Hydrophobic Substituent Effects on Proline Catalysis of Aldol Reactions in Water

Abstract

Derivatives of 4-hydroxyproline with a series of hydrophobic groups in well-defined orientations have been tested as catalysts for the aldol reactions. All of the modified proline catalysts carry out the intermolecular aldol reaction in water and provide high diastereoselectivity and enantioselectivity. Modified prolines with aromatic groups syn to the carboxylic acid are better catalysts than those with small hydrophobic groups (1a is 43.5 times faster than 1f). Quantum mechanical calculations provide transition structures, TS-1a_water and TS-1f_water that support the hypothesis that a stabilizing hydrophobic interaction occurs with 1a.

Introduction

In the past decades, small, metal-free organic molecules have attracted considerable attention as asymmetric organocatalysts.¹ Despite considerable effort, the vast majority of organocatalysts carry out relatively few turnovers; typically more than 10 mol % of catalyst is necessary to achieve reasonable reaction rates and yields. X-ray structures of highly active enzymes show intricate active sites that have shape and chemical complementarity to their substrates and they suggest the hypothesis that the activity of organocatalysts might be improved by creating specific interactions between the organocatalyst and their substrates. In non-aqueous solvents, organocatalysts must utilize specific hydrogen bonding, charge-charge and dipole-charge interactions to drive the association of the organocatalyst with their substrates.² In water, on the other hand, organocatalysts need to make use of the hydrophobic interaction because the high dielectric constant of water screens charge-charge interactions and water competes for hydrogen bonds. The development of organocatalytic asymmetric reactions in aqueous media is an attractive goal because water is an inexhaustable and environmentally friendly solvent.^3,4

Proline was found, in the 1970s, by Hajos and Parrish⁵ and Eder, Sauer and Wiechert⁶ to be an organocatalyst for the intramolecular aldol reaction and, in 2000, by List, Lerner and Barbas to be an organocatalyst for the intermolecular aldol reaction.⁷ In organic solvents, proline requires high loadings (~ 30 mol %) and it is well known that proline does not catalyze the aldol reaction in water.⁸ Since the initial reports of the proline-catalyzed aldol reaction, derivatives of 4-hydroxyproline have attracted considerable interest as catalysts of the asymmetric aldol reaction in water. ⁹ In these previous reports, hydrophobic substituents were attached to the alcohol of 4-hydroxyproline and cyclic ketones were combined with aldehydes and catalyst in an aqueous, biphasic environment formed by the mixture of water and the organic ketone. The organic ketone and the modified proline catalyst have been proposed to assemble into micelles in water due to hydrophobic interactions, excluding water and driving the formation of the enamine.⁸ The enamine is then more soluble in the organic phase where it reacts with the aldehyde to form the carbon-carbon bond and then hydrolysis of the resulting iminium ion takes place. In contrast to this phase-transfer type of mechanism, we explore here the idea that precisely placed hydrophobic groups on modified prolines could create specific hydrophobic interactions with the aldol substrates in water to create organocatalysts that have high activity as well as high diastereo- and enantio-selectivity.

Recently, we have begun exploring the creation of catalytic sites through the rational organization of functional groups on bis-peptide scaffolds. Bis-amino acids are cyclic, stereochemically pure building blocks derived from (2S,4R)-4-hydroxyproline that are combined with other bis-amino acids and α-amino acids through functionalized diketopiperazines to create spiro-ladder oligomers with programmable shape and functionality.¹⁰ Our experience in synthesizing richly functionalized prolines led to the idea of positioning hydrophobic groups that could interact with the substrates of proline aldol catalysis in order to improve diastereoselectivity, enantioselectivity and turnover rate. Herein, we report a series of modified prolines that exhibit excellent diastereoselectivity and enantioselectivity with very low loading (0.5 mol%) in an aldol reaction in water. Moreover, by placing appropriate hydrophobic groups either syn or anti to the carboxylic acid of proline, we demonstrate a structure/catalytic activity relationship together with computational modeling that suggests that a specific hydrophobic interaction between the catalyst and the aldehyde substrate is responsible for an observed rate enhancement of 43.5-fold when compared to a catalyst that cannot form a hydrophobic interaction.

Results and Discussion

Modified proline 1a was synthesized starting from (2S,4R)-4-hydroxyproline 4 in eight steps with an overall yield of 44% (Scheme 1). Our group previously described the first five steps in the synthesis of 5, 6 and their enantiomers.¹⁰ The free amine of 5 was reductively alkylated overnight using benzaldehyde and sodium cyanoborohydride in methanol at room temperature. The resulting N-carboxybenzyl amino acid 7 was combined with 1.05 equiv. phenylisocyanate and 3 equiv. Et₃N at room temperature and stirred overnight. The hydantoin 12 formed cleanly with spontaneous dehydration; the formation of these tetrasubstituted hydantoins is facile under these mild conditions due to assistance by the Thorpe-Ingold effect and the presence of N-alkyl substitution on the amino acid.¹¹ The carboxybenzyl and tert-butyl protecting groups were then removed with 33% HBr in acetic acid and the HBr salt was converted to the TFA salt 1a. Using this reaction sequence starting from 5 and a variety of aldehydes in place of benzaldehyde afforded organocatalysts 1b–1e. Catalysts 2a and 2b, which are epimeric to 1a and 1b at the spiro carbon on the pyrrolidine ring, were prepared analogously starting from 6 (the C4 epimer of 5) and the appropriate aldehyde. Catalysts 1f and 2c were synthesized using similar procedures to 1a–1e, 2a and 2b but no intermediates were isolated. Finally, the organocatalyst 3 was prepared by converting the amino acid 7 to the 1-hydroxy-7-azabenzotriazole (HOAt) ester (not isolated) and combining it with the N-functionalized amino-isobutyric acid 22 to form the hexa-substituted diketopiperazine 17 using a new reaction that we have previously described called “acyl-transfer coupling”.^10b The carboxylbenzyl amine and tert-butyl ester of 17 were deblocked using HBr/AcOH and the salt was then converted to the trifluoroacetate salt 3.

Scheme 1.

Catalyst synthesis.

^aSee reference 10.

To probe the catalytic activity of 1a in the aldol reaction, 1a (10 mol %) was combined with 1 equiv. 4-nitrobenzaldehyde and 10 equiv. cyclohexanone at room temperature in a variety of solvents (Table 1). Solvent screening indicated that water and methanol afford the highest diastereomeric ratios and enantioselectivities (entries 1 and 2, the values are capped at 98% given the practical limits of quantitation by NMR).¹² The reaction catalyzed by 1a was found to be considerably faster in water than in methanol, as the consumption of the starting para-nitrobenzaldehyde remained incomplete in the latter solvent after 12 hours.

Table 1.

Solvent screening.

		Product: SM
Entry	Solvent	Ratioa	drb	eec (%)
1	H2O	>98:2	>98:2	>98
2	MeOH	76:24	>98:2	>98
3	IPA	>98:2	91:9	94
4	t-BuOH	>98:2	90:10	94
5	DMSO	60:40	78:22	89
6	DMF	41:59	79:21	87
7	DCM	97:3	86:14	96
8	MeCN	79:21	88:12	92
9	Toluene	98:2	96:4	96
10	Et2O	>98:2	95:5	93
11	EtOAc	>98:2	94:6	93
12	Hexane	>98:2	95:5	97
13	THF	>98:2	94:6	95
14	Cyclohexanone	>98:2	92:8	93

^a,bThe diastereomeric ratios were determined by 1D ¹H-NMR.

^cThe ee values were determined by HPLC using a Chiralpak AD column.

The scope of aldol catalysis using 1a in water against a variety of aldehydes and ketones as substrates is shown in Table 2. Generally, the reaction between cyclohexanone and aldehydes with an electron-withdrawing substituent was found to proceed to completion with high yield and excellent diastereoselectivity and enantioselectivity (entries 1, 4, 5, 6, and 10). The reaction of cyclohexanone and para-nitrobenzaldehyde catalyzed by 1a (0.5 mol %) provided the aldol product in 96% yield within 78 hours (entry 11). Substituted benzaldehydes containing less electron-withdrawing groups required longer reaction times to achieve good isolated yields, but the diastereo- and enantioselectivities remain excellent (entries 7 and 8). The methoxy-substituted aldehyde (electron-donating) showed poor conversion and slightly lower diastereoselectivity but excellent enantioselectivity was maintained (>98%, entry 9). The reactivity of 1a with cyclopentanone is good (entry 2: 9 hours, 98% yield) with a slight reduction in diastereoselectivity but excellent enantioselectivity. This is noteworthy, considering that there are few reports of catalyzed aldol reactions of cyclopentanone with high levels of diastereo- and enantioselectivity.^{9c, 9i, 9h, 13} Cycloheptanone required 10 mol % loading of 1a to achieve a yield of 77% in 48 hours (entry 3). The diastereoselectivity was modest (anti / syn = 1 : 4) but the enantioselectivity of the major, anti product was good (94% ee).

Table 2.

Substrate screening

Entry	n	R1	Time (h)	Yielda (%)	drb	eec (%)
1	2	p-NO2	12	96	>98:2	>98
2	1	p-NO2	9	98	95:5	>98
3d	3	p-NO2	48	77	80:20	94
4	2	o-NO2	19	92	>98:2	>98
5	2	m-NO2	13	98	>98:2	>98
6	2	p-CN	12	94	>98:2	>98
7	2	p-Br	100	80	97:3	>98
8	2	p-Cl	100	64	97:3	>98
9	2	p-OMe	120	18	92:8	>98
10	2	CO2Me	13	97	98:2	>98
11e	2	p-NO2	78	96	>98:2	>98
12	2	H	100	74	95:5	98

^aIsolated yield.

^bThe diastereomeric ratios were determined by ¹H-NMR of the crude mixture.

^cThe ee values were determined by HPLC using a Chiralpak AD column.

^d10 mol % catalyst.

^e0.5 mol % catalyst.

Catalyst 1a can be recycled, due to its good solubility in water (Table 3), by extracting the completed reaction mixture with hexanes and then adding fresh substrate to the aqueous layer. The high diastereoselectivity and enantioselectivity were maintained through three cycles with an approximately 50% decrease in activity, possibly due to the loss of some catalyst as it partitioned into the organic phase.

Table 3.

Catalyst recycling.

Cycle	Time(h)	Yielda (%)	drb	eec (%)
1st	12	88	>98:2	>98
2nd	15	91	>98:2	>98
3rd	25	80	98:2	>98

^aIsolated yield.

^bThe diastereomeric ratios were determined by ¹H-NMR of the crude product.

^cThe ee values were determined by HPLC using a Chiralpak AD column.

To probe the relationship between catalyst structure and activity, a series of derivatives 1a–1f, 2a–2c, and 3 (Scheme 1), which bear different aromatic groups varying stereochemistry and ring structures at the quaternary carbon, was synthesized. To a first approximation, the activities and selectivities of catalysts 1b–1e are indistinguishable from those of 1a. Switching from the 1a–1b catalysts to the corresponding 2a–2b catalysts involves inverting the C-4 stereocenter, which flips the aromatic side-arm to the opposite side of the proline ring (anti to the carboxylic acid) and leads to a large drop in catalyst rate, requiring over 7 times longer to achieve yields comparable to those of 1a–1b (Table 4). When the benzyl substituent on the hydantoin of 1a was changed to an ethyl group (catalyst 1f), the activity of the catalyst drops to that of 2a and 2b coupled with a slight decrease in diastereoselectivity. Expanding the hydantoin on the quaternary center of 1a to the diketopiperazine 3 (entry 8) demonstrates comparable activity to 1a with a slight reduction in diastereoselectivity. These observations are consistent with the formation of an additional hydrophobic interaction between the benzyl group on the hydantoin of 1a or the similarly placed aromatic group on the hydantoin of 1b–1e and 3 and the aldehyde substrate in the transition state. In toluene, the activity of 1a is almost identical to that of 1f and 2a, which would be expected when the hydrophobic effect is eliminated.

Table 4.

Screening catalysts

Entry	X	Solvent	Time (h)	Yielda (%)	drb	eec (%)
1	1a	H2O	12	96	>98:2	>98
2	1b	H2O	12	98	>98:2	>98
3	1c	H2O	13	99	>98:2	>98
4	1d	H2O	12	98	>98:2	>98
5	1e	H2O	12	97	>98:2	>98
6	2a	H2O	90	87	95:5	>98
7	2b	H2O	90	71	94:6	98
8	3	H2O	13	93	98:2	>98
9	1f	H2O	90	79	97:3	>98
10	2c	H2O	90	43	93:7	98
11	1ad	Toluene	16	98	95:5	97
12	1ad	Toluene	16	94	95:5	96
13	2ad	Toluene	16	90	95:5	89

^aIsolated yield.

^bdr values were determined by crude product NMR.

^cee values were determined by HPLC using a Chiralpak AD column.

^d10 mol % catalyst.

To gain more precise measurements of the relative catalytic rates we selected three catalysts: 1a, 2a and 1f and measured the reaction kinetics of the catalyzed aldol reactions using the procedure described in the experimental section. The relative amounts of product and starting material at a series of time-points were measured using NMR and the data was fit to a first order kinetic model using non-linear optimization. The results demonstrated that catalyst 1a reacts 43.5 times faster than 1f, which has the same stereochemistry as 1a but presents only an ethyl group and that 1a is 9 times faster than catalyst 2a, which is the C4-epimer of 1a (Figure 1).

Figure 1.

Linearized kinetic data for the aldol reactions of cyclohexanone and para-nitrobenzaldehyde catalyzed by 1a, 2a and 1f in water

To better understand the nature of the reaction transition states, quantum chemical computations were performed on the aldol reactions of cyclohexanone and benzaldehyde using 1a and 1f (Table 5). The geometries were optimized by M06-2X/6-31G(d) in the gas phase and using the SMD model with water as solvent. ¹⁴ In vacuum, the difference in free energy of TS-1a and TS-1f is very small (0.2 kcal/mol) (Table 5). This is consistent with the very similar rates of 1a and 1f in toluene (entries 11 and 12 in Table 4). On the other hand, in water, the free energy of activation associated with 1a is more stabilizing than 1f by 2.6 kcal/mol (Table 5). The optimized transition states taking into account the solvation effect of water are illustrated in the Figure 2. The benzyl side-arm of 1a comes into edge-to-face contact with the aromatic ring of the acceptor aldehyde in the List-Houk transition state, affording stabilization of TS-1a_water relative to TS-1f_water. This is in excellent agreement with the experimentally observed 43.5-fold acceleration with the use of 1a in water when compared with 1f. Using transition state theory, the value of ΔΔG^‡ associated with a k_rel of 43.5 is −2.3 kcal/mol at 300 K. The SMD model treats solvent as a dielectric medium with a surface tension term calculated at the solute-solvent boundary that models the hydrophobic effect.¹⁴ These results are consistent with the hypothesis that the observed acceleration is due to a hydrophobic interaction in 1a that is not present in 1f.¹⁵

Table 5.

Gibbs Free energies of transition structures for the aldol addition step relative to isolated reactants and organocatalyst (kcal/mol)^a

	ΔG‡	ΔΔG‡ (1a – 1f)
TS-1avacuum	34.6	0.2
TS-1fvacuum	34.4	—
TS-1awater	24.7	−2.6
TS-1fwater	27.3	—

^aThe geometries were fully optimized by M06-2X/6-31G(d) in vacuum or using the SMD model to account for the solvation effect of water.

Figure 2.

Optimized transition structures for the aldol addition step catalyzed by 1a and 1f in water

In summary, we present a series of proline-based catalysts 1a–1e, and 3 that catalyze the aldol reaction in water with high activity and excellent diastereo- and enantioselectivity. Both enantiomers of these catalysts are synthetically accessible starting from (2S,4R)-4-hydroxyproline¹⁰ making them practical catalysts for the synthesis of stereochemically pure aldol products. Catalysts that present an aromatic group that can interact with the aromatic ring of the aldehyde acceptor through a specific hydrophobic interaction are significantly faster than those that do not. Quantum chemical calculations point to a hydrophobic edge-to-face interaction that stabilizes the transition state of 1a relative to 1f by 2.6 kcal/mol, in excellent agreement with the experimentally observed ΔΔG^‡ of 2.3 kcal/mol. These results suggest that the rational positioning of hydrophobic functional groups can significantly stabilize transition states through hydrophobic interactions in water and that stepwise incorporation of more groups could lead to proline aldol catalysts with even greater activity.

Experimental Section

General procedure for synthesis of catalyst 1a–1e, 2a and 2b

Step 1

To a solution of (3S,5S)-1-((benzyloxy)carbonyl)-5-(tert-butoxycarbonyl)-3-aminopyrrolidine-3-carboxylic acid 5 (2.0 g, 5.5 mmol, 1.0 eq) in methanol (60 mL), benzaldehyde (670 µL, 6.6 mmol, 1.2 eq) was added and stirred for 30 minutes at room temperature. Sodium cyanoborohydride (518 mg, 8.3 mmol, 1.5 eq) was then added and stirred for 20 hours at room temperature until the reaction was complete, as monitored by LC-MS. The reaction mixture was concentrated under vacuum and the resulting residue was dissolved in 50 mL H₂O. The pH was adjusted to 7 by dropwise addition of 2 N HCl and the white precipitate (2.2 g, 4.8 mmol, 88% yield) was obtained by vacuum filtration. This product was used without further purification in the next step. For characterization, the compound was purified by reverse phase C₁₈ high performance liquid chromatography using a 5–95% H₂O/Acetonitrile gradient.

(3S,5S)-3-(Benzylamino)-1-[(benzyloxy)carbonyl]-5-[(tert-butoxy)carbonyl]pyrrolidine-3-carboxylic acid, 7

¹H NMR (500 MHz, d⁶-DMSO, 350 K,δ): 1.38 (9H, s), 2.36 (1H, dd, J= 8.5, 13.3), 2.93 (1H, dd, J= 8.4, 13.3), 3.70 (1H, d, J= 11.4), 4.19 (3H, m), 4.34 (1H, t, J= 8.2), 5.11 (2H, s), 7.30–7.50 (10H, m); ¹³C NMR (125 MHz, d⁶-DMSO, 350 K, δ): 28.0, 28.2, 30.9, 49.0, 55.0, 59.4, 66.7, 81.2, 127.2, 127.8, 128.1, 128.5, 128.7, 137.3, 140.4, 154.2, 170.9, 174.4, 206.2; HPLC analysis (C₁₈ reverse phase, 40 mins, 5–100% H₂O/ACN with 0.1% formic acid): T_r = 17.3 mins; HRMS-ESI: m/z Calculated for C₂₅H₃₁N₂O₆ (M+H)⁺ 455.2177; Found: 455.2182.

Step 2

To a solution of 7 (2.2 g, 4.8 mmol, 1.0 eq), triethylamine (1.4 mL, 10.2 mmol, 2.1 eq) in 130 mL of anhydrous THF, phenyl isocyanate (1.1 mL, 9.7 mmol, 2.0 eq) was added in one portion. The reaction was stirred at room temperature for 20 hours until completion, as determined by LC-MS. The reaction was quenched with saturated NH₄Cl (aq.) (40 mL) and extracted with EtOAc (100 mL + 2x50 mL). The combined organic layers were washed with brine and dried over anhydrous Na₂SO₄. The yellow, solid product was obtained by removal of the solvent in vacuum. The crude product 12 was used directly in the next step without further purification. For characterization, the compound was purified by reverse phase column using 5–95% H₂O/Acetonitrile gradient. Note: diphenyl urea was found in the crude product mixture (detected by LCMS), but it did not interfere with the next step.

7-Benzyl-8-tert-butyl(5S,8S)-1-benzyl-2,4-dioxo-3-phenyl-1,3,7-triazaspiro[4,4]nonane-7,8-dicarboxylate, 12

¹H NMR (500 MHz, d⁶-DMSO, 350 K, δ): 1.34 (9H, s), 2.21 (1H, dd, J= 8.3, 13.8), 2.72 (1H, dd, J=8.6, 13.7), 3.60 (1H, d, J= 11.8), 4.00 (1H, d, J= 11.8), 4.39 (1H, t, J= 8.4), 4.69 (2H, s), 5.10 (2H, s), 7.29–7.43 (11H, m), 7.50 (4H, m); ¹³C NMR (125 MHz, CDCl₃, 298K, δ): 27.8, 36.3, 43.4, 50.3, 58.3, 66.3, 67.6, 82.1, 125.8, 127.7, 127.8, 128.0, 128.1, 128.3, 128.4, 129.0, 129.1, 131.3, 136.0, 136.8, 154.1, 154.8, 170.8, 173.2; IR (neat): 1692, 1402, 1348, 1149, 1124 cm⁻¹; [α]_D¹⁸ = −5.0 (MeOH); HPLC analysis (C₁₈ reverse phase, 40 mins, 5–100% H₂O/ACN with 0.1 % formic acid): T_r = 26.3 mins; HRMS-ESI: m/z Calculated for C₃₂H₃₃N₃NaO₆ (M+Na)⁺ 578.2262; Found: 578.2267.

Step 3

Ph-[Cbz, t-Bu] Pro4 (SS) hydantoin 12 (4.8 mmol) was dissolved in CH₂Cl₂ (20 mL) followed by the addition of 33% HBr/AcOH (40 mL). The reaction was stirred at room temperature for 2 hours. Removal of solvents by vacuum evaporation produced a dark red liquid that was purified by reverse phase using 5–95% H₂O / Acetonitrile gradient to afford the product as a white powder. The amino acid was mixed with 50% TFA/CH₂Cl₂ for 30 minutes with stirring. After concentration under vacuum, the remaining solution (1 to 2 mL volume) was added dropwise to 200 mL diethyl ether/hexane (1:1) and a white precipitate 1a formed (2.0 g, 85% yield from 7). The precipitate was filtered and dried under high vacuum.

(5S,8S)-1-Benzyl-2,4-dioxo-3-phenyl-1,3,7-triazaspiro[4,4]nonane-8-carboxylic acid.TFA, 1a

¹H NMR (500 MHz, d⁶-DMSO, 350 K, δ): 2.48 (1H, m), 2.85 (1H, dd, J= 6.8, 13.7), 3.58 (1H, d, J= 13.4), 3.74 (1H, d, J= 13.4), 4.55 (1H, dd, J= 6.8, 12.3), 4.71 (2H, s), 7.29–7.53 (10H, m); ¹³C NMR (125 MHz, d⁶-DMSO, 350 K, δ): 35.3, 42.8, 49.0, 59.7, 68.4, 127.2, 127.3, 127.6, 128.6, 128.9, 129.1, 132.2, 138.0, 154.9, 169.6, 174.2; IR (neat): 1775, 1718, 1495, 1414, 1181, 1134 cm⁻¹; [α]_D¹⁸ = +10.9 (MeOH); HPLC analysis (C₁₈ reverse phase, 40 mins, 5–100% H₂O/ACN with 0.1 % formic acid): T_r = 11.9 mins; HRMS-ESI: m/z Calculated for C₂₀H₂₀N₃O₄ (M+H)⁺ 366.1448; Found: 366.1450.

Synthetic procedure for catalyst 1b

Starting from Cbz, t-Butyl Pro4(SS) amino acid 5 (728 mg, 2.0 mmol), catalyst 1b was synthesized following the same synthetic procedure of 1a with the exception of substituting 1-naphthylaldehyde for benzaldehyde in step 1.

(3S,5S)-1-[(Benzyloxy)carbonyl]-5-[(tert-butoxy)carbonyl]-3-[(naphthanlen-1-ylmethyl)amino]pyrrolidine-3-carboxylic acid, 8

807mg, 80% crude yield in Step 1; ¹H NMR (500 MHz, d⁶-DMSO, 350 K, δ): 1.38 (9H, s), 2.40 (1H, dd, J=8.1, 13.2), 3.00 (1H, dd, J=8.5, 13.1), 3.75 (1H, d= 11.3), 4.24 (1H, d, J= 11.3), 4.37 (1H, t, J= 7.9), 4.57 (2H, dd, J= 13.2, 22.5), 5.13(2H, s), 7.32–7.39 (5H, m), 7.53–7.64 (3H, m), 7.70 (1H, d), 7.98 (2H, t, J= 9.3), 8.24 (1H, d, J= 8.5); ¹³C NMR (125 MHz, d⁶-DMSO, 350 K, δ): 28.0, 28.1, 37.1, 46.5, 53.4, 59.4, 67.1, 81.8, 124.1, 125.7, 126.5, 127.0, 127.8, 128.3, 129.0, 129.1, 129.7, 130.8, 131.8, 133.9, 137.0, 154.1, 170.6, 171.7; HPLC analysis (C₁₈ reverse phase, 40 mins, 5–100% H₂O/ACN with 0.1 % formic acid): T_r = 19.7 mins; HRMS-ESI: m/z Calculated for C₂₉H₃₃N₂O₆ (M+H)⁺ 505.2333; Found: 505.2343.

7-Benzyl-8-tert-butyl(5S,8S)-1-(naphthalen-1-ylmethyl)-2,4-dioxo-3-phenyl-1,3,7-triazaspiro[4,4]nonane-7,8-dicarboxylate, 13

¹H NMR(500 MHz, d⁶-DMSO, 350 K, δ): 1.20 (9H, s), 2.19 (1H, dd, J= 8.4, 13.7), 2.72 (1H, dd, J= 8.5, 13.7), 3.61 (1H, d, J= 11.8), 4.03 (1H, d, J= 11.8), 4.35 (1H, t, J=8.4), 5.06 (2H, s), 5.18 (2H, s), 7.28–7.33 (5H, m), 7.42–7.45 (1H, m), 7.47–7.63 (8H, m), 7.88 (1H, d), 7.98 (1H, d), 8.21 (1H, d); ¹³C NMR (125 MHz, CDCl₃, 298 K, δ): 27.6, 35.4, 42.0, 49.6, 58.2, 66.1, 67.5, 81.9, 123.1, 125.3, 125.9, 126.1, 126.3, 127.0, 127.8, 128.0, 128.4, 129.0, 129.1, 129.3, 131.0, 131.3, 134.0, 136.0, 154.1, 154.7, 170.5, 173.2. IR (neat): 1694, 1403, 1349, 1149, 1124 cm⁻¹; [α]_D¹⁸ = −3.6 (MeOH); HPLC analysis (C₁₈ reverse phase, 40 mins, 5–100% H₂O/ACN with 0.1 % formic acid): T_r =27.6 mins; HRMS-ESI: m/z Calculated for C₃₆H₃₅N₃NaO₆ (M+Na)⁺ 628.2418; Found: 628.2422.

(5S,8S)-1-(Naphthanlen-1-ylmethyl)-2,4-dioxo-3-phenyl-1,3,7-triazaspiro[4,4]nonane-8-carboxylic acid.TFA, 1b

592mg, 70% yield from 13; ¹H NMR (500 MHz, d⁶-DMSO, 350 K, δ): 2.54 (1H, m), 2.98 (1H, dd, J= 6.7, 13.7), 3.66 (1H, d, J= 13.4), 3.83 (1H, d, J= 13.4), 4.57 (1H, dd, J= 6.7, 12.5), 5.23 (2H, s), 7.42–7.87 (9H, m), 7.88(1H, d, J= 7.8), 7.98 (1H, d, J= 7.5), 8.17 (1H, d, J= 8.5); ¹³C NMR (125 MHz, d⁶-DMSO, 350 K, δ): 34.6, 41.3, 48.4, 59.2, 68.0, 122.8, 123.5, 126.0, 126.5, 126.8, 127.3, 127.8, 128.6, 129.1, 129.2, 130.4, 132.2, 132.8, 133.6, 154.8, 169.0, 174.2; IR (neat): 1775, 1721, 1502, 1415, 1385, 1185, 1138 cm⁻¹; [α]_D¹⁸ = +13.1 (MeOH); HPLC analysis (C₁₈ reverse phase, 40 mins, 5–100% H₂O/ACN with 0.1 % formic acid): T_r = 14.3 mins; HRMS-ESI: m/z Calculated for C₂₄H₂₂N₃O₄ (M+H)⁺ 416.1605; Found: 416.1611.

Synthetic procedure for catalyst 1c

Starting from Cbz, t-Butyl Pro4(SS) amino acid 5 (364 mg, 1.0 mmol), the synthesis of catalyst 1c follows the same procedure as that of catalyst 1a with the following changes: (1) 1-pyrenealdehyde substituted for benzaldehyde in step 1; (2) after the methanol was removed by concentration under vacuum, 10 ml of H₂O was added to the residue and 2 drops of 2N KOH (aq.) was added to assist the dissolution of the amino acid before adjusting the pH to 7 in step 1.

(3S,5S)-1-[(Benzyloxy)carbonyl]-5-[(tert-butoxy)carbonyl]-3-[(pyren-1-ylmethyl)amino]pyrrolidine-3-carboxylic acid, 9

462mg, 80% crude yield in Step 1; ¹H NMR (500HMz, d⁶-DMSO, 350 K, δ): 1.41 (9H, s), 2.74 (1H, dd, J= 8.9, 13.1), 3.06 (1H, dd, J= 8.3, 12.8), 4.03 (1H, d, J= 11.4), 4.33 (1H, d, J= 11.4), 4.42 (1H, t, J= 8.2), 4.97 (2H, dd, J= 13.5, 23.9), 5.15 (2H, s), 7.31–7.40 (5H, m), 8.12 (1H, t, J= 7.7), 8.21 (2H, m), 8.31–8.39 (5H, m), 8.58 (1H, d, J= 9.3); ¹³C NMR (125 MHz, d⁶-DMSO, 350 K, δ): 28.0, 28.1, 36.2, 46.5, 52.8, 59.4, 67.1, 80.0, 123.6, 124.3, 124.6, 126.1, 126.0, 126.2, 126.5, 126.9, 127.7, 127.8, 128.3, 128.6, 128.8, 129.9, 130.8, 131.3, 132.1, 137.0, 154.1, 170.2, 170.8; HPLC analysis (C₁₈ reverse phase, 30 mins, 5–95% H₂O/ACN with 0.1 % formic acid): T_r = 22.5 mins; HRMS-ESI: m/z Calculated for C₃₅H₃₅N₂O₆ (M+H)⁺ 579.2490; Found: 579.2500.

7-Benzyl-8-tert-butyl(5S,8S)-2,4-dioxo-3-phenyl-1-(pyren-1-ylmethyl)-1,3,7-triazaspiro[4,4]nonane-7,8-dicarboxylate, 14

¹H NMR (500 MHz, d⁶-DMSO, 350 K, δ): 1.08 (9H, s), 2.13 (1H, dd, J= 8.5, 13.6), 2.69 (1H, dd, J= 8.5, 13.5), 3.66 (1H, d, J= 11.8), 4.05 (1H, d, J= 11.8), 4.33 (1H, t, J= 8.5), 5.03 (2H, s), 5.49 (2H, dd, J= 16.9, 29.6), 7.26–7.30 (5H, m), 7.43–7.46 (1H, m), 7.52–7.59 (4H, m), 8.10 (1H, m), 8.15–8.20 (3H, m), 8.28–8.34 (4H, m), 8.52 (1H, d, J= 9.3); ¹³C NMR (125 MHz, CDCl₃, 298K., δ): 27.5, 35.4, 42.1, 49.4, 58.1, 66.1, 67.5, 81.8, 122.3, 124.6, 125.0, 125.1, 125.6, 125.7, 125.9, 126.3, 126.6, 127.3, 127.7, 127.9, 128.4, 128.7, 128.8, 129.1, 130.7, 131.3, 131.6, 136.0, 154.1, 154.9, 170.3, 173.2; IR (neat): 1839, 1620, 1486, 1387, 1332, 1116 cm⁻¹; [α]_D¹⁸ = −11.2 (MeOH); HPLC analysis (C₁₈ reverse phase, 40 mins, 5–100% H₂O/ACN with 0.1 % formic acid): T_r = 29.9 mins; HRMS-ESI: m/z Calculated for C₄₂H₃₇N₃NaO₆ (M+Na)⁺ 702.2575; Found: 702.2569.

(5S,8S)-2,4-Dioxo-3-phenyl-1-(pyren-1-ylmethyl)-1,3,7-triazaspiro[4,4]nonane-8-carboxylic acid.TFA, 1c

352mg, 73% yield from 14. ¹H NMR (500 MHz, d⁶-DMSO, 350 K, δ): 2.56 (1H, t, J= 13.2), 3.09 (1H, dd, J= 6.6, 13.5), 3.64 (1H, d, J= 13.6), 3.87 (1H, d, J= 13.6), 4.57 (1H, dd, J= 6.7, 12.8), 5.53 (2H, s), 7.40–7.65 (5H, m), 8.10–8.48 (10H, m); ¹³C NMR (125 MHz, d⁶-DMSO, 350 K, δ): 34.1, 41.7, 48.2, 58.8, 67.8, 123.1, 123.7, 124.3, 124.4, 125.4, 125.8, 125.9, 126.9, 127.3, 127.4, 127.5, 127.9, 128.1, 128.7, 129.2, 130.4, 130.7, 130.9, 131.4, 132.2, 155.0, 168.8, 174.2; IR (neat): 1777, 1719, 1502, 1415, 1186, 1139 cm⁻¹; [α]_D¹⁸ = +3.5 (MeOH); HPLC analysis (C₁₈ reverse phase, 40 mins, 5–100% H₂O/ACN with 0.1 % formic acid): T_r = 17.4 mins; HRMS-ESI: m/z Calculated for C₃₀H₂₄N₃O₄ (M+H)⁺ 490.1761; Found: 490.1768.

Synthetic procedure for catalyst 1d

Starting from Cbz, t-Butyl Pro4(SS) amino acid 5 (364 mg, 1.0 mmol), the synthesis of catalyst 1d follows the same procedure as that of catalyst 1a with the following exceptions: (1) 4-nitrobenzaldehyde substituted for benzaldehyde in step 1; (2) after the methanol was removed by concentration under vacuum, 10 ml of H₂O was added to the residue and 2 drops of 2N KOH (aq.) was added to assist the dissolution of the amino acid before adjusting the pH to 7 in step 1.

(3S,5S)-1-[(Benzyloxy)carbonyl]-5-[(tert-butoxy)carbonyl]-3-([(4-nitrophenyl)methyl]amino)pyrrolidine-3-carboxylic acid, 10

439mg, 88% crude yield in Step 1; ¹H NMR (500 MHz, d⁶-DMSO, 350 K, δ): 1.37 (9H, s), 2.22 (1H, dd, J= 6.8, 13.1), 2.82 (1H, dd, J= 8.8, 13.0), 3.56 (1H, d= 11.2), 4.00–4.08 (3H, m), 4.32 (1H, t, J=7.6), 5.10 (2H, s), 7.30–7.38 (5H, m), 7.68 (2H, d, J= 8.7), 8.19 (2H, d, J= 8.7); ¹³C NMR (125 MHz, d⁶-DMSO, 350 K, δ): 28.0, 28.1, 38.1, 48.3, 54.1, 59.3, 66.9, 81.6, 123.6, 127.8, 128.2, 128.7, 130.2, 137.1, 145.7, 147.7, 154.1, 170.6, 172.9; HPLC analysis (C₁₈ reverse phase, 30 mins, 5–95% H₂O/ACN with 0.1 % formic acid): T_r = 18.7 mins; HRMS-ESI: m/z Calculated for C₂₅H₃₀N₃O₈ (M+H)⁺ 500.2027; Found: 500.2032.

7-Benzyl-8-tert-butyl(5S,8S)-1-[(4-nitrophenyl)methyl)-2,4-dioxo-3-phenyl-1,3,7-triazaspiro[4,4]nonane-7,8-dicarboxylate, 15

¹H NMR (500 MHz, d⁶-DMSO, 350 K, δ): 1.34 (9H, s), 2.21 (1H, dd, J= 8.0, 13.8), 2.83 (1H, dd, J= 8.7, 13.8), 3.61 (1H, d, J= 11.9), 4.07 (1H, d, J= 11.9), 4.42 (1H, t, J= 8.3), 4.82 (2H, dd, J= 17.3, 25.4), 5.09 (2H, s), 7.30–7.34 (5H, m), 7.42 (1H, m), 7.48–7.56 (4H, m), 7.68 (2H, d, J= 8.8), 8.19 (2H, d, J= 8.8); ¹³C NMR (125 MHz, CDCl₃, 298K, δ): 27.8, 37.3, 42.8, 51.6, 58.4, 66.8, 67.8, 82.6, 124.1, 125.7, 128.0, 128.3, 128.5, 129.2, 131.1, 135.8, 144.3, 147.6, 154.1, 154.9, 171.0, 172.7; IR (neat): 1704, 1596, 1516, 1493, 1402, 1340, 1125 cm⁻¹; [α]_D¹⁸ = +0.75 (MeOH); HPLC analysis (C₁₈ reverse phase, 40 mins, 5–100% H₂O/ACN with 0.1 % formic acid): T_r = 25.8 mins; HRMS-ESI: m/z Calculated for C₃₂H₃₂N₄NaO₈ (M+Na)⁺ 623.2112; Found: 623.2111.

(5S,8S)-1-[(4-Nitrophenyl)methyl]-2,4-dioxo-3-phenyl-1,3,7-triazaspiro[4,4]nonane-8-carboxylic acid.TFA, 1d

290mg, 63% yield from 15; ¹H NMR (500 MHz, d⁶-DMSO, 350 K, δ): 2.43 (1H, t, J= 13.2), 2.95 (1H, dd, J=6.6, 13.5), 3.51 (1H, d, J= 13.6), 3.74 (1H, d, J= 13.5), 4.53 (1H, dd, J= 6.6, 12.8), 4.87 (2H, s), 7.41–7.54 (5H, m), 7.72 (2H, d, J= 8.8), 8.24 (2H, d, J= 8.8); ¹³C NMR (125 MHz, DMSO, 350 K, δ): 34.1, 42.6, 48.3, 58.7, 67.5, 124.0, 127.2, 128.3, 128.7, 129.2, 132.0, 145.9, 147.2, 154.9, 168.7, 173.9; IR (neat): 1777, 1718, 1517, 1413, 1345 cm⁻¹; [α]_D¹⁸ = +1.1 (MeOH); HPLC analysis (C₁₈ reverse phase, 40 mins, 5–100% H₂O/ACN with 0.1 % formic acid): T_r = 12.7 mins; HR-MS (M+H)⁺ Calculated for C₂₀H₁₉N₄O₆⁺: 411.1299; found: 411.1303.

Synthetic procedure for catalyst 1e

Starting from Cbz, t-Butyl Pro4(SS) amino acid 5 (364 mg, 1.0 mmol), the synthesis of catalyst 1e follows the same procedure as that of catalyst 1a with the following changes: (1) 4-methoxybenzaldehyde substituted for benzaldehyde in step 1; (2) after the methanol was removed by concentration under vacuum, 10 ml of H₂O was added to the residue and 2 drops of 2N KOH (aq.) was added to assist the dissolution of the amino acid before adjusting the pH to 7 in step 1.

(3S,5S)-1-[(Benzyloxy)carbonyl]-5-[(tert-butoxy)carbonyl]-3-([(4-methoxyphenyl)methyl]amino)pyrrolidine-3-carboxylic acid, 11

411mg, 85% crude yield in Step 1; ¹H NMR (500 MHz, d⁶-DMSO, 350 K, δ): 1.40 (9H, s), 2.38 (1H, dd, J=8.6, 13.3), 2.92 (1H, dd, J= 8.4, 13.2), 3.72 (1H, d, J= 11.4), 3.79 (3H, s), 4.10 (1H, d, J= 12.6), 4.17 (2H,m), 4.35 (1H, t, J= 8.3), 5.12 (2H, s), 6.99 (2H, d, J= 8.7), 7.32–7.42 (7H, m); ¹³C NMR (125 MHz, d⁶-DMSO, 350 K, δ): 28.0, 28.1, 39.1, 48.5, 55.0, 55.7, 59.5, 66.6, 81.1, 114.2, 127.7, 128.1, 128.7, 129.8, 132.3, 137.3, 154.2, 159.0, 171.0, 174.4; HPLC analysis (C₁₈ reverse phase, 30 mins, 5–95% H₂O/ACN with 0.1 % formic acid): T_r = 17.4 mins; HRMS-ESI: m/z Calculated for C₂₆H₃₃N₂O₇ (M+H)⁺ 485.2282; Found: 485.2292.

7-Benzyl-8-tert-butyl(5S,8S)-1-[(4-methoxyphenyl)methyl)-2,4-dioxo-3-phenyl-1,3,7-triazaspiro[4,4]nonane-7,8-dicarboxylate, 16

¹H NMR (500 MHz, d⁶-DMSO, 350 K, δ): 1.36 (9H, s), 2.20 (1H, dd, J= 8.3, 13.8), 2.69 (1H, dd, J= 8.6, 13.7), 3.61 (1H, d, J= 11.8), 3.76 (3H, s), 3.97 (1H, d, J= 11.8), 4.38 (1H, t, J= 8.4), 4.62 (2H, s), 5.09 (2H, s), 6.91 (2H, d, J= 8.7), 7.28–7.38 (7H, m), 7.42 (1H, m), 7.47–7.53 (4H, m); ¹³C NMR (125 MHz, CDCl₃, 298K, δ): 27.8, 36.2, 42.9, 50.2, 55.3, 58.3, 66.2, 67.6, 82.1, 114.4, 125.8, 127.8, 128.1, 128.3, 128.5, 128.9, 129.1, 129.2, 129.4, 131.3, 136.1, 154.1, 154.7, 159.4, 170.8, 173.3; IR (neat): 1682, 1607, 1493, 1399, 1344, 1239, 1121 cm⁻¹; [α]_D¹⁸ = −4.0 (MeOH); HPLC analysis (C₁₈ reverse phase, 40 mins, 5–100% H₂O/ACN with 0.1 % formic acid): T_r = 25.9 mins; HRMS-ESI: m/z Calculated for C₃₃H₃₅N₃NaO₇ (M+Na)⁺ 608.2367; Found: 608.2377.

(5S,8S)-1-[(4-Methoxyphenyl)methyl]-2,4-dioxo-3-phenyl-1,3,7-triazaspiro[4,4]nonane-8-carboxylic acid.TFA, 1e

207mg, 48% yield from 16; ¹H NMR(500 MHz, d⁶-DMSO, 350 K, δ): 2.46 (1H, m), 2.80 (1H, dd, J= 6.7, 13.6), 3.57 (1H, d, J=13.3), 3.70 (1H, d, J=13.4), 3.76 (3H, s), 4.54 (1H, dd, J=6.8, 12.1), 4.62 (2H, s), 6.91 (2H, d, J=8.7), 7.34 (2H, d, J=8.7), 7.39–7.43 (1H, m), 7.49–7.51 (4H, d, J=4.3); ¹³C NMR (125 MHz, d⁶-DMSO, 350 K, δ): 34.5, 42.6, 48.3, 55.6, 58.9, 67.9, 114.6, 127.1, 128.5, 129.0, 129.1, 129.8, 132.3, 154.8, 159.3, 168.6, 173.9; IR (neat): 1775, 1719, 1513, 1414, 1247, 1179, 1135 cm⁻¹; [α]_D¹⁸ = +7.8 (MeOH); HPLC analysis (C₁₈ reverse phase, 40 mins, 5–100% H₂O/ACN with 0.1 % formic acid): T_r = 12.2 mins; HRMS-ESI: m/z Calculated for C₂₁H₂₂N₃O₅ (M+H)⁺ 396.1554; Found: 396.1562.

Synthetic procedure for catalyst 2a

The synthesis of catalyst 2a follows the same procedure as that of catalyst 1a with the change of beginning with Cbz, t-Butyl Pro4(SR) amino acid 6 (364 mg, 1.0 mmol) rather than Cbz, t-Butyl Pro4(SS) amino acid 5.

(3R,5S)-3-(Benzylamino)-1-[(benzyloxy)carbonyl]-5-[(tert-butoxy)carbonyl]pyrrolidine-3-carboxylic acid, 18

341mg, 75% crude yield in Step 1; ¹H NMR (500 MHz, d⁶-DMSO, 350 K, δ): 1.38 (9H, s), 2.58 (1H, dd, J=5.6, 14.1), 2.88 (1H, t, J=9.0), 3.84 (1H, d, J=12.5), 4.07–4.16 (3H, m), 4.51 (1H, t, J=6.7), 5.11 (2H, s), 7.30–7.48 (10H, m); ¹³C NMR (125 MHz, d⁶-DMSO, 350 K, δ): 28.1, 28.2, 39.0, 48.4, 54.7, 59.5, 66.7, 81.2, 127.1, 127.8, 128.1, 128.3, 128.5, 128.7, 137.3, 140.8, 154.5, 171.2, 173.5; HPLC analysis (C₁₈ reverse phase, 40 mins, 5–100% H₂O/ACN with 0.1 % formic acid): T_r = 16.5 mins; HRMS-ESI: m/z Calculated for C₂₅H₃₁N₂O₆ (M+H)⁺ 455.2177; Found: 455.2181.

7-Benzyl-8-tert-butyl(5R,8S)-1-benzyl-2,4-dioxo-3-phenyl-1,3,7-triazaspiro[4,4]nonane-7,8-dicarboxylate, 20

¹H NMR (500 MHz, d⁶-DMSO, 350 K, δ): 1.37 (9H, s), 2.40 (1H, dd, J=6.9, 14.6), 2.77 (1H, dd, J=9.4, 14.3), 3.63 (1H, d, J=12.1), 3.86 (1H, d, J=11.0), 4.35 (1H, broad), 4.60 (2H, broad), 5.05 (1H, d, J=12,1), 7.28–7.35 (11H, m), 7.50 (4H, d, J=4.4); ¹H NMR(500 MHz, CDCl₃, 298K, δ, rotamersobserved): 1.37–1.45 (9H, d), 2.41–2.62 (2H, m), 3.76 (0.5H, d, J=11.8), 3.87 (0.5H, dd, J=11.8), 3.91 (0.5H, d, J=12.1), 3.99 (0.5H, d, J=12.1), 4.14 (0.5H, dd, J=6.8, 9.3), 4.43–4.51 (1.5H, m), 4.76 (0.5H, d, J=15.8), 4.84 (0.5H, d, J=15.8), 4.92 (0.5H, d, J=12.2), 5.11 (0.5H, d, J=12.2), 5.17 (1H, tt, J=12.5), 7.23–7.53 (15H, m); ¹³C NMR (125 MHz, CDCl₃, 298K, δ): 27.8, 27.9, 38.6, (29.9, 44.5, 54.6, 58.9, 59.5, 67.5, 68.4, 69.3, 82.5, 125.8, 127.5, 127.6, 128.0, 128.1, 128.2, 128.3, 128.5, 128.6, 128.9, 129.1, 131.5, 135.9, 136.0, 136.6, 136.7, 153.4, 153.7, 154.7, 154.8, 169.6, 169.8, 171.0, 171.3; IR (neat): 1702, 1492, 1406, 1353, 1139 cm⁻¹; [α]_D¹⁸ = −88.1 (MeOH); HPLC analysis (C₁₈ reverse phase, 40 mins, 5–100% H₂O/ACN with 0.1 % formic acid): T_r = 25.9 mins; HRMS-ESI: m/z Calculated for C₃₂H₃₃N₃NaO₆ (M+Na)⁺ 578.2262; Found: 578.2267.

(5R,8S)-1-Benzyl-2,4-dioxo-3-phenyl-1,3,7-triazaspiro[4,4]nonane-8-carboxylic acid.TFA, 2a

227mg, 63% yield from 20; ¹H NMR (500 MHz, d⁶-DMSO, 350 K, δ): 2.61 (1H, dd, J= 9.7, 14.5), 2.71 (1H, dd, J= 5.9, 14.5), 3.54 (1H, d, J= 13.2), 3.80 (1H, d, J= 13.2), 4.62 (1H, t, J= 6.0), 4.71 (2H, dd, J= 16.9, 51.8), 7.25–7.55 (10H, m); ¹³C NMR (125 MHz, d⁶-DMSO, 350 K, δ): 35.1, 43.1, 49.7, 58.8, 68.0, 127.2, 127.3, 127.7, 128.6, 129.0, 129.1, 132.2, 137.9, 155.0, 169.3, 173.3; IR (neat): 1776, 1714, 1494, 1415, 1185, 1135 cm⁻¹; [α]_D¹⁸ = +13.1 (MeOH); HPLC analysis (C₁₈ reverse phase, 30 mins, 5–95 % H₂O/ACN with 0.1 % formic acid): T_r = 11.6 mins; HRMS-ESI: m/z Calculated for C₂₀H₂₀N₃O₄ (M+H)⁺ 366.1448; Found: 366.1454.

Synthetic procedure for catalyst 2b

The synthesis of catalyst 2b follows the same procedure as that of catalyst 1b with the exception of Cbz, t-Butyl Pro4(SR) amino acid 6 (364 mg, 1.0 mmol), substituted for Cbz, t-Butyl Pro4(SS) amino acid 5.

(3R,5S)-1-[(Benzyloxy)carbonyl]-5-[(tert-butoxy)carbonyl]-3-[(naphthanlen-1-ylmethyl)amino]pyrrolidine-3-carboxylic acid, 19

439mg, 87% crude yield in Step 1; ¹H NMR (500 MHz, d⁶-DMSO, 350 K, δ): 1.41 (9H, s), 2.63 (1H, d, J=13.5), 2.94 (1H, t, J=9.9), 3.89 (1H, d, J=12.4), 4.14 (1H, d, J=12.5), 4.50–4.61 (3H, m), 5.12 (2H, s), 7.30–8.30 (12H, m); ¹³C NMR (125 MHz, d⁶-DMSO, 350 K, δ): 28.0, 28.1, 45.6, 53.8, 59.1, 67.1, 81.9, 124.0, 124.1, 125.7, 126.6, 127.2, 128.0, 128.3, 128.7, 129.1, 129.2, 129.4, 129.6, 130.0, 130.1, 131.8. 134.0. 136.9, 153.9, 170.2; HPLC analysis (C₁₈ reverse phase, 40 mins, 5–100% H₂O/ACN with 0.1 % formic acid): T_r = 19.2 mins; HRMS-ESI: m/z Calculated for C₂₉H₃₃N₂O₆ (M+H) ⁺ 505.2333; Found: 505.2342.

7-Benzyl-8-tert-butyl(5R,8S)-1-(naphthalen-1-ylmethyl)-2,4-dioxo-3-phenyl-1,3,7-triazaspiro[4,4]nonane-7,8-dicarboxylate, 21

¹H NMR (500 MHz, d⁶-DMSO, 350 K, δ): 1.33 (9H, s), 2.41 (1H, dd, J=6.1, 14.6), 2.78 (1H, s), 3.63 (1H, d, J=12.1), 3.87 (1H, broad), 4.18 (1H, broad), 4.72 (1H, broad), 4.88 (1H, broad), 5.13 (1H, s), 7.19 (2H, d, J=6.8), 7.31 (3H, m), 7.39–7.57 (9H, m), 7.86 (1H, d, J=8.2), 7.96 (1H, d, J=7.2), 8.19 (1H, d, J=7.9); ¹H NMR(500 MHz, CDCl₃, 298 K, δ, rotamers observed): 1.31–1.41 (9H, d), 2.14 (0.5H, dd, J=9.7, 14.6), 2.33 (0.5H, dd, J=10.5, 14.7), 2.38–2.49 (1H, m), 3.67 (0.5H, d, J=11.8), 3.81–3.85 (1H, m), 3.91 (0.5H, d, J=12.2), 4.09 (0.5H, d, J=12.2), 4.24 (0.5H, dd, J=5.8, 9.6), 4.87 (1H, dd, J=15.9, 29.5), 5.03 (1H, dd, J=12.3, 15.2), 5.09–5.59 (2H, m), 7.22–7.60 (14H, m), 7.78–7.92 (2H, m), 8.16–8.20 (1H, t, J=6.7); ¹³C NMR (125 MHz, CDCl₃, 298 K, δ): 27.7, 27.8, 38.0, 39.5, 42.7, 43.0, 54.0, 58.9, 59.5, 67.4, 68.6, 69.6, 82.4, 123.0, 123.2, 125.1, 125.2, 125.8, 125.9, 126.2, 126.2, 126.9, 127.1, 127.2, 127.5, 128.0, 128.1, 128.3, 128.4, 128.5, 128.5, 129.1, 129.4, 131.0, 131.1, 131.2, 131.2, 131.5, 131.5, 133.9, 133.9, 136.0, 136.0, 153.5, 154.5, 154.7, 169.5, 169.7, 171.7, 171.8; IR (neat): 1720, 1704, 1647, 1417, 1270 cm⁻¹; [α]_D¹⁸ = −31.4 (MeOH); HPLC analysis (C₁₈ reverse phase, 40 mins, 5–100% H₂O/ACN with 0.1 % formic acid): T_r = 27.5 mins; HRMS-ESI: m/z Calculated for C₃₆H₃₅N₃NaO₆ (M+Na)⁺ 628.2418; Found: 628.2423.

(5R,8S)-1-(Naphthanlen-1-ylmethyl)-2,4-dioxo-3-phenyl-1,3,7-triazaspiro[4,4]nonane-8-carboxylic acid.TFA, 2b

244mg, 53% yield from 21; ¹H NMR (500 MHz, d⁶-DMSO, 350 K, δ): 2.71 (1H, dd, J= 10.0, 14.6), 2.87 (1H, dd, J= 4.8, 14.6), 3.62 (1H, d, J= 13.2), 3.88 (1H, d, J= 13.2), 4.63 (1H, dd, J= 4.9, 10.1), 5.20 (2H, s), 7.35–7.70 (9H, m), 7.89 (1H, d, J= 8.2), 8.00 (1H, d, J= 7.5), 8.13 (1H, d, J= 8.3); ¹³C NMR (125 MHz, d⁶-DMSO, 350 K, δ): 34.8, 41.7, 49.6, 58.7, 68.0, 123.2, 123.6, 126.0, 126.5, 126.7, 127.3, 127.9, 128.6, 129.0, 129.1, 130.5, 132.2, 132.7, 133.6, 154.9, 169.4, 173.7; IR (neat): 1777, 1719, 1502, 1418, 1187, 1137 cm-1; [α]_D¹⁸ = +2.6 (MeOH); HPLC analysis (C₁₈ reverse phase, 30 mins, 5–95 % H₂O/ACN with 0.1 % formic acid): Tr = 14.2mins; HRMS-ESI: m/z Calculated for C₂₄H₂₂N₃O₄ (M+H)⁺ 416.1605; found: 416.1608.

Synthetic procedure for catalyst 1f

To a solution of Cbz, t-Butyl Pro4(SS) amino acid 5 (547 mg, 1.5 mmol, 1.0 eq) in methanol, acetaldehyde (102 µL, 1.8 mmol, 1.2 eq) was added in one portion at room temperature and stirred for 30 minutes, followed by addition of sodium cyanoborohydride (142 mg, 2.3 mmol, 1.5 eq). After the reaction was complete after 20 hours, as monitored by LC-MS, phenyl isocyanate (327 µL, 3.0 mmol, 3.0 eq) was added and the reaction was stirred at room temperature for 40 hours until completion, as indicated by LC-MS. The solvent was removed under vacuum and the residue was dissolved in CH₂Cl₂ (5 mL). To this solution, 33% HBr/AcOH (7 mL) was added and the reaction was stirred for 2 hours at room temperature until the Cbz- and t-Butyl- groups were removed completely, as indicated by LC-MS. After the removal of solvent under vacuum, the residue was purified by reverse phase. The purified compound was added to 50% TFA/CH₂Cl₂ (20 mL) and stirred at room temperature for 30 minutes. After concentration under vacuum, the remaining solution (1 to 2 mL volume) was added dropwise to 100 mL diethyl ether/hexane (1:1) and a white precipitate 1f formed (97 mg, 15% overall yield from 5). The precipitate was filtered and dried under high vacuum.

(5S,8S)-1-Ethyl-2,4-dioxo-3-phenyl-1,3,7-triazaspiro[4,4]nonane-8-carboxylic acid. TFA, 1f

¹H NMR (500 MHz, d⁶-DMSO, 350 K, δ): 1.27 (3H, t, J= 7.0), 2.55 (1H, m), 2.86 (1H, dd, J= 6.8, 13.7), 3.48 (2H, dd, J= 7.2, 14.4), 3.61 (1H, d, J=13.4), 3.72 (1H, d, J=13.4 ), 4.57 (1H, dd, J=6.9, 12.2), 7.39–7.52 (5H, m); ¹³C NMR (125 MHz, d⁶-DMSO, 350 K, δ): 14.7, 35.1, 35.6, 49.9, 59.5, 68.3, 127.0, 128.4, 129.0, 132.4, 154.1, 169.7, 174.2; IR (neat): 1774, 1718, 1502, 1421, 1186, 1138 cm-1; [α]_D¹⁸ = +11.1 (MeOH); HPLC analysis (C₁₈ reverse phase, 30 mins, 5–95 % H₂O/ACN with 0.1 % formic acid): Tr= 7.7 mins; HRMS-ESI: m/z Calculated for C₁₅H₁₈N₃O₄ (M+H)⁺ 304.1292; Found: 304.1293.

Synthetic procedure for catalyst 2c

The synthesis of catalyst 2c follows the same procedure as that of catalyst 1f with the exception of Cbz, t-Butyl Pro4(SR) amino acid 6 (364 mg, 1.0 mmol) substituted for Cbz, t-Butyl Pro4(SS) amino acid 5.

(5R,8S)-1-Ethyl-2,4-dioxo-3-phenyl-1,3,7-triazaspiro[4,4]nonane-8-carboxylic acid. TFA, 2c

60mg, 14% overall yield from 6; 1H NMR (500 MHz, d⁶-DMSO, 350 K, δ): 1.28 (3H, t, J= 7.0), 2.70–2.82 (2H, m), 3.49 (2H, m), 3.63 (1H, d, J= 13.2), 3.82 (1H, d, J= 13.2), 4.76 (1H, dd, J=6.4, 9.5), 7.39–7.51 (5H, m); ¹³C NMR (125 MHz, d⁶-DMSO, 350 K, δ): 15.0, 35.6, 37.3, 52.6, 59.7, 69.2, 126.9, 128.3, 129.0, 132.6, 154.2, 171.4, 173.9; IR (neat): 1775, 1719, 1502, 1423, 1187, 1137 cm-1; [α]_D¹⁸ = +9.6 (MeOH); HPLC analysis (C₁₈ reverse phase, 30 mins, 5–95 % H₂O/ACN with 0.1 % formic acid): Tr= 7.4 mins; HRMS-ESI: m/z Calculated for C₁₅H₁₈N₃O₄ (M+H)⁺ 304.1292; found: 304.1285.

Synthetic procedure for 17

In a flame dried 50 mL round-bottom flask under argon, 1-((Benzyloxy)carbonyl)-5-(tert-butoxycarbonyl)-3-(benzylamino)pyrrolidine-3-carboxylic acid 7 (454 mg, 1.0 mmol, 1 eq) and HOAt (1.36g, 10.0 mmol, 10.0 eq) were dissolved in 18 mL CH₂Cl₂/DMF (2:1) for 10 minutes followed by addition of DIC (188 µL, 1.2 mmol, 1.2 eq). The reaction was stirred at room temperature for 2 hours. To the reaction, a suspension of 1-Naphthyl-Aib-OH (292 mg, 1.2 mmol, 1.2 eq), DIPEA (416 µL, 2.4 mmol, 2.4 eq) in 6 mL DMF was added in one portion. The reaction was stirred at room temperature for 12 hours. To the reaction mixture, DIC (188 µL, 1.2 mmol, 1.2 eq) was added in one portion and the reaction was stirred at room temperature for an additional 12 hours until completion, as determined by LC-MS. The reaction was quenched by the addition of saturated NH₄Cl (aq.) (20 mL) and the product was then extracted with EtOAc (100 mL followed by 2x50 mL). The combined organic EtOAc layers were washed sequentially with saturated NaHCO₃ (aq.), brine, and dried over anhydrous Na₂SO₄. The Na₂SO₄was removed by filtration and the filtrate was concentrated under vacuum to produce a yellow solid crude product. The pure white powder product 17 (440 mg, 67% yield) was obtained after purification by C₁₈ reverse phase chromatography (5–100% Water/Acetonitrile).

2-Benzyl-3-tert-butyl(3S,5S)-6-benzyl-8,8-dimethyl-9-(naphthalen-1-ylmethyl)-7,10-dioxo-2,6,9-triazaspiro[4,5]decane-2,3-dicarboxylate, 17

¹H NMR (500 MHz, d⁶-DMSO, 350 K, δ): 1.35 (9H, s), 1.55 (3H, s), 1.59 (3H, s), 2.35 (1H, dd, J=7.6, 14.1), 2.84 (1H, dd, J=9.2, 14.0), 3.77 (1H, d, J=12.0), 4.08 (1H, d, J=12.0), 4.36 (1H, t, J=7.7), 4.67 (1H, d, J=16.4), 4.85 (1H, d, J=16.4), 5.07 (2H, s), 5.19 (2H., d, J=11.9), 7.20–7.65 (14H, m), 7.81 (1H, d, J=8.3), 7.96 (1H, d, J=8.4). 8.17 (1H, d, J=8.4); ¹³C NMR (125 MHz, CDCl₃, 298K, δ): 26.6, 27.3, 27.7, 39.8, 43.7, 46.9, 55.0, 60.1, 61.8, 66.9, 67.4, 81.9, 122.2, 122.7, 125.4, 125.9, 126.2, 126.5, 127.4, 127.7, 127.8, 127.9, 128.4, 128.9, 129.1, 130.5, 131.8, 133.8, 136.1, 137.4, 154.0, 168.8, 170.6, 171.7; IR (neat): 1698, 1644, 1395, 1345, 1121 cm⁻¹; [α]_D¹⁸ = −2.3 (MeOH); HPLC analysis (C₁₈ reverse phase, 40 mins, 5–100% H₂O/ACN with 0.1 % formic acid): T_r = 28.7 mins; HRMS-ESI: m/z Calculated for C₄₀H₄₄N₃O₆ (M+H)⁺ 662.3225; Found: 662.3234.

Synthetic procedure for catalyst 3

33% HBr/AcOH (7 mL) was added to the solution of compound 17 (440 mg, 0.67 mmol) in CH₂Cl₂ (5 mL) in one portion and the reaction was stirred at room temperature for 2 hours. Removal of solvent under vacuum resulted in dark red liquid that was reverse phase purified with 5–95% Water/Acetonitrile gradient. After lyophilization, the white powder was dissolved in 50% TFA/CH₂Cl₂ (50 mL) and stirred for 30 minutes at room temperature. After concentration under vacuum, the remaining solution (1 to 2 mL volume) was added dropwise to 100 mL diethyl ether/hexane (1:1) and a white precipitate 3 formed (250 mg, 64% yield). The precipitate was filtered and dried under high vacuum.

(3S,5S)-6-Benzyl-8,8-dimethyl-9-(naphthalen-1-ylmethyl)-7,10-dioxo-2,6,9-triazaspiro[4,5]decane-3-carboxylic acid.TFA, 3

¹H NMR (500 MHz, d⁶-DMSO, 350 K, δ): 1.57 (3H, s), 1.59 (3H, s), 2.58 (1H, t, J=13.6), 2.85 (1H, dd, J=7.4, 13.9), 3.64 (1H, d, J=13.1), 3.80 (1H, d, J=13.1), 4.59 (1H, dd, J=7.5, 11.4), 4.74 (1H, d, J=16.6), 4.90 (1H, d, J=16.6), 5.22 (2H, d, J=3.4), 7.23–7.63 (9H, m), 7.82 (1H, d, J=8.2), 7.95 (1H, d, J=7.1), 8.17 (1H, d, J=8.4); ¹³C NMR (125 MHz, d⁶-DMSO, 350 K, δ): 26.4, 26.9, 39.2, 44.3, 46.5, 52.2, 60.3, 61.8, 68.6, 122.8, 123.4, 126.0, 126.1, 126.3, 126.6, 127.2, 127.4, 129.1, 130.5, 133.0, 133.7, 138.2, 168.6, 169.5, 169.8; IR (neat): 1654, 1403, 1363, 1303, 1199 cm⁻¹; [α]_D¹⁸ = +8.8 (MeOH); HPLC analysis (C₁₈ reverse phase, 30mins, 5–95 % H₂O/ACN with 0.1 % formic acid): T_r =16.1 mins; HRMS-ESI: m/z Calculated for C₂₈H₃₀N₃O₄ (M+H)⁺ 472.2231; Found: 472.2236.

Typical procedure for aldol reaction catalyzed by 1a–1f, 2a–2b and 3

To a vial containing 1a (6.3 mg, 0.0132 mmol, 0.01 eq) and 3 mL H₂O, a mixture of cyclohexanone (1.37 mL, 13.2 mmol, 10.0 eq) and 4-nitrobenzaldehyde (200 mg, 1.32 mmol, 1.0 eq) was added. The reaction was stirred at room temperature until completion as determined by TLC. The reaction was quenched by addition of 6 mL saturated NH₄Cl (aq.) and extracted with CHCl₃ (10 mL + 2x5 mL). The combined organic layers were concentrated to afford the crude product. The diastereoselectivity was determined by proton NMR of the crude product. Purification by silica gel chromatography using 10–30% EtOAc/Hexanes afforded the pure aldol reaction product (316 mg, 96% yields). The enantioselectivity was determined by chiral HPLC with a Chiralpak AD column (UV detection set at 254 nm, flow rate of 0.5 mL/min, 20% i-PrOH/hexane as the eluting solvent).

Procedure for catalyst recycling

To a vial containing 1a (6.3 mg, 0.0132 mmol, 0.01 eq) and 3 mL H₂O, mixture of cyclohexanone (1.37 mL, 13.2 mmol, 10.0 eq) and 4-nitrobenzaldehyde (200 mg, 1.32 mmol, 1.0 eq) was added. The reaction was stirred at room temperature until completion, as determined by TLC. Hexanes (6 mL) was added to the reaction mixture and mixed for 5 minutes. The hexanes layer was removed carefully by pipette. To the aqueous layer, a fresh aliquot of cyclohexanone (1.37 mL, 13.2 mmol, 10.0 eq) and 4-nitrobenzaldehyde (200 mg, 1.32 mmol, 1.0 eq) was added for the second recycling round. The recycling procedure was then repeated for a third time. The combined organic layers were evaporated under vacuum and the residue was analyzed by proton NMR with CDCl₃ as solvent to determine the diastereoselectivity. The yield was determined after purification of the crude products by silica gel using 20% EtOAc/hexanes. The enantioselectvity was determined by injection of the purified products into an HPLC with a Chiralpak AD column (UV detection set at 254 nm, flow rate of 0.5 mL/min, 20% i-PrOH/hexane as the eluting solvent).

Typical kinetics experiment for 1a, 2a and 1f catalyzed aldol reaction

To a vial containing 1a (6.3 mg, 0.0132 mmol, 0.01 eq) and 3 mL H₂O, a mixture of cyclohexanone (1.37 mL, 13.2 mmol, 10.0 eq) and 4-nitrobenzaldehyde (200 mg, 1.32 mmol, 1.0 eq) was added. The reaction was stirred vigorously at room temperature. At specific time points (shown in supporting information), 200 µL of the reaction mixture was taken out by pipette and added to a vial containing 1 mL of saturated NH₄Cl (aq.), followed by 600 µL CDCl₃. After shaking the vial and two layers formed, the CDCl₃ layer was taken out by pipette carefully for analysis by 1D H-NMR. The starting material / product ratio was determined by the integration of the aldehyde proton peak (10.11 ppm) versus the proton peak of α position of the hydroxyl group (5.42 ppm for syn products, 4.86 ppm for anti products).

Quantum mechanical calculations

All of the computations were performed in Gaussian 09. ¹⁶ The geometries were fully optimized by M06-2X/6-31G(d) either in the gas phase or using the SMD model to account for the solvation effects of water. The computations in water as solvent were performed using the SMD model parameters¹⁴ along with the IEF-PCM algorithm¹⁷ for bulk electrostatics as implemented in Gaussian 09. ^{14, 16} The reported free energies were computed from partition functions evaluated using a quasiharmonic approximation in both the gas phase and the solution.^14c This approximation is the same as the usual harmonic approximation, except that the vibrational frequencies lower than 100 cm⁻¹ are raised to 100 cm⁻¹.