Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Inorg Chem. Author manuscript; available in PMC 2010 July 20.
Published in final edited form as:
PMCID: PMC2878398

Spontaneous Resolution of Racemic Camphorates in the Formation of Three-Dimensional Metal-Organic Frameworks


An external file that holds a picture, illustration, etc.
Object name is nihms147047f5.jpg

Reported here is a rare example of enantioselective processes between organic racemic ligands (DL-camphorates) with in situ formed chiral metal complexes. Such enantioselectivity leads to a double spontaneous resolution of two pairs of distinctly different racemates (the ΛΔ-Zn(Htea) units and the DL-camphorate ligands) into two chiral 3-D frameworks containing only one enantiommeric form from each racemate. This work demonstrates unique enantioselectivity in the self-assembly of MOFs containing multiple chiral features, which points to a new avenue for the preparation of chiral framework materials and for the resolution of racemates.

Chiral crystalline framework materials are of great interest because of their potential applications in enantioselective processes, however, the synthesis of such chiral materials remains a significant challenge.1, 2 For example, it has not been possible to prepare the enantiopure form of zeolite beta, depsite a lot of efforts. On the other hand, recently developed metal-organic frameworks (MOFs) offer fresh opportunities to study chiral phenomena and to fabricate novel chiral framework materials.37

Currently, the most successful method for the synthesis of crystalline homochiral framework materials is through the use of enantiopure ligands as the crosslinking ligands. 2, 4, 5 To better understand the self-assembly process of chiral MOFs, there have also some studies using racemic ligands in the starting materials so that a comparative study between the self-assembly involving enantiopure ligands and that involving racemic ligands can be made.5a, 5d, 8 Moreover, the synthesis of homochiral MOFs using enantiopure ligands may also be accompanied by the ligand racemization, which further increases the importance of studying self-assembly processes involving racemic ligands.9

Here, we report an unusual enantioselection process in which organic racemic ligands and in situ formed chiral metal complexes undergo enantioselective recognition, leading to double spontaneous resolution of both organic DL-racemates and ΔΛ-complexes during the self-assembly of 3-D MOFs. It is worth noting that while it is not uncommon for organic racemic molecules to undergo spontaneous resolution during crystallization as is well-known in the discovery of chirality in sodium ammonium tartrate,10, 11 such phenomenon has been rarely observed during the synthesis of 3-D MOFs so far. For example, in all earlier studies involving racemic DL-camphoric acid, both D- and L-ligands are incorporated into the same structures, leading to racemic crystals.[5d] The uniqueness of this work is the co-existence of two pairs of enantiomers and their enantioselective recognition leading to the double spontaneous resolution during the self-assembly of 3-D MOFs.

In this work, such a unique enantioselective process occurred during the solvothermal assembly of DL-camphoric acid (= H2cam) with Zn(NO3)2·6H2O in triethanolamine (= H3tea), which resulted in two chiral 3-D frameworks, Zn(Λ-Zn(Htea))(D-cam)) (1DΛ) and Zn(Δ-Zn(Htea))(L-cam) (1LΔ).12, 13 These frameworks contain either all D-Λ form or all L-Δ form. It is found that such enantioselection is promoted by an in situ formed racemic chiral metal complex. Here, triethanolamine acts not only as a solvent, but also as a ligand to generate a racemic building unit via the chelation to Zn2+. The chiral recognition between enantiomers of camphorates and chiral zinc complexes results in the enantioselection and spontaneous resolution of both pairs of racemates (Figure 1).

Figure 1
(a) Two types of distinct racemates with mirror symmetry; (b) the hydrogen-bonded left-handed helix based on Λ-Zn(Htea) units in 1DΛ; (c) the hydrogen-bonded right-handed helix based on Δ-Zn(Htea) units in 1LΔ; (d and e) ...

The asymmetric unit of 1DΛ consists of two independent Zn2+ ions, one D-cam ligand and one Htea ligand. One Zn2+ site (Zn1) possesses a distorted trigonal bipyramidal geometry [ZnO4N] in which three sites are occupied by two O atoms and one N atom from the same Htea ligand and the remaining two sites are completed by two carboxylate O atoms from two D-cam ligands. In comparison, the second Zn2+ site (Zn2) adopts tetrahedral geometry and is coordinated by two O atoms from two Htea ligands and two carboxylate O atoms from two D-cam ligands (Figure 2a).

Figure 2
View of the 1-D chiral connectivity [Zn(Λ-Zn(Htea))]n from achiral sources (a) and the 3-D chiral connectivity [Zn(D-cam)]n from the enantiopure camphorates (b) in 1DΛ.

The dinegatively charged Htea ligand affords two deprotonated O donors and the central N donor to chelate one five-coordinate Zn1 site and also bridges two symmetry-related tetrahedral Zn2 sites. This leads to a [Zn(Λ-Zn(Htea))]n chain along the a axis (Figure 2a). One interesting feature of Λ-Zn(Htea) comes from the non-deprotonated -CH2CH2OH end of the Htea ligand which helps create the chirality of the Λ-Zn(Htea) unit. The orientation of this flexible tail-like -CH2CH2OH end directs the handedness of the central N atom, leading to racemic ΛΔ-Zn(Htea) units (Figure 1a).

Because of the enantioselective interaction with camphorates, the resulting Zn(Htea) units with the same handedness (Λ in 1DΛ and Δ in 1LΔ) are further organized into either left-handed 21 helix in 1DΛ or right-handed 21 helix in 1LΔ through the O–H…O hydrogen bonding interactions (Figure 1b–c). Here, the absolute helicity of the hydrogen bonded helices is determined by the chirality of camphorate ligands. For example, D-cam ligand favors the left-handed helices (Figure 1d), while the L-cam ligand favors the right-handed helices (Figure 1e).

The enantioselective crystallization observed here is believed to come from supramolecular interactions during the self-assembly, and is likely controlled by the α-C atom of the camphorate ligand. As illustrated in Figure 1d and 1e, the flexible hydrogen bonded helices have to adjust their handedness to effectively accommodate the α-C atom of the camphorate ligand. The chiral recognition between hydrogen bonded helices and camphorate ligands also contributes to the final cooperative double spontaneous resolutions.

Within the overall [Zn(Λ-Zn(Htea))(D-cam))]n framework, the 1-dimensional chiral connectivity [Zn(Λ-Zn(Htea))]n from achiral precursors is embedded within the 3-dimensional chiral connectivity [Zn(D-cam)]n from enantiopure chiral precursors (Figure 2b), which is unprecedented in chiral MOFs. In 1DΛ, each [Zn(Λ-Zn(Htea))]n chain is connected to adjacent four chains by the D-cam ligands. Each D-cam ligand is a μ4-linker and each carboxylate group of the D-cam ligand bridges two independent Zn2+ ions. Even without considering the connectivity of the Htea ligands, the D-cam ligands link the Zn2+ sites into a 3-dimensional chiral framework with diamond topology.

To further confirm the enantioselective interaction observed in the DL-camphorate reaction system, we also explored the use of enantiopure D-camphoric acid in the synthesis. Consistent with our prediction, only the Λ-Zn(Htea) form is selected during crystallization, resulting in the formation homochiral form 1DΛ only. Similarly, the use of L-camphoric acids gave homochiral 1LΔ (Figure 3).

Figure 3
The solid state CD spectra of the homochiral samples (red line: 1DΛ; blue line: 1LΔ; dashed line: D-cam).

In conclusion, we present here two chiral 3-dimensional frameworks. Of particular interest is the cooperative double spontaneous resolution of two pairs of distinctly different racemates: the ΛΔ-Zn(Htea) units and the DL-cam ligands (Scheme 1). It is demonstrated that this self-assembly process for the formation of chiral frame work entails cooperative enantioselectivity (Λ-Zn(Htea)↔D-cam and Δ-Zn(Htea)↔L-cam) to produce two enantiopure frameworks. This work represents an interesting example that shows how multiple chiral features in MOFs interact with each other to produce new chemical, structural, and chiral phenomena. Furthermore, this process points to a new avenue for the resolution of racemates, which may have practical value in enantioseparation.

Scheme 1
Spontaneous resolution and cooperative enantioselectivity between two pairs of distinctly different racemates into chiral 3-D frameworks 1DΛ and 1LΔ.

Supplementary Material

supporting info


We thank the support of this work by NIH (X. B. 2 S06 GM063119-05) and NSF (X. B. DMR-0846958). X. B is a Henry Dreyfus Teacher Scholar.


Supporting Information Available: Experiment details and additional structural figures, Powder X-ray diffraction, TGA, and CIF files. This material is available free of charge via the Internet at


1. (a) Férey G. Chem. Soc. Rev. 2008;37:191. [PubMed] (b) Dybtsev DN, Yutkin MP, Peresypkina EV, Virovets AV, Serre C, Férey G, Fedin VP. Inorg. Chem. 2007;46:6843. [PubMed]
2. (a) Kesanli B, Lin W. Coord. Chem. Rev. 2003;246:305. (b) Ma L, Lin W. J. Am. Chem. Soc. 2008;130:13834. [PubMed] (c) Lee SJ, Lin W. Accounts Chem. Res. 2008;41:521. [PubMed]
3. (a) Lin Z, Slawin AMZ, Morris RE. J. Am. Chem. Soc. 2007;129:4880. [PubMed] (b) Lii K-H, Chen C-Y. Inorg. Chem. 2000;39:3374. [PubMed]
4. (a) Seo JS, Whang D, Lee H, Jun SI, Oh J, Jeon YJ, Kim K. Nature. 2000;404:982. [PubMed] (b) Anokhina EV, Go YB, Lee Y, Vogt T, Jacobson AJ. J. Am. Chem. Soc. 2006;128:9957. [PubMed] (c) Vaidhyanathan R, Bradshaw D, Rebilly J-N, Barrio JP, Gould JA, Berry NG, Rosseinsky MJ. Angew. Chem. Int. Ed. 2006;45:6495. [PubMed]
5. (a) Zhang J, Bu X. Angew. Chem. Int. Ed. 2007;46:6115. [PubMed] (b) Zhang J, Chen S, Valle H, Wong M, Austria C, Cruz M, Bu X. J. Am. Chem. Soc. 2007;129:14168. [PubMed] (c) Zhang J, Liu R, Feng P, Bu X. Angew. Chem. Int. Ed. 2007;46:8388. [PubMed] (d) Zhang J, Yao Y-G, Bu X. Chem. Mater. 2007;19:5083. (e) Zhang J, Bu X. Chem. Commun. 2008:444.
6. (a) Murphy DL, Malachowski MR, Campana CF, Cohen SM. Chem. Commun. 2005:5506. [PubMed] (b) Garibay SJ, Stork JR, Wang Z, Cohen SM, Telfer SG. Chem. Commun. 2007:4881. [PubMed]
7. (a) Sun D, Ke Y, Collins DJ, Lorigan GA, Zhou H-C. Inorg. Chem. 2007;46:2725. [PubMed] (b) Ke Y, Collins DJ, Sun D, Zhou H-C. Inorg. Chem. 2006;45:1897. [PubMed]
8. (a) Bailey AJ, Cheetham AK, et al. Angew. Chem. Int. Ed. 2008;47:8634. [PubMed] (b) Kam KC, Young KLM, Cheetham AK. Cryst. Growth & Des. 2007;7:1522.
9. Chen L, Bu X. Chem. Mater. 2006;18:1857.
10. Pérez-García L, Amabilino DB. Chem. Soc. Rev. 2002;31:342. [PubMed]
11. Ye Q, Wang X-S, Zhao H, Xiong R-G. Tetrahedron: Asymmetry. 2005;16:1595.
12. Synthesis of conglomerate Zn(Λ-Zn(Htea))(D-cam)) (1DΛ) and Zn(Δ-Zn(Htea))(L-cam) (1LΔ): Zn(CH3COO)2 (0.1094 g), Na2CO3 (0.0583 g), DL-camphoric acid (0.1037 g), and triethanolamine (1.9582 g) were mixed in a 23ml Teflon cup and the mixture was stirred for 20 min. The vessel was then sealed and heated at 140 °C for 5 days. The autoclave was subsequently allowed to cool to room temperature. The transparent colorless crystals were obtained. The homochiral form Zn(Λ-Zn(Htea))(D-cam)) (1DΛ) or 1LΔ can be obtained by using the similar reaction except d-camphoric acid (or l-camphoric acid) instead of dL-camphoric acid.
13. Crystal data for 1DΛ: C16H27NZn2O7, Mr = 476.13, othorhombic, space group P212121, a = 12.1362(11), b = 12.2752(16) Å, c = 12.7945(14) Å, V = 1906.1(4) Å3, Z = 4, Flack parameter = −0.01(6), R1(wR2) = 0.0768 (0.1782) and S = 1.087 for 1618 reflections with I > 2σ(I). Crystal data for 1LΔ: C16H27NZn2O7, Mr = 476.13, othorhombic, space group P212121, a = 12.1540(5), b = 12.3292(7) Å, c = 12.8100(8) Å, V = 1919.57(18) Å3, Z = 4, Flack parameter = 0.06(4), R1(wR2) = 0.0665 (0.1652) and S = 0.985 for 1809 reflections with I > 2σ(I). The structure was solved by direct methods followed by successive difference Fourier methods. Computations were performed using SHELXTL and final full-matrix refinements were against F2.