Identification of Reaction Sites on Metal-Organic Framework-Based Asymmetric Catalysts for Carbonyl-Ene Reactions

Article abstract of DOI:10.1021/acscatal.8b04827

The characteristics of catalytic sites in metal-organic framework (MOF)-based catalysts could be approximately classified by their location, i.e., the inside of the pore and/or on the surface of the crystal. This classification of catalytic sites in a sin

Full text of DOI:10.1021/acscatal.8b04827

Subscriber access provided by Macquarie University
Article
Identification of Reaction Sites on Metal-Organic Framework-
Based Asymmetric Catalysts for Carbonyl-Ene Reactions
Jeehwan Han, Mi Sun Lee, Praveen K. Thallapally, Min Kim, and Nakcheol Jeong
ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.8b04827 • Publication Date (Web): 12 Feb 2019
Downloaded from http://pubs.acs.org on February 12, 2019
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 34
ACS Catalysis
1
2
3
4
5
6
7
8
9
Identification of Reaction Sites on Metal-Organic
Framework-Based Asymmetric Catalysts for
Carbonyl-Ene Reactions
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Jeehwan Han,† Mi Sun Lee,†, Praveen K. Thallapally,‡ Min Kim,*,§ Nakcheol Jeong*,†
†Department of Chemistry, Korea University, Seoul, 02841, Korea
‡Pacific Northwest National Laboratory, Richland, WA 99352. USA
§Department of Chemistry, Chungbuk National University, Cheongju, 28644, Korea
Abstract
The characteristics of catalytic sites in metal-organic framework (MOF)-based catalysts could be
approximately classified by their location, i.e., the inside of the pore and/or on the surface of the
crystal. This classification of catalytic sites in a single MOF crystal has been widely overlooked.
Particularly, in a chiral MOF, the environment of any specific reaction site will vary depending
on its location. Thus, pin-pointing the reaction site for a MOF-based heterogeneous catalyst is an
intriguing issue. In this study, the active site of a MOF-based catalyst is revealed after a thorough
ACS Paragon Plus Environment
1

ACS Catalysis
Page 2 of 34
1
2
3
4
5
6
investigation comparing substrate size versus reaction rate for two distinct mechanisms of
carbonyl-ene reactions.
7
8
9
Both Zn-mediated stoichiometric carbonyl-ene reactions and Ti-catalyzed carbonyl-ene reactions
were performed separately and compared using homogeneous and heterogeneous media. These
results could provide a clear answer to the question of locating the reaction sites within the MOF.
Through this work, it became evident that the entire MOF crystal is effective, however, the
inside of the pore is an important contributor to having chirality control in the stoichiometric
reaction. In addition, for a catalytic reaction, our findings suggest that the substrate size is mostly
irrelevant as catalysis can simply take place on the surface of the crystals. Thus, the comparison
of the reaction rate and substrate size may not be a valid method to ascertain whether a reaction
occurs inside the MOF or on its surface. This conclusion is further supported by the effect of
particle size on the reaction efficiency and the enantioselectivity along with visualization of the
guest-accessible space using two-photon fluorescence microscopy.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
KEYWORDS Heterogeneous catalysts; Metal-organic frameworks; Carbonyl-ene reaction; Two-
photon fluorescence microscopy; Size selectivity; Enantiomeric excess
Introduction
Metal-organic frameworks (MOFs), also known as porous coordination polymers (PCPs), have
been at the forefront of research during the last decade as a new type of porous material for
storage, separation, and catalytic applications. Because MOFs provide tunable, three-
ACS Paragon Plus Environment
2

Page 3 of 34
ACS Catalysis
1
2
3
4
5
6
7
8
9
dimensional architectures that are ideal for controlling the environment of a catalyst, a significant
amount of work has been done in the area of enantioselective catalytic reactions using chiral
MOFs.1–19 However, the identity of the reaction sites in MOF-based catalysts is still under
debate.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Thiele observed that if the reaction rate is faster than the diffusion rate, then diffusion forces
push the product away from the reaction sites on the surface in porous material-based catalysts.20
Thus, only the surface of the heterogeneous catalyst would be involved in the reactions. In
discussions concerning MOF-based catalysts, this point has been widely overlooked, although a
few reports dealing with this subject have appeared in the literature.21–23
The localization of reactions on the surface of a porous material poses a serious challenge to
the legitimacy of the MOF-catalyst approach. The legitimacy is related to a question regarding
the real advantages of using MOF-based catalysts in place of conventional heterogeneous
catalysts that are typically obtained by the immobilization of discrete catalytic species on a solid
support.24 Recently, two-dimensional metal-organic materials (e.g., metal-organic nanosheets)
have been used as heterogeneous catalytic platforms with full surface utilization and reduced
diffusion issues.25–28 Another issue related to the identification of reaction sites is more critical if
the catalysis considers enantioselective reactions. Because the chiral environment of each cavity
varies according to its location in the MOF crystal, the overall stereoselectivity of the reaction
will greatly depend on the location where the reaction occurs.29
Furthermore, the role of the cavity size in porous materials and the resultant discrimination of
substrates by their size, which is a frequently cited advantage of MOF-based catalysis, have been
overemphasized.24 A significant number of previous reports have used open framework
structures after introducing a bulkier catalytic species to promote the reaction.2,9,12 This
ACS Paragon Plus Environment
3

ACS Catalysis
Page 4 of 34
1
2
3
4
5
6
7
8
9
introduction of the catalytic species substantially reduces the void space in the crystals, and as a
result, the composite becomes unable to accommodate the entire substrate within its crystal.
Thus, the benefit of using MOFs to discriminate between different substrates by the cavity size
should be scrutinized.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In response to this issue, researchers have attempted to provide support for the notion that
reactions occur inside the crystals by comparing the reaction rates of different substrate sizes.
The presence of a threshold point (i.e., the reaction rate sharply decreases as the substrate size
increases) has been considered to be strong evidence of reactions occurring inside the MOF
channels. However, if the reactions are catalytic, then this argument has limitations.
Furthermore, the argument is inappropriate in many cases because the threshold point often
exceeds the cavity size. As is clearly seen in some reports,9,12 substrates apparently larger than
the cavities or channels in the MOFs were easily converted into their corresponding products.
Such substrates underwent the reaction provided that the reacting functionality of the substrate
could make contact with a catalytic site on the surface of the crystals.20 In other words, the
substrates did not necessarily diffuse into the interior of the MOF crystal to undergo catalysis.
Therefore, the cavity size was not directly related to the threshold point.
The identification of reaction sites in MOF catalysts for a given reaction is foundational and
important. The Lewis-acid-promoted carbonyl-ene reaction of 3-methylgeranial and its
congeners (1) to produce cyclic terpenoid analogs (2) was chosen as a reference reaction (Figure
1). This reaction can be facilitated either stoichiometrically with a reagent (e.g., Zn/(S)-
KUMOF-1, KUMOF = Korea University MOF) or catalytically with a catalyst (e.g., Ti/(S)-
KUMOF-1). Both reagents and catalysts have been previously reported for MOF-based
ACS Paragon Plus Environment
4

Page 5 of 34
ACS Catalysis
1
2
3
4
5
6
carbonyl-ene reactions.6 These two characteristic reactions provide us with a method for
identification of the reaction sites based on the nature of the reaction.
7
8
9
The reaction of 1a (the smallest substrate in the family) employs more than a stoichiometric
amount of reagent Zn/(S)-KUMOF and has been previously used to demonstrate that the
reaction must utilize all the reaction sites inside the crystal.6 Further studies using larger
substrates have been suggested to determine the threshold point for this stoichiometric reaction.
However, this reaction was the only one tested at that time for this system. This determination is
more interesting for the catalytic reaction than the previous stoichiometric reaction because this
issue is a subject of real concern for the catalytic reaction. With this preface, the present study
compared the reaction rate as a function of substrate size using two kinds of MOFs to directly
between two distinct reaction mechanisms (Zn/(S)-KUMOF-1 for the stoichiometric reaction
and Ti/(S)-KUMOF-1 for the catalytic reaction).6,30,31
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Two-photon fluorescence microscopy (TPM) has been used to approximately, but decisively,
determine and visualize the guest-accessible volume of the channels after the postsynthetic
introduction of reaction sites to the MOF (i.e., (S)- or (R)-KUMOF-1, Zn/(S)-KUMOF-1, and
Ti/(S)-KUMOF-1) and the substrate diffusion pattern into the modified MOF crystal, which
together make it possible to better understand the situation.2
ACS Paragon Plus Environment
5

ACS Catalysis
Page 6 of 34
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. Four types of catalytic-enantioselective carbonyl-ene reactions in this study.
Experimental
2.1 General considerations (materials and instrumentation)
All materials were obtained from Aldrich and TCI, and used without further purification,
unless otherwise noted. Anhydrous solvents were obtained by distillation over calcium hydride
(dichloromethane, toluene) or sodium/benzophenone (Et2O, THF). Chromatographic purification
of products was accomplished by flash chromatography using Merck silica gel 60 (230-400
mesh) with an appropriate mixture of hexane and ethyl acetate as eluent. Gas chromatography
analyses were performed using an Acme 6000 series system equipped with a flame ionization
detector (FID). High-performance liquid chromatography was conducted using a Jasco LC-1500
Series HPLC system with an ultraviolet detector at 254 nm.
ACS Paragon Plus Environment
6

Page 7 of 34
ACS Catalysis
1
2
3
4
5
6
2.2 Preparation of substrates
7
8
9
The synthetic routes to 1a, 1b, and 1c (Scheme S1) along with all relevant experimental details
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
are given in the supplementary information.
2.3 Preparation of MOF catalysts
2.3.1 Preparation of (S)-KUMOF-1: A reaction mixture in a small vial (4 mL) was prepared
by dissolving Cu(NO3)2⋅3H2O (7.2 mg, 0.030 mmol) and (S)-2,2'-dihydroxy-6,6'-dimethyl-[1,1'-
biphenyl]-4,4'-dicarboxylic acid (9 mg, 0.030 mmol)6 in N,N-diethylformamide (DEF)/MeOH
(1.5 mL/1.5 mL). This vial was sealed with a perforated cap and placed in a larger (20 mL
capacity) vial filled with N,N-dimethylaniline (1.0 mL). The tightly-capped larger vial was
placed in an oven at 65 ºC for 1 day. Blue cubic crystals were obtained in 35% yield based on the
ligand used. The crystals were separated and rinsed with DEF/MeOH (3 mL/3 mL) three times.
Characterization and rigorous crystal structure determination of (S)-KUMOF-1 were reported in
a previous paper.6
2.3.2 Preparation of Zn/(S)-KUMOF-1 and confirmation of hydrogen-metal exchange by
determination of Zn/Cu ratio: Dimethylzinc (0.4 mL, 2 M in dichloromethane, 0.81 mmol) was
added to a suspension of solvent exchanged (S)-KUMOF-1 (102 mg, 0.27 mmol) in
dichloromethane (2 mL) at -78 °C, and the solution was shaken for 3 h at this temperature. The
supernatant was decanted and the resultant Zn/(S)-KUMOF-1 was washed with cold
dichloromethane several times. Inductively coupled plasma atomic emission spectroscopy (ICP-
ACS Paragon Plus Environment
7

ACS Catalysis
Page 8 of 34
1
2
3
4
5
6
AES) was performed after complete removal of dichloromethane by drying under vacuum for 1
day (Table 1).
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 1. Concentration of elements in Zn/(S)-KUMOF-1
Element
Cu
amount (ppm (μmol))
129200 (2033)
Zn
135100 (2066)
2.3.3 Preparation of Ti/(S)-KUMOF-1 and confirmation of hydrogen-metal exchange by
determination of Ti/Cu ratio: A mixture of (S)-KUMOF-1 (24 mg, 0.063 mmol) and Ti(O-iPr)4
(0.2 mL, 1 M in toluene, 0.20 mmol) in toluene (1.5 mL) was shaken for 5 h at 25 °C. The
resultant crystals were washed with cold toluene several times. The Ti(IV) loading in Ti/(S)-
KUMOF-1 was determined by ICP-AES after complete removal of toluene by drying under
vacuum for 1 day (Table 2).
Table 2. Concentration of elements in Ti/(S)-KUMOF-1
Element
Cu
amount (ppm (μmol))
173000 (2722)
Ti
173600 (3625)
2.4 Assessment of channel size
ACS Paragon Plus Environment
8

Page 9 of 34
ACS Catalysis
1
2
3
4
5
6
7
8
9
2.4.1 Incubation of MOFs with pyrene-1-carbaldehyde (A): MOF crystals (roughly 4 mg) were
immersed in a stock solution of A in DMF (1.0 mL of a 0.05 M solution) at room temperature for
2 h (Figure S1).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2.4.2 Measurement of TPM images: The incubated crystals were placed on Paratone-N oil,
which was applied to a glass slide prior to placement of crystals, and then the wet crystals were
blotted dry with tissue paper. Prior to TPM measurement, suitable crystals (75-100 μm in size)
were screened using wide-field fluorescence and white-light transmission microscopes. The
spatial distribution of dyes inside the selected MOFs was imaged with a two-photon confocal
microscope (Leica TCS SP2) equipped with a femtosecond laser (Chameleon, Coherent Inc., λ =
750 nm, 200 fs, 10 mW) and an objective lens (NA 0.5, HC PL Fluotar 20x). By carefully
adjusting the position of the focal plane, we were able to obtain cross-sectional fluorescence
images from the center of the crystals, which reflected the local concentration of dyes. An
internal photomultiplier tube (PMT2) was used to collect the signals into 8 bits unsigned 512 ×
512-pixel images at a scan speed of 400 Hz. Compared with the more conventional one-photon
fluorescence microscopy technique, two-photon microscopy with a near-infrared light source
(750 nm) does not suffer significantly from the attenuation of excitation light, allowing more
reliable information on the 2-dimensional distribution of dyes to be obtained. Full images of the
dye diffusion patterns with time are displayed in the supporting information (Figures S2-S7).
2.5 Carbonyl-ene reaction
2.5.1 General procedure of the homogeneous carbonyl-ene reaction by Zn/(R)-3: A mixture of
(R)-3 (BINOL, 0.089 mmol) and 1 M ZnEt2 in hexane (0.27 mL) was stirred at -78 °C for 1 h.
ACS Paragon Plus Environment
9

ACS Catalysis
Page 10 of 34
1
2
3
4
5
6
7
8
9
The solution was warmed to 0 °C, and 1 (0.089 mmol) in dichloromethane (1.5 mL) was added.
The reaction mixture was stirred for a further 3.5 h at 0 °C, quenched with an aqueous solution of
6 N HCl (3 mL), and extracted with dichloromethane (3 mL × 3). The combined organic layer
was washed with sodium bicarbonate (3 mL × 3) and brine (3 mL × 3). The organic layer was
concentrated under reduced pressure to give the crude product, which was purified by flash
chromatography (n-hexane/ethyl acetate 19:1) to give 2 as pale-yellow oil.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2.5.2 Preparation of Mosher esters of 4 and ee determination: Compound 2 was treated with
(S)-(-)-α-methoxy-α-trifluoromethylphenylacetyl chloride ((S)-(-)-MTPA-Cl) and pyridine in
dichloromethane to give its Mosher ester 4 as a diastereomeric mixture that could be subjected to
gas chromatography for determining optical purity (Scheme S2).
2.5.3 Heterogeneous carbonyl-ene reaction of 1s by Zn/(S)-KUMOF-1
2.5.3.1 Determination of amount of chiral catalyst, (S)-KUMOF-1: The mole number of
catalytic sites was calculated based on the assumption that one catalytic site is present in a Cu
and (S)-2,2'-dihydroxy-6,6'-dimethyl-[1,1'-biphenyl]-4,4'-dicarboxylic acid ligand pair.
2.5.3.2 General procedure for the carbonyl-ene reaction of 1 by Zn/(S)-KUMOF-1: A solution
of 1 (0.089 mmol, in dichloromethane 0.1 mL) was added to a suspension of Zn/(S)-KUMOF-1
(102 mg, 0.27 mmol) in dichloromethane (2 mL) at -78 °C. The reaction mixture was warmed to
0 °C, and stirred for 3.5 h at this temperature. The reaction mixture was then quenched with an
aqueous solution of 6 N HCl (3 mL) and the resultant mixture was filtered through Celite. The
ACS Paragon Plus Environment
10

Page 11 of 34
ACS Catalysis
1
2
3
4
5
6
7
8
9
filtrate was concentrated in vacuo and the residue was purified by flash chromatography (n-
hexane/ethyl acetate 5:1) to give 2 as a pale-yellow oil. The enantiomeric excess (ee) of 2 was
determined as the de of 4 after transformation of 2 to 4 as previously described (Figures S8-S11).
2a was obtained in 92% yield in 3.5 h at 0 °C. The diastereomeric ratio of the MTPA esters of 2a
was measured to be 75:25 (50% ee). 2b was obtained in 52% yield in 3.5 h at 0 °C. The
diastereomeric ratio of the MTPA esters of 2b was measured to be 52.5:47.5 (5% ee).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2.5.4 General procedure for the homogeneous carbonyl-ene reaction by Ti/(R)-3: A mixture of
(R)-3 (BINOL, 0.015 mmol) and 1 M Ti(O-iPr)4 in dichloromethane (15 μL) was stirred at room
temperature for 1 h. The red-brown mixture was cooled to 0 °C, and 1 (0.3 mmol) in
dichloromethane (1.5 mL) was added. The reaction mixture was stirred for 12 h at this
temperature. The resulting solution was filtered through a pad of Celite and the filtrate was
extracted with dichloromethane (3 mL × 3) and washed with sodium bicarbonate (3 mL × 3).
The extract was evaporated under reduced pressure to give the crude product, which was purified
by flash chromatography (n-hexane/ethyl acetate 19:1) to give 2. The ee of 2 was determined as
the de of 4 after the transformation of 2 to 4. 2a was obtained in 90% yield. The diastereomeric
ratio of the MTPA esters (4a) of 2a was measured to be 70.5:29.5 (41% ee). 2b was obtained in
92% yield. The diastereomeric ratio of the MTPA esters (4b) of 2b was measured to be 71.5:28.5
(43% ee). 2c was obtained in 86% yield. The diastereomeric ratio of the MTPA esters (4c) of 2c
was measured to be 60:40 (20% ee).
2.5.5 General procedure for the carbonyl-ene reaction of 1 by Ti/(S)-KUMOF-1: A solution of
1 (0.29 mmol, in dichloromethane, 0.1 mL) was added to a suspension of Ti/(S)-KUMOF-1 (12
ACS Paragon Plus Environment
11

ACS Catalysis
Page 12 of 34
1
2
3
4
5
6
7
8
9
mg, 0.029 mmol) in dichloromethane (1 mL) at 0 °C. The reaction mixture was shaken for 36 h
at this temperature, then the supernatant was collected and the resultant crystalline materials
were washed with dichloromethane. Collection of the supernatant was repeated twice. The
combined organic layer was concentrated in vacuo and the residue was purified by flash
chromatography (n-hexane/ethyl acetate 5:1) to give 2. The ee of 2 was determined as the de of 4
after the transformation of 2 to 4 (Figures S8-S11). 2a was obtained in 85% yield. The
diastereomeric ratio of the MTPA esters (4a) of 2a was measured to be 61.5:38.5 (24% ee). 2b
was obtained in 89% yield. The diastereomeric ratio of the MTPA esters (4b) of 2b was
measured to be 53.5:46.5 (7% ee). 2c was obtained in 83% yield. The diastereomeric ratio of the
MTPA esters (4c) of 2c was measured to be 50:50 (0% ee).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2.5.6 Carbonyl-ene reaction of 1d by Zn/(S)-KUMOF-1: To a suspension of (S)-KUMOF-1
(102 mg, 0.27 mmol) in toluene (2 mL), dimethylzinc (0.4 mL, 2 M in dichloromethane, 0.80
mmol) was added at 0 °C. The resultant mixture was shaken for 3 h at this temperature.
Compound 1d (25 mg, 0.089 mmol) in toluene (0.1 mL) was added to this mixture at -78 °C (for
the preparation of compound 1d, see Supporting Information, Scheme S3). The reaction mixture
was warmed to 0 °C and shaken for 36 h at this temperature, and then quenched with an aqueous
solution of 6 N HCl (3 mL). The resultant mixture was filtered through Celite. The filtrate was
concentrated in vacuo and the residue was purified by flash chromatography (n-hexane/ethyl
acetate 1:1) to give 2d (13.8 mg, 9% yield, Scheme S4). The ee was determined as 0% by HPLC
(Figure S13 and Table S1)
ACS Paragon Plus Environment
12

Page 13 of 34
ACS Catalysis
1
2
3
4
5
6
7
8
9
2.5.7 Carbonyl-ene Reaction of 1d by Ti/(S)-KUMOF-1: A mixture of (S)-KUMOF-1 (24 mg,
0.063 mmol) and Ti(O-iPr)4 (0.3 mL, 1.0 M in dichloromethane 0.3 mmol) in toluene (1.5 mL)
was shaken for 10 h at 25 °C. The Ti/(S)-KUMOF-1 was collected by centrifugation and washed
five times with dichloromethane (1.5 mL). To the Ti/(S)-KUMOF-1 (12 mg, 0.029 mmol) in
toluene (1 mL) 1d (78 mg, 0.29 mmol) in toluene (0.1 mL) was added at 0 °C. The reaction
mixture was shaken for 36 h at this temperature. The supernatant was then collected and the
resultant crystalline materials were washed with ethyl acetate. The collection of the supernatant
was repeated two more times. The combined organic layer was concentrated in vacuo and the
residue was purified by flash chromatography (n-hexane/ethyl acetate 1:1) to give 2d (56 mg,
72% yield, Scheme S4). The ee of 2d was determined to be 0% by HPLC (Figure S14 and Table
S2).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Results and Discussion
3.1 A stoichiometric and homogeneous carbonyl-ene reaction of 1 by Zn catalysts (Lewis cat I,
Zn/(R)–3)
A homogeneous reaction requiring an excess of Lewis acids (up to 6 equiv.) was first studied
to understand the reactivity and nature of the catalyst. This enantioselective carbonyl-ene
reaction was promoted by Zn/(R)-3, which was obtained by mixing (R)-binol 3 and dimethylzinc.
This reaction was initially considered to be catalytic; however, more than the stoichiometric
amount of Lewis acid was required to obtain acceptable results.30,31 Even with a stoichiometric
amount of the reagent, the reaction often stops halfway to completion (77% yield with a 1.5
ACS Paragon Plus Environment
13

ACS Catalysis
Page 14 of 34
1
2
3
4
5
6
7
8
9
equiv. of Zn/(R)–3), and the stereoselectivity is unsatisfactory (52% ee with 1.5 equiv. of
Zn/(R)–3). Six equivalents of Zn/(R)–3 are necessary to achieve both good yields (87%) and high
stereoselectivity (85% ee) of 2a (Table 3).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The rationale for a large excess of the Zn reagent to obtain higher yields and stereoselectivity
is as follows: It suggests that the alkoxy group of product 2 binds to the Zn(II) center after the
chemical transformation is complete because the alkoxy group binds better to metals than a
carbonyl group (I, II, and III in Scheme S5). This newly formed Zn alkoxide III can also act as
a catalyst and can compete with Zn/(R)-3 for the next reaction. The catalytic cycle by III
produces 2 with significantly lower stereoselectivity (Scheme S5). The use of excess Zn/(R)-3
alleviates this unfavorable competition (IV and V in Scheme S5). The presence of a substantial
excess of Zn/(R)-3 increases the chance of reaction with the preferred catalyst. Thus, the
requirement of a large excess of Zn/(R)-3 for this carbonyl-ene reaction is due to minimizing
suicide inhibition; a distinct weakness of this reaction system from a pragmatic point of view
(Scheme S5). As mentioned earlier, this setup (i.e., the requirement of excessive reagent) has
been previously used to support the assertion that all the cavities in the MOF catalyst are reaction
sites.6
Table 3. Dependence of the reaction efficiency on the amount of reagent used
Zn
O
O
H
H
3
Zn/(R)-
O
OMe
CF3
CH2Cl2, 0 oC
O
OH
2a
O
1a
3
Zn/(R)-
4a
entry
equiv. of Zn/(R)-3
t (h)
yield (%)
ee (%)a
ACS Paragon Plus Environment
14

Page 15 of 34
ACS Catalysis
1
2
3
4
5
6
7
8
9
1
2
3
4
0.2
1.5
3
2
2
1
1
20
77
86
87
10
52
80
85
6
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
aThe ee of 2a was determined as the dextrose equivalent (de) of its corresponding 4a for
practical reasons
Next, we investigated the reaction efficiency in terms of the substrate size in homogeneous
reactions with an excess of Zn/(R)-3. All the tested substrates (1a – 1c) afforded their
corresponding products 2 in 1 h with equally high yields and stereoselectivities, regardless of the
substituent size: 87% (82% ee) for 2a, 82% (88% ee) for 2b, and 85% (87% ee) for 2c (Table 4).
There is no significant difference in the reaction rate depending on the substrate size for the
stoichiometric-homogeneous reaction (Lewis Acid Cat I in Figure 1).
Table 4. Enantioselective carbonyl-ene reaction of 1 by Zn/(R)-3
R
R
a
; R = H
H
H
3
Zn/(R)-
b
; R =
H
CH2Cl2, 0 o
1 h
C
O
OH
c
; R =
2
1
entry
substrate
yield (%)
ee (%)a
1
2
3
1a
1b
1c
87
82
85
85
88
87
aThe ee of 2 was determined as the de of its corresponding 4.
ACS Paragon Plus Environment
15

ACS Catalysis
Page 16 of 34
1
2
3
4
5
6
3.2 A stoichiometric and heterogeneous carbonyl-ene reaction of 1 by Zn catalysts (Lewis cat
II, Zn/(S)-KUMOF-1)
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
To explore the corresponding heterogeneous enantioselective carbonyl-ene reaction using
MOF-based catalysts, (S)-KUMOF-1 was selected as a platform. This platform was obtained
from coordination between (S)-2,2'-dihydroxy-6,6'-dimethyl-[1,1'-biphenyl]-4,4'-dicarboxylic
acid and Cu(II). (S)-KUMOF-1 has a NbO-type network topology with a noninterpenetrating
framework.6 The void space of (S)-KUMOF-1 is substantial; the diagonal interatomic distances
between the three sets of opposite water oxygen atoms on Cu(II) of the secondary binding unit
are 23.16 Å; therefore, an imaginary 2 × 2 × 2 nm3 cube can exist within the van der Waals
surfaces of the coordinating water molecules (Figure S5). This is one of the largest void spaces
available among the known chiral MOFs. (S)-KUMOF-1 also provides a large hexagonal
channel with a free aperture diameter of 1.69 nm along the [111] direction.6
Zn/(S)-KUMOF-1 was prepared by treating (S)-KUMOF-1 with dimethyl zinc, as previously
described. The identity of Zn/(S)-KUMOF-1 was confirmed with various spectroscopic
techniques.6 Particularly, the 1.01 Zn/Cu ratio was measured by ICP-AES (Table 1), which
suggests a stoichiometric exchange of the phenolic protons in the MOF ligand with Zn(II) ions.
Next, the guest-accessible volume of MOFs after modification (i.e., Zn-metalation) was
evaluated by two-photon fluorescence confocal microscopy.32 Pyrene-1-carbaldehyde (A) was
employed as a reference dye to assess whether the MOF can accommodate the dye in the
channels (Table S3) because its size (10.4 × 8.2 × 2.8 Å3) is comparable to that of products 2a
(7.8 × 6.0 × 5.2 Å3) and 2b (13.5 × 6.3 × 6.0 Å3) (Table S4). The diffusion rates of the dye into
(S)-KUMOF-1 and Zn/(S)-KUMOF-1 crystals were approximately determined by monitoring
ACS Paragon Plus Environment
16

Page 17 of 34
ACS Catalysis
1
2
3
4
5
6
7
8
9
the intensity of dye-dependent fluorescence from the crystals at regular intervals. As illustrated
in Figure S5(d), the dye gradually diffuses into (S)-KUMOF-1 with increasing incubation time,
and the crystals are uniformly saturated with dye after 2 h. Bright fluorescence is detected
uniformly at the center of the Zn/(S)-KUMOF-1 crystal after incubation with dye for 2 h (Figure
S5(e)). This comparison clearly suggests that there is still plenty of room in Zn/(S)-KUMOF-1
for the diffusion of guest molecules. The amount of dye entrapped by Zn/(S)-KUMOF-1 was
determined to be 325 mg/g (dye/MOF), which is slightly lower than the amount trapped by free
(S)-KUMOF-1 (575 mg/g, dye/MOF). The localization of dye near the surface of Zn/(S)-
KUMOF-1 is apparent, whereas the distribution of dye is uniform in pristine (S)-KUMOF-1.
This observation may be explained by the stronger binding of the carbonyl compound dye A to
Zn(II) in Zn/(S)-KUMOF-1. Further diffusion of the dye is significantly inhibited by the
presence of previously bound dye molecules. Finally, the aperture diameter of the channel after
the introduction of Zn(II) is calculated to be 11.97 Å (Figure S5(f)), which is still adequate for
the access of substrates.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Carbonyl-ene reactions of 1 with heterogeneous reagents (or catalysts) were investigated
(Table 5). While untreated, pristine (S)-KUMOF-1 does not initiate the reaction;6 1a (0.045 M)
satisfactorily undergoes the reaction in the presence of Zn/(S)-KUMOF-1, giving product 2a
with a 92% yield with 50% ee. However, at least 3 equiv. of reagent is required for a complete
reaction.6 Notably, as the reaction progresses, and even after the complete disappearance of 1a
from the solution, product 2a is not detected in the solution. Eventually, neither 1a nor 2a are
detected in the reaction mixture. Product 2a is obtained only after dismantling the MOF crystals
by treating them with aqueous HCl (6 N) solution. This finding suggests that product 2a is
completely entrapped in the MOF pores and remains there during the reaction. Due to the
ACS Paragon Plus Environment
17

ACS Catalysis
Page 18 of 34
1
2
3
4
5
6
7
8
9
necessary destruction of the Zn/(S)-KUMOF-1 crystal for reactivity and selectivity
characterization, no recycling test was attempted for the heterogeneous-stoichiometric reactions.
Although the carbonyl-ene reaction could occur on the surface of MOF crystals, we believe the
portion of the reaction that occurred in that manner is not significant because the yield for the
substrate decreases with increasing substrate size. Additionally, the largest substrate 1c showed
almost no conversion until 20 h after reaction (Table 5). Because an approximate calculation
indicates that the surface/volume ratio is far lower than 1% (Table S5), it is safe to speculate that
the MOF crystals accommodate all the substrates as some form of the product.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 5. Enantioselective carbonyl-ene reaction of 1 by Zn/(S)-KUMOF-1
R
R
a
; R = H
H
H
KUMOF-1
Zn/(S)-
b
; R =
H
CH2Cl2, 0 o
C
O
OH
c
; R =
2
1
entry
substrate t (h)
yield (%)
92
ee (%)a
1
2
3
1a
1b
1c
3.5
3.5
20
50
5
52
NRb
NAc
b
aThe ee of 2 was determined as the de of its corresponding 4; NR: no reaction, under the
detection limit; cNA: not applicable.
There is a distinct difference in the reaction rate as the substrate size increases. The reaction of
1b is halted halfway, even with a prolonged reaction time (3.5 h), giving rise to 2b with a 52%
yield and a negligible stereoselectivity (5%). Furthermore, 1c does not react at all. The presence
of 2a in the catalytic site would naturally reduce the free passage of new substrates. However,
the blockage caused by the presence of 2a is still passable by smaller 1a molecules (Scheme
ACS Paragon Plus Environment
18

Page 19 of 34
ACS Catalysis
1
2
3
4
5
6
7
8
9
1(a)) due to the relatively small size of 2a (7.8 × 6.0 × 5.2 Å3) compared to 1a. As the substrate
size increases, the blockage becomes more significant. The free space in the channels is
significantly reduced, and the incoming substrates are too large to penetrate the smaller channels.
Thus, there is considerable difficulty for the incoming substrates to penetrate and diffuse into the
MOF to reach the free catalytic sites. Although 1b and 2b are larger than 1a and 2a, 2b is not
sufficiently large to obstruct the whole channel (Scheme 1(b)). Thus, the reaction of 1b (13.5 ×
6.3 × 6.0 Å3) gives 2b in 52% yield, but the stereoselectivity of 2b was substantially diminished
to 5% ee. However, 1c is completely resistant to the reaction under the same conditions. Once a
very small amount of 2c (17.8 × 6.7 × 6.5 Å3) forms and remains at the metal centers on the
surface, it completely covers the channels and prevents further reaction between the remaining
substrate 1c and catalytic sites located inside the MOF (Table 5 and Scheme 1(c)). This reaction
is supposed to utilize all the reaction sites; therefore, once a reaction site on the surface of the
MOF participates in the reaction, it substantially blocks the channel and exhibits a clear
discrimination of the substrates by their size by preventing access to available catalytic sites.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 1. Diagrams illustrating the blockage of channels by the products. Molecules in green
are products bound to the metal center, and molecules in pink are substrates to be reacted. For
the reaction by Zn/(S)-KUMOF-1, (a) 1a can penetrate into the MOF without significant
obstruction by 2a. (b) 1b is blocked by the presence of 2b in the channel but still manages to
diffuse into the MOF. (c) Access of 1c is completely blocked because of the initially formed 2c,
which covers the whole channel and blocks it completely
ACS Paragon Plus Environment
19

ACS Catalysis
Page 20 of 34
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
From a previous study and the results mentioned above, it is evident that the chiral
environment of catalytic sites on the surface is not as effective as those inside the crystals in
terms of enantioselectivity.29 The effect of particle size on the reaction efficiency and
stereoselectivity was investigated to verify the reaction sites in the stoichiometric-heterogeneous
Zn/(S)-KUMOF-1 case. Three different sizes of Zn/(S)-KUMOF-1 were prepared; Zn/(S)-
KUMOF-1-(L) with a particle size of >100 μm, Zn/(S)-KUMOF-1-(M) with a particle size of
>20 μm, and Zn/(S)-KUMOF-1-(S) with a particle size of <1 μm. We hypothesized that a larger
particle utilizes more of the inside catalytic sites for the carbonyl-ene reaction and will display
higher stereoselectivity than a small particle. Under the same reaction conditions by catalyst
loading (by weight), Zn/(S)-KUMOF-1-(S) has more surface catalytic sites than Zn/(S)-
KUMOF-1-(M) or Zn/(S)-KUMOF-1-(L) and we expected a proportionate difference in ee.
While the reaction efficiency was almost identical for all three cases, the ee value dramatically
decreased with particle size (Table 6).29 This particle size control experiment clearly
demonstrates again the identification of reaction sites (inside of the pore) of the enantioselective,
stoichiometric-heterogeneous Zn/(S)-KUMOF-1 (Lewis Acid Cat II in Figure 1).
Table 6. Particle size effect on the carbonyl-ene reaction of 1 by Zn/(S)-KUMOF-1
ACS Paragon Plus Environment
20

Page 21 of 34
ACS Catalysis
1
2
3
4
5
6
H
H
Zn cat.
H
CH2Cl2, 0 oC
12 h
O
OH
2a
1a
7
8
9
entry
catalyst
yield (%)
ee (%)a
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
2
3
Zn/(S)-KUMOF-1-(L)
Zn/(S)-KUMOF-1-(M)
Zn/(S)-KUMOF-1-(S)
92
89
91
70
50
0
aThe ee of 2 was determined as the de of its corresponding 4.
3.3 A catalytic and homogeneous carbonyl-ene reaction of 1 by Ti catalysts (Lewis cat III,
Ti/(R)-3)
In parallel with the above study, a truly catalytic carbonyl-ene reaction by Ti/(S)-KUMOF-1
was studied, which represents the case where the reaction could take place entirely on the surface
of the crystal. First, a catalytic-homogeneous version of the carbonyl-ene reaction was performed
(Table 7). The enantioselective carbonyl-ene reaction of 1 by Ti/(R)-3 proceeds uneventfully,
even though the stereoselectivities are uniformly lower than those from the reaction by Zn/(R)-3
(Tables 4 and 7).
As shown by the results summarized in Table 7, the reaction of all the substrates proceeds
smoothly, regardless of the substrate size. 1a provides 2a in high yield (90%) and somewhat
lower stereoselectivity (41% ee). 1b and 1c give rise to 2b in 92% yield with a 43% ee and 2c in
86% yield with a 20% ee, respectively (Table 7). There is no considerable discrimination by
substrate size in terms of the reaction efficiency.
Table 7. Enantioselective carbonyl-ene reaction catalyzed by Ti/(R)-3
ACS Paragon Plus Environment
21

ACS Catalysis
Page 22 of 34
1
2
3
4
5
6
R
R
a
; R = H
H
3
H
Ti/(R)-
b
; R =
CH2Cl2, 0 o
12 h
C
H
O
OH
c
; R =
7
2
1
8
9
entry
substrate
yield (%)
ee (%)a
41
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
2
3
1a
1b
1c
90
92
86
43
20
aThe ee of 2 was determined as the de of its corresponding 4.
3.4 A catalytic and heterogeneous carbonyl-ene reaction of 1 by Ti catalysts (Lewis cat IV,
Ti/(S)-KUMOF-1)
Ti/(S)-KUMOF-1 was prepared by mixing pristine (S)-KUMOF-1 with excess Ti(O-iPr)4, as
previously described.6 This catalyst exhibited somewhat different characteristics from Zn/(S)-
KUMOF-1. The Ti/Cu ratio of Ti/(S)-KUMOF-1 was measured to be 1.30-1.80 by ICP-AES
(Table 2), even after extensive washings. The ratio is constantly slightly higher than the expected
value of 1.0, which indicates that a substantial amount of free Ti(O-iPr)4 is tightly entrapped.
The amount of entrapped dye in the TPM experiment with Ti/(S)-KUMOF-1 is determined to
be 60 mg/g (dye/MOF), which is substantially lower than in the case with Zn/(S)-KUMOF-1. In
addition, most of the entrapped dye is localized on the surface of the MOF crystal and more
visibly than in the case of Zn/(S)-KUMOF-1, as shown in the TPM image (Figure S7(b)). These
observations suggest that the titanium species (Ti(O-iPr)2) attached to the frame occupies or
blocks a considerable portion of the channels and significantly inhibits the diffusion of other dye
molecules. Additionally, it is possible that a substantial amount of extra Ti(O-iPr)4 fills the cavity
ACS Paragon Plus Environment
22

Page 23 of 34
ACS Catalysis
1
2
3
4
5
6
7
8
9
inside the MOF, probably as a dimeric titanium species, and remains there. Considering these
effects, it is apparent that there is little room for the guest substrates inside the MOF, indicating
the plugging of the channel and explaining the difficulty in the penetration of the substrate into
the center of the crystal.32 The aperture of the channel of Ti/(S)-KUMOF-1 is calculated to be
5.73 Å (Figure S7(c)), providing little chance for any substrate to penetrate into the channels;
even the smallest 1a. Thus, the entrapped titanium species would most likely not be involved in
the catalytic cycle of the carbonyl-ene reaction, and the pores are blocked, ensuring that only
surface catalysis is occurring.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Nonetheless, in contrast to the case with Zn/(S)-KUMOF-1, all reactions with 1a – 1c
proceeded smoothly to give the corresponding products 2a – 2c. There is no difference in the
reaction efficiency with different substrate sizes under the reaction conditions when using Ti/(S)-
KUMOF-1, although the reactions require a prolonged reaction time for completion (36 h). The
results of the heterogeneous catalytic reactions are summarized in Table 8. All the tested
substrates (1a – 1c) smoothly undergo the reaction to afford the corresponding products. In
contrast to Zn/(S)-KUMOF-1, there is no substrate discrimination. 1a is transformed into 2a
with an 85% yield and a 24% ee, 1b is transformed into 2b with an 89% yield and a 7% ee, and
1c is transformed into 2c with an 83% yield and a negligible stereoselectivity (~0% ee) (Table
8).
Table 8. Enantioselective carbonyl-ene reaction catalyzed by Ti/(S)-KUMOF-1
R
Ti/(S)-
CH2Cl2, 0 o
36 h
R
a
; R = H
H
H
KUMOF-1
b
; R =
H
C
O
OH
c
; R =
2
1
ACS Paragon Plus Environment
23

ACS Catalysis
Page 24 of 34
1
2
3
4
5
6
7
product obtained from solution product obtained from MOF
entry
substrate
8
9
yield (%)
ee (%)a
yield (%)
2.8
ee (%)
NAb
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
2
3
1a
1b
1c
85
89
83
24
7
0.7
NAb
0
0.2
NAb
aThe ee of 2 was determined as the de of its corresponding 4; bNA: not applicable.
These results have many interesting implications. First, all the reacting substrates do not
necessarily penetrate the crystal to be reacted. In other words, all the reactions take place only on
the surface of the crystals, rather than inside the crystals. Contrary to the case with Zn/(S)-
KUMOF-1, the products obtained at the catalytic site leave the site immediately after the
completion of the transformation. The catalytic sites, particularly those on the surface, would be
regenerated and available for the next reaction of the incoming substrates immediately (Scheme
2). Thus, the reaction of substrates, even those too large to enter the crystals, could also take
place on the surface of crystals. One side of the cavity on the very first layer is potentially open
depending on crystallization; thus, the section of a substrate not participating in the reaction can
be positioned outside of the crystal, extending toward the solvent. Therefore, as long as the
reacting functionality of the molecule reaches a catalytic site, the reaction can proceed without
difficulty. This notion is further confirmed by determining the amount of substrate 1 or product 2
trapped inside Ti/(S)-KUMOF-1 after the reaction is complete. Only 2.8% of product 2a is
recovered from the MOF crystals for the reaction with 1a. Conversely, only traces of products 2b
and 2c are recovered from the reaction with 1b and 1c because they cannot be entrapped at all
(Tables 8, S6-S8, and Figures S15-S17). Second, it may also be inferred that only a limited
ACS Paragon Plus Environment
24

Page 25 of 34
ACS Catalysis
1
2
3
4
5
6
7
8
9
number of catalytic sites among all available sites are utilized, and the reaction time required for
completion is prolonged. Lastly, the reusability of Ti/(S)-KUMOF-1 was studied for the
carbonyl-ene reaction with 1a. The reactivity was not changed by recycling (88% for the first
cycle and 86% for the second cycle), and the morphology was completely retained after
recycling, as confirmed by optical microscopy images and scanning electron microscopy (SEM)
images (Figure S18). Finally, a filtration experiment was performed to confirm the heterogeneity
of the reaction with the recycled sample. When the Ti/(S)-KUMOF-1 catalyst was removed by
filtration at 6 h, the reaction completely stopped with a 52.2% yield (Figure S19) indicating that
there is no free-floating homogeneous catalyst in solution.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 2. Illustration of the situation with Ti/(S)-KUMOF-1. (a)-(c) After the completion of the
reaction, (d) the newly formed 2 leaves the reaction center to yield the site to new incoming
substrates 1
3.5 Visualization of reactions on the surface of the crystals
In a number of previous reports, the MOFs investigated have cavities or channels intrinsically
too small to accommodate the entire molecule.10,13 It is reasonable to suggest that only the very
first layer of the MOF cavities could be utilized for these reactions by trapping a section of the
ACS Paragon Plus Environment
25

ACS Catalysis
Page 26 of 34
1
2
3
4
5
6
7
8
9
substrate. In addition, the inferior stereoselectivity observed is also understandable because the
chirality bias based on the pore environment is diminished on the surface.29 More convincing
evidence for this situation is obtained from the reaction with the dye-containing substrate 1d and
the observed image by TPM (for the preparation of 1d, see the supporting information, Schemes
S3 and S4). The reaction of 1d by Zn/(S)-KUMOF-1 stops at an early stage. Only a trace
amount of 1d is reacted, and the resultant product 2d remains on the surface of the catalyst
crystal. Thus, 2d residing on the surface prevents the diffusion of other substrates (1d) into the
crystal and makes the reaction stop once the exposed channels are covered. After the dismantling
of the crystal, 2d is harvested with a 9% yield and negligible stereoselectivity (Figure S13 and
Table S1). Conversely, with Ti/(S)-KUMOF-1, the reaction of 1d is completed within 36 h to
provide 2d with a 72% yield (0% ee) from the solution (Figure S14 and Table S2). No 2d
accumulates on the catalyst crystals (Table 9 and Figure 2).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 9. Carbonyl-ene reaction with 1d by Zn/(S)-KUMOF-1 and Ti/(S)-KUMOF-1
O2N
O2N
reagent
O
O
O
condition
toluene
*
H
OH
1d
2d
product obtained product obtained
Reaction condition
reagent
from solution
from MOF
entry
time (h) temp (oC) yield (%) ee (%) yield (%) ee (%)
1
2
Zn/(S)-KUMOF-1
12
36
0
NRa
72
NAb
0
9
0
(3 equiv.)
Ti/(S)-KUMOF-1
RT
NRa
NAb
ACS Paragon Plus Environment
26

Page 27 of 34
ACS Catalysis
1
2
3
4
(10 mol%)
5
6
aNR: no reaction, under the detection limit; bNA: not applicable.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. (a) (1) Carbonyl-ene reaction by Zn/(S)-KUMOF-1, showing that dyes are heavily
localized on the surface of the crystal. (2) The TPM image of Zn/(S)-KUMOF-1 after the
reaction. (b) (1) Carbonyl-ene reaction by Ti/(S)-KUMOF-1. (2) The TPM image of Ti/(S)-
KUMOF-1 after the reaction, which shows that the fluorescence from the dyes is detected only
in the surrounding solution, not from the crystal (red line: crystal boundary, scale bar = 50 μm).
Conclusion
The reaction sites of MOF-based catalysts were investigated and vary according to the nature
of the reaction with two representative examples. A reaction requiring more than a stoichiometric
amount of Zn/(S)-KUMOF-1 must utilize all available reaction sites inside the MOF; thus, there
is a clear distinction in reaction efficiency depending on the substrate size. In contrast, a catalytic
reaction by Ti/(S)-KUMOF-1 does not exhibit discrimination of the substrate based on its size
because most reactions take place on the crystal surface. The substrates do not necessarily
diffuse into the center of the crystals because the proximal catalytic sites residing on the surface
ACS Paragon Plus Environment
27

ACS Catalysis
Page 28 of 34
1
2
3
4
5
6
are constantly being regenerated. This detail has been potentially overlooked in the literature
when identifying the causes of MOF-based product selectivity.
7
8
9
Therefore, as long as the reactions being examined are catalytic, the comparison of reaction
rates vs. substrate size could be carefully utilized to discuss whether the reactions take place
inside the MOF or on the surface, and these comparisons should not be used to support the claim
that reactions occur inside the channels of the MOF crystals.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
AUTHOR INFORMATION
Corresponding Author
*E-mail: minkim@chungbuk.ac.kr; njeong@korea.ac.kr
Author Contributions
The manuscript was written through contributions of all authors. All authors have given approval
to the final version of the manuscript.
ASSOCIATED CONTENT
The Supporting Information is available free of charge on the ACS Publication website at DOI:
Additional figures, synthetic procedures, crystal image data, UV data, characterization
data for organic compounds (PDF)
ACKNOWLEDGMENT
We thank Dr. Hye-On Yoon (KBSI) for assisting MS experiments, Mr. Sung Min Shin (KU)
for synthetic protocols of ligands, and Dr. Corinne A. Allen (Department of Energy, USA) for
help in preparing this manuscript. This work was supported by a National Research Foundation
ACS Paragon Plus Environment
28

Page 29 of 34
ACS Catalysis
1
2
3
4
5
6
7
8
9
of Korea (NRF) grant (2011-0016303; Basic Science Research Program NRF-
2016R1C1B1011076 and the Science Research Center NRF-2016R1A5A1009405 for M. K.)
funded by the Korea government (MSIP). Experiments at PAL (beamline 2D) were supported in
part by MEST and POSTECH.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
REFERENCES
(1)
(2)
Yoon, M.; Srirambalaji, R.; Kim, K. Homochiral Metal–Organic Frameworks for
Asymmetric Heterogeneous Catalysis. Chem. Rev. 2012, 112, 1196–1231.
Ma, L.; Falkowski, J. M.; Abney, C.; Lin, W. A Series of Isoreticular Chiral Metal–
organic Frameworks as a Tunable Platform for Asymmetric Catalysis. Nat. Chem. 2010,
2, 838–846.
(3)
Férey, G.; Mellot-Draznieks, C.; Serre, C.; Millange, F.; Dutour, J.; Surblé, S.;
Margiolaki, I. A Chromium Terephthalate-Based Solid with Unusually Large Pore
Volumes and Surface Area. Science 2005, 309, 2040–2042.
(4)
(5)
Inagaki, S.; Guan, S.; Ohsuna, T.; Terasaki, O. An Ordered Mesoporous Organosilica
Hybrid Material with a Crystal-Like Wall Structure. Nature 2002, 416, 304–307.
Fang, Q.-R.; Zhu, G.-S.; Jin, Z.; Ji, Y.-Y.; Ye, J.-W.; Xue, M.; Yang, H.; Wang, Y.; Qiu,
S.-L. Mesoporous Metal–Organic Framework with Rare Etb Topology for Hydrogen
Storage and Dye Assembly. Angew. Chem., Int. Ed. 2007, 46, 6638–6642.
(6)
Jeong, K. S.; Go, Y. B.; Shin, S. M.; Lee, S. J.; Kim, J.; Yaghi, O. M.; Jeong, N.
Asymmetric Catalytic Reactions by NbO-Type Chiral Metal–Organic Frameworks. Chem.
Sci. 2011, 2, 877–882.
ACS Paragon Plus Environment
29