Solution Conformations of Chiral Mannich Bases A Senior Honors ...

1

Solution Conformations of Chiral Mannich Bases

A Senior Honors Thesis Submitted to the Department of Chemistry at Northwestern University in Partial Fulfillment of the Requirements of the Weinberg College of Arts and Sciences Bachelor of Arts Degree in Chemistry with Honors

Paul Bumki Lee

Faculty Advisors: Professors Frederick Northrup and Owen Priest Northwestern University 2145 Sheridan Road, Evanston, IL 60208

April, 2015

2

Acknowledgements I am forever grateful to many individuals who have helped me to develop and grow so much during my four years at Northwestern University. I would first like to thank Daniel Shapiro and Dr. Kimberly Gray for providing me the first opportunity to pursue chemistry research and for their trust in my ability to become an independent researcher. I would also like to thank JC Jacobs as this project originated from his idea of studying the conformations of chiral Mannich bases using variable temperature 1H NMR spectroscopy. His guidance and support from the very beginning helped me to transition smoothly into this project. I would also like to thank Dr. Wu and Dr. Zhang of IMSERC for training me with the NMR instruments and generously guiding me on ways to improve my methods of NMR data collection. I am extremely grateful to Dr. Northrup and Dr. Priest for the countless amounts of time and care that they provided to help me reach the highest level of research possible. The amount of learning that I gained under their guidance was unlike any other moments of my education, and I hope to build on what I learned from them to become a more skilled and knowledgeable researcher. I finally thank my parents for their never-ending love and support and I would like to dedicate this thesis paper and all of my accomplishments at Northwestern to them.

3

Table of Contents Acknowledgements (2) Table of Contents (3) Abstract (4) Chapter1: Introduction (5) Chapter 2: Experimental Procedure (8) 2-1: Synthesis (8) 2-2: NMR Spectroscopy(10) Chapter 3: Room Temperature NMR Spectroscopy (11) 3-1: Chemical Shift Differences (13) 3-1-a. 1-R and 1-S Benzyl Products in CD3OD (13) 3-1-b. 2-R and 2-S Naphtyl Products in CD3OD (14) 3-1-c. 1-R and 1-S Benzyl Products in D2O (15) 3-2: Differences in Separation of Diastereotopic CH2 Doublets (17) 3-3: Line-Width Effects (18) 3-3-a. 1-R and 1-S Benzyl Products in CDCl3 (18) 3-3-b. 2-R and 2-S Naphthyl Products in CDCl3 (20) 3-3-c. 1-R and 1-S Benzyl Products in CD2Cl2 (20) 3-4: OH-NH (21) 3-4-a. Enantiomeric Pairs of Benzyl and Naphthyl Products in CDCl 3 (21) 3-4-b. 1-R and 1-S Benzyl Products in Other Solvents (22) 3-4: Summary (24) Chapter 4: Variable Temperature NMR Spectroscopy (26) 4-1: Variable Temperature NMR Spectroscopy in CDCl3 (26) 4-1-a: 1-R and 1-S Benzyl Products (26) 4-1-b: 2-R ad 2-S Naphthyl Products (32) 4-2: Variable Temperature NMR Spectroscopy in CD2 Cl2 (38) 4-2-a: 1-R and 1-S Benzyl Products (38) 4-3: Variable Temperature NMR Spectroscopy in CD3OD (43) 4-3-a: 1-R and 1-S Benzyl Products (43) 4-3-b: 2-R ad 2-S Naphthyl Products (47) 4-4: 1-R ad 1-S Benzyl Products in D2O (51) 4-4-a: 1-R and 1-S Benzyl Products (51) Chapter 5: Summary and Future Work (56) References (60)

4

Abstract Molecular conformation can have significant effects on the properties of a molecule such as its solubility and reactivity. Variable temperature 1H NMR spectroscopy was used to study the solution conformations of two enantiomeric pairs of chiral Mannich bases: N-(2hydroxylbenzyl)-methylbenzylamine (1-R and 1-S) and N-(2-hydroxylbenzyl)naphthylethylamine (2-R and 2-S). N H HO

N H HO 1-S

1-R

N H HO 2-R

N H HO 2-S

These molecules are capable of intramolecular hydrogen bonding under appropriate temperature and solvent conditions. We observed significant chemical shift and line-width changes in the NMR spectra of these molecules as a function of different solvents (CD2 Cl2, DMSO-d6 , DMF-d7, CDCl3, CD 3OD, and D2O), temperature (a range from -50°C to 130°C, depending on the freezing and boiling points of each solvent), and most surprisingly on whether the molecule was the R- or S-enantiomer. These NMR spectrum changes can be attributed to conformational changes and molecular bond rotation that can be hindered by intramolecular hydrogen bonding. The NMR spectrum changes were more extreme for the heavier naphthyl compound consistent with more hindered rotation. Preliminary computational chemistry results suggest different molecular conformations for the R- and S-enantiomers in some solvents. It is possible that these results may lead to a method of determining the absolute configuration of chiral amines.

5

Chapter 1: Introduction Conformational change in solution is the dynamic process in which a molecule rotates its bond and changes its shape in response to its environment.1 Molecular conformation can have significant effect on molecular reactivity and chemical properties.2,3 For instance, drug inhibitors that undergo intramolecular hydrogen bonding display increased membrane permeability as a result of hindered rotation.3 Several factors can affect molecular conformation. In solution, the solvent forms a cage around a solute that results from favorable interactions.1 The choice of solvent can play large roles in a molecule’s preferred conformation.2 These intramolecular and intermolecular interactions are certain to be temperature dependent and may show the greatest effect on molecular conformation at lower temperature where less energy is available for molecular bond rotation.4 Mannich bases are compounds capable of intramolecular hydrogen bonding between amino and hydroxyl groups forming a six-membered ring under appropriate solvent and temperature conditions.5 Previous variable temperature studies with Mannich bases such as 2diethylaminomethyl-3,4,6-trichlorophenol (Figure 1) showed changes in 1H NMR chemical shifts and line-widths of the methylene ethyl group and OH-NH signals with decreasing temperature as a result of greater actively hindered inversion at the N-atom caused by intramolecular hydrogen bonding.4,5

H O

...

N

Cl

Cl Cl

Figure 1: Structure of 2-diethylaminomethyl-3,4,6-trichlorophenol

6

Stereochemistry may also have significant effect on molecular conformation that can be relevant to drug design and other applications. Chiral drugs are often chiral amines or amide molecules with two non-superimposable R- and S- enantiomers that can cause differences in the chemical and pharmacological effects in living systems, which are also chiral.6 A tragic example of the use of an incorrect enantiomer of a drug was thalidomide, which was marketed during the 1960’s as a treatment for morning sickness for pregnant women.7 This drug was made as a racemic mixture of its R- and S- enantiomers, which are shown below in Figure 2. O

O

O

N

O

N

NH

NH O

O

O

O

Figure 2: Structures of S- and R-thalidomide

It was found that the R-thalidomide was the effective drug while the S-thalidomide caused serious birth defects.7 Therefore, it is important to be able to determine the absolute configuration of chiral amines to ensure that the correct enantiomer of a drug is administered. One of the most common methods for assigning the absolute configuration is to compare the 1H NMR spectra of two diastereomeric derivatives prepared from a substrate of unknown configuration and two enantiomers of a chiral dervatizing agent.8 Mosher’s method involves the reaction of an optically-active amine with both R- and S- (trifluoromethyl)phenylacetylchloride (MTPA-Cl) to form a pair of diastereomeric amides that can be distinguished by NMR spectroscopy in CDCl3 solvent at room temperature.9 Analysis involves observing differences in the chemical shift values of the protons that are close to the stereocenter.9 Based on the Mosher model9 , which is shown below for the S- and R-amide, the protons in the L3 group are shifted

7

upfield for the S-amide relative to the ones for the R-amide because these protons are closer in proximity to the anisotropic shielding caused by the phenyl ring.

Figure 3: Mosher Model15

This shows that preferred conformation of the diastereomers can indicate the absolute stereochemistry of the amine reactant. However, the time-consuming nature of preparing two diasteroemers of a Mosher amide and the high costs of MTPA-Cl are major limitations of this method. While a method has been developed in which a single chiral derivatizing agent derivative is prepared and then NMR spectra of it is taken at room and at lower temperature,8 the large amount of resources and steps required to synthesize these diastereomers present difficulties in the efficiency of determining the absolute configuration of the starting amines. It is thus necessary to develop greater understanding of solvent and temperature effects to search for other methods of determining the stereochemistry of a chiral amine molecule Molecular conformation in solution is dependent on the stereochemistry, solvent, and temperature effects. We propose to expand this work by comparing variable temperature 1H NMR spectra of enantiomeric pairs of N-(2-hydroxylbenzyl)-methylbenzylamine and N-(2hydroxylbenzyl)-naphthylethylamine under a variety of solvent and temperature conditions to study the conformational changes of these chiral Mannich bases in solution.

8

Chapter 2: Experimental Procedure This project has studied NMR spectroscopy of Mannich base compounds and their salts:

R

N H

R= R= X X

,

X= H, OH, OMe, F, Cl, CF CF3 ,NO 22 X=

2-1 Synthesis Much of this work has focused on the enantiomeric pairs of N-(2-hydroxylbenzyl)methylbenzylamine (1-R and 1-S) and N-(2-hydroxylbenzyl)-naphthylethylamine (2-R and 2-S).

N H HO

N H HO 1-S

1-R

N H HO 2-R

N H HO 2-S

Initially, these compounds were synthesized by reacting the chiral (R and S) benzyl amine (3 and 4) with the substituted benzaldehyde (5) to make the aryl aldimines (6) as outlined by Bennett et. al.10 The amine reactant (10 mmol) was dissolved in 5 mL 80% ethyl L-lactate in water followed by addition of 10 mmol of 2-hydroxybenzaldehyde to form the imine. The imine was then reduced to the secondary amine (7) using sodium borohydride.

9

O NaBH4

NH2 + H HO 3 or 4

N H HO

N

Ethyl (-)-L-lactate

HO

5

8 or 9

6 or 7

There were some complications with this reaction scheme that made it difficult to synthesize the desired products in high purity. Sodium borohydride is a strong reducing agent that can reduce the aldehyde reactant before the aldehyde can form the imine intermediate. The difficulty in knowing when to add the sodium borohydride in the mixture led to formation of side-products. In addition, ethyl L-lactate’s high boiling point of 154°C10 caused complications in removing it from the product. While extraction of the insoluble salt was easy, the amine products needed to be purified by column chromatography. A better reaction scheme was needed to ensure the selective reduction of the imine intermediate with fewer steps for purification. An alternative reduction method developed by Carlson, et. al11 was used to overcome the limitations of the above procedure. The chiral amine (3 mmol) and aldehyde reactants (3 mmol) were dissolved in 20 mL dichloromethane (DCM), a solvent with a much lower boiling point, and 2 equivalents of a weaker reducing agent, sodium triacetoxyborohydride, were added to selectively reduce the imine to the secondary amine. O Na (AcO)3BH4

NH2 + H

Dichloromethane

HO 3 or 4

5

N H HO 8 or 9

10

After neutralizing the reducing agent with 25 mL of 5% sodium bicarbonate aqueous solution, the mixture was washed with water in a separatory funnel three times before using a rotovap to evaporate the solvent. A vacuum pump was used to dry the amines further before collecting 1H NMR spectra of the product. If analysis of the corresponding amine salt was desired, the amine could be dissolved in diethyl ether and then protonated to the corresponding HCl salts by exposure to HCl gas, which was generated by the addition of sulfuric acid to sodium chloride in a separate flask. All of these syntheses were time-efficient, easily carried out, and done with commercially available chemicals.

N H HO 8 or 9

HCl (g) Diethyl Ether

N H2 HO 9 or 10

2-2 NMR Spectroscopy 1

H NMR spectra of the chiral amines and their corresponding salts were collected with an

Agilent DD2 500 MHz NMR Spectrometer in the Integrated Molecular Structure Education and Research Center (IMSERC). Room temperature spectra in CDCl3 were used to confirm synthesis of the deisred product, but spectra also were collected under a variety of solvent and temperature conditions to study changes in the solution conformations of these molecules. A closed-cycle refrigeration unit and appropiate 1H NMR probe allowed temperature variation from -50 °C to 130 °C depending on the solvent used. Solvents and the temperature ranges of the solvents that were studied were benzene-d6 (26°C), CD2 Cl2 (-50°C to 40°C), DMSO-d6 (26°C to 130°C), CDCl3 (-50°C to 40°C), DMF- d7 (-50°C to 130°C), CD3OD (-50°C to 40°C), and D2O (15°C to

11

80°C). All spectra were manually re-shimmed at each temperature and referenced to TMS for chemical shift and linewidth. Chapter 3: Room Temperature 1H NMR Spectroscopy The room temperature spectra in CDCl3 for the benzyl and naphthyl R-enantiomers were assigned to confirm the successful synthesis of the chiral Mannich bases as shown in Figures 4 and 5. The diastereotopic methylene protons are substituents of a carbon that has close proximity to a stereocenter and are thus not equivalent to one another. They are observed as two doublets with different chemical shifts in the NMR spectrum.

k c a b a

i, j h N H c HO

e g d f

1-R

Figure 4: 1H NMR Spectrum of 1-R in CDCl3 at Room Temperature

1-R 1H NMR (500 MHz, CDCl3): δ 7.39-7.35 (m, 2H, a), 7.32-7.29 (m, 1H, b), 7.29-7.25 (m, 2H, c), 7.17-7.14 (m, 1H, d), 6.89-6.87 (m, 1H, e), 6.85-6.84 (m, 1H, f), 6.76-6.73 (m, 1H, g), 3.99 (q, J=6.7 Hz, 1H, h), 3.89 (d, J=13.8 Hz, 1H, i), 3.75 (d, J=13.8 Hz, 1H, j), 1.54 (d, J=6.7 Hz, 3H, k).

12

o a

d

f e c

b

m,n l N H g HO

i j h k

2-R

Figure 5: 1H NMR Spectrum of 1-R in CDCl3 at Room Temperature

2-R 1H NMR (500 MHz, CDCl3): δ 7.88-7.86 (m, 1H, a), 7.86-7.81 (m, 2H, b, c), 7.69-7.67 (m, 1H, d), 7.52-7.45 (m, 2H, e, f), 7.45-7.41 (m, 1H, g), 7.18-7.14 (m, 1H, h), 6.87-6.85 (m, 2H, i, j), 6.76-6.72 (m, 1H, k), 3.99 (q, J=6.7 Hz, 1H, l), 3.76 (d, J=13.8 Hz, 1H, m), 3.75 (d, J=13.8 Hz, 1H, n), 1.55 (d, J=6.7 Hz, 3H, o).

Although it is expected that enantiomers should have the same NMR spectra because of their identical chemical and physical properties,7 there were significant differences in the 1H NMR chemical shifts and line-widths for the 1,2-R and 1,2-S benzyl and naphthyl compounds as a function of different solvents and temperatures. Observed differences included changes in the chemical shifts of the alkyl and some aromatic protons with chirality, changes in the chemical shift differences for the diastereotopic methylene protons with chirality, differences in the linewidths of some proton signals with chirality, and changes in chemical shift and line-width for the OH and NH signals. These results will be discussed first for room temperature spectra in different solvents followed by discussion of how such observations change as a function of temperature for spectra acquired in CDCl3, CD2Cl2 , CD3OD, and D2O.

13

3-1. Chemical Shift Differences at Room Temperature (26°C) 3-1-a. 1-R and 1-S Benzyl Products in CD3OD Chemical shift differences of the alkyl protons (methine, methylene, and methyl) and aromatic protons of the enantiomeric pairs have been observed at room temperature for the benzyl products in CD3OD as shown by Figures 6-8.

R

S Figure 6: Methine and Methylene Proton Peaks of 1-R and 1-S Figure 7: Methyl Proton Peaks of 1-R and 1-S Benzyl Products in CD3OD at Room Temperature Benzyl Products in CD3OD at Room Temperature

R

S

Figure 8: Aromatic Proton Peaks of 1-R and 1-S Benzyl Products in CD3OD at Room Temperature

In CD 3OD, the methine and methyl chemical shifts of the 1-R and 1-S benzyl products were more downfield for the S-enantiomer than R-enantiomer as shown in Figures 6 and 7. Both of the methylene chemical shifts of the S-enantiomer were also more downfield than the corresponding methylene peaks of the R-enantiomer. In addition, Figure 8 shows the same trend in which the aromatic protons of the S-enantiomer were more downfield than the aromatic

14

protons of the R-enantiomer. This NMR data suggest differences in the preferred solution conformation of the 1-R and 1-S benzyl compound enantiomers as a result of interaction with the CD3OD solvent. 3-1-b. 2-R and 2-S Naphthyl Products in CD3OD While chemical shift differences were also observed between the 2-R and 2-S naphthyl compound enantiomers in CD3OD, the alkyl and aromatic proton chemical shifts were more downfield for the 2-R naphtyl product than the 2-S as shown in Figures 9-11.

R

S

Figure 9: Methine and Methylene Proton Peaks of 2-R and 2-S Naphthyl Products in CD3OD at Room Temperature

Figure 10: Methyl Proton Peaks of 2-R and 2-S Naphthyl Products in CD3OD at Room Temperature

R

S

Figure 11: Aromatic Proton Peaks of 2-R and 2-S Benzyl Products in CD3OD at Room Temperature

15

The chemical shift differences between the 2-R and 2-S napthyl products were unusual because they were extremely different from what was observed for the 1-R and 1-S benzyl products in CD3OD. While the chemical shifts for the alkyl and aromatic protons were more downfield for the 1-S benzyl product than the 1-R benzyl product, the chemical shifts were more downfield for the 2-R naphthyl product than 2-S naphthyl product. This is one of the few instances in which there was an opposite trend between the chiral naphthyl and benzyl products. Most of the other NMR data, which will be discussed in later sections, suggest that the behaviors for the 2-R and 2-S naphthyl products and 1-R and 1-S benzyl products are similar but have greater effects for the chiral naphthyl products. The differences in the chemical shifts between the 2-R and 2-S naphthyl compound enantiomers also suggest differences in the preferred conformations of the enantiomers as a result of solvent interaction with CD3 OD. 3-1-c. 1-R and 1-S Benzyl Products in D2O Similarly to what was seen for the benzyl products in CD3OD, the alkyl and aromatic proton chemical shifts were more downfield for the 1-S benzyl compound enantiomer than for 1R in D2O as shown in Figures 12-14.

R

S

Figure 12: Methine and Methylene Proton Peaks of 1-R and 1-S Benzyl Products in D2O at Room Temperature

Figure 13: Methyl Proton Peaks of 1-R and 1-S Benzyl Products in D2O at Room Temperature

16

R

S

Figure 13: Aromatic Proton Peaks of 1-R and 1-S Benzyl Products in D2O at Room Temperature

Figure 14: Aromatic Proton Peaks of 1-R and 1-S Benzyl Products in D2O at Room Temperature

The methine, methylene, and methyl proton peaks were more downfield in D2O for the 1S than the corresponding ones for the 1-R benzyl product. Figure 14 also shows that the aromatic chemical shifts in D 2O had a larger downfield shift for the 1-S enantiomer than the 1-R enantiomer. No difference in the alkyl and aromatic proton chemical shifts was observed between the enantiomers in all other solvents (Benzene- d6, CD2Cl2, CDCl3 , DMSO-d6 , and DMF- d7 ) at room temperature. Significant differences in the diastereotopic methylene protons were observed between the two enantiomers in D2O as shown in Figure 12. The methylene protons for the 1-R enantiomer were merged together while the methylene protons for the 1-S enantiomer were further apart from one another. This difference in the separation of the diastereotopic methylene protons will be discussed further in the next section.

17

3-2. Differences in the Diastereotopic Proton Separation at Room Temperature (26°C) Differences in the diastereotopic doublet proton separation between the enantiomers were observed at room temperature for the benzyl products (1-R and 1-S) in CD3 OD and D2O as shown in Table 1 below. Table 1: Separation of Methylene Doublets in Various Solvents at Room Temperature Amine

Solvent

Temperature (°C) Separation of Methylene Doublets (Hz)

1-R

CD3OD

26

74.1

1-S

CD3OD

26

62.8

2-R

CD3OD

26

49.5

2-S

CD3OD

26

64.9

1-R

D2O

26

7.1

1-S

D2O

26

17.4

1-R

CDCl3

26

71.8

1-S

CDCl3

26

71.8

1-R

Benzene-d6

26

94.9

1-S

Benzene-d6

26

94.3

1-R

DMSO-d6

26

43.3

1-S

DMSO-d6

26

44.2

1-R

CD2Cl2

26

54.4

1-S

CD2Cl2

26

55.0

1-R

DMF- d7

26

57.9

1-S

DMF- d7

26

58.3

18

The separation of the diastereotopic methylene protons varied with solvent. Differences in the separation between the 1-R and 1-S benzyl compound enantiomers were only observed in the polar protic solvents CD3OD and D2O. In CD3OD, the difference in the diastereotopic doublet proton separation was smaller for the 1-S benzyl compound enantiomer than for the 1-R. There was an opposite effect in the more polar solvent (D2O) in which the separation between the diastereotopic doublets was larger for the 1-S benzyl compound enantiomer than for 1-R. The separation of the diastereotopic doublet proton peaks was also greater for the 2-S naphthyl compound enantiomer in CD3OD than for 2-R. The data suggest that both polar protic solvents interact heavily with these molecules and give different solvent cage effects that result in different preferred conformation in solution for the R- and S-benzyl enantiomers. There was very different behavior of the separation of the diastereotopic methylene protons between the 1-R and 1-S benzyl compound enantiomers in D2 O that showed even more differences as a function of temperature in Chapter 4. 3-3. Line-width Effects at Room Temperature (26°C) 3-3-a. 1-R and 1-S Benzyl Products in CDCl3 Differences in NMR line-widths were observed at room temperature for the 1-R and 1-S benzyl products in CDCl3 as shown in Figures 15 and 16.

19

R R S

S

Figure 15: Methine and Methylene Proton Peaks of 1-R and 1-S Benzyl Products in CDCl3 at Room Temperature

Figure 16: Methyl Proton Peaks of 1-R and 1-S Benzyl Products in CDCl3 at Room Temperature

The chiral benzyl products 1-R and 1-S showed unusual linewidth broadening in their 1 H NMR spectra in CDCl3. All alkyl signals (methine, methylene, and methyl) were significantly broader for the S-enantiomer (3.3 Hz for methine, 3.16 Hz for methylene, and 3.35 Hz for methyl) than the corresponding ones for the R-enantiomer (1.95 Hz for methine, 2.33 Hz for methylene, and 1.68 Hz for methyl) as shown in Figures 15 and 16. Asymmetric line broadening was observed for the methylene proton peaks. The downfield methylene peaks were broader than the upfield methylene peaks for both enantiomers (2.48 Hz versus 2.18 Hz for 1-R and 3.55 Hz versus 2.77 Hz for 1-S) with a larger effect for the 1-S product. The TMS linewidth for both spectra was 1.40 Hz. 3-3-b. 2-R and 2-S Naphthyl Products in CDCl3 Figures17 and 18 show similar line-width behavior for the chiral naphthyl compounds 2R and 2-S, but the effect is greater than that observed for the benzyl compounds 1-R and 1-S.

20

R

S S Figure 17: Methine and Methylene Proton Peaks of 2-R and 2-S Naphthyl Products in CDCl3 at Room Temperature

Figure 18: Methyl Proton Peaks of 2-R and 2-S Naphthyl Products in CDCl3 at Room Temperature

There was greater line broadening of the methine, methylene, and methyl peaks for the Senantiomer (5.1 Hz for methine, 3.6 Hz for methylene, and 4.0 Hz for methyl) than the Renantiomer (1.7 Hz for methine, 1.97 Hz for methylene, and 1.25 Hz for methyl). In addition, the downfield methylene peaks of the two naphthyl compounds were broader than the upfield methylene peaks (2.13 Hz versus 1.80 Hz for 2-R and 4.08 Hz versus 3.12 Hz for 2-S) with a greater effect for the S-enantiomer than the R-enantiomer as seen clearly for the relative heights of the two doublets on Figure 17. The TMS linewidth for both spectra was 0.83 Hz. The linewidth broadening effect was bigger for the 2-R and 2-S naphthyl products than the 1-R and 1-S benzyl products. These line-width differences suggest different solution conformations for the Rand S-enantiomers and a greater conformation difference for the chiral naphthyl compounds consistent with a more hindered rotation caused by the heavier naphthyl substituent. 3-3-c. 1-R and 1-S Benzyl Products in CD2Cl2 Smaller alkyl line-width differences along with asymmetric line broadening of the methylene protons were observed for the two benzyl compound enantiomers in CD2Cl2 (2.70 Hz versus 2.52 Hz for 1-R and 3.15 Hz versus 2.76 Hz for 1-S), but the effect was still larger for the S-enantiomer as shown in Figures 19 and 20.

21

R

S

Figure 19: Methine and Methylene Proton Peaks of 1-R and 1-S Benzyl Products in CD2Cl2 at Room Temperature

Figure 20: Methyl Proton Peaks of 1-R and 1-S Benzyl Products in CD2Cl2 at Room Temperature

No significant line-width differences were observed for the two 1-R and 1-S benzyl compound enantiomers in other solvents. Solvent obviously plays a significant role in the solution conformations of these compounds. 3-4. Differences in OH-NH Peak at Room Temperature (26°C) 3-4-a. Enantiomeric Pairs of Benzyl and Naphthyl Products in CDCl 3 Figure 21 shows a vertical enlargement of the 1H NMR spectra for the chiral benzyl 1-R and 1-S and naphtyl 2-R and 2-S compounds in CDCl3 .

R *

S *

R S

* *

Figure 21: OH-NH Peaks of R- and S-Benzyl Compounds (Top Two Spectra) and Rand S- Naphthyl Compounds (Bottom Two Spectra) in CDCl3 at Room Temperature.

22

A single broad peak representing the combined OH-NH signal is seen in the 5-7 ppm range for the R- and S-enantiomers. This peak is further upfield and much broader for the R-benzyl compound than the S-benzyl compound which suggests differences in the exchange rate of these protons and consequently the strength of the intramolecular OH-NH hydrogen bonding for the two enantiomers in CDCl3. There were similar differences in the broadening of the OH-NH signals between the enantiomers of the naphthyl products at room temperature. The OH-NH peak at around 5.5-8 ppm was much broader for the 2-R than the 2-S naphthyl products. Both the R- and S- OH-NH peaks for the naphtyl compounds were much broader than the OH-NH peak for the benzyl compounds. Additional differences are observed at lower temperature and will be discussed on the variable temperature section. 3-4-b. 1-R and 1-S Benzyl Products in Other Solvents 1 H NMR spectra acquired in other aprotic solvents showed less extreme differences between the OH/NH signals for the R- and S-enantiomers of the benzyl products as shown in Figures 22-25.

R *

*

S * * Figure 22: OH-NH Peaks of 1-R and 1-S Benzyl Compounds in CD2Cl2 at Room Temperature

Figure 23: OH-NH Peaks of 1-R and 1-S Benzyl Compounds in Benzene- d6 at Room Temperature

23

R *

*

*

*

S *

*

Figure 24: OH-NH Peaks of 1-R and 1-S Benzyl Compounds in DMSO-d6 at Room Temperature

*

*

Figure 25: OH-NH Peaks of 1-R and 1-S Benzyl Compounds in DMF- d7 at Room Temperature

Figure 22 shows the spectra acquired in CD2Cl2 and Figure23 shows the spectra acquired in benzene-d6 . Single broad peaks are observed for OH/NH in both cases. There are minor linewidth differences for the R- and S benzyl compound enantiomers, but these results suggest much more similar intramolecular H-bonding effects for the two enantiomers in these solvents. Figures 24 and 25 show the 1H NMR spectra for 1-R and 1-S in DMSO-d6 and DMF-d7 respectively measured at room temperature. In these solvents, separate broad peaks are observed for OH (about 10 ppm) and NH (about 3-5 ppm). This indicates limited OH/NH proton exchange that may be a consequence of increased interaction between the solvent molecules and the labile protons in these solvents. No OH or NH signals were observed for either benzyl or naphthyl products in the polar protic solvents CD3OD of D2O where the labile protons are expected to be fully exchanged with the solvent.

24

3-5 Summary Chemical shift, line-width, and OH-NH signal differences between the R- and Senantiomers were observed at room temperature. Chemical shift differences and differences in the separation of the diastereotopic methylene peaks were observed only in polar protic solvents CD3OD and D 2O. There was a consistent trend in which the aromatic and alkyl protons of the 1S benzyl compound enantiomer were futher downfield than the corresponding ones for the 1-R. By comparison to the NMR spectra for the corresponding HCl salts of these benzyl products in CD3OD (Figures 26 and 27), protonation of the nitrogen atom causes a significant downfield shift for all alkyl protons in these molecules. It is possible that the observed downfield shift of the alkyl protons for the 1-S amine compared to the 1-R amine may suggest a greater partial positive charge on nitrogen for 1-S due to different solvent interaction. Detailed electronic structure calculations with solvent effect will be necessary to test this hypothesis. R-Salt

R-Salt

R-base

R-base S-Salt

S-Salt

S-base

Figure 26: Methine and Methylene Proton Peaks of 1-R Salt and 1-R (Top Two Spectra) and 1-S Salt and 1-S (Bottom Two Spectra) in CD3OD at Room Temperature

S-base

Figure 27: Methyl Proton Peaks of 1-R Salt and 1-R (Top Two Spectra) and 1-S Salt and 1-S (Bottom Two Spectra) in CD3OD at Room Temperature

25

Differences in the line-width effects and the OH/NH signal between the R- and S-benzyl and naphthyl enantiomers at room temperature were only observed in non-polar solvents. The alkyl proton peaks of the S-enantiomer were broader and more asymmetric (with a broader downfield methylene doublet than the upfield methylene doublet) than the corresponding Renantiomers in CDCl3 and CD2Cl2 . The effects were greater for the naphthyl products. In addition, the OH/NH signal was much broader for the R-enantiomer in CDCl3 than the Senantiomer for both the naphthyl and benzyl compounds with larger effect for the naphthyl product. These effects are consistent with conformational changes that could be attributable to the heavier naphthyl substituent, but electronic structure calculations will be necessary to determine this. Significant differences in the NMR behavior of the chiral benzyl and naphthyl products occurred in protic and aprotic solvents that were more extreme as a function of temperature in Chapter 4.

26

Chapter 4: Variable Temperature NMR Spectrsocopy The observed NMR chemical shift differences between the enantiomers in CD3OD and D2O and the line-width and OH-NH effects observed in CDCl3 and CD2Cl2 at room temperature suggested significant molecular conformational changes dependent on the polarity of the solvents. In this section, variable temperature 1H NMR spectroscopy of the chiral 1-R and 1-S benzyl products and 2-R and 2-S naphthyl products is discussed. Similar trends in these spectroscopic features were observed as a function of temperature for the benzyl and naphthyl products in non-polar solvents CDCl3 and CD2Cl2 and the polar-protic solvents CD3OD and D2O. 4-1. Variable Temperature NMR Spectrsocopy in CDCl3 4-1-a. 1-R and 1-S Benzyl Products Variable temperature 1H NMR spectra are shown for the chiral benzyl products 1-R and 1-S in CDCl3 in Figures 28-33. Significant changes in chemical shifts and line-widths are seen as temperature changes.

Figure 28: Methine and Methylene Proton Peaks of 1-R in CDCl3 at Temperatures from -50°C (Bottom) to 40°C (Top)

Figure 29: Methine and Methylene Proton Peaks of 1-S in CDCl3 at Temperatures from -50°C (Bottom) to 40°C (Top)

27

Figure 30: Methyl Proton Peaks of 1-R in CDCl3 at Temperatures from -50°C (Bottom) to 40°C (Top)

33

11 44

Figure 31: Methyl Proton Peaks of 1-S in CDCl3 at Temperatures from -50°C (Bottom) to 40°C (Top)

22

33

Figure 32: Aromatic Proton Peaks of 1-R in CDCl3 at Temperatures from -50°C (Bottom) to 40°C (Top)

22

Figure 33: Aromatic Proton Peaks of 1-S in CDCl3 at Temperatures from -50°C (Bottom) to 40°C (Top)

1 N H HO 1-R

11 44

2 3 4

28

Figures 34-38 show chemical shifts as a function of temperature for the methine, methyl, the two methylene doublets, and the separation of the diastereotopic methylene peaks respectively. While room temperature NMR spectra showed no differences in these parameters for the two enantiomers in CDCl3, significant differences between the 1-R and 1-S chemical shifts were observed at lower temperatures.

Figure 34: Methine Chemical Shifts of 1-R and 1-S in CDCl3 vs. Temperature

Figure 35: Methyl Chemical Shifts of 1-R and 1-S in CDCl3 vs. Temperature

29

Figure 36: Downfield Methylene Chemical Shifts of 1-R and 1-S in CDCl3 vs. Temperature

Figure 37: Upfield Methylene Chemical Shifts of 1-R and 1-S in CDCl3 vs. Temperature

30

Figure 38: Separation of Methylene Chemical Shifts of 1-R and 1-S in CDCl3 vs. Temperature

The methine, methyl, and the downfield methylene chemical shifts of the 1-R and 1-S benzyl products in CDCl3 were different at low temperature and converged with temperature. Chemical shifts for methine and methyl shifted upfield as temperature increased as shown in Figures 34 and 35. This effect was slightly larger for the 1-S enantiomer than the 1-R enantiomer. For the diastereotopic methylene protons, the downfield doublet (CH2-A) shifted upfield and the upfield doublet (CH2-B) shifted downfield with increasing temperature resulting in decreased separation between the doublets as shown in Figures 36 and 37. The chemical shift differences for the 1-S diastereotopic protons were greater than those for 1-R at lower temperature but they converged to equal separation at temperatures above 5°C as shown in Figure 38. The convergence of the chemical shifts of 1-R and 1-S suggests that the two enantiomers have different solution conformations at lower temperature in CDCl3 but similar conformations at room temperature and above.

31

Much more significant line-width differences between the NMR spectra of 1-R and 1-S benzyl products were observed as a function of temperature in CDCl3 . While slight linebroadening was observed for the methine, methylene, methyl, and some aromatic proton signals for the 1-R enantiomer at lower temperatures, all alkyl NMR signals and some aromatic NMR signals were significantly broader for the 1-S enantiomer especially at temperatures lower than 30°C where the methine quartet and methyl doublets are almost unresolved. The line-widths also broaden again at temperatures above 5°C for the 1-S enantiomer while the line-widths only increase as temperature decreases for the 1-R enantiomer. As discussed in the previous chapter for room temperature spectra, Figures 28 and 29 show asymmetric line broadening for the diastereotopic methylene peaks for both enantiomers. The downfield methylene peaks were broader than the upfield methylene peaks at high temperature for both the 1-R and 1-S products (2.38 Hz versus 1.95 Hz for 1-R at 26°C and 3.15 Hz versus 2.50 Hz for 1-S at 26°C) and at low temperature for only the 1-S enantiomer (5.60 Hz versus 3.85 Hz for 1-S at -30°C). In addition, Figures 32 and 33 show that aromatic protons 1 and 4 along with some of the protons on the other ring merged and broadened for the S-enantiomer at lower temperature while the corresponding aromatic protons for the R-enantiomer remained separated and narrower. These line-width differences suggest differences in the solution conformation of the 1-R and 1-S enantiomers at lower temperature. The variable temperature 1H NMR spectra of the OH and NH signals for 1-R and 1-S in CDCl3 are shown in Figures 39 and 40. Significant changes in the behavior of the OH and NH peaks occurred as temperature changed.

32

Figure 39: OH-NH Signal of 1-R in CDCl3 at Temperatures from -50°C (Bottom) to 40°C (Top)

Figure 40: OH-NH Signal of 1-S in CDCl3 at Temperatures from -50°C (Bottom) to 40°C (Top)

At higher temperature, a single very broad peak is observed for the exchanging OH/NH protons with the peak broader and centered more upfield for 1-R than 1-S as described at room temperature in chapter 3. At temperature lower than -30°C, Figure 39 shows a clear separation of the OH (11.8 ppm) and NH (2.1 ppm) proton signals for 1-R as the exchange slows dramatically. This effect is smaller for the 1-S enantiomer in Figure 37 where the OH and NH peaks may separate at -50 °C, but the peaks are not as distinct as those seen in the 1-R spectrum. 4-1-b. 2-R and 2-S Naphthyl Products Figures 41-45 show smaller differences in the chemical shifts for 2-R and 2-S naphthyl products at low temperature but show similar convergence at higher temperature.

33

Figure 41: Methine Chemical Shifts of 2-R and 2-S in CDCl3 vs. Temperature

Figure 42: Methyl Chemical Shifts of 2-R and 2-S in CDCl3 vs. Temperature

34

Figure 43: Downfield Methylene Chemical Shifts of 2-R and 2-S in CDCl3 vs. Temperature

Figure 44: Upfield Methylene Chemical Shifts of 2-R and 2-S in CDCl3 vs. Temperature

35

Figure 45: Separation of Methylene Chemical Shifts of 2-R and 2-S in CDCl3 vs. Temperature

The methine and methyl chemical shifts for the 2-R and 2-S naphthyl products shifted upfield slightly and the separation of the diastereotopic methylene peaks decreased with temperature, similar to the 1-R and 1-S benzyl products. The convergence of the chemical shifts with temperature suggests that 2-R and 2-S naphthyl products also have different solution conformations at lower temperatures in CDCl3 but with smaller differences compared to the 1-R and 1-S benzyl products. Similar line-broadening behavior was observed for the 2-R and 2-S naphthyl enantiomers but with greater effects than the 1-R and 1-S products as shown in Figures 46-49.

36

Figure 46: Methine and Methylene Proton Peaks of 2-R in Figure 47: Methine and Methylene Proton Peaks of 2-S in CDCl3 at Temperatures from -50°C (Bottom) to 40°C (Top) CDCl3 at Temperatures from -50°C (Bottom) to 40°C (Top)

Figure 48: Methyl Proton Peaks of 2-R in CDCl3 at Temperatures from -50°C (Bottom) to 40°C (Top)

Figure 49: Methyl Proton Peaks of 2-S in CDCl3 at Temperatures from -50°C (Bottom) to 40°C (Top)

37

There was greater line broadening of the methine, methylene, and methyl proton peaks for the 2S enantiomer than the 2-R enantiomer at lower temperature. While the methine and methylene protons for the 2-R enantiomer in Figure 46 only broadened with decreasing temperature, the corresponding protons for the 2-S enantiomer in Figure 47 also showed broadening effects at temperatures greater than 5°C. The 2-R and 2-S naphthyl products also showed slightly different asymmetric line-broadening trend as a function of temperature compared to the 1-R and 1-S benzyl products. While there was no asymmetric line-broadening for the 2-R naphthyl product at any temperature, the 2-S naphthyl product showed greater broadening of the downfield methylene doublet than the upfield methylene doublet (3.62 Hz versus 2.75 Hz for 2-S at 26°C) at higher temperatures. While these line-width differences suggest solution conformation difference for the chiral naphthyl products as a function of temperature, the effects are different from those observed for the benzyl products 1-R and 1-S. Figures 50 and 51 show OH/NH signal behaviors for the 2-R and 2-S naphthyl products, which are qualitatively similar to but larger than those of the benzyl products.

Figure 50: OH-NH Signal of 2-R in CDCl3 at Temperatures from -50°C (Bottom) to 40°C (Top)

Figure 51: OH-NH Signal of 2-S in CDCl3 at Temperatures from -50°C (Bottom) to 40°C (Top)

38

At higher temperature, the OH-NH peak is much broader for the 2-R enantiomer than the 2-S enantiomer. The OH-NH peak for 2-R separated at a slightly higher temperature than the 1-R benzyl product and sharpened from -30°C to -50°C as shown in Figure 50. The difference in the broadness of the 2-R and 2-S OH-NH peaks as a function of temperature suggests different exchange rates that are likely to due to hydrogen bonding. Both the 2-R and 2-S naphthyl OHNH peaks were much broader than the 1-R and 1-S benzyl product signals; this is consistent with greater hindered rotation for the 2-R and 2-S naphthyl products enhancing the conformational differences. At low temperature, the sharpness of the separated OH and NH peaks suggests that the naphthyl compound is in a conformation that does not allow intramolecular hydrogen bonding. A conformation that allows proton exchange requires molecular bond rotation, and the increased mass of the naphthyl substituent compared to the benzyl compound would inhibit bond rotation at low temperature. The differences in the sharpness of the separated OH and NH peaks between the naphthyl enantiomers also suggest different solution conformations at low temperature. 4-2. Variable Temperature NMR Spectrsocopy in CD2Cl2 4-2-a. 1-R and 1-S Benzyl Products The chemical shift plots for 1-R and 1-S in CD2Cl2are shown in Figures 52-56.

Figure 52: Methine Chemical Shifts of 1-R and 1-S in CD2 Cl2vs. Temperature

39

Figure 53: Methyl Chemical Shifts of 1-R and 1-S in CD2Cl2vs. Temperature

Figure 54: Downfield Methylene Chemical Shifts of 1-R and 1-S in CD2Cl2 vs. Temperature

40

Figure 55: Upfield Methylene Chemical Shifts of 1-R and 1-S in CD2 Cl2vs. Temperature

Figure 56: Separation of Methylene Chemical Shifts of 1-R and 1-S in CD2Cl2vs. Temperature

41

The trends in chemical shift differences observed in CD2Cl2are similar to those seen in CDCl3 but the effects are generally smaller. That is, the alkyl proton chemical shifts are different for the 1-R and 1-S enantiomers at lower temperature and converged with increasing temperature, suggesting similar differences in the solution conformation of the two enantiomers in nonpolar solvents. Line-broadening can be observed for the methine, methylene, and methyl signals as a function of temperature in the 1H NMR spectra in Figures 57-60.

Figure 57: Methine and Methylene Proton Peaks of 1- Figure 58: Methine and Methylene Proton Peaks of 1-S R in CD2 Cl2at Temperatures from -50°C (Bottom) to in CD2 Cl2 at Temperatures from -40°C (Bottom) to 26°C 26°C (Top) (Top)

42

Figure 59: Methyl Proton Peaks of 1-R in CD2Cl2 at Temperatures from -50°C (Bottom) to 26°C (Top)

Figure 60: Methyl Proton Peaks of 1-S in CD2Cl2 at Temperatures from -40°C (Bottom) to 26°C (Top)

The alkyl proton line-broadening effects and asymmetric broadening of the CH2 proton peaks are qualitatively similar to, but smaller than those observed for spectra measured in CDCl3. Figures 61 and 62 show the OH/NH NMR signal behaviors for the 1-R and 1-S benzyl products in CD2Cl2 .

Figure 61: Methyl Proton Peaks of 1-R in CD2Cl2 at Temperatures from -50°C (Bottom) to 26°C (Top)

Figure 62: Methyl Proton Peaks of 1-S in CD2Cl2 at Temperatures from -40°C (Bottom) to 26°C (Top)

43

Similarly broad signals, indicative of proton exchange, are seen here, but unlike the spectra recorded in CDCl3 no separation of the OH and NH signals is observed at lower temperatures in CD2Cl2. Differences in the signals for 1-R and 1-S also are smaller than the differences seen in CDCl3. While there are differences in the magnitudes of the observed effects, the qualitatively similar chemical shift, line-width, and OH/NH trends that were observed for spectra measured in CDCl3 and CD2Cl2 suggest similar non-polar solvent effects on the solution conformations of the chiral benzyl and naphthyl compounds. 4-3. Variable Temperature NMR Spectrsocopy in CD3OD 4-3-a. 1-R and 1-S Benzyl Products Figures 63-67 show the alkyl proton chemical shift temperature behavior for 1-R and 1-S in CD3OD. Unlike the spectra acquired in CDCl3, chemical shifts for the 1-R and 1-S alkyl protons were different in CD3OD even at room temperature and this difference becomes larger at lower temperature. The observed trends suggest the alkyl proton chemical shifts might converge, but at temperatures above the CD3OD boiling point.

Figure 63: Methine Chemical Shifts of 1-R and 1-S in CD3OD vs. Temperature

44

Figure 64: Methyl Chemical Shifts of 1-R and 1-S in CD3OD vs. Temperature

Figure 65: Downfield Methylene Chemical Shifts of 1-R and 1-S in CD3OD vs. Temperature

45

Figure 66: Upfield Methylene Chemical Shifts of 1-R and 1-S in CD3OD vs. Temperature

Figure 67: Separation of Methylene Chemical Shifts of 1-R and 1-S in CD3OD vs. Temperature

There are significant differences for the spectra acquired in CD3OD. The methyl proton signals for 1-R and 1-S shifted downfield with temperature in CD3OD as shown in Figure 64 but shifted upfield with temperature in CDCl3. The chemical shift separation of the diastereotopic methylene peaks for 1-R and 1-S in CD3OD increased with temperature as shown in Figure 67 compared to a decrease in this separation with temperature in CDCl3 . In addition, the chemical

46

shift separation for the diastereotopic methylene peaks for 1-S is considerably smaller than that for 1-R in CD3OD, whereas it was slightly larger in CDCl3. Even greater differences in the chemical shift trends with temperature were observed in D2O, which will be discussed later in this chapter. Line broadening effects are even more different in CD3OD as shown in Figures 68 and 69. While there is greater line-broadening of the methine and methylene protons of the 1-S than the 1-R at low temperature, an opposite asymmetric line broadening effect is observed for the methylene protons of the 1-S benzyl enantiomer in CD3OD compared to those in CDCl3 . In CD3OD, the upfield methylene peaks for 1-S are significantly broader than the downfield methylene peaks (8.90 Hz versus 6.65 Hz at -50°C) at low temperatures as shown in Figure 69. For 1-R, no significant asymmetry in the methylene line broadening is observed (Figure 68).

Figure 68: Methine and Methylene Proton Peaks of 1-R in CD3OD at Temperatures from -50°C (Bottom) to 40°C (Top)

Figure 69: Methine and Methylene Proton Peaks of 1-S in CD3OD at Temperatures from -50°C (Bottom) to 40°C (Top)

47

These results suggest that while CD3OD solvent effects still create different molecular conformation effects for the two enantiomers, these effects seem to be very different for the polar protic solvent CD3OD compared to the effects of the nonpolar solvents CDCl3 and CD2Cl2 . 4-3-b. 2-R and 2-S Naphthyl Products Figures 70-76 show the observed alkyl proton chemical shift trends for the 2-R and 2-S naphthyl compounds in CD3OD. While the direction of chemical shift changes with temperature is similar to that for the 1-R and 1-S benzyl compounds, other chemical shift trends are considerably different. As reported in Chapter 3 for room temperature, alkyl chemical shifts for 2-R are downfield from those of 2-S and this difference becomes larger with decreasing temperature for methine, methyl, and the upfield methylene doublet.

Figure 70: Methine Chemical Shifts of 2-R and 2-S in CD3OD vs. Temperature

48

Figure 71: Methyl Chemical Shifts of 2-R and 2-S in CD3OD vs. Temperature

Figure 72: Downfield Methylene Chemical Shifts of 2-R and 2-S in CD3OD vs. Temperature

49

Figure 73: Upfield Methylene Chemical Shifts of 2-R and 2-S in CD3OD vs. Temperature

Figure 74: Separation of Methylene Chemical Shifts of 2-R and 2-S in CD3OD vs. Temperature

50

Figure 75: Methine and Methylene Proton Peaks of 2-R in CD3OD at Temperatures from -50°C (Bottom) to 40°C (Top)

Figure 76: Methine and Methylene Proton Peaks of 2-S in CD3OD at Temperatures from -50°C (Bottom) to 40°C (Top)

The diastereotopic methylene proton behavior is quite different between the benzyl and naphthyl compounds. Unlike the behavior for the benzyl compounds, the spectral separation of the methylene proton signals for 2-S is larger than that for 2-R at all temperatures and this effect increases as temperature decreases. For 2-R, Figure 75 shows the methylene peaks have merged almost completely at -50°C whereas Figure 76 shows 2-S with clearly separated signals for the methylene protons. Line-broadening behavior for 2-R and 2-S in CD3OD is similar to that for 1-R and 1-S. All alkyl peaks broaden with decreasing temperature and this effect is greater for 2-S than 2-R. In addition, the upfield methylene peaks for 2-S are considerably broader than the downfield methylene peaks (4.90 Hz versus 4.77 Hz for 2-R at -50°C and 9.50 Hz versus 9.07 Hz for 2-S at -50°C) as shown on Figure 76 and this effect is even greater for the naphthyl compounds than the benzyl compounds. At -50°C, the ratio of the upfield line-widths to the downfield line-widths is 1.74for the 2-S naphthyl enantiomer compared to a ratio of 1.34 for the 1-S benzyl enantiomer.

51

This asymmetric broadening also seems to be evident for the 2-R enantiomer at low temperatures, but it is difficult to quantify due to significant second order coupling effects as the two methylene peaks merge with decreasing temperature. These line-width differences imply differences in conformation with the heavier naphthyl substituent responsible for more extreme line-broadening effects. 4-4. Variable Temperature NMR Spectrsocopy in D2O 4-4-a. 1-R and 1-S Benzyl Products 1 H NMR spectra also were recorded for the 1-R and 1-S benzyl products in D2O at temperatures from 10 °C to 80 °C. Shimming of these samples was difficult because these compounds were not very soluble in D 2O. Consequently, observation of the line-width effects is less reliable for these spectra. The chemical shift trends for the chiral benzyl products in D2O are significantly different from those in CD3 OD and other solvents as shown in Figures 77-80.

Figure 77: Methine Chemical Shifts of 1-R and 1-S in D2O vs. Temperature

52

Figure 78: Methyl Chemical Shifts of 1-R and 1-S in D2O vs. Temperature

Figure 79: Downfield Methylene Chemical Shifts of 1-R and 1-S in D2O vs. Temperature

53

Figure 80: Upfield Methylene Chemical Shifts of 1-R and 1-S in D2O vs. Temperature

While the alkyl proton chemical shifts for 1-S are downfield of those for 1-R as in CD3OD, the differences in these chemical shifts decrease with decreasing temperatures, an effect opposite to that observed in other solvents. While alkyl proton chemical shifts for 1-S in other solvents showed larger temperature effects than those for 1-R, the D2O chemical shift effects as a function of temperature were larger for the 1-R enantiomer. In addition, the 1-R and 1-S enantiomers showed opposite effects for the separation of the methylene protons, a behavior not seen in any other solvent, as shown in Figures 81-83.

54

Figure 81: Methylene Chemical Shifts of 1-R and 1-S in D2O vs. Temperature

Figure 82: Methine and Methylene Proton Peaks of 1-R in D2O at Temperatures from 15°C (Bottom) to 80°C (Top)

Figure 83: Methine and Methylene Proton Peaks of 1-S in D2O at Temperatures from 15°C (Bottom) to 80°C (Top)

55

For 1-R, the two methylene doublets are almost completely merged at 15°C and their separation increases with increasing temperature. However, for 1-S, the two methylene doublets are separated at 15 °C and merged with increasing temperature. This very different spectral behavior for 1-R and 1-S in D2O as a function of temperature suggests very different conformational changes for these molecules in D2 O compared to molecules in other solvents.

56

Chapter 5: Summary and Future Work Results presented in Chapters 3 and 4 have shown significant differences in the 1H NMR spectra of the pairs of enantiomeric molecules 1-R, 1-S and 2-R, 2-S as a function of solvent, temperature, and stereochemistry. Significant differences have included changes in the chemical shifts of the alkyl and aromatic protons with chirality, different line-broadening effects with chirality, different OH/NH exchange rates, and the chemical shift separation of the diastereotopic methylene protons. We have postulated that the spectral differences between these molecules probably can be attributed to different molecular conformations of the two enantiomers as a function of temperature and solvent. A more detailed explanation is necessary. There were very significant proton chemical shift differences between the 1-R and 1-S benzyl products and the 2-R and 2-S naphthyl products in the polar protic solvents CD3OD and D2O. In Chapter 3, we have speculated that this could be due to the two enantiomers having a different partial positive charge on the nitrogen atom in these molecules as a result of hydrogen bonding interactions between the solute and solvent. Our data for the corresponding HCl salts of these molecules have indicated a significant downfield shift of all alkyl protons for the salt in comparison to the amine. As shown in Chapter 4, the alkyl proton (methine, methylene, and methyl) chemical shift differences measured in CD3OD for the 1-R and 1-S benzyl enantiomers and 2-R and 2-S naphthyl enantiomers become larger at lower temperature. These results are fully consistent with a possible greater partial positive charge on the nitrogen atom in 1-S and 2R as a result of increased interaction with the protic solvent at lower temperatures. Unfortunately, this explanation cannot be applied to the differences in chemical shifts observed for the 1-S and 1-R benzyl enantiomers measured in D2O where the alkyl proton chemical shifts showed a

57

smaller difference at low temperatures. Obviously, a much more detailed study of solution molecular conformations will be necessary to explain the details of these results. Significant line-width differences between the variable temperature NMR spectra of 1-R and 1-S benzyl products and 2-R and 2-S naphthyl products were observed in the non-polar solvents CDCl3 and CD2Cl2 and in the polar protic solvent CD3OD. Greater line-broadening for all alkyl protons was observed with decreasing temperature as discussed in Chapter 4 and these effects were larger for the S-enantiomer than the R-enantiomer. One could postulate that a possible contribution to increased line-broadening for the methine and methylene protons at low temperature could be coupling between these protons and the nearby amine proton (NH) that stays in place longer with the reduced rate of intramolecular and/or intermolecular exchange of the amino and hydroxyl protons. An example of this was seen in NMR spectra of the HCl salt of the 1-R enantiomer in DMSO-d6 (Figures 84-86). Clear evidence of signal broadening and additional coupling is observed for the methine and methylene proton signals (Figure 85) at lower temperatures when the NH2 proton signals are narrow (Figure 84) indicating reduced exchange at corresponding lower temperatures. While additional coupling to NH could be a plausible explanation for signal broadening for methine and methylene proton signals for the amines (1-R, 2-R, 1-S, 2-S) in CDCl3 and CD2 Cl2 at low temperature – and could even explain asymmetric broadening of the two different methylene doublets depending on their relative orientation to the amine proton - it cannot explain all line-broadening effects. The amines showed equivalent line-broadening for the methyl proton signals at low temperature but as shown in Figure 86 for the salt in DMSO-d6, the methyl protons are too far away (four bonds) from the amine protons to experience line-broadening due to coupling at lower temperature. In addition, line broadening of all alkyl signals also occurred with decreasing temperature in

58

CD3OD, a polar protic solvent in which the amine proton is likely to be fully exchanged even at -50 °C. Obviously, a more complex explanation of the observed line-broadening effects will require a more detailed understanding of the solvent effects on molecular conformation.

Figure 84: OH and NH2+ Signals of 1-R Salt in DMSO-d6 at Figure 85: CH and CH2 Peaks of 1-R Salt in DMSO-d6 at Temperatures from 26°C (Bottom) to 130°C (Top) Temperatures from 26°C (Bottom) to 130°C (Top)

Figure 86: CH3 Peaks of 1-R Salt in DMSO-d6 at Temperatures from 26°C (Bottom) to 130°C (Top)

59

Separation of the diastereotopic methylene peaks for 1-R and 1-S and 2-R and 2-S enantiomers varied considerably with solvent, temperature, and stereochemistry. We do not have an explanation for the observed trends in this separation of the diastereotopic methylene protons. Detailed electronic structure studies are needed to provide an understanding of the molecular conformations causing these effects. In future work, electronic structure calculations will be used to help understand the solvent and temperature effects on the molecular conformations of the chiral benzyl and naphthyl enantiomers. A very preliminary consideration of this has been carried out with the MestReNova (Mnova) NMR prediction software which suggested the possibility of different NMR spectra for 1-R and 1-S enantiomers in some solvents. However, this program performs very limited conformational sampling. Much higher level electronic structure calculations will be necessary to understand the differences in the molecular conformation between the enantiomers with different solvent interactions. In addition to electronic structure calculations, the corresponding tertiary amines for 1-R, 1-S, 2-R, and 2-S will be synthesized. NMR studies of these molecules will provide additional information about the preferred conformations of these chiral amine molecules as a function of solvent and temperature without the possibility of OH/NH exchange. Our results indicate solvent and temperature effects play significant roles in determining the solution conformations of chiral Mannich bases that may also be dependent on their stereochemistries. With greater understanding of solution conformations of these chiral molecules in various solvent and temperature conditions, it is possible that these results may eventually lead to an efficient method of determining the absolute configuration of chiral amines.

60

References 1. Bagno, A.; Rastrelli, F.; Saielli, G. NMR Techniques for the Investigation of Solvation Phenomena and Non-Covalent Interactions. Nuc. Mag. Res. Spec. 2005, 47, 41-93. 2. Abraham, M.; Abraham, R.; Acree, W.; Aliev, A.; Leo, A.; Whaley, W. An NMR Method for the Quantitative Assessment of Intramolecular Hydrogen Bonding; Application to Physicochemical, Environmental, and Biochemical Properties. J. Org. Chem. 2014, 79(22), 11075-11083. 3. Labby, K.; Xue, F.; Kraus, J.; Ji, H.; Mataka, J.; Li, H.; Martasek, P., Roman, L.; Poulos, T.; Silverman, R. Intramolecular hydrogen bonding: A potential strategy for more bioavailable inhibitors of neuronal nitric oxide synthase. Bioorg. & Med. Chem. 2012, 20, 2435-2443. 4. Sitkowski, J.; Stefaniak, L. Multinuclear, Variable-Temperature NMR Study of Hydrogen Bonding in Two Ortho-Mannich Bases. J. Phys. Org. Chem. 1995, 8, 463-467. 5. Rospenk, R.; Fritsch, J.; Zundel, G. Solvent Effect on the Proton-Transfer Equilibria and Thermodynamic Data of the Hydrogen Bond in a Mannich Base. J. Phys. Chem. 1984, 88, 321-323. 6. McConathy, J. et al; J Clinical Psych. 2003, 5(2), 70-73. 7. Eriksson, T. et al; Euro Journal of Clinical Pharm. 2001, 57(5), 365-376. 8. Leiro, V.; Seco, J.; Quinoa, E.; Riguera, R. Assigning the Configuration of Amino Alcohols by NMR: A single Derivitization Method. Org. Letters. 2008, 10(13), 2733-2736. 9. Allen, D.; Tomaso, A.; Priest, O. Mosher Amides: Determining the Absolute Stereochemistry of Optically-Active Amines. J. Chem. Ed. 2008, 85(5), 698. 10. Bennett, J.; Charles, K.; Miner, M.; Heuberger, C.; Spina, E.; Bartels, M.; Foreman, T.; Ethyl lactate as a tunable solvent for the synthesis of aryl aldimines. G. Chem. 2009, 11, 166168. 11. Carlson, M.; Ciszewski, J.; Bhatti, M.; Swanson, W.; Wilson, A. A Simple Secondary Amine Synthesis: Reductive Amination Using Sodium Triacetoxyborohydride. J. Chem. Ed. 2000, 77(2), 270-271.