Introduction

One of the puzzles for the chemical origin of life is the origin of biological homochirality, i.e., why life on Earth is based on highly enantioenriched L-amino acids and D-sugars, and not their mirror image molecules (Maddox 1998; Eschenmoser 1999; Bonner 1995; Girard and Kagan 1998; Green et al. 1999; Kondepudi and Asakura 2001; Podlech 2001; Weissbuch et al. 2005). The homochirality of biomolecules should be related to the origin and evolution of life. How life attained the enantiomeric enrichment of biocompounds is one of the greatest subjects to be studied. To date, several mechanisms have been proposed for the origins of chirality, including circularly polarized light (Kuhn and Knopf 1930; Inoue 1992; Bonner and Rubenstein 1987; Feringa and Van Delden 1999), chiral inorganic crystals (Hazen 2004; Hazen and Sholl 2003) such as quartz, spontaneous absolute asymmetric synthesis (Islas et al. 2005; Mislow 2003), etc.

There are achiral organic compounds that crystallize in chiral space groups to form enantiomorphous crystals. Chiral crystallization of achiral organic compounds has been one of the important candidates for the origin of chirality (Kahr and Gurney 2001; Sakamoto 1997; Matsuura and Koshima 2005; Tanaka and Toda 2000). Stereospecific reactions using enantiomorphous crystals as reactants in the solid state have been reported to form enantioenriched organic compounds (Penzien and Schmidt 1969; Green et al. 1979; Claborn et al. 2008). However, this process does not increase the molar amount of chirality.

On the other hand, we continue to study asymmetric autocatalysis (Soai et al. 1995, 2000, 2001, 2004; Shibata et al. 1999; Soai and Kawasaki 2006, 2007, 2008; Bolm et al. 1996; Avalos et al. 2000; Todd 2002; Blackmond 2004; Podlech and Gehring 2005; Mikami and Yamanaka 2003; Caglioti et al. 2005; Gridnev 2006) mediated by chiral crystals composed of achiral compounds (Kawasaki et al. 2005a, 2006a, 2008a, b; Sato et al. 2000, 2004). Chirality in the crystalline state can act as the source of chirality in the addition reaction of dialkylzinc and achiral aldehyde to give a highly enantioenriched product in combination with asymmetric autocatalytic amplification of enantiomeric excess (ee) (Sato et al. 2003a). In this process, the chirality generated by spontaneous crystallization of the achiral molecules transferred to the external organic compound, and then the asymmetric autocatalysis with amplification of ee increased the initial tiny imbalance of chirality to afford a large amount of enantioenriched organic compound (Shibata et al. 1998; Soai et al. 2003; Kawasaki et al. 2005b, 2006b, c, 2009). Asymmetric autocatalysis with amplification of ee should give a strong correlation between the origin of chirality and the homochirality of organic compounds.

We describe here a highly enantioselective synthesis using enantiomorphous crystals of tetraphenylethylene (TPE) 1 as a source of chirality in conjunction with asymmetric autocatalysis (Fig. 1). Enantiomorphous crystals of TPE 1 induced the formation of S- and R-enantiomers of 5-pyrimidyl alkanol 5 in the reaction of diisopropylzinc (i-Pr2Zn) and pyrimidine-5-carbaldehyde 4. In addition, enantiomorphous crystals of achiral tetrakis(p-chlorophenyl)ethylene 2 and tetrakis(p-bromophenyl)ethylene 3 were also subjected to asymmetric autocatalysis and were found to act as chiral initiators (Fig. 1).

Fig. 1
figure 1

The concept of this work: spontaneous chiral crystallization of achiral tetraphenylethylenes followed by asymmetric autocatalysis with amplification of ee utilizing the chiral crystals 1–3 as the source of chirality

Full size image

Results and Discussion

TPE 1 is an achiral organic compound; however, the single-crystal structure belongs to the chiral space group P21 (Hoekstra and Vos 1975). The four phenyl rings are twisted in the same sense relative to the olefinic plane in the crystal lattice. This propeller-like conformation makes TPE 1 chiral, thus the resulting two enantiomeric forms, i.e., P-conformer and M-conformer, differ in the sense of helicity, i.e., the direction of the phenyl-ring torsion. The rotational barriers are sufficiently low that TPE 1 does not possess optical activity in solution (Maeda et al. 1995). TPE 1 shows chirality only in the crystalline state (Fig. 2).

Fig. 2
figure 2

Chiral P- and M-conformations and spontaneous chiral single-crystallization of TPE 1. Solid-state circular dichroism spectra of chiral crystals of TPE 1

Full size image

Single crystals can be grown from toluene solution (Hoekstra and Vos 1975), which is prepared by dissolving 1 g of TPE 1 in 35 mL of toluene at 60°C. After slow evaporation at room temperature for 2 days to 3 days, we obtained clear single crystals of TPE with well-defined crystal face shapes. Solid-state circular dichroism (CD) spectroscopic analysis (Kuroda 2004) of the enantiomorphous single crystals of 1 is shown in Fig. 2. It is possible to distinguish between the enantiomorphs of TPE single crystals 1 using solid-state CD spectroscopic analysis with nujol mulls. One crystal exhibits a positive Cotton effect at 270 nm ([CD(+)270nujol]-1), while the other shows a negative Cotton effect ([CD(–)270nujol]-1). In one specific recrystallization reaction, we obtained five single crystals in which two crystals show a positive Cotton effect and the other three crystals show a negative Cotton effect, i.e., in the same crystallization batch both enantiomorphous single crystals of 1 can be formed.

Furthermore, we observed that TPE 1 crystallizes in hemihedral form to afford enantiomorphous single crystals (Fig. 3). The simplified morphologies are sketched as shown in Fig. 4. Although the absolute structure of the crystal of TPE 1 could not be determined, Miller indices can be determined for the major faces, as in Fig. 4. The chirality of these enantiomorphous single crystals can be determined by comparing the relative sizes of the hemihedral faces. It is especially easily identified by comparison between the sizes of the (110) and (\( 1\overline 1 0 \)) surfaces (hemihedral faces), indicated as red (larger one) and blue (smaller one). If the magnitude of the correlation between the sizes of the hemihedral faces can be determined, the direction of the crystal can be defined unambiguously. When the larger red face is oriented as the upper side, i.e., the small blue face is the lower side, the position of the (\( 10\overline 1 \)) surface relative to the largest (100) face is opposite for the two enantiomorphous crystals. The crystal in which the (\( 10\overline 1 \)) surface is located to the right side of the (100) surface has the plus Cotton effect at 270 nm and the opposite crystal shows the minus Cotton effect in the solid-state CD using nujol. This is the way in which we can recognize the hemihedral crystals of TPE 1. Thus, we can distinguish by eye the enantiomorphs of TPE crystal 1 from the crystal shapes.

Fig. 3
figure 3

Formation of a hemihedral single crystal of TPE 1

Full size image
Fig. 4
figure 4

Hemihedral crystals and Miller indices of the single crystals of TPE 1

Full size image

We performed the asymmetric autocatalysis in the reaction of i-Pr2Zn and pyrimidine-5-carbaldehyde 4 in the presence of enantiomorphous crystals of TPE 1 as shown in Fig. 5. The results of enantioselective addition of i-Pr2Zn to pyrimidine-5-carbaldehyde 4 mediated with chiral crystals of TPE 1 are summarized in Table 1. When aldehyde 4 was treated with i-Pr2Zn in the presence of finely powdered crystal of [CD(+)270nujol]-1, which has a positive Cotton effect in solid-state CD at 270 nm, (S)-pyrimidyl alkanol 5 with 94% ee was produced in 98% yield (Table 1, entry 1). The opposite [CD(–)270nujol]-1, which was ground as fine powder, induced the formation of the opposite (R)-alkanol 5 with 91% ee in 93% yield (entry 2). The same directions of asymmetric induction were observed in entries 3 and 4, i.e., [CD(+)270nujol]- and [CD(–)270nujol]-crystals of TPE 1 catalyzed the production of (S)- and (R)-pyrimidyl alkanol 5 with high ee and high yield, respectively. It should be noted that almost enantiomerically pure alkanol 5 could be obtained by applying an additional three rounds of asymmetric autocatalytic amplification of ee (Sato et al. 2003a) (entries 5 and 6). These results clearly show that the enantiomorphous crystal of achiral hydrocarbon TPE 1 affected the enantioselective addition of i-Pr2Zn to pyrimidine-5-carbaldehyde 4 to afford chiral alkanols 5 with the absolute configuration correlated with that of TPE crystals 1. It should be noted that not all the single crystal 1 exhibit the hemihedrism, but the same correlation between the solid-state CD and the direction of asymmetric induction can be observed even when the non-hemihedral crystal 1 was used as chiral initiator of asymmetric autocatalysis.

Fig. 5
figure 5

Asymmetric autocatalysis with amplification of ee mediated by chiral crystals of TPE 1

Full size image
Table 1 Enantioselective synthesis of 5-pyrimidyl alkanol 5 using enantiomorphs of TPE 1 as a source of chirality in asymmetric autocatalysis
Full size table

Next, we conducted the asymmetric autocatalysis using chiral crystals of tetrakis(p-chlorophenyl)ethylene 2 and tetrakis(p-bromophenyl)ethylene 3 as chiral initiators. It has been reported that compounds 2 and 3 form chiral inclusion complexes 2·p-xylene and 3·p-xylene with the achiral guest compound p-xylene (Tanaka et al. 2000). In the crystal structure, molecules 2 and 3 have a propeller-like structure, the same as TPE 1 (Fig. 6). The absolute structure, i.e., P- or M-configurations of the single crystals 2·p-xylene and 3·p-xylene, could be determined using X-ray single-crystal structure analyses from the refinement of a Flack parameter (Flack 1983).

Fig. 6
figure 6

X-ray single crystal structures of (P)- and (M)-2·p-xylene (thermal ellipsoids: 60% probability)

Full size image

It has also been reported that the chiral arrangement of 2 and 3 in the crystal lattice was stable and is maintained even after removal of the guest solvent (Tanaka et al. 2000). The inclusion solvent, p-xylene, can be removed to form desolvated crystals 2 and 3 under reduced pressure at 95°C (Fig. 7). The crystalline chirality of (P)- and (M)-2 after the removal of p-xylene can be determined using solid-state CD with nujol, as shown in Fig. 7. The crystal of (P)-2, which was obtained from the inclusion crystal of (P)-2·p-xylene by the removal of the inclusion solvent, exhibits a positive Cotton effect at 270 nm, while the crystal of (M)-2 prepared by the desolvation of (M)-2·p-xylene shows a negative Cotton effect.

Fig. 7
figure 7

The formation of desolvated chiral crystals of (P)- and (M)-2 by the removal of p-xylene and the solid-state CD analysis using nujol

Full size image

We utilized both inclusion crystals (P)/(M)-2·p-xylene and 3·p-xylene and crystals after removal of guest solvent (P)/(M)-2 and 3 as the chiral seeds for asymmetric autocatalysis (Fig. 8). In the presence of finely ground powder of these crystals, asymmetric i-Pr2Zn addition to pyrimidine-5-carbaldehyde 4 was performed to give highly enantiomerically enriched 5-pyrimidyl alkanol 5, with the absolute configurations corresponding to that of the chiral crystals composed of tetrakis(p-chlorophenyl)ethylene 2 and tetrakis(p-bromophenyl)ethylene 3.

Fig. 8
figure 8

Asymmetric autocatalysis using chiral crystals of achiral 2 and 3

Full size image

Initially, asymmetric autocatalyses were conducted in the presence of inclusion crystals of 2·p-xylene and 3·p-xylene, as described in Table 2, series I. When i-Pr2Zn addition to pyrimidine-5-carbaldehyde 4 was performed in the presence of inclusion crystals of (M)-2·p-xylene, (S)-pyrimidyl alkanol 5 with 83% ee was formed in 85% yield (Table 2, Series I, entry 1). Using a crystal with the opposite P-chirality, (R)-alkanol 5 with 89% ee was obtained (entry 2). The reproducibility of the sense of enantioselectivity was excellent, as shown in entries 3–6. In the case of using specific single crystals of (M)- and (P)-2·p-xylene as chiral initiators, (S)- and (R)-alkanol 5 were also formed with high ee with the same stereochemical correlation with the above entries, respectively (entries 3 and 4). It should be noted that further consecutive asymmetric autocatalysis gave almost enantiomerically pure alkanol 5 with the corresponding absolute configuration to chiral crystals of 2·p-xylene, as shown in entries 5 and 6, respectively. When the inclusion crystal of tetrakis(p-bromophenyl)ethylene 3 with p-xylene was utilized, the same stereochemical correlation was observed, i.e., inclusion crystal 3·p-xylene with M-helicity induced the production of (S)-alkanol 5 and the P-crystal of 3·p-xylene produced (R)-alkanol 5 (entries 7 and 8). It is worth noting that the application of additional asymmetric autocatalysis can amplify the ee of produced 5-pyrimidyl alkanols 5 to >99.5% ee (entries 9 and 10). In addition, in the cases when crystals 2 and 3 whose guest solvent, p-xylene, was removed under reduced pressure under heat conditions, were utilized as the source of chirality, reproducible results were obtained, as shown in series II. When the crystal of (M)-2 whose inclusion solvent was removed was submitted to the reaction of asymmetric autocatalysis, (S)-5 with 71% ee was formed in 94% isolated yield (entry 11). The origin of this crystal (M)-2 used in entry 11 is the same as the one used in entry 1. In the next entry, (P)-2 mediated the formation of (R)-pyrimidyl alkanol 5 with 85% ee (entry 12). Further additional asymmetric autocatalysis enabled the amplification of ee to achieve almost enantiomerically pure product, as indicated in entries 13 and 14. When a chiral crystal of tetrabromo derivative 3 was employed in the reaction after the removal of p-xylene, (M)-3 induced the production of (S)-5 and (P)-3 initiated the formation of (R)-5, respectively (entries 15–18).

Table 2 Asymmetric autocatalysis utilizing chiral crystals of tetrakis(p-chlorophenyl)ethylene 2 and tetrakis(p-bromophenyl)ethylene 3
Full size table

In this enantioselective reaction, chiral crystals of tetraphenylethylenes 1–3 act as the chiral initiators of asymmetric autocatalysis. The initial addition reaction of i-Pr2Zn to aldehyde 4 should occur on the surface of finely powdered chiral crystals. Since the each crystal faces show the chirality based on the helical structure of tetraphenylethylenes in the crystalline lattice, the direction of the chiral induction should be appeared as averaged chiral effect of the whole powdered crystal surfaces. We assume the possibility of asymmetric π–π interactions between the phenyl rings of the chiral crystals 1–3 and the pyrimidine ring in the aldehyde 4 and/or the initially formed isopropylzinc alkoxide of alkanol 5 on the surface of chiral crystals. Because the molecules of tetraphenylethylenes on the surface of a crystal are chiral, they can differentiate the enantioface of aldehyde 4 when it adsorbed on the surface of the crystal. This enantioface differential adsorption of aldehyde 4 may cause the difference in the reactivity with i-Pr2Zn between the two enantiofaces of aldehyde 4. Thus, an initial asymmetric induction may occur and these small enantioimbalances formed in autocatalyst 5 as isopropylzinc alkoxide were enhanced to high ee by the subsequent asymmetric autocatalysis (Sato et al. 2003b; Blackmond et al. 2001; Gridnev et al. 2004; Buhse 2003; Islas et al. 2005; Lente 2004; Saito and Hyuga 2004; Micskei et al. 2006; Klankermayer et al. 2007; Schiaffino and Ercolani 2008) to afford highly enantioenriched alkanol 5 with the absolute configurations corresponding to the chirality of the submitted crystals 1–3.

Conclusion

As described, the highly enantioselective addition of i-Pr2Zn to pyrimidine-5-carbaldehyde 4 was achieved by utilizing the crystalline chirality of an achiral hydrocarbon, TPE 1 and its tetrachloro and tetrabromo derivatives 2 and 3, which form enantiomorphous organic crystals. These results clearly demonstrate that the chirality of the crystal of tetraphenylethylenes is responsible for the enantioselective addition of i-Pr2Zn to pyrimidine-5-aldehyde 4. This is the first realization of an asymmetric reaction using an enantiomorphous organic crystal of achiral hydrocarbon TPE 1 as the origin of chirality in conjunction with asymmetric autocatalysis. In addition, we showed that chiral crystals of achiral tetrakis(p-chlorophenyl)ethylene 2 and tetrakis(p-bromophenyl)ethylene 3 acted as chiral sources in asymmetric autocatalysis. Spontaneously generated crystalline chirality of achiral organic compounds 1–3, which act as the origin of chirality, have been amplified to afford a large amount of chiral organic compound by the asymmetric autocatalysis with amplification of chirality. These reactions represent one of the chemical processes in which the almost enantiomerically pure organic compound can be obtained based upon spontaneously generated crystal chirality of an achiral organic compound in conjunction with asymmetric autocatalysis.

Experimental Section

General Comments

Reactions and purifications were monitored by thin layer chromatography using Silica gel 60 F254 (pre-coated on aluminum sheet, 0.2 mm thickness, Merck). Chromatographic purification was performed with Silica gel 60 (230–400 mesh, Merck). Solid-state circular dichroism (CD) spectroscopic analyses were recorded as nujol mulls between NaCl plates using Jasco J-820 spectropolarimeter. The NaCl plates were rotated around the optical axis and CD recordings were made for several positions in order to check the reproducibility of the spectra. Diffraction data were obtained on a Rigaku, SCXmini and Bruker AXS, APEX diffractometer using a graphite monochromated Mo Kα radiation.

Synthesis of 1,1,2,2,-Tetraphenylethylene (TPE) 1

To the suspension of zinc (1.8 g, 27.5 mmol, powder) in THF (30 mL) was slowly added dropwise the solution of titanium tetrachloride (1.0 M dichloromethane solution, 13.75 mL, 13.75 mmol) at 0°C under an argon atmosphere and the mixture was allowed to warm to room temperature. The mixture was stirred for 30 min and refluxed for 2.5 h. Then, pyridine (0.8 mL, 9.63 mmol) was added at 0°C. After the mixture was stirred for 20 min, the solution of benzophenone (2.0 g, 11.0 mmol) in THF (10 mL) was added to the reaction mixture and the mixture was refluxed for 15 h. The reaction was quenched with saturated aqueous Rochelle salt at 0°C. The mixture was extracted using ethyl acetate-ether mixed solvent. The combined organic layers were dried over anhydrous magnesium sulfate and evaporated in vacuo. Purification of the residue by silica gel column chromatography (hexane) gave 1,1,2,2-tetraphenylethylene (TPE) 1 (1.48 g, 4.45 mmol) in 81% yield.

Hemihedral Crystallization of TPE 1

TPE 1 (1.0 g) was dissolved in toluene (35 mL) at 60°C. The solution was transfered to sample tube that had been pre-warmed to 40–50°C. The temperature of the solution was gradually lowered at room temperature. The hemihedral single crystal of TPE was appeared, after the slow evaporation of the solvent at room temperature for 2 days to 3 days.

Preparation of the Power-Like Crystal of (M)-2·p-xylene by Seeding Method

Tetrakis(p-chlorophenyl)ethylene 2 (1.0 g) was dissolved in p-xylene (15 mL) at 60°C. The temperature of the mixture was gradually lowered at room temperature for a period of 2 h with stirring. During decreasing temperature, a small piece of single crystal of (M)-2·p-xylene, whose absolute configuration was determined by single crystal X-ray structure analysis from the refinement of a Flack parameter, was added to the solution as the seed crystal in several times, and then, filtered. The resulting precipitate was dried in vacuo to afford the powder-like crystal of (M)-2·p-xylene (ca. 500 mg).

Removal of Inclusion Solvent, p-Xylene

Powder-like crystal of (M)-2·p-xylene was heated at 95°C under reduced pressure for a period of 30 min to form (M)-2 by the removal of p-xylene. TG-DTA measurements show the desolvation of p-xylene from the inclusion crystal of (M)-2.

General Method for Asymmetric Autocatalysis Utilizing TPE 1 (entry 2)

The crystals are ground using agate pestle and mortar before utilizing as a chiral initiator for asymmetric autocatalysis. To a finely powdered crystal of 1 (33.2 mg, 0.1 mmol, particle size: 5–20 μm), toluene (0.3 mL) solution of aldehyde 4 (4.7 mg, 0.025 mmol) was added (finely powdered 1 was not dissolved perfectly in toluene). After the removal of toluene under the reduced pressure with stirring, i-Pr2Zn (0.12 mmol, 0.12 mL, 1.0 M methylcyclohxane solution) was added dropwise over a period of 3 h at 0°C. After the mixture was stirred for 12 h, toluene (0.75 mL) and i-Pr2Zn (0.3 mmol, 0.3 mL, 1.0 M toluene solution) was added at 0°C. A solution of 2 (18.8 mg, 0.1 mmol) in toluene (0.75 mL) was added over a period of 2.5 h at 0°C, and the reaction mixture stirred at 0°C for 1 h. Once again, after toluene (5.0 mL) and i-Pr2Zn (0.8 mmol, 0.8 mL, 1.0 M toluene solution) were added, a solution of 4 (75.3 mg, 0.4 mmol) in toluene (2.0 mL) were added dropwise over a period of 1 h at 0°C. After the mixture was stirred for 3.5 h, the reaction was quenched with a mixture of 30% aqueous ammonia and saturated aqueous ammonium chloride (2/1, v/v) solution (10 mL). The mixture was extracted three times using ethyl acetate. The combined organic layers were dried over anhydrous sodium sulfate and evaporated in vacuo. Purification of the residue by silica gel column chromatography (hexane/ethyl acetate = 3/1, v/v) gave the (R)-alkanol 5 (112.8 mg, 0.486 mmol, 91% ee) in 93% yield. Enantiomeric excess was determined by HPLC using a chiral stationary phase (Daicel Chiralpak IB, eluent 5% 2-propanol in hexane, flow rate 1.0 mL min−1, 254 nm UV detector, retention time 11.1 min for (S)-5, 15.8 min for (R)-5) (Table 1).

General Method for Asymmetric Autocatalysis Using Chiral Crystal of Tetrakis(p-bromophenyl)ethylene 3 (series II, entry 11)

The chiral crystal of (M)-2 was ground into a fine powder using a pestle and mortar (particle size: 10–20 μm), after the removal of p-xylene under reduced pressure at 95°C for a period of 30 min. i-Pr2Zn (0.15 mmol, 0.15 mL, 1.0 M methylcyclohexane solution) was added dropwise to a finely powdered crystal of M-2 (56.6 mg, 0.075 mmol) and aldehyde 4 (4.7 mg, 0.025 mmol) over a period of 2 h at 0°C. After the mixture was stirred overnight, toluene (0.75 mL) and i-Pr2Zn (0.3 mmol, 0.3 mL, 1.0 M toluene solution) was added at 0°C. A solution of 4 (18.8 mg, 0.1 mmol) in toluene (0.75 mL) was added over a period of 1 h at 0°C, and the reaction mixture stirred at 0°C for 2 h. Once again, after toluene (5.0 mL) and i-Pr2Zn (0.8 mmol, 0.8 mL, 1.0 M toluene solution) were added, a solution of 4 (75.3 mg, 0.4 mmol) in toluene (2.0 mL) were added dropwise over a period of 1 h at 0°C. After the mixture was stirred for 2 h, the reaction was quenched with a mixture of 30% aqueous ammonia and saturated aqueous ammonium chloride (2:1, v/v) solution (10 mL). The mixture was extracted using ethyl acetate three times. The combined organic layers were dried over anhydrous sodium sulfate and evaporated in vacuo. Purification of the residue by silica gel column chromatography (hexane/ethyl acetate, 3:1 to 2:1, v/v) gave the (S)-alkanol 5 (114.7 mg, 0.494 mmol, 71% ee) in 94% yield (Table 2).