Rep:Mod:SLimShady

Stefanie Lim CID: 00593047

Modelling using Molecular Mechanics

Hydrogenation of Cyclopentadiene Dimer

Introduction: The Diels-Alder Reaction

The Diels-Alder reaction is one of the most powerful reactions in organic synthesis, involving the reaction of a conjugated diene with a C-C multiple bond to yield an unsaturated 6-membered ring system. It is widely accepted that this occurs via a cycloaddition process with overlap of the pi-orbital systems of the two substrates. However, there is no single universal mechanism for the reaction. In addition to requiring very little energy to create a cyclic system, its selectivity renders the reaction as one of the most useful within organic chemistry. Here, the regioselective dimerisation of cyclopentadiene will be investigated via molecular mechanics to rationalise such principles.

Regioselectivity of Cyclopentadiene Dimerisation

Figure 1 - Dimerisation of cyclopentadiene

At room temperature, cyclopentadiene dimerises via a [4+2] cycloaddition to form dicyclopentadiene of which there are two possible forms: the exo- and the endo-isomer (Figure 1). However, it has been reported in literature that this reaction proceeds to form the endo-isomer exclusively. In an attempt to rationalise this observation, the two isomers were modelled on ChemBio3D and their structures optimised using Allinger's MM2 force-field. The calculated energy values for the two optimised structures are displayed in Table 1.

It is evident below that, energetically, the endo-isomer is the less stable form. Closer inspection of the energy breakdowns of the two isomers show that torsion has the most significant contribution to the overall energy, which is expected due to steric interactions between the bridgehead carbon (yellow) and the hydrogens on either side of the newly formed C-C bond. In the exo-form, the dihedral angle is found to be 155°, corresponding to an anti-periplanar arrangement that is favourable stereoelectronically. On the other hand, the endo-form produces a dihedral angle of 85°, implying the less favourable gauche conformation.

The fact that, despite being higher in energy, the endo-form is observed exclusively leads one to suggest that the dimerisation of cyclopentadiene is kinetically controlled at room temperature. If the temperature were to be increased, it is reasonable to predict the formation of the thermodynamically more stable exo-isomer as the system has more energy to surmount the activation barrier. However, as reactions under kinetic control strongly depend on the transition state, it is difficult to examine this using Allinger's MM2 force-field as the method is not suited for this purpose. It is speculated using Frontier Molecular Orbital Theory (FMO Theory) that preference for the endo-geometry arises from stabilising secondary orbital interactions that are present only in the endo-isomer.

Table 1 - Calculated MM2 energies of isomers of the cyclopentadiene dimer

Interaction Type	Energy contribution (kcal mol-1)		Energy difference (exo - endo) (kcal mol-1)
	exo-isomer (Dimer 1)	endo-isomer (Dimer 2)
Stretch	1.29	1.25	0.04
Bend	20.58	20.85	-0.27
Stretch-Bend	-0.84	-0.84	0.00
Torsion	7.66	9.51	-1.85
Non-1,4 VdW	-1.42	-1.54	0.12
1,4 VdW	4.23	4.32	-0.09
Dipole-Dipole	0.38	0.45	-0.07
Total Energy	31.88	34.00	-2.12

Hydrogenation of the Cyclopentadiene Dimer

Upon hydrogenation of the cyclopentadiene dimer, the reaction is found to initially give the dihydro derivative of the dimer before yielding the tetrahydro derivative. Given that the endo-product is the isomer exclusively formed at room temperature, there are two possible structures for the dihydro derivative: dimers 3 and 4 (Figure 2). The isomer that is most likely to form - assuming energetics for now - depends on the relative energies of the dimers, which can be predicted through the MM2 force-field.

Figure 2 - Hydrogenation of the endo-dimer (2)

Table 2 - Calculated MM2 energies of dihydro-derivatives of the cyclopentadiene dimer

Interaction Type	Energy contribution (kcal mol-1)			Energy difference (Dimer 3 - Dimer 4) (kcal mol-1)
	Dimer 3	Dimer 4	Dimer 5
Stretch	1.28	1.10	1.19	0.18
Bend	19.87	14.52	14.8	5.35
Stretch-Bend	-0.83	-0.55	-0.56	-0.28
Torsion	10.81	12.50	15.24	-1.69
Non-1,4 VdW	-1.23	-1.07	-0.65	-0.16
1,4 VdW	5.63	4.51	6.03	1.12
Dipole-Dipole	0.16	0.14	-	0.02
Total Energy	35.69	31.15	36.09	+4.54

As displayed in Table 2, Dimer 4 is energetically the more stable dihydro derivative, being lower in energy that Dimer 3. It is clear from the MM2 energy breakdown that the major contribution to the energy difference between the two isomers is from bending interactions. Upon closer inspection of the 3D models of the dimers (see Jmol annotations), it can be seen that the alkene sp2-carbon suffers significantly more strain in Dimer 3 than in Dimer 4, with a larger deviation from the ideal bond angle of 120°. In fact, the sp2 bond angle in Dimer 3 is characteristic of an sp3 carbon, highlighting the resultant strain on the structure. Since the bending interaction is the major contribution to stabilisation of Dimer 4 relative to Dimer 3, this idea can also be extended to the original Dimer 2 due to the similarity in energy between Dimers 2 and 3. Additionally, this energy similarity illustrates that the presence of an alkene in this position constitutes a significant proportion of the strain in the molecule, which is only relieved upon hydrogenation to give Dimer 4.

Though the MM2 method allows prediction of the thermodynamic product of the hydrogenation, it is difficult to predict whether this isomer would, in fact, be the predominant form without any kinetic data. It is possible that, like the endo-isomer, Dimer 3 may be the kinetic product and the preferred isomer for certain reaction conditions. However, in order to predict the kinetic product, analysis of the transition state of the reaction will need to be examined, which the MM2 force-field is not designed to model. For this to be done accurately, other computational methods may need to be employed.

For completeness, the MM2 energies of the tetrahydro derivative (Dimer 5) were also calculated to rationalise why this product is only formed after prolonged hydrogenation. The total energy is clearly higher than all the other dimers and, hence, requires harsher conditions to favour the hydrogenation. Torsion is most likely the barrier to reaction due to the evident unfavourable steric interactions between all the hydrogen atoms along the carbon skeleton.

Stereochemistry and Reactivity of an Intermediate in the Synthesis of Taxol

An Introduction to Atropisomerism

A form of stereoisomerism, atropisomerism arises from restricted rotation about a single covalent bond, normally as a result of high torsional energy and steric demand. Where the steric strain barrier to rotation is sufficiently high to impede rotation, isolation of the different conformers becomes possible, but is less facile with increasing temperature. Though first observed in biaryls, this phenomenon is prevalent in macrocyclic structures where rotation about substituted aromatic rings is restricted . Atropisomers form an important class of compounds as they display axial chirality that can equilibrate thermally, unlike other forms of chirality that isomerise chemically. This has significant consequences in drug molecules where the stereochemistry of the active species is vital.

Figure 3 - Atropisomers of a key intermediate in the Paquette's total synthesis of Taxol

One example can be observed in a key intermediate in the total synthesis of the anti-cancer drug, Taxol, proposed by Paquette in 1991. At room temperature, it was found that interconversion occurs between the two forms, with the predominant isomer being the more thermodynamically stable. Using molecular mechanics, the stereochemistry of the major form and the factors that affect this can be predicted.

Stereochemistry of an Intermediate in the Synthesis of Taxol

Upon closer inspection of the structure of the intermediate, it can be seen that each atropisomer can exist in one of two forms due to the presence of cyclohexane: it can exist in either the chair or twist-boat conformation. As the chair conformation is known to be the more energetically stable form, it is expected that the chair forms of each atropisomer will be lower in energy than the corresponding boat forms. This is indeed what is found when the MM2 and MMFF94 energies of the 4 conformations are calculated. It should be noted that the energy values quoted below are based on structures that have been manipulated both manually and by ChemBio3D to produce energy minima.

Table 3 - Calculated MM2 and MMFF94 Energies of Intermediates 9 (chair and twist-boat) and 10 (chair and twist-boat)

Interaction Type	Energy contribution (kcal mol-1)				Difference (9) - (10) (kcal mol-1)
	9 (syn-carbonyl, chair)	9' (syn-carbonyl, twist-boat)	10 (anti-carbonyl, chair)	10' (anti-carbonyl, twist-boat)
Stretch	2.78	2.83	2.62	2.83	0.16
Bend	16.54	11.60	11.34	12.98	5.20
Stretch-Bend	0.43	0.35	0.34	0.39	0.09
Torsion	18.25	25.05	19.67	22.97	-1.42
Non-1,4 VDW	-1.55	-1.59	-2.16	-1.75	0.61
1,4 VDW	13.11	14.81	12.87	14.12	0.24
Dipole-Dipole	-1.72	-1.66	-2.00	-2.03	0.28
Total MM2 Energy	47.84	51.39	42.68	49.51	+5.16
Total MMFF94 Energy	70.53	71.14	60.54	68.87	+9.99

As evident in Table 3, intermediate 10 is shown to be energetically the most stable isomer (by both computational methods), which can be rationalised upon closer inspection of the structures of atropisomers 9 and 10. Firstly, intermediate 9 suffers from unfavourable steric interactions between the carbonyl oxygen and the bridgehead carbon - given their proximity - that is relieved with bond rotation to give intermediate 10. The distance between the two atoms (as shown in the 3D models) increases from 3.23 Å to 4.19 Å going from 9 to 10 and, with this, results in a decrease in repulsive Van der Waals' interactions and a smaller 1,4 VdW contribution to the total energy. It can be argued that given the proximity and relative orientation of the carbonyl to the bridgehead hydrogens, intermediate 9 is stabilised by intramolecular hydrogen bonding. However, due the strained nature of the macrocyclic system, this stabilising effect is expected to be minimal and overridden by the Van der Waals' interactions. Secondly, it can be seen that the largest contribution to the energy difference between atropisomers 9 and 10 is a bending interaction that can be viewed as a result of the deviations from the ideal bond angle on the sp² carbonyl carbon. Intermediate 9 yields a bond angle of 123° which is further away from the ideal sp² angle of 120° than the bond angle of 118° in intermediate 10. Thus, in addition to the effects explained previously, it can be said that stereochemistry of atropisomer 10 is the predominant form of the intermediate in the total synthesis of Taxol.

Hyperstable Alkenes

It has been previously reported that functionalisation of intermediate 10 occurs unexpectedly slowly. In an attempt to rationalise this observation, the energy of the hydrogenated product (10*) was calculated using the same methods (MM2 and MMFF94) and compared to the unsaturated intermediate.

Table 4 - Calculated MM2 and MMFF94 Energies of Intermediates 10 (anti-carbonyl, chair) and 10* (hydrogenated, chair)

Interaction Type	Energy contribution (kcal mol-1)
	10 (anti-carbonyl, chair)	10* (hydrogenated, chair)	Difference (10 - 10*) (kcal mol-1)
Stretch	2.62	2.85	-0.23
Bend	11.34	14.23	-2.89
Stretch-Bend	0.34	0.67	-0.33
Torsion	19.67	22.12	-2.45
Non-1,4 VDW	-2.16	-2.67	0.51
1,4 VDW	12.87	15.89	-3.02
Dipole/Dipole	-2.00	-1.73	-0.27
Total MM2 Energy	42.68	51.36	-8.68
Total MMFF94 Energy	66.13	72.30	-6.17

It is apparent in Table 4 that the hydrogenated product is, in fact, higher in energy than the alkene precursor, illustrating the unusual unreactivity of the unsaturated intermediate. The energy difference is largely due to bending interactions as well as 1,4 VdW interactions, which is expected upon inspection of the 3D models of both species. In intermediate 10, the alkene sp² bond angle is reasonably close to the ideal bond angle of 120°. However, if this is compared to the (now) alkane sp³ bond angle of 116°, the bond is clearly highly strained as the angle is now approaching the angle characteristic of an sp² bond, rather than the expected 109°. Consequently, the deviation from ideality introduces more 1,4 VdW interactions with this change in hybridisation, bringing more hydrogen atoms closer together into the repulsive VdW region. This observation is typical of a phenomenon reported in literature as hyperstable olefins.

Modelling using Semi-Empirical Molecular Orbital Theory

Regioselective Addition of Dichlorocarbene

Using the semi-empirical MOPAC method, the carbenylation of compound 12 with dichlorocarbene was modelled to illustrate the influence of electronic effects - previously disregarded by the MM2 method - on the reactivity of a system. More specifically, the orbitals of the compound will be examined and used to rationalise its observed chemical behaviour. It is known that the carbenylation is regioselective in that it adds onto the syn-alkene of the starting substrate as shown in Figure 4.

Figure 4 - Carbenylation of compound 12 with dichlorocarbene

Figure 5 - σh plane of Cs compound 12

Figure 6 - HOMO calculated using MOPAC/PM6

In order to gain an insight into the electronics of the system, the structure of compound 12 was minimised using the MOPAC/PM6 method and the molecular orbitals (MOs) calculated. As the compound belongs to the point group C_s, it is expected to possess molecular orbitals that are symmetric along the σ_h plane of symmetry (Figure 5). However, upon calculation of the HOMO, it did not produce orbitals with the expected symmetry (Figure 6), suggesting a bug in the programme. Thus, MOPAC/AM1 was used instead to obtain the MOs for compound 12, some of which are displayed below.

Table 5 - Calculated MOPAC/AM1 Molecular Orbitals for Compound 12

Figure 7.1 - LUMO+2	Figure 7.2 - LUMO+1	Figure 7.3 - LUMO
Figure 7.4 - HOMO	Figure 7.5 - HOMO-1	Figure 7.6 - HOMO-2

As illustrated above, the calculated electron density on the HOMO suggests that the two alkenes are inequivalent with the more nucleophilic of the two being the syn-alkene, as predicted. This is in agreement with the observation that dichlorocarbene, being an electrophile, selectively adds to this alkene. The difference in alkene nucleophilicity may be attributed to the interaction of the LUMO+2 orbital, which resembles the C-Cl σ^* orbital, with the pi-system of the anti-alkene. It is possible that electron donation occurs from the alkene into the antibonding C-Cl orbital, reducing the localised C=C electron density and, as a result, the nucleophilicity of this alkene. Additionally, given the electronegativity of the chlorine atom, the energy of the antibonding orbital is lowered and is likely to participate in a stabilising 4-electron interaction with the alkene.

While the regioselectivity of the reaction can be linked to the calculated molecular orbitals of the compound, the stereoselectivity can be explained using sterics: due to the size of the chlorine atoms, dichlorocarbene is likely to attack underneath the molecule to avoid unfavourable steric interactions.

Given that IR spectroscopy is influenced by the electron density within a certain bond, the possibility of the interaction between the antibonding C-Cl bond with the anti-alkene can be investigated using this technique. If compound 12 was to be hydrogenated at the anti-alkene, the pi-electron system of the alkene would be lost and so will the interaction with the C-Cl antibonding orbital. Thus, it can be predicted that, if the interaction occurs, the C-Cl vibration frequencies should higher for the hydrogenated product, given that there is no donation into an antibonding orbital which would inevitably weaken the C-Cl bond. Figures 8 and 9 below display the calculated IR spectra for both the starting and hydrogenated compound.

Figure 8 - IR spectrum of compound 12

Figure 9 - IR spectrum of the monohydrogenated compound

Table 6 - Relevant IR Vibration Frequencies for Compound 12 and its Monohydrogenated Derivative

Vibration Assignment	Vibration Frequency (cm-1)
	Compound 12	Monohydrogenated Derivative
C-Cl stretch	772	780
C=C stretch (syn-)	1761	1754
C=C stretch (anti-)	1741	N/A

The fact that the C-Cl stretching frequency is higher in the monohydrogenated compound supports the possibility of electronic interactions between the antibonding C-Cl bond and the anti-alkene in compound 12. In addition, the lower C=C stretching for the anti-alkene suggests reduced electron density from the alkene, relative to that in the syn-alkene, further illustrating occurence of electronic effects.

Monosaccharide Chemistry: Glycosidation

Figure 10 - Glycosidation Reaction Scheme

In an attempt to quantitatively study the effect of neighbouring group participation, MM2 and MOPAC/PM6 methods were used to calculate the respective energies of molecules A-D, as shown in the figure above. In addition, it can be seen that, for each molecule, the acetate group can approach the anomeric carbon from either the top or bottom face of the ring, resulting in 4 additional conformations of the glycoside. For example, in A, the acetate group approaches from the bottom face. However, in A*, the structure is modelled such that the acetate approaches from the top face. It must be noted, however, given the conformational strain that arises from approaching the anomeric carbon via the opposite face of the ring, these species (A*-D*) are likely to be high in energy and, thus, not observed.

Note: methyl groups were chosen as the R groups on the substituents of the ring due to the fact that it simplifies calculations by elimination of the possibility of hydrogen bonding - as would be the case if hydrogen atoms were chosen - and reducing the number of possible conformations, given its size.

Table 7 - Calculated MM2 and MOPAC/PM6 Energies of Molecules A, A*, B, B*, C, C*, D and D*

Interaction Type	Energy contribution (kcal mol-1)
	A	A*	B	B*	C	C*	D	D*
Stretch	2.41	2.24	3.05	2.41	2.00	2.57	1.71	2.57
Bend	8.61	9.00	13.00	9.35	13.65	17.93	16.60	21.60
Stretch-Bend	0.80	0.80	1.07	0.83	0.73	0.65	0.65	0.60
Torsion	1.74	1.40	1.85	2.17	10.36	6.22	7.00	5.04
Non-1,4 VDW	-1.38	-3.28	4.34	-2.74	-3.18	-4.24	-3.65	-1.92
1,4 VDW	19.63	19.27	18.72	19.45	18.19	19.19	16.14	17.06
Charge-Dipole	-16.69	-0.76	-37.72	-4.41	-7.27	-3.29	-2.83	-4.13
Dipole-Dipole	6.54	3.99	6.92	3.88	-2.37	-1.16	-3.51	-4.12
Total MM2 Energy	21.69	32.65	11.23	30.94	32.13	44.46	32.11	36.70
Total MOPAC/PM6 Energy	-85.75	-69.13	-83.41	-64.78	-91.65	-66.85	-88.73	-67.02

It is clear from Table 7 that both MM2 and MOPAC/PM6 methods support the idea that approach of the anomeric carbon from the opposite face of the ring results in higher energy species (A*-D*). Using the PM6 3D models for molecules A and A* to explain this, it can be seen that approach of the anomeric centre in A results in attack along the optimal Burgi-Dunitz trajectory, with an angle of attack of 104.8°. This is in contrast to the angle of attack in A* of 158.3°, which is far from the Burgi-Dunitz angle of 105°. The same trend is applicable in the case of molecules B and B* and explains the diastereoselectivity of the reaction, avoiding mixtures of products given a starting sugar with a set stereochemistry.

However, while both MOPAC/PM6 and MM2 are consistent in illustrating that molecules A-D are lower in energy than their corresponding A*-D* species, the two methods differ in predicting the distereoselectivity of the reaction i.e. whether axial or equatorial attack is favoured. Between A and B, MM2 suggests that the product formed from axial attack is thermodynamically favoured while MOPAC/PM6 suggests otherwise. Thus, it must be remembered that both methods take into account different parameters when calculating the energies of molecules. As MM2 is designed mainly for hydrocarbons, disregarding electronic interactions between atoms, it is useful as a starting point in highlighting steric effects within a molecule. However, given that the anomeric effect and neighbouring group participation are electronic interactions that are vital to glycosidation, MOPAC/PM6 results are most likely to give more accurate representations of the system.

By comparison of the energies of A with C and B with D, it can be seen that formation of the bond between the acyl oxygen and the anomeric centre to form the intermediate reduces the overall energy of the system. This is expected as the positive charge on the oxygen in the ring can be delocalised and extend onto the adjacent acetate group upon formation of the intermediate. This stabilisation is enough to override the strain that may arise due to the formation of a bicyclic system, as shown by the larger bond angles in C and D compared to A and B. Additionally, comparison of the energies of A and C to the energies of B and D seems to suggest that attack by an equatorial acetate group is more favourable than attack by an axial group. This may be due steric effects, given the size of the acetate groups favouring the equatorial positions. However, the energy difference between the axial and equatorial groups is not significant enough to hinder the neighbouring group effect from occuring for one diastereoisomer.

Mini-project: Regioselective Functionalisation of N-Benzenesulfonyl-3-bromopyrrole

The functionalisation of pyrroles is a useful reaction in the synthesis of many biologically active molecules. Directed lithiation of N-protected pyrroles and reaction with the desired electrophile is a known method of forming 2-substituted pyrroles. However, the C2 versus the C5 regioselectivity of such functionalisations has not been extensively investigated. Here, Fukuda, Ohta, Sudo and Iwao propose conditions for the lithiation of N-benzenesulfonyl-3-bromopyrrole that illustrates such regioselectivity and is employed as a key reaction in the synthesis of the antitumour marine alkaloid lamellarin D.

Figure 11 - Reaction scheme for the regioselective functionalisation of N-benzenesulfonyl-3-bromopyrrole

Analysis of Energies: Thermodynamic vs Kinetic Product

It is important to note the difference in the product distribution with changes in the reaction conditions used, as illustrated in the Figure above. Treatment of the 5-lithiated species with the first condition above yields the C5-substituted pyrrole (5a) as the predominant product. However, upon warming the reaction mixture to 0°C for 30 minutes, the C2-substituted pyrrole (4a) is formed. These observations suggest that 5a is formed as the kinetic product of the reaction while 4a is the thermodynamic product. It is speculated by Fukuda et. al that the functionalisation proceeds via a dynamic equilibrium between the 2- and 5-lithio species, allowing the interconversion of the 2- and 5-substituted products upon warming of the system. In an attempt to rationalise this reactivity, the energies of the two regioisomers, as well as the 2- and 5-lithiated species, were calculated through MM2 and MOPAC/PM6 methods, with the results displayed below.

Table 8 - Calculated MM2 and MOPAC/PM6 Energies of the 2- and 5-lithiated species and isomers 4a and 5a

Interaction Type	Energy contribution (kcal mol-1)				Difference (4a) - (5a) (kcal mol-1)
	2- lithio species	5- lithio species	Isomer 4a (2-substituted pyrrole)	Isomer 5a (5-substituted pyrrole)
Stretch	1.45	1.63	2.07	1.88	0.19
Bend	143.32	144.04	143.70	144.63	-0.93
Stretch-Bend	-0.76	-0.64	-0.49	-0.62	0.14
Torsion	36.62	36.78	37.35	35.84	1.51
Non-1,4 VDW	-2.53	-2.19	-1.31	-2.47	1.16
1,4 VDW	13.98	16.18	13.77	13.11	0.66
Dipole-Dipole	8.39	8.38	12.10	13.94	-1.84
Total MM2 Energy	200.48	204.18	207.17	206.30	+0.87
Total MOPAC/PM6 Energy	-20.07	-17.45	-99.76	-95.24	-4.52

As shown above, the MM2 energies for the two regioisomers do not support the claim that 4a is the thermodynamic product of the reaction, with 5a being lower in energy in both calculations. Closer inspection of the breakdown of the MM2 energies shows that, despite the stabilisation gained through dipole-dipole interactions around the ring, torsion is the main contribution to the positive energy difference between isomer 4a and 5a. This might be understood by considering the steric interactions that occur between the adjacent substituents in 4a relative to 5a, as displayed in the 3D models.

It must be noted, however, that the total MOPAC/PM6 energies of the two regioisomers suggest the opposite trend and do, in fact, support Fukuda et. al's claim of 4a being formed as the thermodynamic product. Not only is 4a lower in energy in this model, the energy difference between the two forms is also significantly larger than that given in the MM2 method. As shown in the 3D models, it can be seen that the minimised structures using the two methods are, in fact, different and may explain the discrepancy between the calculated values. As MM2 is not particularly suited for analysing non-classical species, such as aromatic systems, it is possible that, being a semi-empirical method, PM6 is more applicable to the system in this experiment and, thus, gives a more accurate prediction of the energetics of the products. Thus, it can be suggested that the additional stabilisation may be attributed to electrostatic effects, given the proximity of the electron-withdrawing substituents on the pyrrole ring in 4a relative to 5b.

¹³C NMR Spectroscopy

Due to the difference between the two isomers being the position of functionalisation on the the pyrrole ring, one technique that may be used to characterise the products of the reaction is ¹³C NMR. Given the electron-withdrawing nature of the pyrrole substituents, different arrangements around the ring will affect the electronics and, as a result, will lead to different chemical shifts for the carbons in the ring. It can be expected that, of the two isomers, 4a will result in slightly higher chemical shifts due to the proximity of the substituents increasing the deshielding effect on the attached carbons. Thus, the ¹³C NMR of the two isomers were predicted using Gaussian on the optimised structures of the isomers and compared with those reported in literature.

Figure 12 - 13C NMR spectrum for isomer 4a

Figure 13 - 13C NMR spectrum for isomer 5a

Figure 14 - 13C NMR assignments for isomers 4a and 5a

Table 9 - ¹³C NMR data for isomers 4a and 5a; Reference = TMS mPW1PW91/6-31G(d,p) CDCl3 GIAO

Carbon Assignment	Isomer 4a			Carbon Assignment	Isomer 5a
	Peak Chemical Shift (δ / ppm)				Peak Chemical Shift (δ / ppm)
	Calculated value	Experimental value	Difference (experimental - calculated)		Calculated Value	Experimental value	Difference (experimental - calculated)
19	51.0	52.1	1.1	19	51.5	52.1	0.6
3	110.7	109.7	-1.0	2	110.3	99.1	-11.2
1	119.7	115.3	-4.4	3	118.8	124.8	6.0
4	123.5	123.3	-0.2	4	121.4	125.2	3.8
12, 14*	124.5	126.8	2.3	1**	129.7	127.8	-1.9
2	125.3	127.9	2.6	12, 14*	126.2	128.4	2.2
11, 15*	127.2	129.0	1.8	11, 15*	126.9	129.0	2.1
13	131.2	134.1	2.9	13	132.9	134.2	1.3
10	137.3	138.7	1.4	10	134.8	138.2	3.4
16**	159.8	159.3	-0.5	16**	159.3	158.2	-1.1

Note: the chemical shifts of the starred (*) entries above are those that have been averaged to account for the fact that the spectrum is obtained for a solvated molecule at room temperature for the experimental system. In this case, the molecule is free to rotate, rendering certain carbon atoms as being equivalent. Here, the readings for the ortho and for the meta carbons on the phenyl ring are averaged to account for this rotation and give a single shift. This is a reasonable correction as this produces a spectrum that agrees with the experimental data. The double starred (**) entries are those that have been corrected due to carbons being connected to 'heavy' atoms.

As displayed in the table above, there is very good agreement between the calculated and experimentally obtained NMR data for both isomers, with deviations of approximately only 3 ppm. This supports the proposed structures of the products formed in the reactions and, hence, validates the product distribution reported in the literature. While most of the carbon environments appear to produce very similar chemical shifts between the two isomers, one point to note is the difference in the shift for the C3 carbon attached to bromine on the pyrrole (carbon assignment 1). In 4a, it is significantly more deshielded due to the proximity of the electron-withdrawing groups on the ring. In 5a, the ester group is located on the other side of the ring and, hence, interacts less strongly with the electronics of the bromine and sulfonyl group on the pyrrole.

IR Spectroscopy

Like ¹³C NMR, IR spectroscopy is another relevant technique for verifying the identities of products 4a and 5a produced in the functionalisation of the pyrrole derivative. Again, a predicted IR spectrum is calculated for the two isomers using Gaussian, based on their optimised structures.

Figure 15 - IR spectrum for isomer 4a

Figure 16 - IR spectrum for isomer 5a

Table 10 - IR absorption frequencies for isomers 4a and 5a

Vibration Assignment	Absorption frequencies (cm-1)
	Isomer 4a			Isomer 5a
	Calculated value	Experimental value	Difference	Calculated value	Experimental value	Difference
C=O stretch	1787 (1716)	1720	-67 (4)	1808 (1736)	1729	-79 (-7)
C-H asym. bend	1446	1452	6	1487	1449	-38
C-H sym. bend	1364	1360	-4	1367	1386	19
C-N stretch	1294 (1242)	1257	-37 (15)	1241 (1191)	1215	-26 (24)
C-H asym. in-plane bend	1156	1174	18	1177	1188	11
C-H sym. in-plane bend	1152	1144	-8	1149	1141	-8

Note: the values in brackets are the frequencies after assuming 8% error.

As shown by the difference between the experimental and calculated values, it is known that the errors in the predicted wavenumbers for stretches are generally too high, resulting in relatively large discrepancies between the two frequencies. Thus, after estimating the error of such frequencies to be approximately 8%, it seems that the predicted frequencies of the isomers agree with those reported in literature. However, despite such agreement, it is difficult to conclusively differentiate the two isomers based on IR spectroscopic data since both compounds contain the same functionalities and this is reflected in the similarity between the two spectra. Hence, while useful in confirming the successful synthesis of the compounds, IR spectroscopy does not provide sufficient information about structure of the molecule to fully determine the regioselectivity of a reaction.