and α-methylhydrocinnamic acid as unknowns, definitive identification by usual spectroscopic methods required more careful analyses of the spectroscopic data, along with the use of additional analytical techniques. By solely relying on the usual 1D-NMR spectroscopy (1H-NMR, 13C-NMR), or even 2D-NMR spectroscopy (COSY, HSQC), students' structures could only be partially solved. Additional spectroscopic methods such as HMBC and mass spectrometry (EI-MS) were needed for the complete spectral assignments, thus making these "simple" compounds ideal case studies for the introduction of HMBC, EI-MS and basic computational chemistry at the undergraduate level.

Keywords

Nuclear Magnetic Resonance (NMR) Spectroscopy, Mass Spectrometry (MS), Computational Chemistry, Structure Elucidation.


Contents

Introduction

Methods

Results

Discussion

Conclusion

Supporting Materials

References

Appendix


Introduction

In first and second year organic chemistry classes, students are often exposed to a series of analytical methods to learn about the identification of synthesized compounds. In general, students start by recording melting points (mp) and infrared (IR) spectra of their compounds and are given the 1H-NMR spectrum to complete their identification. Occasionally, the chemical formula, chemical ionization mass spectrum (CI-MS) or elemental analysis is provided to the student as a resource to facilitate their assignments.

We, as instructors, have all witnessed that, as soon as students are introduced to 1H-NMR spectroscopy and become acquainted with the spectrometer use, it rapidly becomes the sole analytical method, often at the expense of all other methods. One may be led to think that 1H-NMR spectroscopy has no limitation in distinguishing compounds and that it is a universal analytical method that can solve it all.

In this paper, we will introduce a very simple identification exercise based on the spectral comparison of two isomers of phenylbutyric acid (phenylbutanoic acid): 3-phenylbutyric acid (PBA) and α-methylhydrocinnamic acid (MHCA) (2-methyl-3-phenylpropanoic acid) (Figure 1). The students are asked for the unequivocal differentiation and spectral assignment of the structure of these compounds based on the knowledge and analytical data that is typically studied in their first few years of organic chemistry.

The questions that we are asking the students are the following: i) Can you unequivocally assign each spectrum to each compound and can you explain your assignment rationale? ii) Do you understand the strengths and limitations of each of the analytical techniques you used? iii) Did you need to learn other analytical techniques, and how did they help?

Figure 1. Chemical structure of α-methylhydrocinnamic acid (MHCA) and 3-phenylbutyric acid (PBA).

 

Methods

3-Phenylbutyric acid and α-methylhydrocinnamic acid were purchased from Sigma-Aldrich and used without purification. Melting points were recorded using a Barnstead Mel-Temp capillary melting point apparatus and are uncorrected. 1H-NMR and 13C-NMR were recorded on a Bruker 400 or 500 MHz spectrometers and were referenced to residual solvent peaks (CDCl3, 7.26 ppm for 1H-NMR, and 77.1 ppm for 13C-NMR). The mass spectra were recorded on a Varian 4000 GC/MS/MS spectrometer. IR spectra were recorded on a PerkinElmer UATR Two Fourier transform spectrophotometer and the UV spectra were recorded on a Varian Cary 100 Bio spectrophotometer.

 

Results

3-Phenylbutyric acid (PBA): C10H12O2, white solid, mp = 36-38 °C. 1H-NMR (400 MHz, CDCl3) δ = 10.17 (Very broad signal, no reliable integration possible); 7.31 (m, 2H); 7.22 (m, 3H); 3.28 (sextet resembling signal, 1H); 2.68 (dd, J = 15.5, 6.8 Hz, 1H); 2.58 (dd, J = 15.5, 8.2 Hz, 1H) 1.33 (d, J = 7.0 Hz, 3H) ppm. 13C-NMR (101 MHz, CDCl3) δ = 178.11, 145.50, 128.65, 126.79, 126.60, 42.55, 36.24, 21.94 ppm. IR (ATR, neat) ν = 3091-2569 (broad band), 1698, 1430, 1279, 1215, 927, 755, 691 cm-1. UV (DCM) λmax (ε) = 258 nm (187 M-1 cm-1). EI-MS m/z = 164 (Molecular Ion), 118, 105 (Base Peak), 79, 77.

α-Methylhydrocinnamic acid (MHCA): C10H12O2, white solid, mp = 38-40 °C. 1H-NMR (400 MHz, CDCl3) δ = 11.14 (Very broad signal, no reliable integration possible); 7.30 (m, 2H); 7.20 (m, 3H); 3.08 (dd, J = 13.3, 6.3 Hz, 1H); 2.77 (sextet resembling signal, 1H); 2.68 (dd, J = 13.4, 8.0 Hz, 1H) 1.18 (d, J = 6.9 Hz, 3H) ppm. 13C-NMR (101 MHz, CDCl3) δ = 182.00, 139.10, 129.09, 128.51, 126.52, 41.24, 39.39, 16.59 ppm. IR (ATR, neat) ν = 3086-2557 (broad band), 1698, 1453, 1209, 1249, 1230, 941, 740, 702 cm-1. UV (DCM) λmax (ε) = 258 (145 M-1 cm-1). EI-MS m/z = 164 (Molecular Ion), 118, 91 (Base Peak), 77, 65.

Discussion

We are exposing the second year students to 1H-NMR spectra of both 3-phenylbutyric acid and α-methylhydrocinnamic acid with a simple task, to assign all the peaks in the 1H-NMR spectra and determine which spectrum corresponds to which compound. This proved to be a challenging exercise for the students on many levels. Looking at the proton spectra (Figure 2) and the structures, the presence of diastereotopic protons [1] (see Appendix) may be the first difficulty encountered in the assignment for the students. Secondly, the chemical shifts and peak shapes of all the protons in the spectra are nearly identical and appear in the same relative spectral region, making the differentiation between the two compounds almost impossible. The simplicity of the 1H-NMR spectra with their limited information and limited amount of signals thus demonstrated the need for additional analytical data.

Students are given each compound in an unlabelled vial and, under an instructor's supervision, are asked to differentiate these two compounds. They will be provided access to any needed spectroscopy technique, external literature and all available software in a chemistry department. This allows the students to embark on a series of measurements, starting with simple experiments such as elemental analysis and melting points. [For the sake of this article and for clarity, we will assign all the spectra to the respective compounds; a set of spectra and their Free Induction Decays (FIDs) will be provided in the supplementary material for instructors and training purposes.]

Figure 2. Superimposed 1H-NMR spectra of PBA (blue) and MHCA (red) in CDCl3, showing relevant regions.

Differentiation using elemental analysis. As the compounds are constitutional isomers (see Appendix), the students should quickly abandon the elemental analysis technique since it will give identical results in both cases. While it is a valuable technique in many cases, it is not helpful in this instance since it only provides the partial chemical composition of the two compounds.

Differentiation using melting point. Simply taking the melting point of these compounds will indicate that the students are dealing with two different compounds as there are two different melting point ranges: 36-38 °C and 38-40 °C. The melting points of the two isomers are similar and almost overlapping. Therefore, the students would have to be certain that the melting points are taken accurately on a well-calibrated apparatus and that no impurities are present in their samples. The subsequent question to be asked to the student is: can you base your entire identification on melting points that are in such close proximity to each other? The student should conclude that they cannot accurately distinguish between the two compounds and can then move on to other types of spectroscopy, such as IR and UV spectroscopy. Both of these techniques are quick and allow for the immediate recording of spectra while requiring limited data processing.

Differentiation using IR spectroscopy. The stacked IR spectra of both compounds (Figure 3) show limited differences between the compounds. The near-identical functional group identification (carbonyl groups visible at 1698 and 1699 cm-1, hydroxyl groups of carboxylic acids between 3050 and 2400 cm-1, aromatic (C=C) stretches between 1600 and 1450 cm-1) and the fingerprint region offer very little help to differentiate between the compounds. Because the compounds have identical "building blocks", IR spectroscopy is not a suitable technique to differentiate between these structural isomers. Nonetheless, it is extremely helpful in providing the students with information on the functional groups present in both molecules.

Figure 3. Stacked IR spectrum (ATR, neat) of PBA (blue) and MHCA (red).

Differentiation using UV spectroscopy. UV spectroscopy can be useful in certain cases when structural isomers have different absorption maxima due to conjugation. However, overlaying the UV spectra (Figure 4) show that both compounds have identical absorption profiles and a λmax of 258 nm (benzene π → π* secondary transition) with slightly different molar extinction coefficient (ε) values. A quick literature search can provide the students with the significance of the observed spectra, thus allowing them to correlate the UV spectra to the possible structures [1].

For benzene, three aromatic π → π* electronic transitions take place due to electron-electron repulsion and symmetry considerations. These transitions correspond to the first primary band, the second primary band and the secondary band. The first primary band occurs at 184 nm and is a spin allowed transition but cannot be seen in standard UV experiments. The second primary band typically has a much larger molar absorptivity (ε = 7400) than those observed in the MHCA and PBA spectra.

Comparatively, the benzene spectrum has the secondary band displaying a molar absorptivity of 230 M-1 cm-1, which is similar to the molar absorptivities we observe in the MHCA and PBA spectra. The secondary band at 258 nm is the largest peak observed in the UV spectra of both PBA (ε = 187 M-1 cm-1) and MHCA (ε = 146 M-1 cm-1). This corresponds to an electronically allowed (π to π*) and symmetry forbidden transition based on the symmetry of the lowest energy anti-bonding molecular orbitals. Typically, most of the intricate structure is seen in this secondary band, however this is lost when using a polar solvent, such as DCM. This changes the relative energies of distinct orbitals causing a loss of vibrational fine structure and explains why shoulders are seen in the spectra rather than distinct bands.

Figure 4. Superimposed UV spectra of PBA (blue) and MHCA (red) in dichloromethane (0.05 M concentration, ambient temperature).

With the first analytical techniques giving almost identical results, the students are now ready to turn to using NMR spectroscopy as a method to solve their problem. As both compounds are soluble in CDCl3, the students can easily prepare their NMR samples and collect a NMR 'full package' to save time. This package will consist of the following spectroscopy techniques: 1H-NMR, 13C-NMR, 13C-APT (attached proton test) and DEPT (distortionless enhancement by polarization transfer), COSY (correlation spectroscopy), HSQC (heteronuclear single quantum correlation) and HMBC (heteronuclear multi-bond correlation). However, with all these spectra on hand, students should be warned that they might not need all of them to solve the problem, that they might need to learn about more advanced NMR techniques to interpret them properly and that there can never be too much evidence to validate the structure of your final compound [2-5].

Differentiation using 1D-NMR spectroscopy. At the second year organic chemistry level, students are generally taught to look at the following features in each NMR spectrum: the shape of the peaks, their relative ratio of integrations and their chemical shifts. It is only when more advanced NMR techniques and more complex compounds are introduced that the students are made aware of the importance of the coupling constants. In this exercise, both compounds have identical multiplets with integrations of 2H and 3H in the aromatic region (7.31-7.20 ppm) consistent with the presence of monosubstituted benzene rings, and 3H doublets with similar chemical shifts and coupling constants (1.33 ppm, J = 7 Hz for PBA and 1.18 ppm, J = 6.9 Hz for MHCA) consistent with the presence of a shielded methyl group (-CH3). The region between 2.50 and 3.50 ppm, corresponding to benzylic protons or protons adjacent to carbonyl groups, contains the remainder of the peaks in the spectrum and includes a sextet-looking signal and a pair of doublets of doublets (dd) for both compounds (Figure 5).

Interestingly, this region shows several unexpected features for such simple molecules. Upon closer examination of both spectra, the sextet resembling signal was determined to be a multiplet resulting from the presence within this signal of two different coupling constants, which gave evidence to the presence of diastereotopic protons in both molecules.

The diastereotopic protons (dd) have different chemical shifts but almost identical coupling constants while exhibiting a very strong "roof effect" [1] (see Appendix), a phenomenon that is a common second order effect [1] (see Appendix), usually only observed in low-field spectra.

Figure 5. Side by side, truncated 1H-NMR region (3.50 to 2.50 pm) for PBA (blue) and MHCA (red).

The information gathered by the 1H-NMR in this case is substantial, as the students can now put together most of the building blocks of their compounds: a carboxylic acid (-COOH, confirmed by IR and 1H-NMR), a monosubstituted benzene ring (-C6H5, confirmed by 1H-NMR), a methyl group (-CH3, confirmed by 1H-NMR) connected to a methine group (>CH-) and possibly a methylene group (-CH2-) carrying two diastereotopic protons, the lattermost still requiring corroboration by complementary NMR techniques.

Students can also gather additional information from the 1H-NMR including the chemical shifts of each peak and the coupling constants that are found to be associated within each peak. At the second year level, students may differentiate between unambiguous compounds using the "predict the 1H-NMR shift" function on ChemDraw®. They may also calculate or predict coupling constants of the minimized structures using the Karplus equation on the MestReJ software, or its modified versions, such as those developed by Altona et al. [6-8]. The ChemDraw® prediction is generally a good indicator of the relative chemical shifts of all protons in the molecule and the program's coupling constant predictions provide a good basis of the coupling trends.

ChemDraw 1H-NMR Predictions. As seen in Table 1, the stark differences between the predicted and the observed chemical shifts are only seen for one of the diastereotopic protons, where a difference of ±0.15 and ±0.20 ppm is observed for MHCA and PBA, respectively. However, such predictions should only be used to provide students with an idea of the chemical shifts differences of the protons within the same molecule, and should not be used as evidence to differentiate between the two molecules due to the large predictions differences.

Table 1. Summary of the predicted (ChemBioDraw®, Level: Ultra, Version 13.0.2.3020) and observed chemical shifts for PBA and MHCA.

   Predicted Observed Difference
 PBA δ CH3: 1.25 ppm
δ CH2: 2.62 & 2.38 ppm
δ CH: 3.20 ppm
δ CH3: 1.33 ppm
δ CH2: 2.68 & 2.58 ppm
δ CH: 3.28 ppm
±0.08 ppm
±0.06 & ±0.20 ppm
±0.08 ppm
 MHCA δ CH3: 1.12 ppm
δ CH2: 3.08 & 2.83 ppm
δ CH: 2.85 ppm
δ CH3: 1.18 ppm
δ CH2: 3.08 & 2.68 ppm
δ CH: 2.77 pm
±0.06 ppm
±0.00 & 0.15 ppm
±0.08 ppm

 

Structure Minimization and Coupling Constant Predictions. PBA and MHCA molecules were built on GaussView (5.0) and optimized with tight convergence along various parameters. The ground state Hartree-Fock method, using the 6-311G basis set, was deemed most appropriate for the undergraduate level; the optimization time averaged approximately 1 hour and gave consistent results (Figure 6). (Note: after optimizing structures, it was found that the enantiomers of both structures provided different values for the dihedral angles. All values reported are averages obtained from the enantiomeric structures). The dihedral angles (Φ), as well as the angles between the diastereotopic protons, were obtained from the minimized structures. The values for the dihedral angles were then inputted into the MestReJ software and the corresponding 3J coupling constants were obtained. This procedure was repeated multiple times on several computers for both molecules, with the average angles and coupling constants summarized in Table 2.

Figure 6. Minimized Structures for PBA and MHCA showing the relevant dihedral angles [9a].

Table 2. Summary of the predicted (MestReJ) and observed dihedral angle and coupling constants for PBA and MHCA in CDCl3. (HLA: Haasnoot-de Leeuw-Altona; DAD: Díez-Altona-Donders) [6-10].
 

 

J and Φ values for MHCA

 

J and Φ values for PBA

 Karplus Prediction (Hz)
3Jbc = 1.699 Hz (69.6°)
3Jac = 10.14 Hz (174.6°)
2Jab = Not Predicted (107.1°)
3Jab = 2.17 Hz (67.45°)
3Jac = 10.13 Hz (174.2°)
2Jbc = Not Predicted (107.1°)

 Altona Prediction (Hz)

(HLA-General)
3Jbc = 1.46 Hz (69.6°)
3Jac = 14.47 Hz (174.6°)
2Jab= Not Predicted (107.1°)
3Jab = 1.81 Hz (67.45°)
3Jac = 14.45 Hz (174.2°)
2Jbc = Not Predicted (107.1°)

 Altona Prediction (Hz)
(HLA-Chemical Groups)
3Jbc = 2.17 Hz (69.6°)
3Jac = 12.78 Hz (174.6°)
2Jab= Not Predicted (107.1°)
3Jab = 2.26 Hz (67.45°)
3Jac = 12.81 Hz (174.2°)
2Jbc = Not Predicted (107.1°)

 Altona Prediction (Hz)
(DAD-Chemical Groups)
3Jbc = 12.66 Hz (69.6°)
3Jac = 12.66 Hz (174.6°)
2Jab= Not Predicted (107.1°)
3Jab = 2.28 Hz (67.45°)
3Jac = 12.66 Hz (174.2°)
2Jbc = Not Predicted (107.1°)

 Observed (Hz)
3Jbc = 6.30 Hz
3Jac = 8.00 Hz
2Jab = 13.3 Hz
3Jab = 6.80 Hz
3Jac = 8.20 Hz
2Jbc = 15.5 Hz

Figure 7. All possible staggered conformations of PBA.

The coupling constants calculated differ somewhat in the quantitative results provided depending on which equation was used, but all give the same qualitative models. However, an important concern regarding the minimization of the two compounds is that there are multiple conformations for each compound (each one having its own dihedral angles), which all partially contribute to the overall observed coupling constant. Ideally, the students will minimize and calculate the relative energies of each conformation, extract their respective dihedral angles, and calculate the coupling constants of the protons in all conformations (Figure 7). The students will then take a weighted average of the variety of coupling constants based on their respective stabilities in order to obtain a more accurate coupling constant prediction. However, such an endeavor is more time consuming and requires more knowledge of the software than is feasibly taught in an undergraduate laboratory experiment.

We elected to focus on the minimization of the entire structure to explain the presence of diastereotopic protons in the spectra as a consequence that each compound has three different, minimizable staggered conformers. In any case, not all the conformations will contribute equally to the coupling constant and, under the time constraint for the lab exercise, the 'overall' minimization of both compounds appeared to be the ideal method to get a clear idea of their preferred conformations. Therefore the values predicted for the coupling constants are those calculated from the most stable conformation for each compound, obtained with a reasonable level of computation. The 1H-NMR spectra and predictions gave an appreciable amount of information regarding the partial structure of the compounds and provided insight into some correlations and couplings, however other NMR spectroscopy techniques such as 13C-NMR and APT are needed to continue the structure elucidation.

Differentiation using 13C-NMR. The differentiation between the two compounds by comparing 13C chemical shifts is a difficult task despite using both the chemical shift prediction method and the actual 13C-NMR spectra. It is hard to see a clear difference by looking at the stacked 13C-NMR superimposed plot of both isomers (Figure 8). A few carbon peaks have different chemical shifts, but the students should ask themselves whether such differences give a major hint on how to confidently assign the structure of each isomer? At the second or third year undergraduate level, 13C-NMR interpretation is mainly focused on the number of carbons present in the molecule and where these peaks are located in one of the four regions of the spectra that contain characteristic carbon peaks: carbonyl groups region, unsaturated and aromatic ring carbons region, saturated carbons affected by electronegative atoms region, and saturated carbons unaffected by their surrounding region. In this case, the 13C-NMR spectra of each compound will provide the students with the confirmation of the presence of a carbonyl group (at 178 and 182 ppm), the presence of three carbons unaffected by any electronegative atoms and a benzene ring, which was already verified by the 1H-NMR. Other types of 13C-NMR spectroscopy available to students are the DEPT and APT experiments. After reading about the strengths and limitations of each type of experiment, students will have the choice of picking one technique over the other to further continue their investigation.

Figure 8. Stacked 13C-NMR spectra of MHCA (red, top) and PBA (blue, bottom) in CDCl3, showing relevant regions.

APT and DEPT experiments. Distortionless enhancement by polarization transfer (DEPT) and attached proton test (APT) are two different kinds of 13C-NMR experiments. The APT experiment is simpler than the DEPT experiments, and allows for the separation of the carbons unattached to protons (quaternary carbons) and CH2 signals from CH and CH3 signals. There are several types of DEPT experiments: DEPT-45, DEPT-90, and DEPT-135. In a DEPT-135 experiment, all carbons in the molecule that are attached to a proton will be observed; however, the phase (i.e. positive or negative peaks) of the carbon will differ depending on whether it has an odd or even number of hydrogen atoms attached. APT, although a less sensitive technique than the DEPT, is often preferred since it shows all carbon signals at once, unlike the DEPT experiment which suppresses quaternary carbons and requires three different experiments (DEPT-45, -90 and -135) to yield the same result.

The APT spectra of both compounds are very helpful in determining the different types of carbons present in the molecule (primary, secondary, tertiary or quaternary). The carbonyl carbon (either 178 or 182 ppm) was reconfirmed, along with the presence of the quaternary carbon of the benzene ring (either 145.5 or 139.1 ppm). More information can be extrapolated from the APT spectra, but it will be suggested to the students that interpreting the APT and the HSQC spectra side by side will provide more valuable information and help to save time and energy.

Differentiation using 2D-NMR - 1H-1H Correlation Spectroscopy (COSY). COSY is one of the most frequent two-dimensional experiments used by students. This experiment charts the proton spectrum on the vertical and horizontal axes which gives rise to a third dimension displaying the intensity of the overlapping signals. This helps simplify complex spectra by showing which protons couple together. The student will obtain a spectrum that consists of peaks that are in a diagonal line and these peaks will correspond to the same proton peak on each axis. The cross peaks (off-diagonal peaks) that are found show the peaks that are correlated together by their spin-spin coupling, indicating protons that are coupled together [11].

Figure 9. Overlay of truncated COSY spectra (4.0 to 0.5 ppm on both the horizontal and vertical axes) for MHCA (red) and PBA (blue).

Although valuable, the information extracted from the COSY spectra on simple molecules, such as PBA and MHCA, is limited (Figure 9). The aromatic protons are only coupled to each other as expected in a monosubstituted benzene ring (spectral region omitted in Figure 9). The interesting part of the spectral analysis is that the COSY may show, in both cases, the coupling between a methyl group and methine and a methylene group in the following order: CH3-CH-CH2-. The COSY spectrum needs to be further complemented with an HSQC experiment to prove this potential sequence, as well as to confirm the presence of diastereotopic protons.

Heteronuclear Single Quantum Coherence (HSQC). HSQC is a sensitive technique used to identify specific protons that are attached to specific carbons. Similar to the COSY technique, HSQC experiments are two-dimensional, with the difference being that HSQC charts a proton spectrum on one axis and a carbon spectrum on the other. The cross peaks that result from this experiment show correlation between the carbon atoms and the specific proton atoms that are attached. This allows the students to match proton NMR peaks to their specific carbon NMR peaks, or vice versa, often helping further assignment of NMR peaks to the corresponding molecule [11].

Since the APT spectra has already provided the students with information regarding the different carbon types present in the molecules, the HSQC will serve as definitive proof of the presence of diastereotopic protons as they both couple to the same carbon. For the MHCA spectrum, protons at 3.08 & 2.68 ppm couple with the same carbon at 39.39 ppm, while for the PBA spectrum, the protons at 2.58 & 2.68 ppm couple with the same carbon at 41.55 ppm (Figure 10).

Figure 10. Overlay of truncated HSQC spectra (4.0 to 0.5 ppm & 60 to 10 ppm) for PBA (blue) and MHCA (red).

At this point in the exercise, it is important to allow the students to write all the possible structures and to provide them with a chart thereby allowing them to write all the peaks they have assigned in an effort to limit mistakes or repetition in their work (Table 3). It is also important for the students to assign a single set of labels for all carbons and protons once they have decided on the two possible final structures to avoid any confusion. Most of the peaks may now be assigned with the data collected thus far; however, while the data accumulated will allow for the correct labelling of both molecules, their differentiation will remain unclear.

Table 3. Summary and mock chart of the information gathered by the students mid-point through the exercise.

 

 

 1H and 13C chemical shifts (ppm)

 1H and 13C chemical shifts (ppm)
Ha: 10.17 ppm; carbonyl: 178.11 ppm
Hb & Hb': 2.68 & 2.58 ppm; Cb: 42.55 ppm
Hc: 3.28 ppm; Cc: 36.24 ppm
Hd: 1.33 ppm; Cd: 21.94ppm
He: 7.22 ppm; Ce: 126.79 ppm
Hf: 7.31ppm; Cf: 128.65 ppm
Hg: 7.22 ppm; Cg: 126.60 ppm
Ch: 145.50 ppm
Ha: 11.14 ppm; carbonyl: 182.00 ppm
Hb & Hb': 3.08 & 2.68 ppm; Cb: 39.39 ppm
Hc: 2.77 ppm; Cc: 41.24 ppm
Hd: 1.18 ppm; Cd: 16.59 ppm
He: 7.20 ppm; Ce: 129.09 ppm
Hf: 7.30 ppm; Cf: 128.51ppm
Hg: 7.20 ppm; Cg: 126.52 ppm
Ch: 139.10 ppm

Most of the peaks can be correctly assigned through careful analysis of the data despite the presence of a few peaks that may prove more challenging due to overlapping signals. However, in order to confidently differentiate the compounds, the students are to be made aware of another 2D-NMR technique: HMBC.

Heteronuclear Multiple Bond Correlation (HMBC). HMBC is a sensitive, long-ranged technique that allows the student to identify specific protons that are attached to specific carbon atoms. In contrast to the HSQC experiment, HMBC can observe heteronuclear correlations over several bonds. This enables the students to ascertain which atoms are in proximity to each other through bonds. HMBC utilizes a two-dimensional spectrum with a proton spectrum on one axis and a carbon spectrum on the other. The resultant cross-peaks show which protons and carbons are attached over several bonds and show correlations that demonstrate primarily 2JCH and 3JCH connectivities [11]. The possible couplings observed in an HMBC spectrum and the expected/predicted couplings for PBA and MHCA are shown in Figure 11. HMBC spectra may seem overwhelming at first due to its multitude of couplings; however, it can become an easy task if the instructor ensures that the students make several predictions of expected cross-peaks to see before embarking on the interpretation of the spectra. With the instructor's advice, the students will start their interpretation from a well-defined set of peaks in their spectra. In both the MHCA and PBA compounds, the diastereotopic protons will offer an ideal starting point.

Figure 11. Possible HMBC coupling observed for the diastereotopic protons in MHCA with 5 possible couplings (left) and in PBA with 4 possible couplings (right). The centre figure demonstrates how the proton on the methine group can either 2JCH or 3JCH couple to adjacent carbons through bond.

The difference between the HMBC spectra, albeit subtle, is clearly shown, as there are 5 couplings in the case of MHCA versus 4 couplings in the case of PBA (Figure 11). Combined with all the previous findings, this analysis aids in the assignment of the structures to their corresponding spectra. The superimposed PBA and MHCA HMBC spectra show all the diastereotopic protons couplings present in the molecules and, in the MHCA spectra, the presence of the extra interaction (shown in the dashed rectangle) thus putting the differentiation problem to rest (Figure 12).

In addition, a three-bond coupling between the methyl protons (Hd, δ = 1.18 ppm) and the carbonyl carbon (Ca, δ = 182.00 ppm) may be predicted in the MHCA spectrum. This is not seen for PBA because the methyl hydrogen (Hd, δ = 1.33 ppm) to carbonyl carbon (Ca, δ = 178.11 ppm) coupling would have to occur across four bonds, which is only expected in conjugated systems. Upon analysis of the HMBC spectra, a cross signal between the methyl protons and the carbonyl carbon is observed in the MHCA spectrum and is absent in the PBA spectrum. Furthermore, a three-bond coupling between the methyl protons (Hd, δ = 1.33 ppm) and the quaternary aromatic carbon (Ch, δ = 145.50 ppm) is found in the PBA spectrum but not the MHCA spectrum as it would correspond to a four-bond coupling between the methyl protons (Hd, δ = 1.18 ppm) and the quaternary aromatic carbon (Ch, δ = 139.10 ppm).

The above technique shows a clear difference between MHCA and PBA. At this point, the students should be exposed to mass spectrometry and fragmentation patterns as additional techniques to validate their structures and spectral assignments.

Figure 12. Superimposed truncated HMBC (4.0 to 2.0 ppm & 200 to 0 ppm) spectra for PBA (blue) and MHCA (red) with labelled observed couplings. The solid brackets show the diastereotopic protons coupling peaks present in both molecules, while the dashed rectangle shows the extra coupling present in MHCA.

Differentiation using mass spectrometry fragmentation patterns. When the students have exhausted the NMR techniques available to them, they will need to confidently confirm their structure determination by utilizing mass spectrometry. The students should be reminded here that, generally speaking, more than one analysis is needed to prove structures and that it is always better to collect more evidence to substantiate their argument. In general, the first mass spectrometry techniques undergraduate students are exposed to are chemical ionization and electron ionization mass spectrometry, with each technique having its advantages and limitations.

Chemical Ionization - Mass Spectrometry (CI-MS): While this method is an informative method to visualize the molecular ion (M+H+ = 165) for both isomers, it is not a very useful technique for structure elucidation as CI-MS fragmentation patterns are often difficult to interpret [1-4].

Electron Ionization - Mass Spectrometry (EI-MS): This is a second mass spectrometry technique that demonstrates a clear difference between the two isomers. When both spectra are recorded, the same molecular ion is observed for both compounds (m/z =164) while the base peaks for each compound are dramatically different (m/z = 91 and 105). The compounds also shared an identical fragment with an m/z of 118 (Figure 13).

With the molecular ion present in both spectra, the students should focus on the fragmentation patterns of both compounds. To understand the fragmentation pattern of these molecules, a close look at the parent molecule, 3-phenylpropanoic acid, and its fragmentation peaks and patterns was necessary. The main fragment for 3-phenylpropanoic acid and its ester had an m/z of 104, which corresponds to the well-documented expulsion of formic acid or methyl formate [12, 13]. The common peak at m/z of 118 is also due to the expulsion of formic acid (loss of 46) observed in the PBA and MHCA spectra. The base peaks for both compounds can be rationalized using the well-documented fragmentation of benzene derivatives (e.g. toluene and cumene), where the formation of the benzyl cation spontaneously rearranges to the tropylium (cycloheptatrienyl) ion (m/z = 91). When the side chain attached to the benzene immediately branches, as in the case of PBA, the formation of the methyltropylium ion is observed (m/z = 105). Another advantage from observing the formation of the tropylium or methyltropylium ions is that these ions both exhibit characteristic fragmentations of their own through the loss of an ethyne moiety. This results in a commonly observed peak corresponding to either the cyclopentadienyl (m/z = 65) and methylcyclopentadienyl (m/z = 79) cations in the spectra of substituted benzene rings (Figure 14).

Figure 13. Stacked EI-MS spectra of PBA (blue, top) and MHCA (red, bottom).

Figure 14. Observed fragments for PBA (top) and MHCA (bottom) in EI-MS [14,15].

Conclusion

Through the identification and differentiation of the simple molecules of 3-phenylbutyric acid and α-methylhydrocinnamic acid, students are shown the limitations of routine undergraduate characterization experiments, such as techniques using infrared spectroscopy, and several 1D- and 2D-NMR techniques. The difficulty in distinguishing between the two 'simple' compounds facilitates the introduction of mass spectrometry and the more complex HMBC NMR experiment. Due to the relative simplicity to explain and carry out these techniques in an undergraduate course, this case study is recommended to introduce more advanced characterization techniques to undergraduate students early on in their studies. This exercise also has the benefit of exposing students to a practical, realistic problem they may encounter in their chemistry careers.

Acknowledgements

We would like to thank the Simon Fraser University chemistry department for its financial and logistical support. We would also like to thank Dr. Charles Walsby for his help with the minimization softwares, and Dr. Hamel Tailor for the critical reading of the manuscript.

Supporting Materials: Complete processed spectral data for both compounds are provided in the supplementary material as pdf files. The NMR FIDs are available from the corresponding author: nmerbouhsfu.ca.


References

1. Pavia, D. L.; Lampman, G. M.; Kriz, G. S.; Vyvyan, J. R., "Introduction to spectroscopy, 5th ed.", Brooks/Cole, Belmont (CA), 2009.
2. Silverstein, R. M.; Webster, F. X.; Kiemle, D. J., "Spectrometric identification of organic compounds, 7th ed.", John Wiley & Sons, Hoboken (NJ), 2005.
3. Lambert, J. B., "Organic structural spectroscopy", Prentice Hall, Upper Saddle River (NJ), 1998.
4. Crews, P.; Rodríguez, J.; Jaspars, M., "Organic structure analysis", Oxford University Press, New York, 1998.
5. Breitmaier, E., "Structure elucidation by NMR in organic chemistry: a practical guide. 3rd rev. ed.", Wiley, Chichester (England), 2002.
6. Haasnoot, C. A. G.; Deleeuw, F. A. A. M.; Altona, C., Tetrahedron, 1980, 36 (19), 2783-2792.
7. Donders, L. A.; Deleeuw, F. A. A. M.; Altona, C., Magn. Reson. Chem., 1989, 27 (6), 556-563.
8. Altona, C.; Ippel, J. H.; Hoekzema, A. J. A. W.; Erkelens, C.; Groesbeek, M.; Donders, L. A., Magn. Reson. Chem., 1989, 27 (6), 564-576.
9. a) Dihedral angles were calculated using the Gaussian-09W electronic structure program, May 2011.
b) Navarro-Vazquez, A.; Cobas, J. C.; Sardina, F. J.; Casanueva, J.; Diez, E., J. Chem. Inf. Model., 2004, 44 (5), 1680-1685.
10. Cookson, R. C.; Crabb, T. A.; Frankel, J. J.; Hudec, J., Tetrahedron, 1966, 7, Suppl. 355-390.
11. Miller, V. R., "Use of HSQC, HMBC, and COSY in Sophomore Organic Chemistry Lab", in: Soulsby, D.; Anna, L. J.; Wallner, A. S., editors, "NMR Spectroscopy in the Undergraduate Curriculum", American Chemical Society, 2013, p. 103-128 (ACS Symposium Series, vol. 1128).
12. Ballantine, J. A.; Curtis, R. F., Org. Magn. Res., 1970, 3, 1215-1217.
13. Rosca, S. I.; Stan, R.; Ungureanu, E. M.; Stanciu, G.; Rosca, S., U.P.B. Sci. Bull., Series B., 2008, 70, 77-84.
14. Resink, J. J.; Venema, A.; Nibbering, N. M. M., Org. Mass. Spectrom., 1974, 9 (10), 1055-1058.
15. Amick, A. W.; Hoegg, E.; Harrison, S.; Houston, K. R.; Hark, R. R.; Reingold, I. D.; Barth, D.; Letzel, M. C.; Kuck, D., Int. J. Mass. Spectrom., 2012, 316, 206-215.
16. Bruice, P. Y., "Organic chemistry. 7th ed.", Pearson/Prentice Hall, Upper Saddle River (NJ), 2014.

Appendix: Useful Definitions [1,16]
- Constitutional/structural isomers: molecules that have the same molecular formula, but different connectivity of their atoms.
- Diastereotopic protons/hydrogens: two hydrogen atoms bonded to a single carbon
atom that when one hydrogen is replaced with a deuterium, result in a pair of diastereomers.
- Roof/leaning effect: a second-order NMR spectra effect in which, when two peaks are
strongly coupled, the intensities of the peaks in a NMR signal are tilted upwards in the
direction towards the NMR peak that it has the coupled spin.
- Second-order spectra: strong coupling effects most commonly observed when the
chemical shifts between two groups of protons are similar to the magnitude of the
coupling constant between them.