Development of Heterogeneous Enantioselective Catalysts using Chiral Metal-Organic Frameworks (MOFs)

Summary

Here, we present a protocol for active site validation of metal-organic framework catalysts by comparing stoichiometric and catalytic carbonyl-ene reactions to find out whether a reaction takes place on the inner or outer surface of metal-organic frameworks.

Abstract

Substrate size discrimination by the pore size and homogeneity of the chiral environment at the reaction sites are important issues in the validation of the reaction site in metal-organic framework (MOF)–based catalysts in an enantioselective catalytic reaction system. Therefore, a method of validating the reaction site of MOF-based catalysts is necessary to investigate this issue. Substrate size discrimination by pore size was accomplished by comparing the substrate size versus the reaction rate in two different types of carbonyl-ene reactions with two kinds of MOFs. The MOF catalysts were used to compare the performance of the two reaction types (Zn-mediated stoichiometric and Ti-catalyzed carbonyl-ene reactions) in two different media. Using the proposed method, it was observed that the entire MOF crystal participated in the reaction, and the interior of the crystal pore played an important role in exerting chiral control when the reaction was stoichiometric. Homogeneity of the chiral environment of MOF catalysts was established by the size control method for a particle used in the Zn-mediated stoichiometric reaction system. The protocol proposed for the catalytic reaction revealed that the reaction mainly occurred on the catalyst surface regardless of the substrate size, which reveals the actual reaction sites in MOF-based heterogeneous catalysts. This method for reaction site validation of MOF catalysts suggests various considerations for developing heterogeneous enantioselective MOF catalysts.

Introduction

MOFs are considered a useful heterogeneous catalyst for chemical reactions. There are many different reported uses of MOFs for enantioselective catalysis^{1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19}. Still, it has yet to be determined whether the reactions take place on the inner or outer surface of the MOFs. Recent studies have raised questions concerning the utilization of the available surface and reduced diffusion^20,21,22,23. A more striking issue is that the chiral environment varies with the location of each cavity in the MOF crystal. This heterogeneity of the chiral environment implies that the stereoselectivity of the reaction product depends on the reaction site²⁴. Thus, designing an efficient enantioselective catalyst requires identification of the location where the reaction would take place. To do so, it is necessary to ensure that the reaction occurs either only on the inner surface or only on the outer surface of the MOF while leaving the interior intact. The porous structure of MOFs and their large surface area containing chiral environment active sites can be exploited for enantioselective catalysis. For this reason, MOFs are excellent replacements of solid-supported heterogeneous catalysts²⁵. The use of MOFs as heterogeneous catalysts needs to be reconsidered if the reaction does not occur inside them. The location of the reaction site is important, as well as the size of the cavity. In porous materials, the size of the cavity determines the substrate based on its size. There are some reports of MOF-based catalysts that overlook the cavity size issue²⁵. Many MOF-based catalysts introduce bulky catalytic species (e.g., Ti(O-iPr)₄) to the original framework structure^3,8,13. There is a change in the cavity size when bulky catalytic species are adopted in the original framework structure. The reduced cavity size caused by the bulky catalytic species makes it impossible for the substrate to fully diffuse into the MOFs. Thus, discrimination of substrate size by the cavity size of the MOFs needs to be considered for these cases. The catalytic reactions by MOFs often make it difficult to support evidence of reactions taking place inside the MOF cavity. Some studies have shown that substrates larger than the MOF cavities are converted to the expected products with ease, which seems contradictory^8,13. These results can be interpreted as a contact between the functional group of the substrate and catalytic site initiating the catalytic reaction. In this case, there is no need for the substrate to diffuse into the MOFs; the reaction occurs on the surface of the MOF crystals²⁶ and the cavity size is not directly involved in the discrimination of the substrate based on its size.

To identify the reaction sites of MOFs, a known Lewis-acid promoted carbonyl-ene reaction was selected². Using 3-methylgeranial and its congeners as substrates, four types of enantioselective carbonyl-ene reactions (Figure 1) were studied²⁷. The reactions, which have been previously reported, were classified into two classes: a stoichiometric reaction using a Zn reagent and catalytic reactions using a Ti reagent²⁷. The reaction of the smallest substrate requires a stoichiometric amount of Zn/KUMOF-1 (KUMOF = Korea University Metal-Organic Framework); it has been reported that this reaction takes place inside of the crystal²⁷. Two kinds of MOFs were used in this method, Zn/KUMOF-1 for the stoichiometric reaction and Ti/KUMOF-1 for the catalytic reaction. Owing to the distinct reaction mechanisms of these two kinds of MOFs, a comparison between the reaction rate versus substrate size is possible^2,28,29. The effect of particle size on the carbonyl-ene reaction with Zn/KUMOF-1²⁷ demonstrated that, as seen in the previous report, the chiral environment of the outer surface was different from the inner side of the MOF crystal²⁴. This article demonstrates a method that determines the reaction sites by comparing the reactions of three kinds of substrates with two classes of catalysts and the effect of particle size as reported in the previous paper²⁷.

Protocol

1. Preparation of (S)-KUMOF-1 crystals in three sizes

NOTE: Each step follows the experimental section and supplementary information of previous reports^2,24,27. Three different sizes of (S)-KUMOF-1 were prepared: large (S)-KUMOF-1-(L), medium (S)-KUMOF-1-(M), and small (S)-KUMOF-1-(S) with particle sizes >100 μm, >20 μm, and <1 μm, respectively. When out of the solvent, (S)-KUMOF-1 dismantles. Therefore, the crystals should always be kept wet while in use.

1. Synthesis of small size (S)-KUMOF-1-(S)
   1. In a 10 mL cell, dissolve Cu(NO₃)₂ ∙ 3H₂O (0.2 mg, 0.0008 mmol) and (S)-2,2'-dihydroxy-6,6'-dimethyl-[1,1'-biphenyl]-4,4'-dicarboxylic acid² (0.24 mg, 0.0008 mmol) in 4 mL of DEF/MeOH (DEF = N,N-diethylformamide, 1/1, v/v).
      NOTE: It is best to use newly prepared DEF and MeOH (methanol). (S) in (S)-KUMOF-1 means that the stereochemical configuration of the ligand used in KUMOF synthesis is S.
   2. Cap the reaction cell with a PTFE (polytetrafluoroethylene) cap and place it into a microwave reactor (65 °C, 100 psi, 50 W, 20 min).
      NOTE: To obtain the required number of crystals, repeat the above steps (1.1.1. and 1.1.2.) several times.
   3. Whisk gently with a small spatula to float the obtained blue cubic crystals (45% yield).
   4. Pour the floating crystals on filter paper, and wash 3x with 3 mL of hot DEF.
   5. Exchange the solvent 3x with 3 mL of anhydrous dichloromethane (DCM) for storage.
      NOTE: Every step requiring DCM in the protocol is DCM distilled over CaH₂.
2. Synthesis of medium size (S)-KUMOF-1-(M)
   1. Dissolve Cu(NO₃)₂ ∙ 3H₂O (7.2 mg, 0.030 mmol) in 1.5 mL of MeOH and (S)-2,2'-dihydroxy-6,6'-dimethyl-[1,1'-biphenyl]-4,4'-dicarboxylic acid (9 mg, 0.030 mmol) in 1.5 mL of DEF.
      NOTE: The compounds and solvents mentioned are for one vial set. Scaling up is needed to obtain the required number of MOFs for catalytic use. Multiply the scales in this step and make stock solutions for each compound. Then subdivide the stock solutions into each vial.
   2. Combine the two solutions in a 4 mL vial.
   3. Cover the 4 mL vial with PTFE tape and punch the cover with a needle to make a hole.
   4. Put this small vial into a 20 mL vial and add 1.0 mL of N,N-dimethylaniline into the space between the small and large vials.
   5. Cap the large vial tightly and place in an oven at 65 °C for 1 day.
   6. Whisk gently with a small spatula to float the obtained blue cubic crystals.
   7. Pour the floating crystals on a filter paper and wash 3x with DEF/MeOH (3 mL/3 mL).
      NOTE: After pouring the floating crystals, tilt the vial above the filter paper. Then eject the solvent with a syringe to wash down every crystal remaining in the vial.
   8. Exchange the solvent 3x with 3 mL of anhydrous DCM for storage.
3. Synthesis of large size (S)-KUMOF-1-(L)
   1. Use the same procedure as in section 1.2, except at step 1.2.3, leave the 4 mL vial open.
      NOTE: The yield of the obtained crystal is based on the ligand used. The yield for the medium and large size (S)-KUMOF-1 were almost the same (35% yield) after final washing.

2. Preparation of Zn/(S)-KUMOF-1 in three sizes

NOTE: Each step follows the experimental section and supplementary information of previous reports^2,24,27.

1. Add dimethylzinc (0.68 mL, 1.2 M in toluene, 0.81 mmol) to a suspension of (S)-KUMOF-1 (102 mg, 0.27 mmol) in 2 mL of DCM at -78 °C and shake 3 h at this temperature.
   CAUTION: All steps at -78 °C are done with a cryogenic cooling bath (dry ice with acetone). Always be careful when handling this equipment.
   NOTE: All shaking procedures are done using a plate shaker (180 rpm).
2. Decant the supernatant and wash with 3 mL cold DCM several times for complete removal of unreacted dimethylzinc.
   NOTE: Three sizes of Zn/KUMOF-1 are required for the carbonyl-ene reaction. Follow the same steps as described for the three sizes of KUMOF-1. The number of catalytic sites is calculated assuming that one catalytic site is present in a Cu and a ligand pair. For this reason, the Zn/Cu and Ti/Cu ratios of the prepared crystals were determined as in the previous report using inductive coupled plasma atomic emission spectroscopy (ICP-AES)²⁷. The amounts of Zn and Ti reagents used in this protocol were the same as those used in our previous study²⁷.

3. Preparation of Ti/(S)-KUMOF-1 in three sizes

NOTE: Each step follows the experimental section and supplementary information of previous reports^2,24,27.

1. Add Ti(O-iPr)₄ (59 μL, 0.20 mmol) to a suspension of (S)-KUMOF-1 (24 mg, 0.063 mmol) in 2 mL of DCM and shake for 5 h at room temperature.
2. Decant the supernatant and wash with 3 mL of cold DCM several times for the complete removal of residual Ti(O-iPr)₄.

4. Carbonyl-ene reaction using the prepared MOFs

NOTE: Prepare a series of substrates according to the method described in our previous report²⁷. All three substrates are used individually in each carbonyl-ene reaction except for the particle size effect determination, in which only the smallest substrate (1a) is used²⁷. Each step follows the experimental section and supplementary information of previous reports^2,24,27.

1. Heterogeneous stoichiometric carbonyl-ene reaction by Zn/(S)-KUMOF-1.
   1. Add the substrate solution (0.089 mmol) in 0.1 mL of DCM to a suspension of Zn/(S)-KUMOF-1 (102 mg, 0.27 mmol) in 2 mL of DCM at -78 °C.
   2. Warm the reaction mixture slowly to 0 °C and shake for 3.5 h at this temperature.
   3. Quench the reaction mixture with 3 mL of an aqueous solution of 6 N HCl.
   4. Filter the resultant mixture through a diatomaceous silica pad.
   5. Concentrate the filtrate in vacuo and purify the residue by flash chromatography (n-hexane/ethyl acetate 10:1).
      NOTE: Silica gel 60 (230–400 mesh) and an appropriate n-hexane/ethyl acetate mixture as the eluent are used for flash chromatography. The product is a pale yellow oil. Optical purity of all products in this protocol were determined as described previously²⁷. The same procedure should be followed for the three sizes of Zn/(S)-KUMOF-1.
2. Heterogeneous catalytic carbonyl-ene reaction by Ti/(S)-KUMOF-1.
   1. Add the substrate solution (0.29 mmol) in 0.1 mL of DCM to a suspension of Ti/(S)-KUMOF-1 (12 mg, 0.029 mmol) in DCM (2 mL) at 0 °C, and shake for 36 h at this temperature.
   2. Collect the supernatant and wash the resultant crystals 3x with 3 mL of DCM.
   3. Concentrate the collected supernatant in vacuo and purify the residue by flash chromatography (n-hexane/ethyl acetate 10:1).

Results

The enantioselective carbonyl-ene reaction using the Zn reagent is stoichiometric because of the difference in the binding affinities of the alkoxy and carbonyl groups to the metal (Figure 2). For this reason, the substrates were converted into the products at the reaction site and remained there. The desired products were obtained by dismantling the crystals, as detailed in section 4 of the protocol. The results of the heterogeneous enantioselective carbonyl-ene reaction of substrates by Zn...

Discussion

After the synthesis of (S)-KUMOF-1, crystals in some vials seem to be powdery and are not appropriate for use in catalysis. Therefore, proper crystals of (S)-KUMOF-1 need to be selected. The yield of (S)-KUMOF-1 is calculated using only those vials in which it was successfully synthesized. When withdrawn from the solvent, (S)-KUMOF-1 dismantles. Therefore, the crystals should always be kept wet. For this reason, weighi...

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
Acetone	Daejung	1009-4110
Analytical Balance	Sartorius	CP224S
Copper(II) nitrate trihydrate	Sigma Aldrich	61194
Dichloromethane	Daejung	3030-4465
Dimethyl zinc	Acros	377241000
Ethyl acetate	Daejung	4016-4410
Filter paper	Whatman	WF1-0900
Methanol	Daejung	5558-4410
Microwave synthesizer	CEM		Discover SP
Microwave synthesizer 10 mL Vessel Accessory Kit	CEM	909050
N,N-Diethylformamide	TCI	D0506
N,N-Dimethylaniline	TCI	D0665
n-Hexane	Daejung	4081-4410
Normject All plastic syringe 5 mL luer tip 100/pk	Normject	A5
Pasteur Pipette 150 mm	Hilgenberg	HG.3150101
PTFE tape	KDY	TP-75
Rotary Evaporator	Eyela	243239
Shaker	DAIHAN Scientific	DH.WSO04010
Silica gel 60 (230-400 mesh)	Merck	109385
Synthetic Oven	Eyela	NDO-600ND
Titanium isopropoxide	Sigma Aldrich	87560
Vial (20 mL)	SamooKurex	SCV2660
Vial (5 mL)	SamooKurex	SCV1545