supporting online material for - science · 2008. 2. 12. · supporting online material for...

www.sciencemag.org/cgi/content/full/319/5865/939/DC1

Supporting Online Material for

High-Throughput Synthesis of Zeolitic Imidazolate Frameworks and

Application to CO2 Capture

Rahul Banerjee,* Anh Phan, Bo Wang, Carolyn Knobler, Hiroyasu Furukawa, Michael O’Keeffe, Omar M. Yaghi*

*To whom correspondence should be addressed.

E-mail: [email protected] (R.B.); [email protected] (O.M.Y.)

Published 15 February 2008, Science 319, 939 (2008)

DOI: 10.1126/science.1152516

This PDF file includes:

SOM Text S1 to S5 Figs. S1 to S86 Tables S1 to S23

1

Materials and Methods (160 pages) for:

High-Throughput Synthesis of Zeolitic Imidazolate Frameworks and

Application to CO2 Capture

Rahul Banerjee1*, Anh Phan1, Bo Wang1, Carolyn Knobler1, Hiroyasu Furukawa1,

Michael O’Keeffe2, Omar M. Yaghi1*

1Center for Reticular Chemistry at California NanoSystems Institute, Department

of Chemistry and Biochemistry, University of California, Los Angeles, 607 East

Charles E. Young Drive, Los Angeles, California 90095, USA. 2Department of

Chemistry and Biochemistry, Arizona State University, Tempe, Arizona 85287,

USA. * E-Mail: [email protected] or [email protected]

Table of Contents

Section S1. Detailed synthesis procedures for ZIFs including multi-gram scale products, and experimental and simulated PXRD patterns. Elemental microanalysis for ZIF-60, 61, 68 to 70, 74 and 76.

Section S2. Representative optical photographs of ZIFs crystals (0.1 to 1 mm) and single crystal x-ray diffraction data collection, structure solution and refinement procedures.

Section S3. Thermal stability of ZIFs and the thermal gravimetric analysis (TGA) data.

Section S4. Chemical stability of ZIFs in refluxing organic and aqueous media and the corresponding PXRD patterns.

Section S5. Porosity of ZIFs and carbon dioxide (CO2) selectivity and capacity studies.

2

Section S1: Detailed synthesis procedures for ZIFs including multi-gram scale

products, and experimental and simulated PXRD patterns. Elemental microanalyses

for ZIF-60, 61, 68 to 70, 74 and 76.

General Synthetic Procedures: Benzimidazole, 2-methylimidazole, 2-ethylimidazole,

5-chloro benzimidazole, purine and cobalt nitrate hexahydrate were purchased from the

Aldrich Chemical Co. and imidazole, N,N-dimethylformamide (DMF), N-

methylpyrrolidinone (NMP) were purchased from the Fisher Scientific International Inc.

2-nitroimidazole was purchased from 3B Scientific Corporation. N,N-diethylformamide

(DEF) was obtained from BASF Corporation. Zinc nitrate tetrahydrate was purchased

from EM Science. All starting materials were used without further purification. All

experimental operations were performed in air.

The fully integrated high throughput synthetic apparatus consists of a number of

components, including experimental design and execution software, robotic dispensing

and handling hardware, an automated high speed optical microscope for visualizing the

crystals/materials inside the wells, and the x-ray diffractometer for the data analysis and

structural interpretation. All ZIFs were synthesized by using this combinatorial

experimentation utilizing a 96-well glass plate as the reaction vessel. This design allows

us to use 0.300 mL reactant volume per well. A solution of imidazole(s) in DMF/DEF

(0.075–0.20 M) and a solution of a hydrated metal salt (usually nitrate) in DMF/DEF

(0.075–0.20 M) were used as stock solutions. This concentration range is used to avoid

contamination between two sets of reactions and to avoid blockage of the liquid handler

3

by precipitation of the reactant. DMF and DEF have been chosen as the solvents of

synthesis due to their high boiling points, which are suitable for the solvothermal

synthesis. A lower boiling solvent may result in the precipitation of the reactants rather

than of the product. After the 96-well glass plate was loaded with mixtures of stock

solutions dispensed by a programmed liquid handler (Gilson, model 215), it was covered

with a PTFE (poly tetrafluroethylene) sheet, sealed by fastening the sheet with a metal

clamp, then heated in an oven and allowed to react solvothermally for 24–96 h. Highly

crystalline materials were obtained by combining the requisite hydrated metal salt and

imidazole-type linkers. The very smooth interior of these 96 well glass-plates helps to

form diffraction quality single crystals. After reaction, the products were examined under

an automated optical microscope, isolated, and characterized by single-crystal x-ray

diffraction. The isolation of the sample array is accomplished in parallel by sonication

and transfer in a custom-designed shallow metal plate which allows the presence of small

a amount of solvent during the PXRD data collection. Samples were analyzed by a

Bruker D8 DISCOVER high throughput PXRD machine with a horizontal x-y stage for

automated analysis and equipped with an image plate detector system. The data

collection mode is in transmission geometry. The data collection time was 3–6 min per

sample. A structure database was set up that contains the simulated PXRD patterns of all

reported ZIFs. When a PXRD pattern of a new ZIF has been collected it can be checked

with this database in order to determine whether, if the material under study is a new

phase or it falls in the category of the ZIFs published already.

4

Thus far, ZIF-62, 64, 65, 67, 71, 72, 73, 75, and 77 were only produced in quantities

sufficient for a single crystal x-ray structure determination because these products were

produced only on a microscopic scale. The remaining ZIFs (ZIF-60, 61, 68, 69, 70, 74,

and 76), were prepared in quantities that allowed bulk characterizations (elemental

analysis, infrared spectroscopy and powder x-ray diffraction). In particular, ZIF-61, 68,

69, 70, and 76 were scaled up to gram quantities for a detailed investigation of their

porosity, thermal stability and chemical stability. The micro-synthesis conditions are

scalable to 10 gram scale by using the same reaction conditions in a larger volume (20

mL vial) and phase pure ZIF materials can thus be obtained. In this section, we report

detailed synthetic procedures for ZIF-2 to ZIF-77, which were discovered by the high

throughput method utilizing the 96-well micro-plates. It is noteworthy that the phase

purity of these mixed ligand ZIFs has always been under scrutiny since there is a

possibility of phase separation by those two individual ligands used for the mixed ligand

ZIF synthesis. To prove the phase purity of the bulk sample the PXRD and EA of the as-

synthesized and the activated samples are presented here. A comparison calculated/found

EA of the activated samples proved the phase purity of the sample. Activation conditions

of the respective samples and the corresponding TGA traces are described in a later

section of the manuscript. A detailed comparison of the PXRD patterns of the as-

synthesized and the activated samples with respect to their simulated PXRD patterns are

also presented. The ratio of these two ligands has been obtained from their respective

single crystal data.

5

(ZIF-2 crb): Zn(IM)2. 0.18 mL imidazole stock solution (0.15 M (1 M = 1 mol dm-3),

2.7 × 10-5 mol) and 0.12 mL Zn(NO3)2·4H2O stock solution (0.15 M, 1.8 × 10-5 mol)

were mixed together. After the glass plate was loaded with mixtures of stock solutions

dispensed by a programmed liquid handler (Gilson, model 215), it was covered with a

PTFE sheet, sealed by fastening the sheet with a metal clamp, then heated in an oven at

85 °C and allowed to react solvothermally for 72 h. The product was in the form of

prism-shaped single crystals. The unit cell dimensions of the ZIF thus obtained match

with the unit cell dimensions of ZIF-2 previously reported.

(ZIF-4 cag): Zn(IM)2. 0.18 mL imidazole stock solution (0.15 M, 2.7 × 10-5 mol) and

0.60 mL Zn(NO3)2·4H2O stock solution (0.15 M, 0.90 × 10-5 mol) were mixed together.

After the glass plate was loaded with mixtures of stock solutions dispensed by a

programmed liquid handler (Gilson, model 215), it was covered with a PTFE sheet,

sealed by fastening the sheet with a metal clamp, then heated in an oven at 100 °C and

allowed to react solvothermally for 72 h. The product was in the form of prism-shaped

single crystals. The unit cell dimensions of the ZIF thus obtained matches with the unit

cell dimensions of ZIF-4 previously reported.

(ZIF-8 sod): Zn(mIM)2. To a 0.225 mL 2-methylimidazole stock solution (0.20 M, 4.5 ×

10-5 mol), 0.075 mL Zn(NO3)2·4H2O stock solution (0.20 M, 1.5 × 10-5 mol) were added.




6

allowed to react solvothermally for 72 h. The product was in the form of cube-shaped



(ZIF-10 mer): Zn(IM)2. To a 0.20 mL imidazole stock solution (0.20 M, 4.0 × 10-5 mol)

in DMF 0.075 mL Zn(NO3)2·4H2O stock solution (0.20 M, 1.5 × 10-5 mol) were added.





single crystals. The unit cell dimensions of the ZIF thus obtained match with the unit cell

dimensions of ZIF-10 previously reported.

(ZIF-11 rho): Zn(bIM)2. To a 0.20 mL benzimidazole stock solution (0.20 M, 4.0 × 10-5

mol) in DEF 0.075 mL Zn(NO3)2·4H2O stock solution (0.20 M, 1.50 × 10-5 mol) in DEF

was added. After the glass plate was loaded with mixtures of stock solutions dispensed by

a programmed liquid handler (Gilson, model 215), it was covered with a PTFE sheet,





7

(ZIF-12 rho): Co(bIM)2. To a 0.20 mL benzimidazole stock solution (0.20 M, 4.0 × 10-5

mol) in DEF 0.075 mL Co(NO3)2·6H2O stock solution (0.20 M, 1.5 × 10-5 mol) in DEF

were added. After the glass plate was loaded with mixtures of stock solutions dispensed

by a programmed liquid handler (Gilson, model 215), it was covered with a PTFE sheet,





(ZIF-14 ana): Zn(eIM)2. To a 0.20 mL 2-ethylimidazole stock solution (0.20 M, 4.0 ×

10-5 mol) in DMF 0.075 mL Zn(NO3)2·4H2O stock solution (0.20 M, 1.50 × 10-5 mol) in

DMF was added. After the glass plate was loaded with mixtures of stock solutions




prism-shaped single crystals. The unit cell dimensions of the ZIF thus obtained match

with the unit cell dimensions of ANA structure (ZIF-14) previously reported.

(ZIF-20 lta): Zn(Pur)2. To a 0.18 mL purine stock solution (0.20 M, 3.6 × 10-5 mol) in

DMF 0.075 mL Zn(NO3)2·4H2O stock solution (0.20 M, 1.5 × 10-5 mol) in DMF were

added. After the glass plate was loaded with mixtures of stock solutions dispensed by a



8




(ZIF-21 lta): Co(Pur)2. To a 0.19 mL purin stock solution (0.20 M, 3.8 × 10-5 mol) in

DMF 0.075 mL Co(NO3)2·6H2O stock solution (0.20 M, 1.5 × 10-5 mol) in DMF was





single crystals. The unit cell dimension of the ZIF thus obtained matches with the unit

cell dimension of ZIF-21 previously reported previously reported.

9

(ZIF-60 mer): Zn(IM)1.5(mIM)0.5. 0.18 mL imidazole stock solution (0.15 M, 2.7 × 10-5

mol) and 0.060 mL 2-methylimidazole stock solution (0.15 M, 0.90 × 10-5 mol) was

mixed together. To this solution was added 0.060 mL Zn(NO3)2·4H2O stock solution

(0.15 M, 0.90 × 10-5 mol). After the glass plate was loaded with mixtures of stock


with a PTFE sheet, sealed by fastening the sheet with a metal clamp, then heated in an

oven at 85 °C and allowed to react solvothermally for 72 h. The product was in the form

of prism-shaped single crystals.

Elemental analysis C13H9N8Zn2 = Zn(IM)1.5(MeIM)0.5·(Me2NH)(H2O)3: Calcd. C, 33.17;

H, 6.55; N, 22.76. Found C, 33.28; H, 6.19; N, 22.13.

Elemental analysis (activated) C13H9N8Zn2 = Zn(IM)1.5(MeIM)0.5: Calcd. C, 37.44; H,

3.38; N, 26.87; Zn, 31.36. Found C, 36.97; H, 3.19; N, 27.13; Zn, 32.06.

FT-IR : (KBr 4000–400 cm-1): 3445(br), 3134(w), 2599(w), 2528(w), 2503(w), 1683(s),

1602(m), 1505(w), 1250(w), 1163(w), 955(w), 756(w), 675(w).

10

0

2000

4000

6000

8000

10000

12000

14000

16000

18000

2 12 22 32

Experimetal

Simulated

2θ (degree)

Figure S1. Comparison of the experimental PXRD pattern of as-prepared ZIF-60 (top)

with the one simulated from its single crystal structure (bottom). The very high degree of

correspondence between the patterns indicates that the bulk material has the same

structure as the single crystal.

(ZIF-61 zni): Zn(IM)(mIM). 0.12 mL imidazole stock solution (0.15 M, 1.8 × 10-5 mol)

and 0.12 mL 2-methylimidazole stock solution (0.15 M, 1.8 × 10-5 mol) was mixed

together. To this solution was added 0.060 mL Zn(NO3)2·4H2O stock solution (0.15 M,

0.90 × 10-5 mol). After the glass plate was loaded with mixtures of stock solutions



11

100 °C and allowed to react solvothermally for 96 h. The product was in the form of rod-

shaped single crystals.

Elemental analysis C7H9N4Zn = Zn(IM)(MeIM): Calcd. C, 38.99; H, 3.74; N, 25.99; Zn,

30.34. Found C, 39.17; H, 3.39; N, 26.13; Zn, 29.98.


1499(m), 1474(w), 1316(w), 1174(w), 1008(w), 837(w), 675(w), 420(s).

0

10000

20000

30000

40000

50000

60000

70000

2 12 22 32

Experimental

Simulated

2θ (degree)





12

(ZIF-62 cag): Zn(IM)1.75(bIM)0.25. 0.15 mL imidazole stock solution (0.20 M, 3.0 × 10-5

mol) and 0.075 mL benzimidazole stock solution (0.20 M, 1.5 × 10-5 mol) was mixed






prism-shaped single crystals.

(ZIF-64 dft): Zn(IM)2. To a 0.27 mL imidazole stock solution (0.20 M, 5.4 × 10-5 mol)

0.030 mL Zn(NO3)2·4H2O stock solution (0.20 M, 0.060 × 10-5 mol) was added. After the

glass plate was loaded with mixtures of stock solutions dispensed by a programmed

liquid handler (Gilson, model 215), it was covered with a PTFE sheet, sealed by fastening

the sheet with a metal clamp, then heated in an oven at 100 °C and allowed to react

solvothermally for 72 h. The product was in the form of rod-shaped single crystals.

(ZIF-65 sod): Co(nIM)2. To a 0.25 mL 2-nitroimidazole stock solution (0.20 M, 5.0 ×

10-5 mol), 0.050 mL Co(NO3)2·6H2O stock solution (0.20 M, 1.0 × 10-5 mol) was added.





single crystals.

13

(ZIF-67 sod): Co(mIM)2. To a 0.225 mL 2-nitroimidazole stock solution (0.20 M, 4.5 ×

10-5 mol), 0.075 mL Co(NO3)2·6H2O stock solution (0.20 M, 1.5 × 10-5 mol) was added.





single crystals.

(ZIF-68 gme): Zn(nIM)(bIM). 0.180 mL 2-nitroimidazole stock solution (0.20 M, 3.6 ×

10-5 mol) and 0.060 mL benzimidazole stock solution (0.20 M, 1.2 × 10-5 mol) was mixed






yellow colored prism-shaped single crystals.

Elemental analysis C10H9N5O2Zn = Zn(NO2IM)(PhIM)·(DMF)(H2O)2: Calcd. C, 36.10;

H, 3.33; N, 21.05. Found C, 35.47; H, 2.89; N, 21.83.

Elemental analysis (activated) C10H9N5O2Zn = Zn(NO2IM)(PhIM): Calcd. C, 40.49; H,

2.38; N, 23.62; O, 10.79; Zn, 22.05. Found C, 40.09; H, 2.12; N, 23.60; O, 11.44; Zn,

21.95.

14


2523(m), 2319(w), 1678(w), 1479(w), 1367(w), 1245(w), 1174(w), 909(s), 741(w),

669(w), 552(m).

0

5000

10000

15000

20000

25000

30000

35000

2 12 22 32

Experimental

Simulated

2 θ (degree)





(ZIF-69 gme): Zn(nIM)(5cbIM). 0.12 mL 2-nitroimidazole stock solution (0.20 M, 2.4

× 10-5 mol) and 0.12 mL 5-chlorobenzimidazole stock solution (0.20 M, 2.4 × 10-5 mol)

was mixed together. To this solution was added 0.060 mL Zn(NO3)2·4H2O stock solution

15




oven at 100 °C and allowed to react solvothermally for 72 h. The product was in the form

of yellow colored prism-shaped single crystals.

Elemental analysis C10H8N4O2ClZn = Zn(NO2IM)(5ClPhIM)·(H2O)4: Calcd. C, 32.78;

H, 4.44; N, 17.65. Found C, 31.89; H, 4.39; N, 17.13.

Elemental analysis (activated) C10H8N4O2ClZn = Zn(NO2IM)(5ClPhIM): Calcd. C,

36.28; H, 1.83; N, 21.16; O, 9.67; Cl, 10.74; Zn, 19.75. Found C, 35.90; H, 1.81; N,

20.76; O, 9.82; Zn, 19.51.


2660(m), 2523(w), 2319(w), 1667(w), 1470(w), 1364(w), 1291(w), 1240(s), 1174(w),

807(w), 659(m), 603(w).

16

0

5000

10000

15000

20000

25000

2 12 22 322θ(degree)

Experimetal

Simulated





17

(ZIF-70 gme): Zn(IM)1.13(nIM)0.87. 0.12 mL 2-nitroimidazole stock solution (0.20 M,

2.4 × 10-5 mol) and 0.12 mL imidazole stock solution (0.20 M, 2.4 × 10-5 mol) was mixed






prism-shaped single crystals.

Elemental analysis C6H5.13N4.87O1.74Zn = Zn(IM)1.10(NO2IM)0.90(DMF)(H2O)4: Calcd. C,

35.30; H, 4.44; N, 20.59. Found C, 35.17; H, 4.39; N, 20.13.

Elemental analysis (activated) C6H5.13N4.87O1.74Zn = Zn(IM)1.10(NO2IM)0.90: Calcd. C,

29.93; H, 2.14; N, 28.34; O, 11.57; Zn, 27.16. Found C, 29.77; H, 2.12; N, 28.36; O,

11.90; Zn, 27.01.


2518(m), 2329(w), 1678(w), 1510(w), 1372(w), 1168(w), 1102(w), 669(s).

18

0

5000

10000

15000

20000

25000

2 12 22 32

Experimetal

Simulated

2θ (degree)





19

(ZIF-71 rho): Zn(dcIM)2. To a 0.24 mL 4,5-dichloroimidazole stock solution (0.075 M,

1.8 × 10-5 mol) 0.060 mL Zn(NO3)2·4H2O stock solution (0.075 M, 0.45 × 10-5 mol) was




allowed to react solvothermally for 96 h. The product was in the form of block-shaped

single crystals.

(ZIF-72 lcs): Zn(dcIM)2. To a 0.24 mL 4,5-dichloroimidazole stock solution (0.20 M,

4.8 × 10-5 mol) 0.060 mL Zn(NO3)2·4H2O stock solution (0.20 M, 1.2 × 10-5 mol) in DEF

was added. After the glass plate was loaded with mixtures of stock solutions dispensed by

a programmed liquid handler (Gilson, model 215), it was covered with a PTFE sheet,


allowed to react solvothermally for almost 6 days. The product was in the form of prism-

shaped single crystals.

(ZIF-73 frl): Zn(nIM)1.74(dmbIM)0.26. 0.12 mL 2-nitroimidazole stock solution (0.20 M,

2.4 × 10-5 mol) and 0.12 mL 5,6-dimethylbenzimidazole stock solution (0.20 M, 0.90 ×

10-5 mol) was mixed together. To this solution was added 0.060 mL Zn(NO3)2·4H2O

stock solution (0.15 M, 0.90 × 10-5 mol). After the glass plate was loaded with mixtures

of stock solutions dispensed by a programmed liquid handler (Gilson, model 215), it was

covered with a PTFE sheet, sealed by fastening the sheet with a metal clamp, then heated

20

in an oven at 85 °C and allowed to react solvothermally for 72 h. The product was in the

form of prism-shaped single crystals.

(ZIF-74 gis): Zn(nIM)(dmbIM). 0.18 mL 2-nitroimidazole stock solution (0.20 M, 3.6

× 10-5 mol) and 0.060 mL 5,6-dimethylbenzimidazole stock solution (0.20 M, 1.2 × 10-5

mol) was mixed together. To this solution was added 0.060 mL Zn(NO3)2·4H2O stock

solution (0.15 M, 0.90 × 10-5 mol). After the glass plate was loaded with mixtures of

stock solutions dispensed by a programmed liquid handler (Gilson, model 215), it was

covered with a PTFE sheet, sealed by fastening the sheet with a metal clamp, then heated

in an oven at 85 °C and allowed to react solvothermally for 5 days. The product was in

the form of rod-shaped single crystals.

Elemental analysis C12H13N5O2Zn = Zn(NO2IM)(5,6Me2PhIM): Calcd. C, 39.28; H,

5.36; N, 28.65. Found C, 39.47; H, 4.39; N, 27.13.


1469(m), 1342(w), 1286(w), 1092(w), 939(w), 802(w), 669(w), 603(s).

21

0

1000

2000

3000

4000

5000

6000

2 12 22 32

Experimental

Simulated

2θ (degree)





22

(ZIF-75 gis): Co(nIM)(dmbIM). 0.12 mL imidazole stock solution (0.15 M, 1.8 × 10-5

mol) and 0.12 mL 5,6-dimethylbenzimidazole stock solution (0.15 M, 0.90 × 10-5 mol)

was mixed together. To this solution was added 0.060 mL Co(NO3)2·6H2O stock solution




oven at 85 °C and allowed to react solvothermally for 5 days. The product was in the

form of pink colored rod -shaped single crystals.

(ZIF-76 lta): Zn(IM)(cbIM). 0.15 mL imidazole stock solution (0.15 M, 2.25 × 10-5

mol) and 0.075 mL 5-chlorobenzimidazole stock solution (0.15 M, 1.13 × 10-5 mol) was

mixed together. To this solution was added 0.075 mL Zn(NO3)2·4H2O stock solution




oven at 65 °C and allowed to react solvothermally for 5 days. The product was in the

form of hexagon-shaped single crystals.

Elemental analysis C8H13N5Zn = Zn(IM)1.5(5ClPhIM)0.5·(H2O)4: Calcd. C, 33.96; H,

5.57; N, 18.05. Found C, 33.27; H, 5.39; N, 18.13.

Elemental analysis (activated) C8H13N5Zn = Zn(IM)1.5(5ClPhIM)0.5: Calcd. C, 39.28; H,

5.36; N, 28.65. Found C, 39.47; H, 4.39; N, 27.13.


1469(m), 1342(w), 1286(w), 1092(w), 939(w), 802(w), 669(w), 603(s).

23

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

2 7 12 17 22 27 32 2θ (degree)





(ZIF-77 frl): Zn(nIM)2. To a 0.26 mL 2-nitroimidazole stock solution (0.20 M, 5.2 × 10-

5 mol) 0.060 mL Zn(NO3)2·4H2O stock solution (0.20 M, 1.2 × 10-5 mol) was added.



24



single crystals.

25

Section S2. Representative optical photographs of ZIFs crystals (0.1 to 1 mm) and

single crystal x-ray diffraction data collection, structure solution and refinement

procedures.

General Data Collection and Refinement Procedures:

Data were collected on a Bruker SMART APEXII three circle diffractometer equipped

with a CCD area detector and operated at 1200 W power (40 kV, 30 mA) to generate Cu

Kα radiation (λ = 1.5418 Å). The incident x-ray beam was focused and monochromated

using Bruker Excalibur Gobel mirror optics. Crystals of ZIF-60, 61, 62, and 70 were

mounted in flame sealed capillaries containing a small amount of mother liquor to

prevent desolvation during data collection. Photographic images of wells containing

crystals (0.1 to 1.0 mm) are shown in Fig. S8. All crystals of other ZIFs reported in the

paper were mounted on nylon CryoLoops (Hampton Research) with Paraton-N (Hampton

Research).

Initial scans of each specimen were performed to obtain preliminary unit cell parameters

and to assess the mosaicity (i.e. breadth of spots between frames) of the crystal to select

the required frame width for data collection. In every case frame widths of 0.5° were

judged to be appropriate and full hemispheres of data were collected using the Bruker

APEX21 software suite to carry out overlapping φ and ω scans at three different detector

(2θ) settings (2θ = 28, 60, 100°). Following data collection, reflections were sampled

from all regions of the Ewald sphere to redetermine unit cell parameters for data

integration and to check for rotational twinning using CELL_NOW2. In no data collection

was evidence for crystal decay encountered. Following exhaustive review of collected

frames the resolution of the dataset was judged. Data were integrated using Bruker

26

APEX2 V 2.13 software with a narrow frame algorithm and a 0.400 fractional lower limit

of average intensity. Data were subsequently corrected for absorption by the program

SADABS4. The absorption coefficient (μ) ranges between 1-6 for most of the ZIFs

reported in this paper. However it is note worthy that μ is based on the atomic contents

and this is uncertain for most of the structures diffuse electron density of the guest

molecule, this number is imprecise. Space group determination and tests for merohedral

twinning were carried out using XPREP3. ZIF-68, 69, and 70 were found to be

merohedrally twinned. This could possibly be the reason for their higher R–factor

although all the atomic positions were defined. In all other cases the highest possible

space group was chosen and no indication of merohedral twinning was observed.

All structures were solved by direct methods and refined using the SHELXTL 975

software suite. Atoms were located from iterative examination of difference F-maps

following least squares refinements of the earlier models. Final models were refined

anisotropically (if the number of data permitted) until full convergence was achieved.

However, several structures contained disordered components within the frameworks. In

these cases occupancies X and (1 – X) and equivalent displacement parameters were

assigned to each pair of atoms in the disordered groups and X was refined as an

independent free variables. Hydrogen atoms were placed in calculated positions (C—H =

0.93 Å) and included as riding atoms with isotropic displacement parameters 1.2–1.5

times Ueq of the attached C atoms. Modeling of electron density within the voids of the

frameworks did not lead to identification of guest entities in all structures due to the

disordered contents of the large pores in the frameworks. This challenge, which is typical

of highly porous crystals that contain solvent filled pores, lies in the raw data where

27

observed strong (high intensity) scattering becomes limited to ~1.0 Å at best, with higher

resolution data present but weak (low intensity). As is a common strategy for improving

x-ray data, increasing the exposure time of the crystal to x-rays did not improve the

quality of the high angle data in these cases, as the intensity from low angle data

saturated the detector and minimal improvement in the high angle data was achieved.

Additionally, diffuse scattering from the highly disordered solvent in the void spaces

within the crystal and from the capillary used to mount the crystal contributes to the

background noise and the 'washing out' of high angle data. The only optimal crystals

suitable for analysis were generally small and weakly diffracting, and unfortunately,

larger crystals, which would usually improve the quality of the data, presented a lowered

degree of crystallinity and attempts to optimize the crystal growing conditions for large

high-quality specimens has not yet been fruitful. For ZIF-60, 61, 62, and 70 data have

been collected at elevated temperatures (-15 °C, rather than -120 to -100 °C). For other

ZIFs reported in this paper it was found that data collection at -120 °C was optimal for

obtaining the best data. In such cases the modeling of the disordered guest entities

becomes intractable because they become frozen into highly disordered arrays. Thus,

electron density within void spaces, which could not be assigned to any definite guest

entity, was modeled as isolated oxygen atoms, and the foremost errors in all the models

lies with assignment of guest electron density. To prove the correctness of the atomic

positions in the framework the application of the SQUEEZE6 routine of A. Spek had been

performed. However atomic co-ordinates for the “non-SQUEEZE” structures have also

been presented and the CIFs were also submitted for the cases where the program

SQUEEZE has been employed. The assignment and refinement of the metal-organic ZIF

28

framework atoms was unambiguous, as judged by the resulting bond and angle metrics

which are chemically accurate and precise values. All structures were examined using the

Adsym subroutine of PLATON7 to assure that no additional symmetry could be applied to

the models. All ellipsoids in ORTEP diagrams are displayed at the 30% probability level

unless noted otherwise. For all structures we note that the elevated R-values are

commonly encountered in MOF crystallography, for the reasons expressed above, by us

and other research groups 8-17

1. Bruker (2005). APEX2. Version 5.053. Bruker AXS Inc., Madison,Wisconsin,

USA.

2. Sheldrick, G. M. (2004). CELL_NOW. University of Göttingen, Germany.

Steiner, Th. (1998). Acta Cryst. B54, 456–463.

3. Bruker (2004). SAINT-Plus (Version 7.03). Bruker AXS Inc., Madison,

Wisconsin, USA.

4. Sheldrick, G. M. (2002). SADABS (Version 2.03) and TWINABS (Version

1.02).University of Göttingen, Germany.

5. Sheldrick, G. M. (1997). SHELXS ‘97 and SHELXL ‘97. University of Göttingen,

Germany.

6. WINGX

7. Spek, A. L. (2005) PLATON, A Multipurpose Crystallographic Tool, Utrecht

University, Utrecht, The Netherlands.

8. Dakin, L. A., Ong P. C., Panek, J. S., Staples, R. J. & Stavropoulos, P.

Organometallics 19, 2896–2908 (2000).

29

9. Noro, S., Kitaura, R., Kondo, M., Kitagawa, S., Ishii, T., Matsuzaka, H. &

Yamashita, M. J. Am. Chem. Soc. 124, 2568–2583 (2002).

10. Eddaoudi, M., Kim, J., Vodak, D., Sudik, A., Wachter, J., O’Keeffe, M. & Yaghi,

O. M. Proc. Natl. Acad. Sci. U.S.A. 99, 4900–4904 (2002).

11. Heintz, R. A., Zhao, H., Ouyang, X., Grandinetti, G., Cowen, J. & Dunbar, K. R.

Inorg. Chem. 38, 144–156 (1999).

12. Biradha, K., Hongo, Y. & Fujita, M. Angew. Chem. Int. Ed. 39, 3843–3845

(2000).

13. Grosshans, P., Jouaiti, A., Hosseini, M. W. & Kyritsakas, N. New J. Chem, (Nouv.

J. Chim.) 27, 793–797 (2003).

14. Takeda, N., Umemoto, K., Yamaguchi, K. & Fujita, M. Nature (London) 398,

794–796 (1999).

15. Eddaoudi, M., Kim, J., Rosi, N., Vodak, D., Wachter, J., O’Keeffe, M. & Yaghi,

O. M. Science 295, 469–472 (2002).

16. Kesanli, B., Cui, Y., Smith, M. R., Bittner, E. W., Bockrath, B. C. & Lin, W.

Angew. Chem. Int. Ed. 44, 72–75 (2005).

17. Cotton, F. A., Lin, C. & Murillo, C. A. Inorg. Chem. 40, 478–484 (2001).

30

Fig. S8. Representative optical photographs of zeolitic imidazolate frameworks (ZIFs)

crystals (0.1 to 1 mm) found in the wells after reacting reagents using the high-

throughput method.

31

ZIF-60 (MER─ Tetragonal)

Experimental and Refinement Details for ZIF-60

A colorless prismatic crystal (0.25 × 0.20 × 0.18 mm3) of ZIF-60 was placed in a 0.4 mm

diameter borosilicate capillary along with a small amount of mother liquor. The capillary

was flame sealed and mounted on a SMART APEXII three circle diffractometer

equipped with a CCD area detector and operated at 1200 W power (40 kV, 30 mA) to

generate Cu Kα radiation (λ = 1.5418 Å) while being flash frozen to 258(2) K in a liquid

N2 cooled stream of nitrogen. A total of 36256 reflections were collected of which 2911

were unique and 2404 of these were greater than 2σ(I). The range of θ was from 2.29 to

59.36°. Analysis of the data showed negligible decay during collection. The structure was

solved in the tetragonal I4/mmm space group, with Z = 4, using direct methods. Atoms

C2 and C4 were found to be disordered and were refined isotropically with the occupancy

for each group modeled as its own independent free variable (X, 1 – X). All other non-

hydrogen atoms were refined anisotropically with hydrogen atoms generated as spheres

riding the coordinates of their parent atoms. Modeling of electron density within the

voids of the frameworks did not lead to identification of guest entities in all structures

due to the lowered resolution of the data. The attempts made to model the guests (solvent

molecules) did not lead to identification of guest entities in all structures due to the

limited periodicity of the solvent molecules in the crystals. Since the solvent is not

bonded to the framework this can be expected for the MOF structures. Many atomic co-

ordinates that have been attributed to solvent molecules lie on a special position.

However, very high displacement parameters, high esd’s and partial occupancy due to the

disorder make it impossible to determine accurate positions for these solvent molecules.

32

Thus, electron density within void spaces which could not be assigned to any definite

guest entity was modeled as isolated carbon and oxygen atoms, and the foremost errors in

all the models lies with assignment of guest electron density. To prove the correctness of

the atomic positions in the framework the application of the SQUEEZE routine of A.

Spek has been performed. However atomic co-ordinates for the “non-SQUEEZE”

structures are also presented. The unit cell of ZIF-60 contains 1.5 imidazole and 0.5 2-

methyl imidazole (3:1) per zinc. Final full matrix least-squares refinement on F2

converged to R1 = 0.0360 (F >2σF)) and wR2

= 0.1157 (all data) with GOF = 1.093. For

the structure where the SQUEEZE program has not been employed, final full matrix

least-squares refinement on F2 converged to R1

= 0.0674 (F >2σF)) and wR2 = 0.2605 (all

data) with GOF = 1.059. When only framework atoms are included in the latter structure

factor calculation, the residual electron density in the F-map is located within the pores of

ZIF-60.

33

Table S1. Crystal data and structure refinement for ZIF-60. Empirical formula C87 H56 N36 O4 Zn8

Formula weight 2192.80

Temperature 258(2) K

Wavelength 1.54178 Å

Crystal system Tetragonal

Space group I4/mmm

Unit cell dimensions a = 27.2448(3) Å α = 90°

b = 27.2448(3) Å β = 90°

c = 19.2254(3) Å γ = 90°

Volume 14270.6(3)

Z 4

Density (calculated) 1.021

Absorption coefficient 1.839

F(000) 4408

Crystal size 0.25 × 0.20 × 0.18 mm3

Theta range for data collection 2.29- 59.36

Index ranges -30 <= h <= 24, -30 <= k <= 28, -21 <= l <= 19

Reflections collected 36256

Independent reflections 2911 [Rint= 0.0247]

Completeness to theta = 59.36° 99.7%

Absorption correction Semi-empirical from equivalents

Refinement method Full-matrix least-squares on F2

Data / restraints / parameters 2911/0/146

Goodness-of-fit on F2 1.059

Final R indices [I>2sigma(I)] R1 = 0.0674, wR2 = 0.2421

R indices (all data) R1 = 0.0761, wR2 = 0.2605

Largest diff. peak and hole 0.634 and -0.444 e.Å-3

34

Table S2. Crystal data and structure refinement for ZIF-60 (SQUEEZE). Empirical formula C13 H14 N8 Zn2





Space group I4/mmm


b = 27.2448(3) Å β = 90°

c = 19.2254(3) Å γ = 90°

Volume 14270.6(3)

Z 16



F(000) 3328.0

Crystal size 0.25 × 0.20 × 0.18 mm3


Index ranges -30 <= h <= 24, -30 <= k <= 28, -21 <= l <= 19











35

Figure S9. ORTEP drawing of an asymmetric unit of ZIF-60, excluding the guest entities.

36

ZIF-61 (ZNI─ Tetragonal)



diameter borosilicate capillary along with a small amount of mother liquor. The capillary

was flame sealed and mounted on a SMART APEXII three circle diffractometer


generate Cu Kα radiation (λ = 1.5418 Å) radiation while being flash frozen to 258(2) K in

a liquid N2 cooled stream of nitrogen. A total of 17067 reflections were collected of

which 1443 were unique and 1335 of these were greater than 2σ(I). The range of θ was

from 3.76 to 64.55°. Analysis of the data showed negligible decay during collection. The

structure was solved in the tetragonal I41/acd space group with Z = 2 using direct

methods. Atoms C1, C5, and C6 were found to be disordered and were refined

anisotropically with the occupancy for each group modeled as its own independent free

variable (X, 1 – X). All other non-hydrogen atoms were also refined anisotropically with

hydrogen atoms generated as spheres riding the coordinates of their parent atoms.

Modeling of electron density within the voids of the frameworks did not lead to

identification of guest entities in all structures due to the lowered resolution of the data.

The unit cell of ZIF-61 contains one imidazole and one 2-methylimidazole (1:1) per zinc.

Final full matrix least-squares refinement on F2 converged to R1 = 0.0523 (F >2σF)) and

wR2 = 0.1339 (all data) with GOF = 1.148.

37

Table S3. Crystal data and structure refinement for ZIF-61. Empirical formula C13 H14 N8 Zn2





Space group I41/acd


b = 23.4803(3) Å β = 90°

c = 12.5545(3) Å γ = 90°

Volume 6921.6(3)

Z 2



F(000) 3312.0

Crystal size 0.21 × 0.18 × 0.16 mm3


Index ranges -25 <= h <= 27, -26 <= k <= 27, -12 <= l <= 14











38

Figure S10. ORTEP drawing of the asymmetric unit of ZIF-61, excluding the guest entities.

39

ZIF-62 (CAG─ Orthorhombic)



diameter borosilicate capillary with a small amount of mother liquor. The capillary was

flame sealed and mounted on a SMART APEXII three circle diffractometer equipped

with a CCD area detector and operated at 1200 W power (40 kV, 30 mA) to generate Cu

Kα radiation (λ = 1.5418 Å) radiation while being flash frozen to 258(2) K in a liquid N2

cooled stream of nitrogen. A total of 17208 reflections were collected, of which 2873



solved in the orthorhombic Pbca space group with Z = 2 using direct methods. Modeling

of electron density within the void of the frameworks leads to identification of one DMF

guest molecule in the asymmetric unit. These imidazole linkers have no crystallographic

symmetry. The contents per 2 Zn (independent) = 4 linkers; 3 are unsubstituted imidazole

with no disorder and the fourth is 0.375(11) benzimidazole and 0.625(11) imidazole.

Note that this fourth position can contain a maximum of 50% benzimidazole because a

center of symmetry in this space group would bring the benzene rings related by this

center into contact with each other. No such limitation exists for the unsubstituted

imidazole. With the exception of the 4 partial occupancy benzene carbon atoms and the

solvent atoms, all non-hydrogen atoms have been refined anisotropically. All hydrogen

atoms have been placed in geometrically located positions and their displacement

parameters are tied to those of the attached carbon atoms. The unit cell of ZIF-62

contains 1.812 imidazole and 0.187 benzimidazole per zinc. Final full matrix least-

40

squares refinement on F2 converged to R1 = 0.0590 (F >2σF)) and wR2

= 0.1890 (all data)

with GOF = 1.139.

41





Crystal system Orthorhombic

Space group Pbca


b = 15.6620(14) Å β = 90°

c = 18.2073(19) Å γ = 90°

Volume 4466.2(7)

Z 2



F(000) 1810

Crystal size 0.20 × 0.15 × 0.12 mm3


Index ranges -16 <= h <= 16, -14 <= k <= 16, -19 <= l <= 19






Data / restraints / parameters 2873/11/ 236





42

Figure S11. ORTEP drawing of two zinc atoms surrounded by seven linkers of ZIF-62 excluding the guest entities. Imidazole in the center is shared by two crystallographically different zinc atoms. This central imidazole is the major (60%) occupant.

Figure S12. ORTEP drawing of two zinc atoms surrounded by seven linkers of ZIF-62 excluding the guest entities. Benzimidazole in the center is shared by two crystallographically different zinc atoms. This central benzimidazole is the minor (40%) occupant.

43

ZIF-64 (CRB─Monoclinic)


A colorless needle shaped crystal (0.25 × 0.14 × 0.12 mm3) of ZIF-64 was placed in a 0.7

mm diameter nylon CryoLoops (Hampton Research) with Paraton-N (Hampton

Research). The loop was mounted on a SMART APEXII three circle diffractometer



N2 cooled stream of nitrogen. A total of 17208 reflections were collected, of which 2873



solved in the monoclinic P2/n space group with Z = 16 using direct methods. Atoms C6

and C21 were found to be disordered and were refined anisotropically with the

occupancy for each group modeled as its own independent free variable (X, 1 – X). All

other non-hydrogen atoms were refined anisotropically with hydrogen atoms generated as

spheres riding the coordinates of their parent atoms. Although the unit cell has a = c and

all angles 90˚, Rint is 0.4189 for a tetragonal cell and 0.1546 for a monoclinic cell.

XPREP prefers a monoclinic cell. The 4 independent Zn atoms of the monoclinic space

group drop to 2 in the space group P4/n, but in the later space group the model refines

poorly (R ∼ 28%) and the disordered carbon atoms of the imidazole retain their disorder.

An inspection of hkl with b as the unique axis has h 0 0 not equal to 0 0 l. Also, although

there are no strong 0 k 0 reflections, the odd k are not entirely absent. The n-glide absence

(h 0 l, h 1 l ≠ 2n) are required for both space groups to be tested. Modeling of electron

density within the void of the frameworks leads to identification of 2.5 DMF guest

44

molecules in the asymmetric unit. These imidazole linkers have no crystallographic

symmetry. The unit cell of ZIF-64 contains two imidazole linkers per zinc. All hydrogen

atoms have been placed in geometrically located positions and their displacement

parameters are tied to those of the attached carbon atoms. Final full matrix least-squares

refinement on F2 converged to R1 = 0.1116 (F >2σF)) and wR2

= 0.2573 (all data) with

GOF = 1.083.

45

Table S5. Crystal data and structure refinement for ZIF-64. Empirical formula C7.75 H9.25 N4.56 O0.56 Zn




Crystal system Monoclinic

Space group P2/n


b = 9.906(2) Å β = 90.00(2)°

c = 21.110(4) Å γ = 90°

Volume 4414.6(14)

Z 16



F(000) 1955

Crystal size 0.25 × 0.14 × 0.12 mm3


Index ranges -21 <= h <= 21, -10 <= k <= 9, -21 <= l <= 17






Data / restraints / parameters 2873/0/ 497





46

Figure S13. ORTEP drawing of four unique zinc atoms and eight unique imdazole linkers of ZIF-64 excluding the guest entities.

47

ZIF-65 (SOD─ CUBIC)



diameter nylon CryoLoops (Hampton Research) with Paraton-N (Hampton Research).

The loop was mounted on a SMART APEXII three circle diffractometer equipped with a

CCD area detector and operated at 1200 W power (40 kV, 30 mA) to generate Cu Kα

radiation (λ = 1.5418 Å) while being flash frozen to 153(2) K in a liquid N2 cooled

stream of nitrogen. A total of 7062 reflections were collected of which 264 were unique

and 257 of these were greater than 2σ(I). The range of θ was from 3.62 to 36.79°.

Analysis of the data showed negligible decay during collection. The structure was solved

in the cubic I-43m space group, with Z = 2, using direct methods. All non-hydrogen

atoms were refined anisotropically with hydrogen atoms generated as spheres riding the

coordinates of their parent atoms. Modeling of electron density within the voids of the

frameworks did not lead to identification of guest entities in all structures due to the

lowered resolution of the data. Since the solvent is neither bonded to the framework nor

tightly packed into the voids, this can be expected for the MOF structures. The unit cell

of ZIF-65 contains two 2-nitroimidazole linker per cobalt. Final full matrix least-squares



GOF = 1.113. When only framework atoms are included in the structure factor

calculation, the residual electron density in the F-map is located within the pores of ZIF-

65.

48

Table S6. Crystal data and structure refinement for ZIF-65. Empirical formula C36 H24 Co6 N36 O24




Crystal system Cubic

Space group I-43m


b = 17.2715(4) Å β = 90°

c = 17.2715(4) Å γ = 90°

Volume 5152.2(2)

Z 2



F(000) 1692

Crystal size 0.20 × 0.16 × 0.12 mm3

Theta range for data collection 3.62 to 36.79°

Index ranges -13 <= h <= 13, -13 <= k <= 13, -12 <= l <= 13



Completeness to theta = 36.79° 100%








49


50

ZIF-67 (SOD─ CUBIC)










in the cubic I-43m space group, with Z = 4, using direct methods. All non-hydrogen



frameworks did not lead to identification of guest entities in any of the structures due to

the lowered resolution of the data. The attempts made to model the guests (solvent

molecules) did not lead to identification of guest entities in any of the structures due to

the limited periodicity of the solvent molecules in the crystals. Since the solvent is neither

bonded to the framework nor tightly packed into the voids, this can be expected for the

MOF structures. Thus, electron density within void spaces which could not be assigned to

any definite guest entity was modeled as isolated oxygen atoms, and the foremost errors

in all the models lie with the assignment of guest electron density. To assess the

correctness of the atomic positions in the framework, the application of the SQUEEZE

51

routine of A. Spek has been performed. However, atomic co-ordinates for the “non-

SQUEEZE” structures are also presented. The unit cell of ZIF-67 contains two 2-

methylimidazole linkers per cobalt. Final full matrix least-squares refinement on F2


= 0.0825 (all data) with GOF = 1.118. For

the structure where the SQUEEZE program has not been employed, final full matrix

least-squares refinement on F2 converged to R1 = 0.0724 (F >2σF)) and wR2

= 0.2138 (all

data) with GOF = 1.193. When only framework atoms are included in the latter structure

factor calculation, the residual electron density in the F-map is located within the pores of

ZIF-67.

52






Space group I-43m


b = 16.9589(3) Å β = 90°

c = 16.9589(3) Å γ = 90°

Volume 4877.45(15)

Z 4



F(000) 1420

Crystal size 0.20 × 0.16 × 0.12 mm3


Index ranges -14 <= h <= 15, -15 <= k <= 14, -14 <= l <= 15











53

Table S8. Crystal data and structure refinement for ZIF-67 (SQUEEZE). Empirical formula C8 H10 Co N4





Space group I-43m


b = 16.9589(3) Å β = 90°

c = 16.9589(3) Å γ = 90°

Volume 4877.45(15)

Z 4



F(000) 1420

Crystal size 0.20 × 0.16 × 0.12 mm3


Index ranges -14 <= h <= 15, -15 <= k <= 14, -14 <= l <= 15











54


55

ZIF-68 (GME─ HEXAGONAL)







stream of nitrogen. A total of 30723 reflections were collected of which 1382 were

unique and 1209 of these were greater than 2σ(I). The range of θ was from 1.91 to

40.66°. Analysis of the data showed negligible decay during collection. Space groups

P63/mmc (hexagonal) and P-31c (trigonal) were suggested by XPREP with very similar

CFOM (7.60 vs 7.32). The structure was solved in the hexagonal P63/mmc space group,

with Z = 24, using direct methods. Atoms Zn1, N2, N3, C1, and C11 were refined

anisotropically. All other non-hydrogen atoms were refined isotropically with hydrogen

atoms generated as spheres riding the coordinates of their parent atoms. ZIF-68 is

composed of one 2-nitroimidazole and one benzimidazole per Zn. An asymmetric unit

contains one half nitroimidazole (linker with mirror symmetry, no disorder), one half

benzimidazole (linker with mirror symmetry, no disorder), one half nitroimidazole (linker

with mirror symmetry), and one half benzimidazole (linker with a 2-fold axis). The

attempts made to model the guests (solvent molecules) did not lead to identification of

guest entities in any of the structures due to the limited periodicity of the solvent

56

molecules in the crystals. Since the solvent is neither bonded to the framework nor tightly

packed into the voids, solvent disorder can be expected for the MOF structures. Thus,

electron density within void spaces which could not be assigned to any definite guest

entity was modeled as isolated carbon atoms, and the foremost errors in all the models

lies with the assignment of guest electron density. To assess the correctness of the atomic

positions in the framework, the application of the SQUEEZE routine of A. Spek has been

performed. It should be noted that the precision of this model is low; however, the

structure is reported to display the framework for ZIF-68 as isolated in the crystalline

form. Other supporting characterization data (vide infra Materials and Methods) are

consistent with the crystal structure. Final full matrix least-squares refinement on F2

converged to R1 = 0.1367 (F >2σF)) and wR2 = 0.4772 (all data) with GOF = 2.374.

When only framework atoms are included in the latter structure factor calculation, the

residual electron density in the F-map is located within the pores of ZIF-68.

57

Table S9. Crystal data and structure refinement for ZIF-68. Empirical formula C7.06 H4.94 N3.53 O1.59 Zn0.71




Crystal system Hexagonal

Space group P63/mmc


b = 26.6407(4) Å β = 90°

c = 18.4882(4) Å γ = 120°

Volume 11363.6(3)

Z 24



F(000) 3600

Crystal size 0.24 × 0.18 × 0.16 mm3


Index ranges -22 <= h <= 22, -22 <= k <= 23, -14 <= l <= 16











58

Figure S16. ORTEP drawing of zinc atom surrounded by four linkers of ZIF-69 excluding the guest entities.

59













with Z = 24, using direct methods. Atoms N1, C12, O3 and N9 were found to be

disordered and were refined anisotropically with the occupancy for each group modeled

as its own independent free variable (X, 1 – X). All other non-hydrogen atoms were

refined anisotropically with hydrogen atoms generated as spheres riding the coordinates

of their parent atoms. ZIF-69 is composed of one 2-nitroimidazole and one

monochlorobenzoimidazole per Zn. An asymmetric unit contains one half nitroimidazole

(linker with mirror symmetry, no disorder), one half monochorobenzoimidazole (linker

with mirror symmetry, no disorder), one half nitroimidazole (linker with mirror

symmetry, disorder with separable atoms for NO2 and HCCH), and one half

60

monochlorobenzoimidazole (linker with a 2-fold axis, disorder with separable atoms for

N and for Cl). The attempts made to model the guests (solvent molecules) did not lead to

identification of guest entities in any of the structures due to the limited periodicity of the

solvent molecules in the crystals. Since the solvent is neither bonded to the framework

nor tightly packed into the voids, solvent disorder can be expected for the MOF

structures. Thus, electron density within void spaces which could not be assigned to any

definite guest entity was modeled as isolated carbon atoms, and the foremost errors in all

the models lies with the assignment of guest electron density. To assess the correctness of

the atomic positions in the framework, the application of the SQUEEZE routine of A.

Spek has been performed. However, atomic co-ordinates for the “non-SQUEEZE”

structures are also presented. It should be noted that the precision of this model is low;

however, the structure is reported to demonstrate the nature of the framework of ZIF-69.

Other supporting characterization data (vide infra Materials and Methods) agree with the

structure. Final full matrix least-squares refinement on F2 converged to R1 = 0.0716 (F

>2σF)) and wR2 = 0.1978 (all data) with GOF = 1.011. For the structure where the

SQUEEZE program has not been employed, final full matrix least-squares refinement on

F2 converged to R1 = 0.0930 (F >2σF)) and wR2

= 0.3169 (all data) with GOF = 1.671.



61

Table S10. Crystal data and structure refinement for ZIF-69. Empirical formula C12.58 H6 Cl N5.25 O2.50 Zn





Space group P63/mmc


b = 26.0840(18) Å β = 90°

c = 19.4082(18) Å γ = 120°

Volume 11435.7(15)

Z 24



F(000) 4446

Crystal size 0.24 × 0.18 × 0.16 mm3


Index ranges -23 <= h <= 22, -23 <= k <= 23, -16 <= l <= 17











62

Table S11. Crystal data and structure refinement for ZIF-69 (SQUEEZE). Empirical formula C10.25 H6 Cl N5.25 O2.75 Zn





Space group P63/mmc


b = 26.0840(18) Å β = 90°

c = 19.4082(18) Å γ = 120°

Volume 11435.7(15)

Z 24



F(000) 4158

Crystal size 0.24 × 0.18 × 0.16 mm3


Index ranges -23 <= h <= 22, -23 <= k <= 23, -16 <= l <= 17











63

Figure S17. ORTEP drawing of zinc atom surrounded by four linkers of ZIF-69 excluding the guest entities.

64




diameter borosilicate capillary along with a small amount of mother liquor. The sealed

capillary was mounted on a SMART APEXII three circle diffractometer equipped with a








with Z = 4, using direct methods. Atoms C6 and C7 were found to be disordered and

were refined anisotropically with the occupancy for each group modeled as its own

independent free variable (X, 1 – X). All other non-hydrogen atoms were refined

anisotropically with hydrogen atoms generated as spheres riding the coordinates of their

parent atoms. The attempts made to model the guests (solvent molecules) did not lead to

identification of guest entities in any of the structures due to the limited periodicity of the

solvent molecules in the crystals. Since the solvent is neither bonded to the framework

nor tightly packed into the voids, solvent disorder can be expected for the MOF

structures. Thus, electron density within void spaces which could not be assigned to any

definite guest entity was modeled as isolated carbon atoms, and the foremost errors in all

the models lie with the assignment of guest electron density. “Solvent” is 4 atoms refined

65

as carbon atoms of undefined solvent, located in a void of the framework. To assess the

correctness of the atomic positions in the framework, the application of the SQUEEZE

routine of A. Spek has been performed. However, atomic co-ordinates for the “non-

SQUEEZE” structures are also presented. The tetrahedrally coordinated Zn is surrounded

by four linkers; one imidazole with 2-fold symmetry and one 2-nitroimidazole with

mirror symmetry have no disorder. The two remaining positions have both imidazole and

2-nitroimidazole as linkers, always with mirror symmetry. In one position the 2-

nitroimidazole occupancy is 29.8% and in the other position this occupancy is 44.2%.

The framework contains 1.13 imidazole and 0.87 2-nitroimidazole per Zn. It should be

noted that the precision of this model is low; however, the structure is reported to

describe the framework of ZIF-70 can be isolated in crystalline form. Other supporting

characterization data (vide infra Materials and Methods) agree with the crystal structure.

Final full matrix least-squares refinement on F2 converged to R1 = 0.0580 (F >2σF)) and

wR2 = 0.1944 (all data) with GOF = 1.129. For the structure where the SQUEEZE

program has not been employed, final full matrix least-squares refinement on F2


= 0.2560 (all data) with GOF = 1.132.



66

Table S12. Crystal data and structure refinement for ZIF-70. Empirical formula C6 H5.13 N4.87 O1.74 Zn





Space group P63/mmc


b = 27.0111(9) Å β = 90°

c = 18.0208(9) Å γ = 120°

Volume 11386.5(10)

Z 4



F(000) 3028

Crystal size 0.24 × 0.18 × 0.16 mm3


Index ranges -30 <= h <= 31, -31 <= k <= 30, -21 <= l <= 20











67

Table S13. Crystal data and structure refinement for ZIF-70 (SQUEEZE). Empirical formula C6 H5.13 N4.87 O1.74 Zn





Space group P63/mmc


b = 27.0111(9) Å β = 90°

c = 18.0208(9) Å γ = 120°

Volume 11386.5(10)

Z 4



F(000) 2924

Crystal size 0.24 × 0.18 × 0.16 mm3


Index ranges -30 <= h <= 31, -31 <= k <= 30, -21 <= l <= 20











68

Figure S18. ORTEP drawing of zinc atom surrounded by four linkers of ZIF-70 excluding the guest entities and the major component imidazole in the disordered position.

Figure S19. ORTEP drawing of a zinc atom surrounded by four linkers of ZIF-70 excluding the guest entities and the minor component 2-nitroimidazole in the disordered positions.

69

ZIF-71 (RHO─ CUBIC)








unique and 782 of these were greater than 2σ(I). The range of θ was from 1.55 to 36.83°.


in the cubic Pm-3m space group, with Z = 48, using direct methods. All non-hydrogen


coordinates of their parent atoms. Since absorption corrections (SADABS) were

ineffectual for improving the data quality of data this data has been processed without an

absorption correction. Modeling of electron density within the voids of the framework

did not lead to identification of guest entities in all structures due to the lowered

resolution of the data and a limited periodicity of the solvent molecules in the crystals.

Since the solvent is not bonded to the framework this can be expected for the MOF

structures. Many atomic co-ordinates that have been attributed to solvent molecules lie on

a special position. However, very high displacement parameters, high esd’s and partial

occupancy due to the disorder make it impossible to determine accurate positions for

these solvent molecules. Thus, electron density within void spaces which could not be

assigned to any definite guest entity was modeled as isolated carbon and oxygen atoms,

70

and the foremost errors in all the models lie with assignment of guest electron density. To

prove the correctness of the atomic positions in the framework the application of the

SQUEEZE routine of A. Spek has been performed. However atomic co-ordinates for the

“non-SQUEEZE” structures are also presented. The unit cell of ZIF-71 contains two 4,5-

dichloroimidazole per zinc. Final full matrix least-squares refinement on F2 converged to

R1 = 0.0424 (F >2σF)) and wR2 = 0.1045 (all data) with GOF = 1.022. For the structure

where the SQUEEZE program has not been employed, final full matrix least-squares



GOF = 1.073. When only framework atoms are included in the latter structure factor

calculation, the residual electron density in the F-map is located within the pores of ZIF-

71.

71

Table S14. Crystal data and structure refinement for ZIF-71. Empirical formula C7.30 H2 Cl4 N4 O0.03 Zn





Space group Pm-3m


b = 28.5539(2) Å β = 90°

c = 28.5539(2) Å γ = 90°

Volume 23280.7(3)

Z 48



F(000) 8259

Crystal size 0.22 × 0.20 × 0.19 mm3


Index ranges -22 <= h <= 22, -22 <= k <= 22, -22 <= l <= 22




Absorption correction None







72

Table S15. Crystal data and structure refinement for ZIF-71 (SQUEEZE). Empirical formula C6 H2 Cl4 N4 Zn





Space group Pm-3m


b = 28.5539(2) Å β = 90°

c = 28.5539(2) Å γ = 90°

Volume 23280.7(3)

Z 48



F(000) 7872

Crystal size 0.22 × 0.20 × 0.19 mm3


Index ranges -22 <= h <= 22, -22 <= k <= 22, -22 <= l <= 22




Absorption correction None







73


74

ZIF-72 (LCS─ CUBIC)










in the cubic Ia-3d space group with Z = 2 using direct methods. Due to the low data to

parameter ratio atom C1 has been refined isotropically with the attached hydrogen atom

riding the coordinates of its parent atom. All other non-hydrogen atoms were refined

anisotropically. Absorption correction by SADABS decreases R(int) to 0.0235. The

absorption coefficient has a rather high value, μ = 10.268. A final ratio of 8.2 for

reflections to parameters was achieved. The unit cell of ZIF-72 reveals that the unit cell

contains two 4,5-dichloroimidazoles per zinc [Zn(45DCIM)2]. Final full matrix least-

squares refinement on F2 converged to R1 = 0.0172 (F >2σF)) and wR2 = 0.0371 (all data)

with GOF = 1.185.

75

Table S16. Crystal data and structure refinement for ZIF-72. Empirical formula C6 H2 Cl4 N4 Zn





Space group Ia-3d


b = 19.6544(2) Å β = 90°

c = 19.6544(2) Å γ = 90°

Volume 7592.40(13)

Z 2



F(000) 3936.0

Crystal size 0.18 × 0.16 × 0.15 mm3


Index ranges -18 <= h <= 17, -17 <= k <= 18, -18 <= l <= 17











76


77

ZIF-73 (FRL─ Orthorhombic)



diameter nylon Cryo-Loops (Hampton Research) with Paraton-N (Hampton Research).







solved in the orthorhombic Ibam space group, with Z = 2, using direct methods. In the

framework two types of Zn occur in the ration of 4 to 1, both tetrahedrally linked to

imidazole via N. Zn(1) is surrounded by three 2-nitroimidazole linkers, but the fourth

position of the tetrahedron is occupied by either 2-nitroimidazole (35%) or 5,6

dimethylbenzimidazole (65%). Zn(2) is surrounded by four 2-nitroimidazole linkers. The

framework contains Zn, 2-nitroimidazole and 5,6-dimethylbenzimidazole in the ratio

1:1.74:0.26. Note that 5,6-dimethylbenzimidazole is found in only one position of the

framework. As a result N4, C7, C9, C10, and C11 have been refined anisotropically with

a 65% (X) site occupancy factor and atoms C14, N8, N17 and O18 have been refined

anisotropically with a 35% (1 – X) occupancy. All other non-hydrogen atoms were

refined anisotropically with hydrogen atoms generated as spheres riding the coordinates

of their parent atoms. Modeling of electron density within the voids did not lead to the

identification of a guest molecule. Many atomic co-ordinates that have been attributed to

78

solvent molecules lie on a special position. However, very high displacement parameters,

high esd’s and partial occupancy due to the disorder make it impossible to determine

accurate positions for these solvent molecules. Thus, electron density within void spaces

which could not be assigned to any definite guest entity was modeled as isolated carbon

and oxygen atoms, and the foremost errors in all the models lie with assignment of guest

electron density. To prove the correctness of the atomic positions in the framework the

application of the SQUEEZE routine of A. Spek has been performed. However, atomic

co-ordinates for the “non-SQUEEZE” structures are also presented. The unit cell of ZIF-

73 contains 0.26 5,6-dimethyl benzimidazole and 1.74 2-nitroimidazole per zinc. Final

full matrix least-squares refinement on F2 converged to R1 = 0.0322 (F >2σF)) and wR2 =

0.0882 (all data) with GOF = 1.115. For the structure where the SQUEEZE program has

not been applied, final full matrix least-squares refinement on F2 converged to R1

=

0.0396 (F >2σF)) and wR2 = 0.1185 (all data) with GOF = 1.146.

79

Table S17. Crystal data and structure refinement for ZIF-73. Empirical formula C75.60 H57 N57.40 O34.80 Zn10





Space group Ibam


b = 22.7960(4) Å β = 90°

c = 24.9704(5) Å γ = 90°

Volume 6242.1(2)

Z 2



F(000) 3186

Crystal size 0.29 × 0.16 × 0.12 mm3


Index ranges -12 <= h <= 12, -25 <= k <= 26, -28 <= l <= 28











80

Table S18. Crystal data and structure refinement for ZIF-73 (SQUEEZE). Empirical formula C82.60 H57 N61.40 O38.80 Zn10





Space group Ibam


b = 22.7960(4) Å β = 90°

c = 24.9704(5) Å γ = 90°

Volume 6242.1(2)

Z 2



F(000) 2982

Crystal size 0.29 × 0.16 × 0.12 mm3


Index ranges -12 <= h <= 12, -25 <= k <= 26, -28 <= l <= 28











81

Figure S22. ORTEP drawing of an asymmetric unit of ZIF-73 excluding the guest entities and the minor component 2-nitroimidazole. C8, N4, C7, C9, C11, and C10 are part of the major component ligand, 5,6-dimethylbenzimidazole that links Zn1 atoms .

Figure S23. ORTEP drawing of an asymmetric unit of ZIF-73 excluding the guest Entities and the major component ligand, 5,6-dimethylbenzimidazole. C8, N8, C14, and N18 are atoms of the minor component ligand 2-nitroimidazole that links Zn1 atoms.

82

ZIF-74 (GIS─ Tetragonal)


A colorless prismic crystal (0.24 × 0.21 × 0.19 mm3) of ZIF-74 was placed in a 0.7 mm








solved in the tetragonal I41/a space group, with Z = 8, using direct methods. Atoms O1

and O2 were found to be disordered and were refined anisotropically with the occupancy

for each group modeled as its own independent free variable (X, 1 – X). All other non-

hydrogen atoms were refined anisotropically with hydrogen atoms generated as spheres

riding the coordinates of their parent atoms. It is probable that N5, which is bonded to the

disordered oxygen atoms O1 and O2, also could be modeled in two sites. However,

overlap for these N5 positions are too great and as a result N5 has been described as a

single atom with full occupancy (and consequently a rather high Ueq). C10, is expected

to have Ueq values similar to the rest of the five membered imidazole ring rather than

similar to the Ueq values for N5 of the NO2 ring. Modeling of electron density within the

voids of the frameworks lead to the identification of a DMF guest disordered about two

positions which has been refined isotropically with the hydrogen atoms for the CHO

functionality omitted. The unit cell of ZIF-74 contains one 5,6-dimethyl benzimidazole

83

and one 2-nitro imidazole (1:1) per zinc. Final full matrix least-squares refinement on F2

converged to R1 = 0.0440 (F >2σF)) and wR2 = 0.1292 (all data) with GOF = 1.097.

84






Space group I41/a


b = 16.6703(2) Å β = 90°

c = 21.6026(6) Å γ = 90°

Volume 6003.2(3)

Z 8



F(000) 2936

Crystal size 0.24 × 0.21 × 0.19 mm3


Index ranges -19 <= h <= 16, -16 <= k <= 18, -24 <= l <= 21











85


86

ZIF-75 (GIS─ Tetragonal)


A pink colored prismic crystal (0.24 × 0.21 × 0.19 mm3) of ZIF-75 was placed in a 0.7

mm diameter nylon Cryo-Loops (Hampton Research) with Paraton-N (Hampton






44.98°. Analysis of the data showed negligible decay during collection. However a high μ

= 9.195 and a low data to parameter ratio (6.08) is notable. The structure was solved in

the tetragonal I41/a space group, with Z = 4, using direct methods. Atoms O1 and O2

were found to be disordered and were refined isotropically with the occupancy for each

group modeled as its own independent free variable (X, 1 – X). All other non-hydrogen



frameworks does not lead to the identification of any guest molecule. Thus, electron

density within void spaces, which could not be assigned to any definite guest entity, was

modeled as isolated oxygen and carbon atoms, and the foremost errors in all the models

lies with assignment of guest electron density. The unit cell of ZIF-75 contains one 5,6-

dimethylbenzimidazole and one 2-nitroimidazole (1:1) per zinc. Final full matrix least-

squares refinement on F2 converged to R1 = 0.0574 (F >2σF)) and wR2 = 0.1476 (all data)

with GOF = 1.037.

87






Space group I41/a


b = 16.6695(2) Å β = 90°

c = 21.5797(5) Å γ = 90°

Volume 5996.4(2)

Z 4



F(000) 2816

Crystal size 0.24 × 0.21 × 0.19 mm3


Index ranges -14 <= h <= 15, -14 <= k <= 15, -19 <= l <= 18











88


89

ZIF-76 (LTA─ Cubic)










solved in the cubic P-43m space group, with Z = 24, using direct methods. Although a

partial structural solution was found in space group Pm-3m but a better refinement was

achived in space group P-43m. All non-hydrogen atoms except zinc were refined

isotropically with hydrogen atoms generated as spheres riding the coordinates of their

parent atoms. The attempts made to model the guests (solvent molecules) did not lead to

identification of guest entities in all structures due to the limited periodicity of the solvent

molecules in the crystals. Since the solvent is neither bonded nor fits tightly in the

framework disordered solvent molecules can be expected for the MOF structures. Many

atomic co-ordinates that have been attributed to solvent molecules lie on a special

position. However, very high displacement parameters, high esd’s and partial occupancy

due to the disorder make it impossible to determine accurate positions for these solvent

molecules. Thus, electron density within void spaces which could not be assigned to any

definite guest entity has been modeled as isolated carbon and oxygen atoms, and the

90

foremost errors in all the models lie in the assignment of the guest electron density. ZIF-

76 is composed of 1.5 imidazole and 0.5 monochloro-benzimidazole per Zn. The larger

linker 5-chlorobenzimidazole occurs in two different linker positions, at half occupancy

in each. In these positions, the other half occupant is imidazole and the imidazole cycle

perfectly overlaps the imidazole moiety of the larger linker. The contents of the crystal

agree well with the elemental analysis of the bulk sample (3 imidazole and 1

chlorobenzimidazole tetrahedrally coordinated to a Zn atom). The benzimidazole must

have half occupancy (or less) because full occupancy in either position would require two

crystallographically related benzimidazole linkers to approach each other too closely. The

linkers in the remaining two positions are fully occupied by unsubstituted imidazole. The

framework is further complicated by a disorder involving three of the four linkers.

Rotation of 60 degrees about one of the N-Zn bonds (N1–Zn1) gives a second nitrogen

triangle involving the other three Zn-N bonds. Still another disorder results because the

chlorine atom of the chlorobenzene is randomly (50%) bonded to either of the two

possible ring carbon atoms. Three of the four imidazole positions are located about

mirror planes; two chlorine atom positions of an imidazole are related by mirror

symmetry, although of course only one of these positions is filled for any given

imidazole. It should be noted that the precision of this model is low; however, the

structure is reported to describe the ZIF-76 framework. Other supporting characterization

data (vide infra Materials and Methods) agree with crystal structure. To asses the

correctness of the atomic positions in the framework the application of the SQUEEZE

routine of A. Spek has been applied. However atomic co-ordinates for the “non-

SQUEEZE” structures are also presented. The unit cell of ZIF-76 contains 1.5 imidazole

91

and 0.5 5-chlorobenzimidazole imidazole per zinc. Final full matrix least-squares

refinement on F2 converged to R1 = 0.0854 (F >2σF)) and wR2 = 0.2376 (all data) with

GOF = 1.136. For the structure where the SQUEEZE program has not been employed,

final full matrix least-squares refinement on F2 converged to R1 = 0.1108 (F >2σF)) and

wR1 = 0.3072 (all data) with GOF = 1.542. When only framework atoms are included in

the latter structure factor calculation, the residual electron density in the F-map is located

within the pores of ZIF-76.

92

Table S21. Crystal data and structure refinement for ZIF-76. Empirical formula C7.69 H3.58 Cl0.42 N2.78ZN0.67





Space group P-43m


b = 22.6702(3) Å β = 90°

c = 22.6702(3) Å γ = 90°

Volume 11651.08(18)

Z 24



F(000) 3466

Crystal size 0.25 × 0.20 × 0.18 mm3


Index ranges -17 <= h <= 17, -17 <= k <= 17, -17 <= l <= 17











93

Table S22. Crystal data and structure refinement for ZIF-76 (SQUEEZE). Empirical formula C5.83 H3.58 Cl0.42 N2.67





Space group P-43m


b = 22.6702(3) Å β = 90°

c = 22.6702(3) Å γ = 90°

Volume 11651.08(18)

Z 24



F(000) 3036

Crystal size 0.25 × 0.20 × 0.18 mm3


Index ranges -17 <= h <= 17, -17 <= k <= 17, -17 <= l <= 17











94

Figure S26. These figures depict two of the many Zn tetrahedron forms, each containing 3 imidazole linkers and 1 chlorobenzimidazole linker. Note that in one case Zn is tetrahedrally surrounded by N1, N2, N3, and N4, and chlorobenzimidazole contains N1. In the other figure, Zn is tetrahedrally surrounded by N1, N2’, N3’ and N4’, and chlorobenzimidazole contains N2’. The imidazole containing N1 is present in both figures, in one as chlorobenzimidazole and in the other as imidazole. Chlorine is present in one of the two mirror-related positions.

95

ZIF-77 (FRL─ Orthorhombic)


A pink colorless cubic crystal (0.24 × 0.21 × 0.20 mm3) of ZIF-77 was placed in a 0.7

mm diameter nylon Cryo-Loops (Hampton Research) with Paraton-N (Hampton






54.23°. Analysis of the data showed negligible decay during collection. However a high μ

= 3.058 (because μ is based on atomic contents and this is uncertain because of the guest

this number is quite imprecise) and a rather low data to parameter ratio (8.23) should be

noted. The structure was solved in the orthorhombic Ibam space group, with Z = 2, using

direct methods. All non-hydrogen atoms in the framework were refined anisotropically

with hydrogen atoms generated as spheres riding the coordinates of their parent atoms.

Modeling of electron density within the voids of the frameworks leads to the

identification of one DMF as a guest molecule. However the other electron density within

void spaces, which could not be assigned to any definite guest entity, was modeled as

oxygen and carbon atoms, and the foremost errors in all the models lie with assignment

of guest electron density. The unit cell of ZIF-77 contains two 2-nitroimidazole links per

zinc. Final full matrix least-squares refinement on F2 converged to R1 = 0.0434 (F >2σF))

and wR2 = 0.1267 (all data) with GOF = 1.190.

96






Space group Ibam


b = 22.3469(12) Å β = 90°

c = 24.9087(14) Å γ = 90°

Volume 6192.4(6)

Z 2



F(000) 3224

Crystal size 0.24 × 0.21 × 0.20 mm3


Index ranges -11 <= h <= 11, -22 <= k <= 23, -26 <= l <= 22











97


98

Section S3. Thermal stability of ZIFs and the thermal gravimetric analysis

(TGA) data.

All samples were run on a TA Instruments Q-500 series thermal gravimetric

analyzer with samples held in platinum pans in a continuous flow nitrogen atmosphere.

Samples were heated at a constant rate of 5 °C/min during all TGA experiments.

Figure S28. The overlay of TGA traces of as-synthesized, solvent-exchanged, and

evacuated (activated) samples of ZIF-60.

The TGA traces of the as-synthesized ZIF-60 sample and the sample after being

immersed in dry methanol solvents at ambient temperature for three days are shown in

Figure S28. Clearly, treatment with dry methanol simplified the thermogravimetric

behavior of ZIF-60 significantly, indicative of effective solvent-exchange. In particular,

99

in the TGA trace of methanol-exchanged ZIF-60 sample, the initial gradual weight-loss

step of 34.3% till 100 °C were replaced by a small initial step (16.7% of weight loss ) at

100 °C temperature, a plateau to 550 °C and a sharp weight loss from that point onwards

indicates the decomposition of the material.

Figure S29. The overlay of TGA traces of as-synthesized and evacuated (activated)

samples of ZIF-61.

The TGA traces of as synthesized ZIF-61 sample. A flat plateau to 300 °C and a

very small weight loss (5.2%) indicated non-porous nature of the material. The as

synthesized sample was evacuated at room temperature for 10 hours to remove the

adhering DMF solvent from the surface. A flat plateau to 550 °C and the subsequent

decomposition of the material indicated the activation of the sample was complete.

100



The TGA traces of an as synthesized ZIF-68 sample and a sample after being


Figure S30. Clearly, treatment with methanol simplified the thermogravimetric behavior

of ZIF-68 significantly, indicative of effective solvent-exchange. In particular, in the

TGA trace of methanol-exchanged ZIF-68 sample, the original gradual weight-loss step

of 21.3% up to 168 °C was replaced by a small initial step (4.5%) at 118 °C temperature,

a plateau up to 400 °C and a sharp weight loss from that point onwards indicates the

decomposition of the material. The sample was activated by heating it at 85 °C under

vacuum for 72 hours. A flat plateau up till 400 °C and a subsequent decomposition of the

material indicated the activation of the sample was complete.

101



The TGA traces of as synthesized ZIF-69 sample and a sample after being




TGA trace of methanol-exchanged ZIF-69 sample, the original gradual weight-loss step

of 10.6% up to 164 °C was replaced by a small initial step (4.1%) at 80 °C temperature, a

plateau till 400 °C and a sharp weight loss from that point onwards indicates the


vacuum for 72 hours. A flat plateau to 400 °C and a subsequent decomposition of the


102



The TGA traces of an as synthesized ZIF-70 sample and a sample after being




TGA trace of a methanol-exchanged ZIF-70 sample, the original gradual weight-loss step

of 27.2% up to 170 °C were replaced by a small initial step (15.4%) at ambient

temperature (48 °C), a plateau till 350 °C and a sharp weight loss from that point onwards

indicates the decomposition of the material. The sample was activated by heating it at 85

°C under vacuum for 72 hours. A flat plateau to 350 °C and a subsequent decomposition

of the material indicated the activation of the sample was complete.

103



The TGA traces of as-synthesized ZIF-76 sample and a sample after being




TGA trace of a methanol-exchanged ZIF-68 sample, the original gradual weight-loss step

of 21.9% up to 164 °C were replaced by a small initial step (2%) at ambient temperature,

a plateau up to 550 °C and a sharp weight loss from that point onwards indicates the


vacuum for 72 hours. A flat plateau up till 550 °C and a subsequent decomposition of the


104

0

5000

10000

15000

20000

25000

30000

35000

2 12 22 32

Activated

Simulated

2 θ (degree)

Figure S34. PXRD patterns of activated and simulated ZIF-68. The PXRD pattern of the

activated sample has been collected after the sorption studies. Y-axis of the graph shows

the relative intensity.

105

0

2000

4000

6000

8000

10000

12000

14000

2 12 22 322θ(degree)

Activated

Simulated




106

0

5000

10000

15000

20000

25000

2 12 22 32

Activated

Simulated




107

Section S4. Chemical stability of ZIFs in refluxing organic and aqueous media and

the corresponding PXRD patterns.

ZIF-61, 68, 69, and 70 were tested for their stability in benzene, methanol and water.

These solvents were chosen to compare the relative effects of non-polar to polar solvents.

The tests were performed at room temperature, 50 °C and at the boiling point of each

solvent (methanol 65 °C, benzene 80 °C and water 100 °C) for up to 7 days. The

structural stability of the frameworks was monitored by aliquoting portions of the

samples for PXRD analysis after every 24 h period. These conditions generally reflect the

extreme operational parameters of typical industrial chemical processes. ZIF samples

were immersed in the desired solvent for 1–7 days at ambient temperature, 50 °C, and at

the boiling point of each medium. During this process, samples were periodically

observed under an optical microscope and found to be insoluble under each of these

conditions. PXRD patterns collected for each sample at designated intervals showed that

the solid samples of ZIF-61, 68, 69, and 70 maintained their full crystallinity and were

clearly impervious to the boiling organic solvents for 7 days. Hydrothermal stability of

ZIF-61, 68, 69, and 70 is superior to those of original MCM and SBA types of ordered

mesoporous silica. This exceptional resistance to hydrolysis is outstanding among metal–

organic solids. The plausible explanations for this type of resistance to hydrolysis can be

attributed to the hydrophobic pore and surface structure of ZIFs that likely repels water

molecules to attack ZnN4 units; and most stable bonding between IM and Zn(II)/Co(II)

among N-donor ligands (on the scale of metal-complex formation constants).

108

ZIF-61 Stability Test: Benzene at RT

0

5000

10000

15000

20000

25000

2 12 22 32 42

As Prepared

1day

3 day

5 day

7 day

2θ (degree)

Figure S37. PXRD patterns of ZIF-61 collected during stability test in benzene at room

temperature. The framework structure of ZIF-61 was unchanged after 7 days. Y-axis of

the graph shows the relative intensity.

109

ZIF-61 Stability Test: Benzene at 50 °C

0

5000

10000

15000

20000

25000

30000

2 12 22 32 42

As Prepared

1day

3 day

5 day

7 day

2θ (degree)

Figure S38. PXRD patterns of ZIF-61 collected during stability test in benzene at 50 °C.

The framework structure of ZIF-61 was unchanged after 7 days. Y-axis of the graph

shows the relative intensity.

110


0

5000

10000

15000

20000

25000

30000

2 12 22 32 422θ (degree)

As Prepared

1day

3 day

5 day

7 day

Figure S39. PXRD patterns of ZIF-61 collected during stability test in refluxing benzene.



111

ZIF-61 Stability Test: MeOH at RT

0

5000

10000

15000

20000

25000

2 12 22 32 42

As Prepared

1day

3 day

5 day

7 day

2θ (degree)

Figure S40. PXRD patterns of ZIF-61 collected during stability test in methanol at room



112

ZIF-61 Stability Test: MeOH at 50 °C

0

5000

10000

15000

20000

25000

30000

2 12 22 32 42

As Prepared

1day

3 day

5 day

7 day

2θ (degree)

Figure S41. PXRD patterns of ZIF-61 collected during stability test in methanol at 50

°C. The framework structure of ZIF-61 was unchanged after 7 days. Y-axis of the graph


113


0

5000

10000

15000

20000

25000

30000

2 12 22 32 422θ (degree)

As Prepared

1day

3 day

5 day

7 day

Figure S42. PXRD patterns of ZIF-61 collected during stability test in refluxing

methanol. The framework structure of ZIF-61 was unchanged after 7 days. Y-axis of the

graph shows the relative intensity.

114

ZIF-61 Stability Test: H2O at RT

0

5000

10000

15000

20000

25000

2 12 22 32 42

As Prepared

1day

3 day

5 day

7 day

2θ (degree)

Figure S43. PXRD patterns of ZIF-61 collected during stability test in water at room



115

ZIF-61 Stability Test: H2O at 50 °C

0

5000

10000

15000

20000

25000

30000

2 12 22 32 42

As Prepared

1day

3 day

5 day

7 day

2θ (degree)

Figure S44. PXRD patterns of ZIF-61 collected during stability test in water at 50 °C.



116


0

5000

10000

15000

20000

25000

30000

2 12 22 32 422θ (degree)

As Prepared

1day

3 day

5 day

7 day

Figure S45. PXRD patterns of ZIF-61 collected during stability test in refluxing water.

The framework structure of ZIF-61 was unchanged after 7 days. X-axis of the graph


117


0

2000

4000

6000

8000

10000

12000

14000

16000

18000

2 12 22 32 42

As Prepared

1day

3 day

5 day

7 day

2θ (degree)




118


0

2000

4000

6000

8000

10000

2 12 22 32 42

As Prepared

1day

3 day

5 day

7 day

2θ (degree)


The framework structure of ZIF-68 was unchanged after 7 days. X-axis of the graph


119


0

2000

4000

6000

8000

10000

12000

2 12 22 32 422θ (degree)

As Prepared

1day

3 day

5 day

7 day




120


0

2000

4000

6000

8000

10000

12000

14000

16000

18000

2 12 22 32 42

As Prepared

1day

3 day

5 day

7 day

2θ (degree)




121


0

2000

4000

6000

8000

10000

2 12 22 32 42

As Prepared

1day

3 day

5 day

7 day

2θ (degree)




122


0

2000

4000

6000

8000

10000

12000

2 12 22 32 422θ (degree)

As Prepared

1day

3 day

5 day

7 day




123


0

2000

4000

6000

8000

10000

12000

14000

16000

18000

2 12 22 32 42

As Prepared

1day

3 day

5 day

7 day

2θ (degree)




124


0

2000

4000

6000

8000

10000

2 12 22 32 42

As Prepared

1day

3 day

5 day

7 day

2θ (degree)




125


0

2000

4000

6000

8000

10000

12000

2 12 22 32 422θ (degree)

As Prepared

1day

3 day

5 day

7 day




126


0

2000

4000

6000

8000

10000

12000

14000

16000

2 12 22 32 422θ (degree)

As Prepared

1day

3 day

5 day

7 day




127


0

2000

4000

6000

8000

10000

2 12 22 32 422θ (degree)

As Prepared

1day

3 day

5 day

7 day




128


0

2000

4000

6000

8000

10000

12000

14000

16000

2 12 22 32 422θ (degree)

As Prepared

1day

3 day

5 day

7 day




129


0

2000

4000

6000

8000

10000

12000

14000

16000

2 12 22 32 422θ (degree)

As Prepared

1day

3 day

5 day

7 day




130


0

2000

4000

6000

8000

10000

2 12 22 32 422θ (degree)

As Prepared

1day

3 day

5 day

7 day




131

ZIF-69 Stability Test: MeOH at 65°C

0

2000

4000

6000

8000

10000

12000

14000

16000

2 12 22 32 422θ (degree)

As Prepared

1day

3 day

5 day

7 day




132


0

2000

4000

6000

8000

10000

12000

14000

16000

2 12 22 32 422θ (degree)

As Prepared

1day

3 day

5 day

7 day




133


0

2000

4000

6000

8000

10000

2 12 22 32 422θ (degree)

As Prepared

1day

3 day

5 day

7 day




134


0

2000

4000

6000

8000

10000

12000

14000

16000

2 12 22 32 422θ (degree)

As Prepared

1day

3 day

5 day

7 day




135


0

5000

10000

15000

20000

25000

30000

35000

40000

45000

2 12 22 32 42

As Prepared

1day

3 day

5 day

7 day

2θ (degree)




136


0

2000

4000

6000

8000

10000

2 12 22 32 42

As Prepared

1day

3 day

5 day

7 day

2θ (degree)




137


0

2000

4000

6000

8000

10000

2 12 22 32 42

As Prepared

1day

3 day

5 day

7 day

2θ (degree)




138


0

5000

10000

15000

20000

25000

30000

35000

40000

45000

2 12 22 32 42

As Prepared

1day

3 day

5 day

7 day

2θ (degree)




139


0

2000

4000

6000

8000

10000

2 12 22 32 42

As Prepared

1day

3 day

5 day

7 day

2θ (degree)




140


0

2000

4000

6000

8000

10000

2 12 22 32 42

As Prepared

1day

3 day

5 day

7 day

2θ (degree)




141


0

5000

10000

15000

20000

25000

30000

35000

40000

45000

2 12 22 32 42

As Prepared

1day

3 day

5 day

7 day

2θ (degree)




142


0

2000

4000

6000

8000

10000

2 12 22 32 42

As Prepared

1day

3 day

5 day

7 day

2θ (degree)




143


0

2000

4000

6000

8000

10000

2 12 22 32 42

As Prepared

1day

3 day

5 day

7 day

2θ (degree)




144

Section S5. Porosity of ZIFs and carbon dixoide (CO2) selectivity and capacity studies.

All low-pressure gas-sorption experiments (up to 1 atm) were performed on a

Quantachrome Autosorb-1 automatic volumetric instrument. In light of the TGA results

shown in the previous section, ZIF-68, 69, and 70 were evacuated in the following way

prior to gas-sorption analysis. The as-synthesized ZIF samples were immersed in dry

methanol at ambient temperature for 72 h, evacuated at ambient temperature for 24 h,

then at an elevated temperature (85 °C) for 48 h. ZIF samples thus obtained were

optimally evacuated, as evidenced by their well-maintained PXRD patterns and the long

plateau (ambient temperature to 350 °C) in their TGA traces, shown in Figures S30 and

S31 and S32.

The architectural rigidity and consequently the permanent porosity of evacuated

ZIF-68, 69, and 70 were unequivocally proven by gas-sorption analysis. Type I nitrogen

adsorption isotherm behavior was observed for all three of these ZIFs (Fig. 3A and S73),

which reveals its microporous nature. Apparent surface areas of 1,970 m2 g-1 (Langmuir

model, P/P0 = 0.06–0.30, the linearity of fitting 0.9999) and 1,730 m2 g-1 (Brunauer–

Emmett–Teller (BET) model [K. S. Walton, R. Q. Snurr, JACS 129, 8552 (2007)], P/P0 =

0.03–0.06, the linearity of fitting 0.9996) for ZIF-70 were obtained by using the data

points on the adsorption branch. Similarly, apparent surface areas of 1,220 m2 g-1

(Langmuir model, P/P0 = 0.04–0.30, the linearity of fitting 0.9999) and 1,090 m2 g-1

(BET model, P/P0 = 0.01–0.04, the linearity of fitting 0.9999) for ZIF-68, and 1,070 m2

g-1 (Langmuir model, P/P0 = 0.07–0.30, the linearity of fitting 0.9999) and 950 m2 g-1

(BET model, P/P0 = 0.01–0.04, the linearity of fitting 0.9997) for ZIF-69 were obtained

145

by using the data points on the adsorption branch respectively. These surface areas

surpass the highest values reported for zeolites, mesoporous silicas (MCMs, SBAs) and

ZIFs reported so far.

0

50

100

150

200

250

300

350

400

450

500

-0.05 0.05 0.15 0.25 0.35 0.45 0.55 0.65 0.75 0.85 0.95

ZIF-70 N2 AdsorptionZIF-70 N2 DesorptionZIF68 N2 AdsorptionZIF 68 N2 DesorptionZIF69 N2 AdsorptionZIF69 N2 Desorption

Upta

ke c

m3/

g ST

P

P/P0

Figure S73. The N2 gas-sorption isotherms for prototypical ZIF-68, 69, and 70 measured

at 77 K. The filled and open circles represent adsorption and desorption branches,

respectively. STP represents standard temperature and pressure.

146

CO2 and CO isotherms of ZIF-68, 69, and 70 at room temperature (273 K) are

shown below in Figure S74-S76. A nice fitting of each CO2 isotherm adsorption branch

has been calculated based on Toth’s isotherm model (Figure S77-S82).

Toth Model: M = Mmax·B(1/n)·P/(1 + B × P)1/n……………………….(1)

M, gas uptake (mmol/g); Mmax, maximum gas uptake (mmol/g); B and n, fitting

constants; Henry’s law constant K =dNdPP →0

= B1/ n ⋅ Mmax…………………….(2)

Henry’s law selectivity (upper limit selectivity), gas component i over j Sij

Si,j = Ki/Kj……………………………………………………………...(3)

147

Figure S74. The CO2 (red) and CO (black) gas-sorption isotherms for prototypical ZIF-

68 at 273 K. The filled and open circles represent adsorption and desorption branches,

respectively. The CO2 uptake at 760 torr is higher than that of CO suggesting a stronger

interaction between the framework and the CO2 molecules.

148



respectively. The CO2 uptake at 760 torr is eight times higher than that of CO suggesting

a stronger interaction between the framework and the CO2 molecules.

149



respectively. The CO2 uptake at 760 torr is eight times higher than that of CO suggesting

a stronger interaction between the framework and the CO2 molecules.

150

Figure S77. CO2 isotherm adsorption branch of ZIF-68 (red filled circles) and fitting

based on Toth isotherm model (red open circles); Mmax, maximum uptake; B and n, fitting

constants; X2 fitting error; K, Henry’s law constant.

ZIF-68 CO2

B = 0.01607 Mmax = 17.74 n = 0.5462 X2 = 5.809×10-4

K = 7.010 mmol g-1 atm-1

151

Figure S78. CO isotherm adsorption branch of ZIF-68 (red filled circles) and fitting



ZIF-68 CO

B = 0.001541 Mmax = 20.65 n = 0.6069 X2 = 1.717×10-6

K = 0.3648 mmol g-1 atm-1

152




ZIF-69 CO2

B = 0.006563 Mmax = 10.22 n = 0.7113 X2 = 2.579×10-4

K = 6.625 mmol g-1 atm-1

153




ZIF-69 CO

B = 0.001581 Mmax = 17.78 n = 0.6051 X2 = 5.968×10-6

K = 0.3176 mmol g-1 atm-1

154




ZIF-70 CO2

B = 0.08928 Mmax = 525.7 n = 0.2288 X2 = 7.516 ×10-4

K = 10.41 mmol g-1 atm-1

155




ZIF-70 CO

B = 0.003418 Mmax = 8.226 n = 0.5661 X2 = 3.891 ×10-6

K = 0.2751 mmol g-1 atm-1

156

The gas-separation property of ZIF-68, 69, and 70 were tested by breakthrough

experiments (see Figure S86 for schematic representation of the separation apparatus)

using a CO2/CO (about 50:50 v/v) gas mixture. 1.1 g of activated ZIF-68, 1.4 g of

activated ZIF-69 and 1.2 g of activated ZIF-70 were packed into a stainless-steel column

(0.46 inner-diameter and 8 cm length) in a glove box respectively. The columns were

then attached to gas-separation apparatus built as shown in Figure S86. Helium gas was

initially purged into the sample columns. All the experiments were carried out at room

temperature. The gases (20 psi) were premixed in 1:1 ratio and dosed into the column at a

flow rate of 20 mL min−1. The relative amounts of the gases passing through the column

were monitored on a Hiden Analytical HPR20 gas analysis system detecting ion peaks at

m/z+ 44 (CO2) and 12 (CO) (see Figures S83-S85). The relative intensity of each gas

component was normalized to the same level when purging gas mixtures through bypass

before they passed through the column.

157

0 50 100 150 200 250 300 350 400

Time (s)

Rel

ativ

e in

tens

ity

COCO2

Figure S83. Breakthrough curves of CO2 (blue) and CO (black) for ZIF-68 using a

CO2/CO gas mixture. ZIF-68 has not only a permanent porosity, but also shows gas-

separation ability. The relative intensities of each gas passing through the ZIF-68-packed

column were obtained using a mass spectrometer to detect ion peaks at m/z+ = 44 (CO2)

and 12 (CO).

158

0 0.5 1 1.5 2 2.5 3 3.5

Time (m)

Re

lati

ve

in

ten

sity

CO

CO2





and 12 (CO).

159

0 20 40 60 80 100 120 140 160 180 200Time (s)

Rel

ativ

e in

tens

ity

COCO2





and 12 (CO).

160

v1a

v2a

g2ag1a

Gas 1

v1b

v2b

g2bg1b

Gas 2

HIDEN Gas MS

v6

v7

c1

f1

PC

v3 vent

v1c

v2c

g2cg1c

Gas 3

f2 f3

v4 v5

v = valve

f = flowmeter (rotameter)

g = pressure gauge

c = sample cell

Gas Separation/Breakthrough Manifold: Mass Spec Detector

Mass Spec samples continuously from effluent stream; 1-4 data points/minute

Allows for detection of “IR-invisible” materials, e.g. Cl2V6, V7 switch between sample cell c1 and bypass tube to establish baseline signal, calibrate concentration measurement via software

Figure S86. Schematic representation of gas separation/breakthrough apparatus.

supporting online material for - science · 2008. 2. 12. · supporting online material for...

Documents