close

Вход

Забыли?

вход по аккаунту

?

Metal -organic frameworks: Their microwave synthesis and applications as adsorbents for preconcentration

код для вставкиСкачать
METAL-ORGANIC FRAMEWORKS: THEIR MICROWAVE SYNTHESIS AND
APPLICATIONS AS ADSORBENTS FOR PRECONCENTRATION
BY
ZHENG NI
B.Sc., University o f Science and Technology o f China, 1999
M.S., National University o f Singapore, 2001
DISSERTATION
Submitted in partial fulfillment o f the requirements
for the degree o f Doctor o f Philosophy in Chemistry
in the Graduate College o f the
University o f Illinois at Urbana-Champaign, 2007
Urbana, Illinois
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
UMI Number: 3269994
INFORMATION TO USERS
The quality of this reproduction is dependent upon the quality of the copy
submitted. Broken or indistinct print, colored or poor quality illustrations and
photographs, print bleed-through, substandard margins, and improper
alignment can adversely affect reproduction.
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if unauthorized
copyright material had to be removed, a note will indicate the deletion.
®
UMI
UMI Microform 3269994
Copyright 2007 by ProQuest Information and Learning Company.
All rights reserved. This microform edition is protected against
unauthorized copying under Title 17, United States Code.
ProQuest Information and Learning Company
300 North Zeeb Road
P.O. Box 1346
Ann Arbor, Ml 48106-1346
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
C
e r t if ic a t e
of
C
o m m it t e e
A
p p r o v a l
University of Illinois at Urbana-Champaign
Graduate College
February 23, 2007
We hereby recommend that the thesis by:
ZHENG NI
Entitled:
METAL-ORGANIC FRAMEWORKS: THEIR MICROWAVE SYNTHESIS
AND APPLICATIONS AS ADSORBENTS FOR PRECONCENTRATION
Be accepted in partial fulfillment o f the requirements for the degree of:
Doctor of Philosophy
Signatures:
Director o f Research- Richard I. Masel
Head ofD eparm ent - Sieven C. Zimmerman
Committee on Final Examinatioj
Chairperson- Rii
Commil
Committee Member - Mark A. Shannon
[. Masel
r - Kennel
iuslick
Committee Member -
Committee Member -
igory S. Girolami
Committee Member *
Required for doctoral degree but not for master’s degree
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
© 2007 by Zheng Ni. All rights reserved
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
ABSTRACT
Metal-Organic Frameworks (MOFs) have been a focus o f intense activity in
recent years because o f their many unique features, such as extremely high porosities and
tailorable molecule cavities. In this thesis work, I have explored different aspects o f MOF
concerning their syntheses and applications.
In the first part o f this work, a detailed investigation has been carried out to
explore the potential o f using MOFs as adsorbent for trapping and preconcentration on a
portable micro gas detector. Previously people have tried nanotubes, Tenax and different
active carbons as adsorbents for this purpose, but neither their adsorption capacities nor
selectivities could meet the requirements for the micro device. In this study, IRMOF1, a
well-known MOF materials, was tested as an adsorbent for preconcentration for the first
time using dimethly methylphosphonate (DMMP) as a test case. DMMP is a simulant of
nerve agent. We find that DMMP is selectively adsorbed on IRMOF1 and is easily
released upon heating to 250 “C. Concentration gains o f more than 5000 were observed
for DMMP with a 4-s sampling time. Sorption capacities are 0.95 g o f DMMP/g of
IRMOF1. By comparison, dodecane shows a preconcentration gain o f ~5 under similar
conditions. These results demonstrate that MOFs can be quite useful in selective
preconcentrators.
In the second part, we have developed a new method for rapid synthesis o f MOFs,
which we named “microwave-assisted solvothermal synthesis” (MASS). So far most o f
the reported MOF syntheses were either solvothermal or hydrothermal syntheses, which
took from half a day to few weeks. Other effective syntheses such as evaporation and
iii
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
diffusion methods may took even longer time. Here we show that MASS method allows
many well known MOF crystals to be synthesized in under a minute. The properties o f
the crystals made by MASS method are o f the same quality as those produced by the
standard solvothermal method, but the synthesis is much more rapid and resulting MOF
crystal are no longer dependent on the initial nucleus and wall conditions. The
homogeneous effects o f microwave could create a uniform seeding condition, therefore
the size and shape o f the crystals can be well controlled by simply changing a few
reaction conditions.
MASS method provides us a simple and fast approach to quickly build a library o f
other new MOFs. In the third part, we demonstrate the syntheses o f 14 new MOF
materials based on the MASS method. Each has been tested by TGA to explore its
sorption behavior with 4 different vapors. The crystals o f some new MOFs have also
been grown under a similar condition using solvothermal method. The structures for 3 o f
our new MOFs are given based on the single X-ray analysis, and their sorption features
are explained based on the structure information.
IV
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Acknowledgements
First and foremost I would like to express my sincere gratitude to Prof. Richard
Masel. He has been continuously encouraging and guiding me over the five years and
boost my research work into a level I would never imagined before. Some basic attitudes
and methodology that I learned from him will shape the habits o f my future research
work.
I shall never forget my labmates for the great company. Lu Chang, Jason Ganley,
Fred Thomas, Su Ha and Craig Miesse are acknowledged for their generous help during
my early years in the lab. Nickolas Ndiege, Robert Larson, Kun-lun Chu, Adarsh Radadia,
Vaidyanathan Subramanian, Hae-Kwon Jeong and Byunghoon Bae are acknowledged for
their effort on many collaboration projects. Matt Luebbers, Ihwhan Oh, Chelsea Monty,
Zachary Dunbar, John Haan and Tianjiao Wu make the group more and more interesting.
I also want to give my special thanks to three undergraduates who has worked for me to
expedite the discovery o f many new interesting materials. They are Mike Boyd, Ladonna
Adams and Savanah Albert.
I want to thank Professor Mark Shannon, Professor Keith Cadwallader and
Professor Kenneth Suslick for helpful discussions and suggestions during my research
development. They also generously provided me the access to use those expensive
instruments in their lab to collect some important data.
I would also like to mention people in our service shops and many campus
laboratories, who allowed my research went smoothly. Especially Terasa Wieckowska
from X-ray lab, Bill Knight and Rodger Smith from Machineshop, Jerome Baudry from
v
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Computing lab, Jim Wentz and Ben Fisher from electronic shop, Julio Soares from Laser
lab and Vania Petrova from SEM lab.
Last and not least, I am in great dept to my wife, Xiaodong Chen. She has always
been supportive o f my plans, trying to help me in every way with great affection and
intelligence. I would also like to thank my parents Zhenhua Ni and Yiren Tian. Their
encouragement and assistance never take a break.
VI
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Table of Contents
List o f T ables................................................................................................................................. x
List o f Figures...............................................................................................................................xi
Chapter 1 Introduction to Metal Organic Framework (M O F).................................................1
1.1 Basic Concepts o f M O F s....................................................................................................... 1
1.2 Overview o f Solvothermal and Hydrothermal Synthesis................................................10
1.3 Introduction o f MOF A pplications.....................................................................................11
1.4 Introduction o f Preconcentrators......................................................................................... 19
1.5 Scope o f this T hesis............................................................................................................. 21
1.6 References..............................................................................................................................25
Chapter 2 Using MOF as Adsorbents for Trapping and Preconcentration o f Organic
Phosphonates................................................................................................................................ 30
2.1 Introduction............................................................................................................................30
2.2 Experimental......................................................................................................................... 33
2.3 Results and Discussions.......................................................................................................37
2.4 Conclusions........................................................................................................................... 47
2.5 References..............................................................................................................................47
Chapter 3 Rapid Production o f Metal-Organic Frameworks via Microwave-Assisted
Solvothermal Synthesis.............................................................................................................. 49
vii
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
3.1 Introduction............................................................................................................................49
3.2 Experimental......................................................................................................................... 52
3.3 Results and Discussions.......................................................................................................54
3.4 Conclusions............................................................................................................................68
3.5 References..............................................................................................................................70
Chapter 4 New MOFs Discovery and Their Sorption Behavior M easurem ents................72
4.1 Introduction............................................................................................................................72
4.2 Experimental......................................................................................................................... 74
4.3 Results and Discussions.......................................................................................................79
4.4 Conclusions............................................................................................................................92
4.5 References..............................................................................................................................92
Chapter 5 Structure Characterization o f Some New M OFs.................................................. 94
5.1 Introduction............................................................................................................................94
5.2 Experimental......................................................................................................................... 96
5.3 Results and D iscussions....................................................................................................... 98
5.4 Comparison o f water sorption between IRMOF1 and ZnM OF3..................................106
5.5 Conclusions..........................................................................................................................109
5 .6 References............................................................................................................................ 110
Chapter 6 Pack MOFs in a MEMS D evice............................................................................ 111
6.1 Introduction........................................................................................................................I l l
viii
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
6.2 Experimental........................................................................................................................114
6.3 Results and Discussions..................................................................................................... 117
6.4 References............................................................................................................................ 123
Appendix A: SEM Operating Instructions for MOF Powder Observation........................124
Appendix B: MOF Crystal Mounting Procedures for Single X-ray Structure
Analysis.......................................................................................................................................127
Appendix C: Preconcentrator Operating Instructions...........................................................130
Curriculum V ita e .......................................................................................................................134
ix
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
List O f Tables
Table 4.1 Physical characterizations o f selected MOF com pounds.....................................82
Table 4.2 The sorption capacity o f four vapors in new metal-organic compounds........... 90
Table 5.1 Crystal and structure refinement data for Z nM O F l-3..........................................99
Table 5.2. Selected bond lengths
(A) and angles (°) o f Z n M O F l......................................101
Table 5.3. Selected bond lengths
(A) and angles (°) o f ZnM O F2......................................104
Table 5.4 Selected bond lengths
(A) and angles (°) o f ZnM O F3....................................... 106
Table 5.5. A structure comparison o f IRMOF1 and ZnM O F3........................................... 107
x
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
List Of Figures
Figure 1.1 A list o f coordination numbers and their corresponding geom etries.................. 3
Figure 1.2 A m otif design based on octahedral metal node and linear linkers. When the
amount o f metal nodes increased, 3D MOFs m otif should be predominantly obtained....... 5
Figure 1.3 The square Cu 2(carboxylate )4 SBUs and octahedralZri4 0 (carboxylate )6SBUs .
.......................................................................................................................................................... 7
Figure 1.4 Cubic Structures o f IRM OF1-16.............................................................................. 8
Figure 1.5 A trigonal prismatic SBU and chiral channel structure reported by Kim and
co-workers.....................................................................................................................................14
Figure 1.6 [Cu(4 ,4 ’-bipy)2(SiF6)] and comparison o f its methane sorption isotherm with
that o f zeolite 5A ..........................................................................................................................16
Figure 1.7 Molecule structures o f some common CWA and explosive..............................24
Figure 1.8 Molecule structures o f some interferences and simulants for CWA and
explosive that will be discussed in this thesis work................................................................25
Figure 2.1 The typical preconcentration measurement setup for (a) sampling and (b)
thermal desorption...................................................................................................................... 31
Figure 2.2 A preconcentrator design based on Carbotrap adsorbents developed by Zeller ..
........................................................................................................................................................32
Figure 2.3 The structure o f a typical MOF molecule. The spheres are metal clusters, the
rods are organic linkages. In IRMOF1, the spheres are ZruO clusters, while the linkages
are BDC ion (
O
O
C
-
P
h
-
C
O
O
33
Figure 2.4 A picture o f Valcon E rotor packed 5 IRMOF1 crystals in trap slot sitting on a
U.S. penny. This device was used to make the measurements reported in Figure 2 .5 ......35
xi
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Figure 2.5 Schematic diagram o f the experimental procedure used to measure the gain of
the preconcentrator, (a) First a solution contain ppb levels o f DMMP are fed into a
groove in a Valeo gas sampling valve containing MOF crystals (red), (b) Next gas is fed
into an empty tube (blue), (c) Then both grooves are rotated against blank openings and
the valve is heated to desorb the MOF. (d) Finally, the DMMP desorbed from the MOF
is injected into the colum n......................................................................................................... 36
Figure 2.6 XRPD patterns o f (a) IRMOF-1 prepared by the solvothermal synthesis of
Eddaoudi, M. et al (green) and a simulated XRPD curve based on the published structure
o f Eddaoudi, M. (red)..................................................................................................................38
Figure 2.7 A comparison o f the GC chromatograms and MS spectra o f A) DMMP
directly injected into a GC/MS (dotted line) and B) DMMP trapped and thermally
desorbed (solid line). Notice that there are no satellite peaks.............................................. 39
Figure 2.8 A TGA spectrum taken by saturating a sample o f IRMOF1 with DMMP or
toluene and then heating at 10 °C/min to desorb the DMMP or toluene. Notice that a
tremendous amount o f DMMP desorbs almost 1 gram o f DMMP per gram o f adsorbent...
........................................................................................................................................................40
Figure 2.9 A TGA spectrum taken by saturating a sample o f Tenax with DMMP or
toluene vapor and then heating at 10 °C/min to desorb the D M M P..................................41
Figure 2.10 A TGA spectrum taken by saturating a sample o f 20/40 mesh Carbotrap with
DMMP or toluene vapor and then heating at 10 °C/min to desorb the DM M P................42
Figure 2.11 A comparison o f the GC chromatograms produced by exposing the IRMOF1
in a groove in a Valeo sample valve to a gas stream containing 105 ppb o f DMMP
Xll
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
followed by thermal desorption to that from an empty groove containing 105 ppb o f
D M M P..........................................................................................................................................43
Figure 2.12 A comparison o f the GC chromatograms produced by exposing the IRMOF1
in a groove in a Valeo sample valve to a gas stream containing 642 ppb o f DMMP
followed by a thermal desorption (red) to that from an empty groove containing 651 ppm
o f DMMP (black)........................................................................................................................ 44
Figure 2.13 A comparison o f the GC chromatograms produced by exposing the IRMOF1
in a groove in a Valeo sample valve to a gas stream containing 2030 ppb o f dodecane
followed by thermal desorption to that from an empty groove containing 2030 ppb of
dodecane...................................................................................................................................... 45
Figure 2.14 A comparison o f the GC chromatograms produced by exposing the Tenax
TA in a groove in a Valeo sample valve to a gas stream containing 105 ppb o f DMMP
followed by thermal desorption to that from an empty groove containing 105 ppb of
DMMP .........................................................................................................................................46
Figure 3.1 Micro MOF synthesis by stirring assisted solvothermal synthesis................... 51
Figure 3.2 A comparison o f MASS product and conventional solvothermal product o f
IRM OF1........................................................................................................................................52
Figure 3.3 (a) Enlarged SEM image o f micro IRMOF-1 (b) SEM image o f micro
IRMOF-1 (c) SEM image o f micro IRMOF-2 (d) SEM image of micro IRM OF-3......... 55
Figure 3.4 XRPD patterns o f (a) IRMOF1, (b) IRMOF2 and (c) IRMOF3 prepared by the
solvothermal synthesis (black), by MASS synthesis (green), and a simulated curve based
on the published structure o f Eddaoudi, M. (red)....................................................................56
Figure 3.5 SEM image o f MOF38 synthesized by MASS method.......................................58
xiii
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Figure 3.6 XRPD patterns of IRMOF-38 prepared by microwave-assisted solvothermal
synthesis (black), and a simulated XRPD curve based on the published structure o f
Eddaoudi, M. (red)...................................................................................................................... 59
Figure 3.7 Microwave synthesis of IRMOF2 in Zn:L molar ratio o f a) 4:3 b) 6:1 and c)
5:4 at concentration o f [L]=0.02mM. Notice the pure product of IRMOF2 is only
observed when Zn:L=5:4........................................................................................................... 62
Figure 3.8 Microwave synthesis of IRMOF7 (molar ratio o f Zn:L=5:4) in concentration
o f a) [L]=53.8 mM, b) [L]=26.9 mM and c) [L]=0.538 m M ................................................ 63
Figure 3.9 XRPD patterns o f IRMOF-7 prepared by microwave-assisted solvothermal
synthesis (black), and a simulated XRPD curve based on the published structure o f
Eddaoudi, M. (red)...................................................................................................................... 64
Figure 3.10 SEM images o f IRMOF-1 produced in the BDCFL concentration o f a)50.0
mM, b)25.0 mM, c)12.5 mM, d)6.25 mM, e)3.13 mM, f)1.56 mM, g)0.78 mM, h)0.39
mM, i)0.20 mM (concentration o f Zn precursor is reduced correspondently to keep the
M:L ratio constant.).................................................................................................................... 65
Figure 3.11 Estimated average size o f IRMOF-1 micro-crystal versus concentration of
BDCH2 in the reactant solution................................................................................................ 66
Figure 3.12 Enlarged SEM images o f IRMOF-1 synthesized from concentration o f
BDCH2 at (a) 0.05M and (b) 0.0002M.................................................................................... 66
Figure 3.13 Proposed crystal nucleation mechanism for a) conventional solvothermal
synthesis and b) microwave assisted solvothermal synthesis............................................... 69
Figure 4.1 List o f organic linkers that were applied as building blocks in this thesis work..
........................................................................................................................................................73
xiv
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Figure 4.2 SEM images o f 14 new metal-organic materials..................................................80
Figure 4.3 XRPD patterns of 14 new metal-organic materials............................................. 81
Figure 4.4 EDX spectrum o f ZnM OF2.................................................................................... 83
Figure 4.5 IUPAC classification o f adsorption isotherm s.....................................................85
Figure 4.6Contours o f constant local density o f adsorbed Ar molecules for several values
o f the pore loading (Monte Carlo computer simulations for the pore size 18.2x54.6x A3).
N at is the number o f argon molecules adsorbed. These local densities have been averaged
along the direction o f the pore axis and thus show where adsorption is occurring in a
cross-sectional view down the pore axis.................................................................................. 86
Figure 4.7 IR spectrum o f IR M O F1......................................................................................... 88
Figure 4.8 IR spectrum o f ZnM O F3......................................................................................... 88
Figure 4.9 Desorption o f toluene, DMMP, nitrobenzene and water vapor from CoMOF2
........................................................................................................................................................ 89
Figure 4.10 Sorption o f toluene, DMMP, nitrobenzene and water in 14 MOFs and
IRM OF1........................................................................................................................................91
Figure 5.1 Crystals o f some new MOFs grown by traditional solvothermal synthesis
95
Figure 5.2 An ORTEP diagram o f the repeating unit in ZnM OFl. H atoms are not shown.
100
Figure 5.3 The 2D sheet in Z nM O Fl......................................................................................100
Figure 5.4 The 2D sheet in ZnM O Fl, a view down from crystallographic c a x is
101
Figure 5.5 An ORTEP diagram o f the repeating unit in ZnMOF2. H atoms are not shown.
102
Figure 5.6 A cube packing diagram o f Z nM O F2................................................................. 103
xv
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Figure 5.7 An ORTEP diagram o f the repeating unit in ZnMOF3. H atoms are not shown.
...................................................................................................................................................... 104
Figure 5.8 A cube packing diagram o f Z nM O F3................................................................. 105
Figure 5.9 Thermal desorption o f H 2O vapor from IRMOF1 and ZnM OF3.....................108
Figure 5.10 The differential curves o f water desorption TGA curves from IRMOF1 and
Z nM O F3.....................................................................................................................................109
Figure 6.1 One o f the designs for MEMS based preconcentrator with MOF coated inside.
112
Figure 6.2 SEM images of silicalite crystal on glass prepared by reacting 3-chloropropyl
tethered glass plates with silicalite microcrystals for 2 min under the conditions o f
sonication....................................................................................................................................113
Figure 6.3 The model o f anchoring a typical IRMOF1 building unit to a carboxylic acidterminated SAM and optical microscope and a AFM image o f a selectively grown film of
IRMOF1 on a patterned SAM o f 16-mercapto-hexadecanoic acid and 1H,1H, 2H, 2Hperfluorododecane thiol on A u(l 11) for the mother solution at 25 °C .............................. 115
Figure 6.4 A MOF based preconcentrator with powder packed on posted reactor design
118
Figure 6.5 A silicon wafer coated with MOF powders by epoxy....................................... 119
Figure 6.6 A glass substrate coated with MOF powders by CP-TES SAMs.....................120
Figure 6.7 A silicon wafer coated with MOF powders by ultrasonication........................121
Figure 6.8 The dielectrophoresis of CuM OFl on an interdigital Au plate after
a) 0 sec, b) 30 sec, c) lm in and d) 5 m in ............................................................................... 122
Figure A .l The final look o f a MOF crystal mounted in solvent...................................... 129
xvi
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Figure A.2 The schematic diagram o f the purge and trap setup for the preconcentration
measurement.............................................................................................................................. 130
Figure A.3 Schematic diagram of the experimental procedures for measuring the gain o f
the preconcentrator. A) Firstly, a helium gas containing ppb levels o f DMMP are fed into
a groove in a Valeo gas sampling valve containing MOF crystals. B) Secondly, gas is fed
into an empty tube. C) Then both grooves are rotated against blank openings and the
valve is heated to desorb the vapors. D) Finally, the DMMP desorbed from the MOF is
injected into the column............................................................................................................132
xvii
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Chapter 1 Introduction to Metal Organic Framework (MOF)
In recent, remarkable progress has been made in the area o f molecular inorganicorganic hybrid compounds. Among them, the synthesis and characterization o f infinite
ID, 2D and 3D Metal Organic Frameworks (MOFs) has been an area o f rapid growth.
MOFs now have taken an important position in the whole porous materials family and
added many novel features that would advance the traditional zeolite and active carbon
materials. In this chapter, a brief introduction o f M OF’s basics and their developments is
given. Then the synthesis methods for MOFs are covered, with the emphasis on the
method o f solvothermal or hydrothermal, which became dominant approaches in recent
years for building interesting MOF structures. Then the technical applications o f MOF
are focused in the area o f sorption, catalysis, optics and etc. Since no one has report the
applications o f using MOFs as adsorbents in gas chromatography, a brief review of
traditional adsorbents for preconcentration in gas chromatography is given here as the
background for our pioneer research efforts. Finally, we intend to address the scope of
this thesis work and how these research efforts fit in the context o f the development o f
the MOFs and preconcentrators.
1.1 Basic Concepts of MOFs
MOFs are a class o f coordination polymers with backbones constructed from
metal ion and ligand and to form infinite network o f one-, two- and three-dimension in
space .1 The crystal structures in terms o f their framework contain two major components,
‘metal nodes’ and ‘organic linkers’. These are defined as starting reagents with which the
1
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
principal framework o f the coordination polymer is constructed. The important
characteristics o f nodes and linkers are their coordination numbers and coordination
geometries. These features will affect the overall structures o f the framework. MOFs may
also contain auxiliary components when necessary. These auxiliary components can be
blocking ligands, counteranions, and nonbonding guests or template molecules.
Node
Transition metal ion are often applied as versatile nodes in the construction of
coordination polymers. Depending on the metal and its oxidation state, coordination
numbers can range from 2 to 7. Lanthanide ions have even larger coordination numbers
from 7 to 10. These coordination numbers give rise to various coordination geometries.
Common seen geometries reported in previous literature include: linear, T- or Y-shaped,
tetrahedral, square-planar, square-pyramidal, trigonal-bipyramidal, octahedral, trigonalprismatic, pentagonal-bipyramidal, and the corresponding distorted forms (Figure l . l ).2'4
When the metal node is coordinatively unsaturated, the vacant sites could be utilized for
chemical adsorption and heterogeneous catalysis .5,6 Besides transition metals, metal
cluster can also act as a node unit sometimes.
Linker
Organic linkers, on the other hand, create a wide variety o f linking sites with
tuned binding strength and directionality. Linkers normally concerns with bi-functional
ligands. Sometimes multi-functional linkers for MOF also appeared in the literature (such
as TPP and trimesic ligand applied in our work). The shapes o f these ligands will further
2
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
number of
functional sites
connector
linker
<0
Figure 1.1 A
geometries .7
list of coordination numbers
and their corresponding
diversify the geometry o f the whole framework structures. Based on the type o f the atoms
on the binding sites, MOF linkers can generally be classified into three categories. They
are halides, nitrogen linkers and oxygen linkers. Halides (F, Cl, Br and I) are the smallest
and simplest linkers. Some MOF structures bridged by halogen have been extensively
investigated because o f their physical properties .8,9 Nitrogen linkers include ligands with
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
cyano or pyridyl at terminal. These types o f linkers also form strong coordination
bonding with metals as halides, and have been extensively investigated since 1980s.
Some famous examples among them are pyrazine, 4,4’-bipyridyl(bipy) and 4,4'biphenyldicarbonitrile .10-13 Oxygen linkers mainly refer to the carboxylate ligands. This
types o f ligands have been extensively studied in recent years due to their two advantages:
first, they are anionic ligands. Therefore, the resulting frameworks are usually neutral,
which can be isolated from the counter ions and mother liquors and demonstrate
interesting sorption features; second, carboxylic groups facilitate the formation of
secondary building unites (SBUs) which will be introduced in next session. The SBUs
allows more control over the crystal structures, so that desirable and predictable
molecular architectures can be achieved. Some famous examples o f oxygen linkers are
benzenedicarboxylic acid (BDCH 2) and 4,4’-biphenyldicarboxylic acid. (BPDCH 2) .14,15
Design of Motifs
Various combinations o f connectors and linkers mentioned in the previous section
afford various specific structural motifs.
Figure 1.2 illustrates that if a octahedral
coordinated node and linear linkers are given, ID, 2D and 3D motifs can be easily
designed just by changing the molar ratio between them. However the control over the
dimensionality o f the products in realty has rarely been achieved. The difficulties for
researchers to achieve desired m otif are due to the following reasons: first, the final MOF
structures are not only affected by the M/L molar ratios, coordination geometries o f node
and shapes of linkers, but also significantly affected by the experimental conditions such
4
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
ID
Metal centre
*
Figure 1.2 A m otif design based on octahedral metal node and linear linkers.
When the amount o f metal nodes increased, 3D MOFs m otif should be
predominantly obtained .16
as solvent conditions, concentrations, temperatures and even the dust and inner wall
conditions of the container. Second, MOF formation is a process o f self-assembly and
self-recognition o f the linkers and nodes molecules in solution. When a given design and
condition favors more than one coordination possibilities, the resulting structures are
more impossible to be predicted.
Therefore, the answer to the question ‘Are crystal
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
structures predictable?’, recently asked by Dunitz, is still ‘no ’.17 MOF structures cannot
yet be completely predicted a priori from composition alone, a situation that John
Maddox, former editor o f Nature, describes as ‘a continuing scandal’.18
However, a more recently been developed concept, ‘secondary building units’
(SBU), leads to increased structural predictability and control in MOF construction,
which sheds more light on the actual MOF crystallization processes .19 SBU is a concept
borrowed from aluminosilicate zeolite chemistry. It is a primary building units o f the
ligands and metallic ions, whose chemical composition can be easily modified without
losing structural robustness. When a suitable SBU and an appropriate linker are devised,
the reaction under suitable experimental conditions should lead to the targeted
frameworks .20 Figure 1.3 shows examples o f a square SBU and an octahedral SBU.
These two SBUs will be the focus o f my research work in Chapter 5.
The design o f rigid frameworks based on such SBUs was first demonstrated in
MOF-5, 14 also know as IRMOF-1 in the later publications .15 The design principles
employed for IRMOF1 will be discussed in here to show that SBUs have intrinsic
geometric properties that facilitate network design and help us to address the issues of
network synthesis and robustness.
In IRMOF1 (Figure 1.4.1), Zn4 0 (C 02)6 units containing four Z 114O tetrahedral
with a common vertex and six carboxylate C atoms that define an octahedral SBU are
joined together by benzene links. This leads to a cubic network in which the vertices are
the octahedral SBUs and the edges are the benzene struts. In practice, this compound was
prepared from Zn(II) and BDC acid under conditions pre-determined to yield the
octahedral SBU in situ. Because the SBU and benzene links are relatively large and rigid
6
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
entities, the structure produced has exceptional porosity (indicated by its sorption), and
stability (indicated by thermal analysis and single crystal X-ray diffraction studies on the
completely evacuated framework ) .14,19
The exceptional stability o f IRMOF1 can be understood by comparing its basic
network, composed o f single atom vertices, with the actual structure o f IRM OF1, which
has cationic clusters at those vertices. The basic network has no resistance to shear if the
links are considered to be universal joints. However, in the actual IRMOF 1 structure, the
cationic clusters have a truncated tetrahedral envelope, and the rigidly planar O 2C-C 6H 4CO 2 linkers have a planar slat envelope. The linkage o f these two groups produces an
inherently rigid structure held together by mutually perpendicular hinges.
R
O
O
R
Figure 1.3 The
square
Cu 2(carboxylate )4
SBUs
and
octahedral
Zn4 0 (carboxylate )6 SBUs.7
7
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
This approach, based on the concept o f SBUs, has been useful in rationalizing the
topologies o f MOF structures,15 and more importantly, it has allowed the synthesis and
use o f a large number o f inorganic and organic SBUs with varying geometries (Figure 1.4)
In many o f these cases, identifying the reaction conditions that produce an SBU with a
specific geometry in situ means that the addition o f a rigid organic SBU will result in the
formation o f a predetermined network. In other words, with this strategy it is now
possible to control the overall coordination number o f the inorganic and organic SBUs,
and therefore the need to identify the networks that are expected to form from different
geometric shapes becomes particularly acute.
Figure 1.4 Cubic Structures o f IRMOF1-16
8
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Yesterday, Today and Tomorrow
From the point of view in broad sense, MOF m otif design and synthesis, is one
part o f crystal engineering. Crystal engineering is a term coined by Schmidt,
21
that, under
appropriate conditions, the chemical components o f a reactive mixture self-assemble via
a series of molecular recognition events. The pioneering work of Desiraju22 and Etter23 on
organic crystals have demonstrate frameworks assembled via strong hydrogen bonds
(tens o f hundreds of kJ/mol). MOF structures represent a new type o f infinite network
based on the much stronger and highly directional coordinative interactions (several
hundreds o f kJ/mol) between metal centers and organic ligands, thus combining the
properties o f purely organic and inorganic compounds. People now are aiming to design
and synthesis even more robust frameworks based on the covalent bonding (about
thousands o f kJ/mol) and very few successful examples of COFs (covalent organic
frameworks) have recently been reported in 2005.24
Cambridge Crystal Structure Database
The Cambridge Crystal Structure Database offers an invaluable resource about
crystal engineering to crystal engineers. Now available on CD-ROM, it enables all know
organic structures, and all inorganic structures (containing carbon) to be examined for
structural or functional features in few minutes, and to be imported into Cerius2 for
visualization and detailed examination. Apart from the value for re-evaluating know
structures, it enables structures to be readily compared, using Etter’s topological
analysis,25 and many hypotheses to be tested using extant data, without the need for new
structural determinations.
9
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
1.2 Overview of Solvothermal and Hydrothermal Synthesis
This section provides a background for the microwave assisted solvothermal
synthesis (MASS) we invented in this thesis work. In chapter 3 and chapter 4, we will
demonstrate that MASS method can combine the advantages from both the solvothermal
method and microwave method, and can produce the MOF crystals in much shorter time
with high repeatability and high yield.
Solvothermal and hydrothermal methods were borrowed from geological
formation o f natural porous zeolites,
0f\
and were intensely adopted for synthesis o f novel
carboxylate-metal MOFs in recent years. A solvothermal or hydrothermal synthesis
involves the reaction o f precursors in closed vessels above the normal boiling point o f the
water. Typical reaction temperatures range from 80-400 °C. The elevated temperatures
thermodynamically favor the formation o f complex structure o f lower enthalpy, higher
entropy and lower symmetry. The solvothermal/hydrothermal synthesis o f zeolites is
considered to obey Ostwald’s law o f successive reactions in which an initial metastable
kinetic phase transforms to a more thermodynamically stable phase with increased
reaction time. The typical illustrative example is the formation o f Zeolite A; the initial
kinetic phase converts to solid at longer reaction times. At significantly long duration, the
thermodynamically most stable products, silica and alumina, are recovered.27 The
decreased viscosity o f solvent at elevated temperatures corresponds to increased diffusion
rate, which should increase the crystal growth rate eventually.
Since 1990s, most synthesis adopted solvothermal or hydrothermal methods,
which took 'A day to multi-weeks processing time. Some important examples will be
given at the beginning part o f Chapter 3.
10
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
There are two drawbacks o f using these methods: first, synthesis time for
solvothermal can be extended to days and weeks; another difficulty, ubiquitous
throughout solid-state chemistry, is controlling the dimensionality o f the resulting
materials. Dimensionality o f the final structure is highly dependent on solvent,
temperature, wall conditions and even the small particles in solvent. The precise
mechanisms and processes occurring during solvothermal/hydrothermal process are not
well understood. The reactions are complicated, multi-component systems performed in
physical reactors that are not readily amenable to traditional mechanistic probing
experimental methodologies.
Computational studies using molecular modeling have been reported in recent.
Ferey and collaborators have developed a computer-based strategy that, for a given welldefined building block and an organic ligand, generates all possible 3D structure within a
predefined space group.28 This approach, already successful with zeolite-type materials,
enabled them to find the structure o f a highly complex MOF by simply comparing
simulated and experimental powder X-ray diffraction patterns.28 This advance throws
new light onto the difficult task o f rational structure prediction, even though the authors
recognize its many limitations.
1.3 Introduction of MOF Applications
Nano-porous compounds have attracted the attention of chemists, physicists, and
materials scientists because o f interest in the novel phenomena in nano-sized cavities.
There is also commercial interest in their application in separation, storage, and
heterogeneous catalysis. MOFs have been a focus o f intense as novel nanoporous
11
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
materials. Compared with nanotubes, activated carbon and zeolites, MOF have some
unique features due to its structural flexibility.
These unique features include:
1. highest porosity and surface area
2. open structures with almost no inaccessible void space
3. flexible cavities shape and size
4. tunable functionality o f the pore
There are also few disadvantages. The major defect is that inclusion o f the organic
ligand in frameworks cause a relative low thermal stability when compared with zeolite.
Therefore it can not be used for applications requiring high temperature (above 300-500
°C), for example, application as high temperature catalysts. In this section, we will
introduce the applications o f MOFs in some major areas.
Chirality.
Chiral metal-organic frameworks are interesting because they could be applied to
enantioselective separations, especially so because o f the difficulties in preparing
enantiopure traditional zeolite frameworks. The fact that metal-organic frameworks are
prepared in a modular or building block fashion means that chirality can be introduced
simply by choosing chiral building blocks, most conveniently as the organic ligands.
Rosseinsky et al. reported a Ni frameworks linked by trimesic acid ligand, which
contained diols coordinated at the nickel centres.
9Q
The 1,2-ethanediol-containing
network had monodentate diols, and exhibited fourfold interpenetration. Despite this
interpenetration, pseudo-tetrahedral pores o f diameter 12
A
accounted for 28% o f the
12
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
crystal volume, and were interconnected to form continuous channels. The helical nature
o f the networks meant that the crystals were inherently chiral, bulk samples consisting of
50:50 ratios o f enantiomeric crystals. Use o f racemic 1,2-propanediol as the alcohol led
to a closely-related helical structure, but with chelating diols at the nickel centres and
only twofold network interpenetration. Individual crystals o f this material were again
found to grow as single enantiomers, and, interestingly, each crystal contained only one
enantiomer o f the diol. This suggested that the diol co-ligand had exerted a templating
effect on the handedness o f the coordination networks. This clearly pointed to the
possibility of harnessing the chirality o f organic building blocks to produce enantiopure
porous metal-organic frameworks.
Kim and coworkers used an enantiopure ligand derived from D-tartaric acid to
prepare a porous framework on reaction with Zn(N 0 3 ) 2.3° The distorted trigonal prismatic
SBU in Figure 1.5 was adopted, and the structure exhibited chiral channels. Pyridinium
groups pointed into the channels, and these could undergo exchange o f protons for other
cations. The chirality o f the material was manifested in its ability to discriminate to some
degree between enantiomers o f [Ru(2,2’-bipy)3]2+. Immersion o f the framework in a
methanolic solution o f racemic [Ru(2,2’-bipy)3]2+ led to a change in the color o f the
crystals from colorless to yellow, and a 66% enantiomeric excess o f the A form o f the
complex was found to have been included. The material was further used to demonstrate
enantioselective catalysis (see in catalysis section).
13
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Figure 1.5 A trigonal prismatic SBU and chiral channel structure reported by
Kim and co-workers.30
14
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Gas Sorption
Sorption o f gases has been used to investigate the effective porosity and apparent
surface area o f frameworks, and to probe the possibility o f future applications in gas
storage. For example, Yaghi reported N2 sorption by IRMOF1 at 78 K and found a
reversible type I isotherm similar to isotherms observed for most zeolites.14 Strikingly,
Kitagawa reported that Cu(4,4’bipy)2(SiF6) was able to sorb very large quantities o f
methane.31 In fact, relative to framework weight, more methane was adsorbed by this
material than by zeolite 5A, which was taken as the optimum conventional zeolite for
methane sorption (see Figure 1.6).
A great impetus for studying this kind o f sorption is the need for low-pressure
storage media for gases such as methane and hydrogen, due to the potential applications
o f the latter as clean fuels. Subsequently, Yaghi et al. reported still greater weight-forweight methane sorption o f 240 cm3 g_1 in a Zn-carboxylate framework.15 When
compared to a conventional gas cylinder, at a pressure o f 205 atm, the results suggested
that this material could store 70% as much methane in the same volume, but at a lower
(safer) pressure o f 36 atm.
Very surprisingly, adsorbed oxygen molecules were recently investigated within a
metal-organic framework at 90 K.32 The 0 2 molecules were located by X-ray
crystallography and found to exist as ( 0 2)2 van der Waals dimers at an intermolecular
distance o f 3.28(4) A, close to that in solid 0 2, despite the study being conducted at well
above the pure 0 2 freezing point o f 54.4 K. This illustrates the fact that MOF are
providing unusual new microenvironments.
15
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
[{CusiF4(414^bpy]2]n3
Zeolite aft
mmol -gr1
0
§
10
15
20- 2:5 30
P /a lm
35
40
*
Figure 1.6 [Cu(4,4’-bipy)2(SiF6)] and comparison o f its methane sorption
isotherm with that o f zeolite 5A
16
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Ion exchange and solubility
When crystals o f the silver-bipyridyl polymer [Ag(4,4’-bipy)(N 03)]33 were
immersed in aqueous solutions o f anions such as PF6 , MoO4 , BF4- or S 0 4 2- they were
found to undergo anion exchange, i.e. the aqueous anion, which was in excess, was
incorporated into the crystals and the crystal's nitrate anions were released into solution.
For PF6-, the exchange was almost complete after 6 hours, and the crystals had become
opaque, with XRPD giving a new diffraction pattern. Re-exposure to a nitrate-containing
solution gave back transparent crystals with the original XRPD pattern. This process was
later studied by microscopy, including TEM and AFM, and with spectroscopic
examination o f the supernatant.34 The exchange process was concluded to occur via the
solution phase, i.e. there was dissolution o f the original polymer and subsequent
crystallization o f the new anion-exchanged polymer, rather than ion diffusion and
exchange within an inert solid framework (as occurs with aluminosilicate zeolites). Since
coordination polymers are normally obtained as crystals from solution, and good quality
X-ray crystal structures imply that in the solid the chains are very long indeed,
coordination polymers have often been assumed to be effectively insoluble. This is often
supported by visual observation o f crystal stability on immersion in given solvents. This
latest work shows how important it is to examine the supernatant solution. Clearly the
stability o f coordination polymers to leaching will be a critical point with regard to
potential applications in ion-exchange or water/solvent purification.
17
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Nonlinear optical (NLO) properties
Lin et al. exploited the fact that diamondoid networks lack centers o f symmetry to
prepare two materials which exhibited second order nonlinear optical (NLO) behaviour.35
Odd numbers o f interpenetrating diamondoid networks are also non-centrosymmetric,
and the latter is a requirement for second order NLO activity. Frameworks exhibiting 3and 5-fold interpenetration based on Zn(II) or Cd(II) and mixed pyridine-carboxylate
ligands, were prepared, and found in one case to give a second order NLO response of
three times that o f potassium dideuterophosphate (KDP).
Catalysis.
Catalysis is often cited as a desirable characteristic o f metal-organic frameworks,
and a justification for general research into these materials. However, there are still
relatively few reports o f actual catalytic activity, and detailed investigations are
particularly lacking. In 1994 Fujita reported that the channelled Cd framework [Cd(4,4’•3 Z"
bipy) 2(N 0 3 ) 2] would catalyze the cyanosilylation o f aldehydes.
In particular, treatment
o f benzaldehyde and cyanotrimethylsilane with a powdered suspension o f [Cd(4,4’bipy) 2(N 0 3 ) 2]
in
dichloromethane
at
40
°C
over
24
hours
gave
2-
(trimethylsiloxy)phenylacetonitrile in 77% yield. Importantly, these workers had
considered that the reaction might be occurring due to partial dissolution o f the
framework into catalytically active soluble discrete complexes. Accordingly, they tested
the activity o f a supernatant dichloromethane phase and found it gave no activity,
supporting the contention that catalysis was occurring at the solid. Also, no activity was
18
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
observed for the separate framework component precursors Cd(N 03)2 and 4,4’-bipy.
Some selectivity was observed in the catalytic behavior, in that 2-tolualdehyde was a
more active substrate than 3-tolualdehyde, and although a- and P-naphthaldehydc were
good substrates, there was almost no conversion o f 9-anthraldehyde. This selectivity was
taken to be due to the limited cavity size o f the material. The selective inclusion o f orthodihalobenzenes in preference to meta or para isomers was also reported, supporting the
material's selective inclusion characteristics. The cadmium centres are distorted
octahedral with N 4O 2 coordination sets, and, if coordination to Cd is involved in the
catalysis, it implies that there is dissociation o f the terminal nitrate anions to give active
Lewis acid sites. The enantiopure Zn-based framework o f Kim and coworkers30 (see
above) has chiral channels with pyridinium functional groups protruding into the channel
space. Accordingly, these researchers tested for catalytic activity in transesterification
reactions. They observed catalytic activity and even a modest ca. 8% enantiomeric excess
in the product when racemic l-phenyl-2-propanol was used as the solvent. There was also
evidence that the pyridinium groups o f the channels could be alkylated with, for example
methyl iodide, and this protected nonacidic framework showed little or no activity.
1,4 Introduction of Preconcentrators
A preconcentrator is a device to collect, concentrate, and delivers analyte sampled
from air for analysis with a detector. In this section we will introduce some basic facts in
gas chromatograph and some traditional adsorbents for preconcentration.
19
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Preconcentrator and Purge and Trap Thermal Desorption
A gas or vapor detector is normally operated by pneumatically sampling ambient
air and analyte mixture into a chemical detector. For a detection system that has
insufficient sensitivity to detect the analyte at trace concentrations level, an optional
preconcentrator device can be used to augment the performance o f the detector. The gas
or vapor preconcentrator serves the function o f collecting and concentrating analyte over
a period of sampling time, followed by a thermal release o f the analyte as a concentrated
wave into the detector.37'43 The advantages o f the detector in incorporating a
preconcentrator device are:
enhanced sensitivity and improved selectivity. The
preconcentrator is a generic frontend modification to analytical systems and can be used
with a variety o f analytical
systems, including gas
chromatographs,44’46 mass
spectrometers,47 ion mobility spectrometers (IMS), and MEMS-based chemical sensors.48
Desirable features o f the preconcentrator device include the capability o f operation at
high flow rates, thermal heating with short-time constants, and selective collection for the
analyte(s) o f interest.
Preconcentration Adsorbents
An ideal preconception material should meet at least two requirements: during
adsorption, the adsorbent materials should have significant adsorption capacity to
accumulate large amount of analyte molecules inside preconcentrator chamber; during
thermal desorption, the analyte molecules should be desorbed intact from the adsorbent.
That is, the heat applied to a sample should be enough to volatilize the organic
compounds without degrading them and without producing unwanted artifacts from the
adsorbent itself.
20
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
In 1970, a novel thermally stable porous polymer named Tenax-GC (poly[2,6diphenyl-p-phenylene oxide]) was first introduced for use as a gas chromatographic
stationary phase.49 This material has a suitable adsorptive capacity towards volatile and
semivolatile organic compounds and a low affinity for water vapor. Exploiting these
features, researchers soon began to apply Tenax as a preconcentration adsorbents for
isolating and concentrating volatile organics in food, environmental, and biochemical
samples.50"52 Until mid
1990s,
many
other types
o f adsorbent materials
for
preconcentration have been investigated.
To summarize, these porous materials include: polymer (Tenax, polystyrene, or
polyurethane foams); carbon (graphitized carbon black, charcoal, or carbon sieves); silica
gel and alumina. Polymeric materials sometimes must be cleaned before used to remove
residual monomers. For example, Tenax is especially useful for preconcentration because
it is hydrophobic, does not retain water, but has a relative low sorption capacity. Carbon
materials have an excellent open porosity, but they have a disordered structure and
sometimes irreversibly adsorb certain classes o f analytes. Silica gel and alumina have
high capacity but take up water vapor.
More recently developed adsorbents especially for MEMS based detectors
will be introduced at the beginning o f Chapter 2.
1.5 Scope of this Thesis
Basically, my thesis work includes two parts: The first part is about using new
adsorbents, MOFs, to solve an old problem, insufficient preconcentration in GC detector.
21
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Second part is about using an existent technology, microwave, to solve an old problem,
slow and unstable synthesis of MOF.
First Problem
Preconcentrators are a key component in many of the systems now being
deployed for homeland defense.
The preconcentrator collects target molecules and
presents them to a spectrometer for analysis.
The Department o f Homeland Security
(DHS) has placed refrigerator sized preconcentrators in every subway station in
Washington DC and New York.
They would like to put a preconcentrator in every
subway turnstile, bus and subway car, but cannot at present because the preconcentrators
are too big.
The walk in sniffers used in airports use a smaller preconcentrator, (only 2
ft3), but then the preconcentrators need to be replaced on a regular basis.
A key reason for the difficulty is that purge-and-trap preconcentrators is old
technology.
People are using the same materials to trap molecules that they were using
50 years ago. The common adsorbents only adsorb 1% or so o f their weight in target
molecules.
Further, the adsorbents are not selective.
They adsorb cigarette smoke,
diesel fumes, and many other molecules, and these other molecules can displace the
target molecules from the surface o f the adsorbent.
People get around that problem by
putting a huge amount o f adsorbent in the preconcentrator, so there is room to trap
everything.
That makes the preconcentrator quite large. In our research work, we
demonstrate, for the first time, that MOFs are selective for the molecules o f interest
(explosives, chemical weapons agents) and have much higher adsorption capacity than
any molecule that had been used before. (Figure 1.7 and Figure 1.8)
22
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Second Problem
Another part o f my work related to the improvement for MOF synthesis. Previous
MOF growth methods, such as evaporation, diffusion and solvothermal, will all take few
hours to few months for the crystallization process. One would get a milligram or two per
day. Such situations not only slow down the paces o f the MOF research, but also made
MOFs too expensive for many real industrial practices. We developed a much better
synthesis approach, microwave assisted solvothermal synthesis (MASS).
A few days
growth period now can be reduced to less than one minute. Particle sizes were much
more uniform than with previous methods. Yields increased from milligrams per day to
grams per day.
Arrangement of Chapters
My research efforts in this thesis are arranged as follows: chapter 2 introduced the
work o f using IRMOF 1 as adsorbent for preconcentrator, its performance was described
and compared with traditional adsorbents; chapter 3 illustrated the major advantages of
microwave assisted approach for MOF syntheses, several well-known MOF structures
were reproduced by microwave and verified based on the XRPD comparison; In chapter
4 we focused on the synthesis o f a library o f new MOFs prepared by MASS methods,
some basic physical properties such as porosities and sorption behaviors were measured
and compared; in chapter 5 we used single X-ray technique and determined 3 structures
o f our new MOFs, an interesting sorption behavior for water vapor was compared and
was discussed on the structure basis. In chapter 6 we showed some efforts to incorporate
23
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
MOF powders into the MEMS based microGC device. Some preliminary results were
provided for a future investigation.
Common CWA and explosives:
HaC
)+ -
D
h/ *
— M—
o— B
cHjCHi
I
H*C
0 A (tubus)
OB(win)
PH,
H.C-
GD(aomao)
r
" i
-CHa
KjC-
CH,
K>
GF (cjdobwjl win)
r
vx
TNT (brinilretohuna)
Figure 1.7 Molecule structures o f some common CWA and explosive.
24
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Simulants
CHi
H £-
HCjH
MMP(dffUlfc|lDMbrl|fcioqtaaHl0 P M P tlf lu o ly lw lt|lp to f h o a l4
Hftrabi
Interferences:
rHi
blm
Figure 1.8 Molecule structures o f some interferences and simulants for CWA
and explosive that will be discussed in this thesis work.
1.6 References
1.
Stuart, L. J. Chem. Soc. Rev., 2003, 32, 276-288.
2.
Khlobystov A. N.; Blake, A. J.; Champness, N. R.; Lemenovskii, D. A.; Majouga,
A. G.; Zyk, N. V.; Schroder, M. Coord. Chem. Rev. 2001, 222, 155-192.
3.
Munakata, M. Adv. Inorg. Chem. 1998, 46, 173-303.
4.
Kitagawa, S.; Munakata, M. Trends Inorg. Chem. 1993, 3, 437-462.
25
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
5.
Pan, L.; Adams, K. M.; Hernandez, H. E.; Wang, X.; Zheng, C.; Hattori, Y.;
Kaneko, K. J. Am. Chem. Soc. 2 0 0 3 ,125, 3062-3067
6.
Reineke, T. M.; Eddaoudi, M.; Fehr, M.; Kelley, D.; Yaghi, O. M. J. Am. Chem.
Soc. 1 9 9 9 ,121, 1651-1657
7.
Kitagawa, S.; Kitaura, R.; Noro, S. Angew. Chem.Int. Ed. 2004, 43, 2334-2375.
8.
Okamoto, H.; Yamashita, M. Bull. Chem. Soc. Jpn. 1998, 71, 2023-2039.
9.
Clark, R. J. H. Chem. Soc. Rev. 1 9 9 0 ,19, 107-131.
10.
Hagrman, P. J.; Hagrman, D.; Zubieta, J. Angew. Chem. 1999, 111, 2798-2848;
Angew. Chem. Int. Ed. 1999, 38, 2638-2684.
11.
Kitagawa, S.; Kondo, M. Bull. Chem. Soc. Jpn. 1998, 71, 1739-1753.
12.
Moulton, B.; Zaworotko, M. J. Chem. Rev. 2 0 0 1 ,101, 1629-1658.
13.
Zaworotko, M. J. Chem. Commun. 2001, 1-9.
14.
Li, H.; Eddaoudi, M.; O'Keeffe, M.; Yaghi, O. M. Nature 1999, 402, 276-279.
15.
Eddaoudi, M.; Kim, J.; Rosi, N.; Vodak, D.; Wachter, J.; O'Keeffe, M.; Yaghi, O.
M. Science 2002, 295, 469-472.
16.
Paz, F. A.; Klinowski, J. Chemistry & Industry 2006, 21-23
17.
Dunitz , J. Chem. Common 2003, 545
18.
Maddox, J., Nature 1998, 335, 201
19.
Eddaoudi, M.; Moler, D. B.; Li, H.; Chen, B.; Reineke, T. M.; O'Keeffe, M.;
Yaghi, O. M. Acc. Chem. Res. 2001, 34, 319-330.
20.
Yaghi, O. M.; O'Keeffe, M.; Ockwig, N. W.; Chae, H. K.; Eddaoudi, M.; Kim, J.
Nature 2003, 423, 705-714
21.
Schmidt, G. J. Pure. Appl. Chem. 1971, 27, 847
26
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
22.
Desiraju, G. in “Crystal engineering: the design o f organic solids”, Amsterdam:
Elsevier, 1989
23.
Etter, M. J. Phys. Chem. 1991, 95, 4601
24.
Cote, A. P.; Benin, A. I.; Ockwig, N. W.; O ’Keeffe, M.; Matzger, A. J.; Yaghi, O.
M. Science, 2005, 310, 1166-1170.
25.
Etter, M. C. Ace. Chem. Res. 1990, 23, 120.
26.
Barrer FRS, R. M. in “Hydrothermal synthesis o f zeolites”, Academic Press: New
York, 1982
27.
Francis, R. J.; O ’Hare, D. J. Chem. Soc., Dalton Trans. 1998, 3133-3148
28.
Ferey, G.; Mellot-Draznieks, C.; Serre, C.; Millange, F. Acc. Chem. Res. 2005, 38,
217
29.
Kepert, C. J.; Prior T. J.; Rosseinsky, M. J. J. Am. Chem. Soc., 2 0 0 0 , 122, 5158
30.
Seo, J. S.; Whang, D.; Lee, H.; Jun, S. I.; Oh, J.; Jeon Y. J.; Kim, K. Nature, 2000,
404, 982
31.
Noro, S.-I.; Kitagawa, S.; Kondo M., Seki, K. Angew. Chem. Int. Ed., 2000, 39,
2081
32.
Kitaura, R.; Kitagawa, S.; Kubota, Y.; Kobayashi, T. C.; Kindo, K.; Mita, Y.;
Matsuo, A.; Kobayashi, M.; Chang, H. C.; Ozawa, T. C.; Suzuki, M.; Sakata M.;
Takata, M. Science, 2002, 298, 2358
33.
Yaghi, O. M.; Li, H. L. J. Am. Chem. Soc., 1 9 9 6 , 118, 295
34.
Khlobystov, A. N.; Champness, N. R.; Roberts, C. J.; Tendler, S. J. B.; Thompson,
C.; Schroder, M. CrystEngComm, 2002, 4, 426
27
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
35.
Evans, O. R.; Xiong, R.-G.; Wang, Z. Y.; Wong G. K.; Lin, W. Angew. Chem.
Int. Ed., 1999, 38, 536
36.
Fujita, M.; Kwon, Y. J.; Washizu S.; Ogura, K. J. Am. Chem. Soc., 1 9 9 4 ,116,
1151
37.
Matzke, C. M.; Kottenstette, R. J.; Casalnuovo, S. A.; Frye-Mason, G. C.; Hudson,
M. L.; Sasaki, D. Y.; Manginell, R. P.; Wong C. C. Proc. SPIE, 1998, 3511, 262268.
38.
Mowry, C.; Morgan, C.; Baca, Q.; Manginell, R.; Kottenstette, R.; Lewis, P.;
Frye-Mason, G. Proc. SPIE, 2002,4 5 75, 83-90.
39.
Lu, C.-J.; Whiting, J.; Sacks, R. D.; Zellers, E. T. Anal. Chem., 2003 , 75, 14001409.
40.
Potkay, J. A.; Driscoll, J. A.; Agah, M.; Sacks, R. D.; Wise, K. D. Proc. IEEE
MEMS, 2003, 395-398.
41.
Correia, J. H.; de Graaf, G.; Kong, S. H.; Bartek, M.; Wolffenbuttel, R. F. Sens.
Actuators A, Phys., 2000, 82, 191-197.
42.
Geschke, O.; Klank, H.; Telleman, P. in “Microsystem Engineering o f Lab-on-aChip Devices”. Hoboken, NJ: Wiley, 2004.
43.
Butler, M. A.; Frye-Mason, G. C.; Osbourn, G. C. Proc. Space, 7th Int. Conf.
Expo. Eng., Construction, Operations and Bus. Space, 2000, 476—481.
44.
Manginell R. P.; Frye-Mason, G. C. U.S. Patent 6 1 71 378, 2001.
45.
Tian, W.-C.; Pang, S. W.; Lu, C.-J.; Zellers, E. T. J. Microelectromech. Syst.,
2 0 0 3 ,12, 264-272.
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
46.
Zellers, E. T.; Steinecker, W. H.; Lambertus, G. R.; Agah, M.; Lu, C.-J.; Chan, H.
K. L.; Potkay, J. A.; Obomy, M. C.; Nichols, J. M.; Astle, A.; Kim, H. S.; Rowe,
M. P.; Kim, J.; da Silva, L. W.; Zheng, J.; Whiting, J. J. Proc. Solid-State Sens.,
Actuator and Microsyst. Workshop, Hilton Head, SC, 2004, 61-66.
47.
S. V. Krishnaswamy and C. B. Freidhoff, U.S. Patent 5 4 8 1 1 1 0 , 1996.
48.
Tian, W. -C .; Pang, S. W. J. Vac. Sci. Technol. B, 2003, 21, 274.
49.
Van Wijk, R. J. Chromatogr. Sci., 1970, 8, 418-420.
50.
Zlakis, A.; Lichtenstein, H. A.; Tishbee, A. Chromatographia, 1973, 6, 67-72.
51.
Bellar, T. A.; Lichtenberg, J. J., J. Am. Water Works Assoc., 1974, 66, 739-744.
52.
Bertsch W.; Chang, R. C.; Zlatkis, J. J. Chromatogr. Sci., 1 9 7 4 , 12, 175-182.
29
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Chapter 2 Using MOF as Adsorbents for Trapping and
Preconcentration of Organic Phosphonates
2.1 Introduction
Traditional bench-top gas chromatography (GC) system can not meet increasing
requirements for real time analysis and rapid determination o f trace contaminations in the
air. Therefore, the development of portable micro gas chromatography system (//GC) has
been recently focused due to their wild applications in biomedical surveillance,
environmental monitoring, food inspection, industrial emission mapping, explosive and
chemical warfare agents (CWA) detection and etc. In most o f these applications a
preconcentrator in GC system is required to improve the resolution o f the GC column or
to enhance the sensitivity of the detector in the vicinity where the concentrations of
analytes can be as low as parts per billion (ppb) ranges.
The typical preconception measurement setup is shown in Figure 2 .1.1 The
operations begin with an accumulation step o f low-concentration organic vapor during
sample air flows through the preconcentrator at ambient condition, followed by a thermal
desorption step to release analytes into a small volume thereby increasing the effective
concentration to increase sensitivity o f detector, and finally an injection step to create a
sharp peak to facilitate separations in GC columns.
The pChemLab® developed at Sandia National Laboratories incorporates a micro
preconcentrator design with an adsorbent film coated on a heated membrane. Mitra and
his coworkers developed a similar design by spin coating a thin film o f commercially
available gas chromatography stationary phase,2 OV17 (50% polymethyl-50% phenylphase) as their adsorbent layer. However, these designs need to conquer the delamination
30
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
trouble during each thermal recycle, and they can not provide a substantial large
adsorption capacity. Zeller and his coworkers have developed another preconcentrator
design by packing Tenax polymers or a series o f carbon-based materials within (Figure
2.2).1,3-5 The porous carbon-based adsorbents have surface area from 100-1200 m2/g,
which make these preconcentrators reach >5000 gain o f many organic vapors at 100 ppb
concentration. But their responses are based on a 10 min sampling time which is difficult
to meet a real time analysis in most applications, and these adsorbents do not distinguish
specific target from a mixture o f interferences.
PRfLCONCENTRATOR-FOCUSER
PRECONCEKTRATOR-FOCUSER
CARRIER
GAS
NEEDLE
VALVE
CARRIER
GAS
NEEDLE
VALVE
COLUMN
COLUMN
0
U 'T ™
PUMP
0* * - ™
INLET
VALVE COMMERCIAL GC
HP-GC 6890
SIX-WAY
VALVE
COMMERCIAL GC
HP-GC S8«0
(a)
(b>
Figure 2.1 The typical preconcentration measurement setup for (a) sampling and
(b) thermal desorption.
The objective o f this chapter is to evaluate the potential o f metal-organic
frameworks (MOF) as novel adsorbents for preconcentration o f organic phosphonates.
MOFs are a new class o f materials, where metal centers and bridging organic ligands are
arranged in a three-dimensional skeleton as indicated in Figure 2.3. They have surface
areas ranging from 1000 to 5400 m2/g, tailorable polarity and pore size, and high thermal
31
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
6-8
stability. '
Applications discussed in the literature include catalysts for alkynes
conversion,9 gases separation,9 and the storage o f hydrogen,10 methane,11 and ammonia.
We have been interested in these molecules because the can be tailored to selectively
adsorb molecules o f interest. Thus, they could be quite useful in the preconcentrators
used for the detection o f trace impurities; however, at this point, there are no published
reports of using MOFs in preconcentrators.
Figure 2.2 A preconcentrator design based on Carbotrap adsorbents developed
by Zeller.
In this chapter, we explored the M OF’s performance as an adsorbent for
dimethylmethyl phosphonate (DMMP) preconcentration. DMMP is a molecule that is
commonly used as a stimulant o f never agents. It has polarity and volatility similar to that
of sarin, but is much safer to use. We are using a MOF developed by Yaghi et. Al.,
IRMOF1.12 IRMOF1 has a high porosity at 2900 m2/g and linking groups that have a
polarity similar to those o f the phosphonates. Thus, it is a reasonable candidate as an
adsorbent for phosphonates.
32
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
We find that IRM 0F1 is an excellent selective adsorbent for phosphonates.
We observe preconcentration gains over 5000 for DMMP with sampling times o f 4 s. By
comparison, if we instead use Tenax TA as adsorbent, the gain is only 2. Furthermore,
IRMOF1 is selective: dodecane shows a gain o f only ~5. Thus, it appears that MOFs are
excellent candidates as adsorbents for preconcentrators.
Figure 2.3 The structure o f a typical MOF molecule.
clusters, the rods are organic linkages.
The spheres are metal
In IRMOF1, the spheres are ZruO
clusters, while the linkages are BDC ion ( OOC-Ph-COO ).
2.2 Experimental
IRMOF1 crystals were synthesized according to the reported procedures.11 A
0.200
g
aliquot
of
Zn(N 0 3 ) 2' 6 H 20
(0.672
mmol)
and
0.084
g
of
1,4-
benzenedicarboxylate acid ligand (0.50 mmol) were dissolved in 20 mL o f N .W diethylformamide. The solution was stirred for 10 min and then sealed in a closed vessel.
The vessel was heated at 90 °C for 20 h. Small light yellow crystals were collected after
the solvothermal treatment (yield in 40%). X-ray powder diffraction (XRPD) patterns
33
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
were collected on Bruker General Area Detector Diffraction System. Solvated crystals
used for XRPD measurements were transferred along with their mother liquor in to a 0.7mm capillary tube.
Surface Measurement was performed on a PulseChemiSorb 2705 (Micrometritics)
BET machine. IRMOF1 crystals were ground to 325 mesh and then degassed under
vacuum at 150 °C for 1 day before the measurement.
Breakdown Test was performed to find out whether DMMP can be adsorbed on
IRMOF1 and whether it can be desorbed without decomposition. IRMOF1 crystals were
first treated under vacuum at 150 °C for 1 day. Then 5 mg of the IRMOF1 was packed
into a 3-mm glass GC liner with both ends blocked by glass wools. The capillary was
then installed in a programmable temperature vaporizer injector (CIS4, Gerstel, Germany)
in an Agilent 5973N GC/MS system. A 10-mL sample o f DMMP vapor and air mixture
was injected through the sample inlet at 50 °C. The sample is kept at 50 °C for 10 min and
then heated to 250 °C at 10 °C/s to release the vapor into the GC analytical column
(Rtx5ms, 20 m x 0.25 mm i.d. x 0.5 junl film, Restek). A parallel measurement was also
carried out with an empty glass GC liner capped by glass wools on both ends.
Next we measured the sorption capacity o f IRMOF1 by saturating IRMOF1 with
DMMP and toluene vapors, respectively. We then loaded the samples into a
thermogravametric analyzer and measured the weight loss during thermal desorption. In
details, IRMOF1 crystals were ground to 325 mesh. The solvates were completely
removed by heating the solid at 150 °C under vacuum. A 5-mg sample o f IRMOF1 was
then packed into a 3-mm glass GC liner as descrbed earlier. Organic solvent vapor was
mixed by 15 seem He gas through a saturator and then passed through the sample
34
R ep ro d u ced with p erm ission o f the copyright ow ner. Further reproduction prohibited w ithout perm ission.
capillary at 50 °C. After 10 /j L o f solvent loss was observed from saturator, the samples
were removed from the system. Thermal gravimetric analyses were then carried out on a
Perkin— Elmer TGA7 instrument heating from 25 to 300 °C at 10 °C/min.
Next, a real-time preconcentration measurement was carried out using a novel
purge-and-trap system. A few cubic shaped crystals (size - 1 0 0 //m) or fine MOF
powders were mounted in one o f the four 0.06-juL grooves on a standard Valcon E rotor
from Valeo Instruments.(Figure 2.4) The rotor was then installed inside the injection
valve and controlled by a multiposition electronic actuator. Error! Reference source not
found, shows the schematic configuration used for the performance test. The organic
vapor was introduced through a bubbler in an Isotemp water bath and diluted to the
appropriate concentration. This sample gas was then connected to the sampling inlet port
of the valve. A flame ionization detector (FID) was connected to the injection outlet port
for data collection. Helium gas was applied as both the carrier gas and injection gas at
pressures o f 20 and 10 psi, respectively.
Figure 2.4 A picture o f Valcon E rotor packed 5 IRMOF1 crystals in trap slot
sitting on a U.S. penny. This device was used to make the measurements
reported in Figure 2.5.
35
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Injection Valve
Injection Valve
e a rn e r
C arrier
Gas
Injection
Gas
Injection
G as
(b)
(a)
Injection Valve
C arrier
Gas
0<§)i
a
Injection Valve
V ent
C arrier
Gas
FID
Injection
Gas
(c)
(d)
Figure 2.5 Schematic o f the experimental procedure used to measure the gain o f
the preconcentrator, (a) First a solution contain ppb levels o f DMMP are fed
into a groove in a Valeo gas sampling valve containing MOF crystals (red), (b)
Next gas is fed into an empty tube (blue), (c) Then both grooves are rotated
against blank openings and the valve is heated to desorb the MOF. (d) Finally,
the DMMP desorbed from the MOF is injected into the column.
36
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
In a typical measurement, the MOF adsorbent trap was treated by heating at 250
°C while purging with pure helium gas (Error! Reference source not found.d). The
adsorbents were confirmed to be cleaned when the FID signal dropped to the baseline.
The trap was then actuated to a sampling position as shown in Error! Reference source
not found.a. A gas sample containing DMMP vapor at a few hundred ppb level passed
through the trap at 30 °C for 4-60 s. Next, the trap was actuated to a sealed position;
meanwhile, an empty reference groove underwent the same sampling procedure as that o f
the MOF trap (Error! Reference source not found.b). Then both trap and reference
were sealed in the valve system and heated to 250 °C for thermal desorption (Error!
Reference source not found.c), followed by trap injection (Error! Reference source
not found.d) and reference injection (Error! Reference source not found.a),
respectively.
2.3 Results and Discussions
Figure 2.6 compares the XRPD patterns o f IRMOF1 prepared by the solvothermal
synthesis and a simulated curve based on the published structure o f Eddauoudi, M .12 The
collected XRPD pattern compared well with a simulated curve based on the previously
reported IRMOF1 structure. A further BET measurement shows our crystals had a
surface area o f 650 m 2/g.
37
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Figure 2.6 XRPD patterns o f (a) IRMOF-1 prepared by the solvothermal
synthesis o f Eddaoudi, M. et al (green) and a simulated XRPD curve based on
the published structure of Eddaoudi, M. (red ) . 12
Figure 2.6 compares the XRPD patterns of IRMOF1 prepared by the solvothermal
synthesis and a simulated curve based on the published structure o f Eddauoudi, M .12 The
collected XRPD pattern compared well with a simulated curve based on the previously
reported IRMOF1 structure. A further BET measurement shows our crystals had a
surface area o f 650 m 2/g.
38
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
m
1®
m fr
m /x
Figure 2.7 A comparison of the GC chromatograms and MS spectra o f A) DMMP
directly injected into a GC/MS (dotted line) and B) DMMP trapped and thermally
desorbed (solid line). Notice that there are no satellite peaks.
Figure 2.7 shows the result o f breakdown test. Desorption peaks from IRMOF1
and the empty capillaries are compared in the figure. A sharp DMMP peak comes out at
-17.50 min for both capillaries. No “breakdown” peaks were observed during the entire
measurement. This result shows that DMMP vapor can be effective captured by IRMOF1
at 50 °C, and there is no reaction during the adsorption/desorption process.
39
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
£
CD
2
D M M P loss
15
Toluene loss
ce
%
05
■e
h
20
70
170
120
220
270
T e m p e r a t u r e ( C)
Figure 2.8
A TGA spectrum taken by saturating a sample o f IRMOF1 with
DMMP or toluene and then heating at
10 °C/min to desorb the DMMP or
toluene. Notice that a tremendous amount o f DMMP desorbs - almost 1 gram o f
DMMP per gram o f adsorbent.
Figure 2.8 shows the TGA results. Notice that lg o f IRMOF1 can adsorb 0.95 g
o f DMMP while it only adsorbs -0.1 g o f toluene. The desorption o f toluene does not
show any clear features, suggesting that the toluene is only weakly bound. In contrast, the
TGA spectrum o f DMMP shows an inflection point at -140 °C. Differentiation o f the
data and using an analysis as outlined in M asel 13 indicates that there is a single binding
site for DMMP in IRMOF1 with a binding energy o f -1 9 kcal/mol.
40
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
102
100
DMMP
- to lu e n e
135
185
235
285
T e m p e ra tu re (C|
Figure 2.9 A TGA spectrum taken by saturating a sample of Tenax with DMMP
or toluene vapor and then heating at 10 °C/min to desorb the DMMP.
We also made parallel experiments using 60/80mesh Tenax TA and 20/40 mesh
Carbotrap as adsorbents shown in Figure 2.9 and Figure 2.9. We find that much less
DMMP was adsorbed in Tenax and Carbotrap than in IRMOF1. When adsorbents were
saturated, only 0.013 g o f DMMP / g o f adsorbents were observed from Tenax TA, and
only 0.02 g o f DMMP / g o f adsorbents were observed from Carbotrap. This compares to
0.95 g DMMP / g o f adsorbents for IRMOF1. Clearly, IRMOF1 has much higher
adsorption capacity for DMMP than Tenax or Carbotrap.
41
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
10 2
100
— toluene
DMMP
85
■|
35
285
T em p era tu r* (C)
Figure 2.10
A TGA spectrum taken by saturating a sample o f 20/40 mesh
Carbotrap with DMMP or toluene vapor and then heating at
10 °C/min to
desorb the DMMP.
Figure 2.11 shows the results o f an experiment where 30 seem o f 107 ppb DMMP
flowed into the trap for 1 min, the trap was then sealed and heated to 250 °C, and then
desorbed DMMP was injected into an Agilent 5973N GC/MS. A second pulse created by
following the same procedure with an empty slot is also shown. Notice, there is obviously
a large preconcentration gain although quantification is not possible, since we are outside
o f the linear response region o f the detector.
42
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
3 5E+05
Groove with MOF
2 0E-HJ5
1 5E+05
1 OE 105
Empty Groove
0 0E+00
100
Figure 2.11
150
250
300
350
400
A comparison o f the GC chromatograms produced by
exposing the IRMOF1 in a groove in a Valeo sample valve to a gas stream
containing 105 ppb o f DMMP followed by thermal desorption to that from an
empty groove containing 105 ppb o f DMMP.
In order to quantify our results, we compared the GC spectrum o f 642 ppb
DMMP with a trap and DMMP vapor at 651 ppm with no trap. A sampling time o f 4 s
was used for both experiments, and the sample gas flow rate was 150 seem for each case.
The MOF was ground to 325 mesh so the desorbed peak shows much less tailing effect
than that in Figure 2.11. Figure 2.12 compares the peaks seen in both samples. Notice
that the GC peak measured by adsorbing 642 ppb DMMP in IRMOF1 for 4 s and then
43
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
desorbing is ~5 times larger than from the empty groove measured with 651 ppm DMMP.
Therefore, we conclude that preconcentration gains o f over 5000 are possible using only
5 jug o f IRMOF1 and a 4-s sample time.
120000
Groove with MOF
100000
60000
40000
Empty Groove
20000
Time (Second)
Figure 2.12
A comparison o f the GC chromatograms produced by
exposing the IRMOF1 in a groove in a Valeo sample valve to a gas stream
containing 642 ppb o f DMMP followed by thermal desorption (red) to that from
an empty groove containing 651 ppm o f DMMP (black).
In order to quantify our results, we compared the GC spectrum o f 642 ppb
DMMP with a trap and DMMP vapor at 651 ppm with no trap. A sampling time o f 4 s
was used for both experiments, and the sample gas flow rate was 150 seem for each case.
The MOF was ground to 325 mesh so the desorbed peak shows much less tailing effect
44
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
than that in Figure 2.11. Figure 2.12 compares the peaks seen in both samples. Notice
that the GC peak measured by adsorbing 642 ppb DMMP in IRMOF1 for 4 s and then
desorbing is ~5 times larger than from the empty groove measured with 651 ppm DMMP.
Therefore, we conclude that preconcentration gains o f over 5000 are possible using only
5 pg o f IRM OF1 and a 4-s sample time.
1200
Groove with MOF
1000
800
(5
600
.e>
<si
Empty Groove
100
i
200
,
1 _
0
M r
(
1
1
1
I
5
10
15
20
----------------------------------- —
1
25
"’
30
1
35
—
1
40
45
5]
-200
Time (Second)
Figure 2.13
A comparison o f the GC chromatograms produced by
exposing the IRMOF 1 in a groove in a Valeo sample valve to a gas stream
containing 2030 ppb o f dodecane followed by a thermal desorption to that from
an empty groove containing 2030 ppb o f dodecane.
45
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Figure 2.13 shows the results o f a similar experiment with dodecane vapor. In this
case, we passed 100 seem of a gas stream containing 2028 ppb dodecane through the trap
for 1 min and repeated the same measurements procedures. Notice that the dodecane peak
is only enhanced by a factor o f 5 when the preconcentrator is used compared to a gain o f
over 5000 for DMMP. Thus, it is clear that while IRMOF 1 adsorbs DMMP very rapidly,
it only slowly adsorbs dodecane, presumably because DMMP can have strong dipoledipole interactions with the framework while the dodecane only interacts via van der
Waals forces.
25000 T
Groove with Tenax
20000
15000
Empty Groove
10000
5000
100
120
Time (Second)
Figure 2.14
A comparison o f the GC chromatograms produced by
exposing the Tenax TA in a groove in a Valeo sample valve to a gas stream
containing 105 ppb o f DMMP followed by thermal desorption to that from an
empty groove containing 105 ppb o f DMMP.
46
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
We have also made same measurements using Tenax TA rather than IRMOF 1 in
the trap as shown in Figure 2.14. In this case, we only observe a concentration gain for
DMMP o f a factor o f 2 compared to 5000 with IRMOF 1. Clearly, the IRMOF 1 is much
more effective in a DMMP preconcentrator than Tenax TA.
2.4 Conclusions
In summary, a metal organic framework can be applied as good adsorbents for
preconcentrators in micro GC systems. These adsorbents have a tremendous capacity:
almost 1 g o f DMMP/g o f adsorbent, equivalent to -0 .7 g o f DMMP/mL o f trap.
Furthermore, the adsorption process is selective. We observed >5000 preconcentration
gain for DMMP and only 5 gain for toluene. Low gains were observed with Tenax TA
and Carbotrap. These results indicate that MOF molecules are effective adsorbents to be
used in preconcentrators.
2.5 References
1.
Tian, W. -C .; Chan, H. K. L.; Lu, C. -J.; Pang, S. W.; Zellers, E. T. J. o f
Microelectromechanical Systems,. 2005, 14, 498.
2.
Kim, M.; Mitra, S. J. o f Chromatography A, 2003, 996, 1-11
3.
Tian, W. -C .; Pang, S. W.; Lu, C.-J.; Zellers, E. T. J. o f Microelectromechanical
Systems, 2 0 0 3 ,12, 264.
4.
Lu, C.-J.; Zellers, E. T. Anal. Chem. 2001, 73, 3449-3457.
5.
Tian, W. —C.; Pang, S. W. J. Vac. Sci. Technol. B, 2003, 21, 274.
47
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
6.
Yaghi, O. M.; O ’Keeffe, M.; Ockwig, N. W.; Chae, H. K.; Eddaoudi, M.; Kim, J.
Nature, 2003, 423, 705-714.
7.
Eddaoudi, M.; Moler, D. B.; Li, H.; Chen, B.; Reineke, T. M.; O ’Keeffe, M.;
Yaghi, O. M. Acc. Chem. Res., 2001, 34, 319-330.
8.
Rowsell, J. L. C.; Yaghi, O. M. Microporous and Mesoporous Materials, 2004,
73, 3-14.
9.
Eddaoudi, M.; Kim, J.; Rosi, N.; Vodak, D.; Wachter, J.; O ’Keeffe, M.; Yaghi, O.
M. Science, 2002, 295, 469-472.
10.
Rosi, N. L.; Eckert, J.; Eddaoudi, M.; Vodak, D. T.; Kim, J.; O ’Keeffe, M.;
Yaghi, O. M. Science, 2003, 300, 1127.
11.
Eddaoudi, M.; Kim, J.; Rosi, N.; Vodak, D.; Wachter, J.; O ’Keeffe, M.; Yaghi, O.
M. Science, 2002, 295, 469.
12.
Li, H.; Eddaoudi, M.; O ’Keeffe, M.; Yaghi, O. M. Nature, 1999, 402, 276.
13.
Masel, R. I. in “Principles o f Adsorption and Reaction on Solid Surfaces”, John
Wiley & Sons, Inc., New York, 1996, 513.
14.
Ni, Z.; Masel, R. I. J. Am. Chem. Soc., 2006, 128, 12394.
48
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Chapter 3 Rapid Production of Metal-Organic Frameworks via
Microwave-Assisted Solvothermal Synthesis
3.1 Introduction
In this chapter we present a new synthetic approach that we named “microwaveassisted solvothermal synthesis” (MASS), which allows high quality metal-organic
framework crystals to be synthesized in under a minute. The properties o f the crystals
made by the microwave-assisted process are o f the same quality as those produced by the
standard solvothermal process, but the synthesis is much more rapid. Although the
microwave method usually cannot yield crystals with a size big enough for single X-ray
analysis, its homogeneous effects could create a uniform seeding condition; therefore the
size and shape o f the crystals can be well controlled and the synthesis cycle can be
largely shortened for many practical applications.
Borrowing from the geological formation o f natural porous materials - zeolites ,1
hydrothermal and solvothermal methodology 2"4 were adopted as major synthesis
approaches for the discovery o f new interesting MOF structures in recent years. So far all
o f the reported syntheses were ‘A day to multiday procedures. Yaghi and co-workers
produced a large number o f MOFs via solvothermal synthesis. The process too 1-7 days .6
Lin and co-workers created a nonlinear optically active MOF material through a multiday
hydrothermal synthesis .7 Wiliams and co-workers reported a thermal stable structure
through a 12 h solvothermal synthesis.
Q
Kim and co-workers reported several
catalytically active homochiral metal-organic materials through a two-day liquid
diffusion method 9 or solvothermal method .10 Suslick and co-workers reported polar
49
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
PIZA-1 structure by a two-day solvothermal synthesis ,11 and a nonpolar PIZA-4 structure
through one-week deprotonating vapor diffusion .12
Microwave-assisted processes have been used to produce small metal and oxide
particles. Such processes can involve heating a solution with microwaves for a period of
an hour or more to produce nanosized crystals o f metal. Typically, particle sizes are 15
nm. The expectation with microwave heating is that small particles, on the order o f 10-15
nm, will form.
13
Microwave synthesis to give 5-20 nanometer sized particles o f oxides is
also known .14
In Chapter 2 we found even less than mg scale o f MOF adsorbents would give
more than 5000 gain for a preconcentrator, therefore our original goal for MOF synthesis
is to find a way to produce MOF crystals in small size so they can be easily packed in the
a micro device for a enhanced performance. Several efforts such as ultrasonic and stirring
has been tried to create more nucleation sites during traditional solvothermal synthesis to
reduce the size o f crystals down to micro size. We could get small IRMOF 1 crystals in
one hour as shown in Figure 3.1, but the shapes are irregular and sizes could vary from
submicron to 20 pm in one pot, which is difficult to be manipulated for the followed
packing process. (Packing issues will be introduced in the last chapter o f the thesis.)
When microwave was introduced into the process, we found the resulting IRMOF 1 can
be created in 30sec with well defined cubic shape and uniformed size, shown in Figure
3.2. This discovery then led us to further optimize the synthesis conditions to some other
well-known MOF structures under microwave systems and revealed that MASS could be
a fast, easy and general method for most MOF synthesis.
50
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
In the scope o f this chapter: we report on the assembly o f five known MOFs,
namely IRMOF 1, IRMOF2, IRMOF3, IRMOF7 and MOF38 through a rapid microwaveassisted methodology. To get micro crystals with uniform size and identical morphology,
detailed conditions such as molar ratio and concentration has been considered and
optimized. We also demonstrate that crystal size can be varied from micrometer down to
submicrometer scale by manipulating the concentration o f the reactant solution. Finally, a
mechanism has been proposed to explain the phenomenons we observed in microwave
synthesis.
Figure 3.1 Micro MOF synthesis by stirring assisted solvothermal synthesis.
51
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Figure 3.2 A comparison o f MASS product and conventional solvothermal
product o f IRMOF 1
3.2 E xperim ental
In general, for the synthesis o f microcrystals o f IRMOF-n, a mixture o f metal
precursor and correspoinding spacing ligand was dissolved in A, A ’-diethyl form amide
(DEF) solvent. To create a homogeneous seeding environment, the mixture is thoroughly
stirred for 15 min to get a clear solution.
In a typical synthesis of IRMOF 1, an exact amount of Zn(N 0 3 )2' 6 H 2 0 (0.2 g,
0.67 mmol) and 1,4-benzenedicarboxylate acid (BDCH 2) (0.083g, 0.50mmol) are
dissolved in 10ml o f the A A ’-diethylformamide (DEF) resulting in a clear solution. An
amount of 1 mL o f the solution was sealed in a 4-mL Pyrex sample vial. The vial was
then placed inside a hood behind a blast shield and heated by a microwave synthesizer
(model 520A from Resonance Instrument Inc.) at 150 W for 25 s. A yellow suspension
formed after the microwave treatment as shown in Figure 3.2. The product was rinsed
(centrifuged and re-dispersed in DEF by sonicating) for 3 times before analysis.
52
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
The syntheses for rest o f MOFs are very similar to that o f IRMOF 1. Therefore in
the followed descriptions, only amounts o f precursors, solvents, and microwave
information is included.
IRMOF2 synthesis: Exact amount o f 2-bromoterephthalic acid, (2 -BrBDCH 2)
(0.040 g, 0.160 mmol), and zinc nitrate tetrahydrate, Zn(N 0 3 )2.6 H 20 , (0.0594 g, 0.20
mmol), were dissolved in 15 mL diethylformamide. The solution was then sealed with a
Pyrex sample vial and heated with a household microwave oven (800W) for a reaction
time o f 80 sec. A yellow suspension formed after the microwave treatment.
IRMOF3 synthesis: Exact amount o f 2-aminoterephthalic acid, (2-AminoBDCH2) (0.2 g, 0.67 mmol), and zinc nitrate tetrahydrate, Zn(N 0 3 ) 2.6 H 20 , (0.0913 g,
0.504 mmol), were dissolved in a mixture o f 39 mL diethylformamide and 3ml ethanol.
The solution was then sealed with a Pyrex sample vial and heated with a household
microwave oven (800W) for a reaction time o f 60 sec. An orange suspension formed
after the microwave treatment.
IRMOF7 synthesis: Exact amount o f 1,4 naphthalene dicarboxylic acid, (NDCFE)
(0.116 g, 0.538 mmol), and zinc nitrate hexahydrate, Zn(N 0 3 )2.6 H 2 0 , (0.2 g, 0.673
mmol), were dissolved in a 20 mL diethylformamide. The solution was then sealed with a
Pyrex sample vial and heated at 150 W for a reaction time o f 60 sec.
MOF38 synthesis: Exact amount o f zinc nitrate hexahydrate, Zn(N 0 3 ) 2.6 H 2 0 (0.1
g, 0.336 mmol) and 1,3,5-benzene-tricarboxylic acid (BTCH 3) (0.039g, 0.187 mmol)
were dissolved in 10 mL o f DEF. The solution was then sealed with a Pyrex sample vial
and heated with a household microwave oven (800W) for a reaction time o f 60 seconds.
53
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
All XRPD spectrums were recorded in a Bruker General Area Detector
Diffraction System in X-ray Lab. All the MOF powder samples were soaked in the DEF
solvent during the process o f data collections.
3.3 Results and Discussions
The powders after microwave treatment usually suspend in the solvent for a few
minutes. They eventually drop onto the bottom o f the vial and leave a clear solvent layer
on top. A further SEM studies show all o f these powders have well defined shape and
size ranging from sub-micron to a few micron. Figure 3.3 shows the SEM images of
IRMOF 1-3 powders produced through microwave assisted solvothermal method. They
all have a cubic shape, indicating their structures may contain a cubic crystal symmetry.
These powders also have a uniform size with a very narrow distribution range.
To verify that these powders have same framework structures as those in previous
reports, an XRPD spectrum for each sample was collected and was compared with the
simulated spectrum calculated from the published structures. For the IRMOF 1 sample,
we also compare the two spectrums with the spectrum o f conventional synthesized
powders from Chapter 2. Our results (Figure 3.4a) demonstrate that both conventional
synthesis and microwave assisted synthesis produce the same IRMOF 1 compound as
reported. After optimization o f the synthesis conditions, which will be discussed in next
section, we demonstrated that IRMOF2 and IRMOF3 can also be synthesized in a similar
microwave conditions as shown in Figure 3.4b and Figure 3.4c. The difference is that
microwave synthesized crystals have 2-3 orders o f magnitude smaller size than those
from solvothermal synthesis, and they can be produced rapidly within under a minute,
54
R ep ro d u ced with p erm ission o f the copyright ow ner. Further reproduction prohibited w ithout perm ission.
which is 2-4 orders o f magnitude shorter time than any existing conventional crystal
growth methods. The yields were also high. Syntheses o f the IRMOF 1-3 all have yield
above 90%.
Figure 3.3 (a) Enlarged SEM image o f micro IRM OF-1 (b) SEM image o f micro
IRM OF-1 (c) SEM image o f micro IRMOF-2 (d) SEM image o f micro
IRMOF-3
55
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
U!
b)
Figure 3.4 XRPD patterns o f (a) IRMOF 1, (b) IRMOF2 and (c) IRMOF3
prepared by the solvothermal synthesis (black), by MASS synthesis (green),
and a simulated curve based on the published structure o f Eddaoudi, M. (red).
56
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
As we know, the crystal formation usually starts from a favored nucleation
conditions. Such nucleation condition can be a dust particle or crystal seed inside the
solution, or a suitable geometry on the inner wall o f the container. These conditions are
trivial but could dramatically affect the result o f the final products. Therefore, many
reported MOF crystals synthesis were difficult to be repeated and could even be events of
serendipities. In chapter 2, our experience for the synthesis o f IRMOF1 through the
solvothermal method shows: sometimes two vials o f solvent prepared from same batch
and heated at same temperatures would give total different results.
The MASS method, however, shows highly repeatable results. This unique
advantage o f MASS method drove us to further investigate if it can reproduce some used
to be reported as “non-repeated” MOF structures.
Yaghi and coworkers once reported that MOF38, Zn 30 (BTC)2(HTEA )2 (where
BTC=l,3,5-benzenetricarboxylate, HTEA=triethylammonium) contains an interesting
porous 3D network connected by trigonal Zn prisms SBU and triangle SBU (secondary
building unit). Repeated attempts to reproduce its synthesis, however, were reported to be
“unsuccessful”.
In our experiment design, we start from the same precursors, zinc nitrate and
benzenetricarboxylic acid(BTC), and optimized the solvent conditions to be suitable for
microwave synthesis. The resulting powders revealed to have shapes in hexagonal or
trigonal plates (Figure 3.5). We then compared the experimental data with the calculated
XRPD spectrum from the publication (Figure 3.6). We found the major peaks between
the two spectrums matched well. Few peak locations have slight deviations and we
believe these are due to the different solvate we modified in our experiment.
57
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
As a conclusion, microwave assisted solvothermal synthesis can produce MOF
crystals with well defined shape and uniformed size. The synthesis usually takes only
about 1 minute. But the results are highly repeatable with a high yield. We demonstrate
that MASS could be applied as a general method for MOF synthesis.
Figure 3.5 SEM image o f MOF38 synthesized by MASS method.
58
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
!
—
'
Figure 3.6 XRPD
patterns
o f MOF38
prepared
by
microwave-assisted
solvothermal synthesis (black), and a simulated XRPD curve based on the
published structure o f Eddaoudi, M. (red).
O ptim ization of M icrowave Synthesis Conditions
The formation o f crystals can be considered as process o f self-recognition and
self-assembly o f free molecules in solvent. The natures and interaction o f reactant
molecules, such as coordination geometry, charge balance and shapes o f the ligands,
usually play an important role for the structure determinations. But the crystal formation
may also be affected by the seeding conditions and experimental conditions such as
temperature, solvent conditions, concentration, reaction time and etc. How exactly each
factor affect the crystallization process is still complicated and uncertain.
59
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Because o f these reasons, a resulting crystal form may have a different component
ratio from the nature or stoichiometry o f the reactant molecules. In some other cases, a
suitable synthesis condition can favor the crystal nucleation of two or more solid state.
Such phenomenon is called polymorphism, which will make the prediction o f the
structure to be more difficult.
Above phenomenons also happened during the MOF crystal growth in microwave
systems. Therefore detailed synthesis conditions need to be carefully optimized to obtain
the expected structure. One advantage o f the microwave method is that crystal nucleation
is no longer depends on the seeding condition or wall conditions. The rest o f the
conditions such as temperature, microwave power, metal/ligand ratio and concentration
can all be easily measured and controlled in microwave systems. The process is rapid and
the result is highly repeatable. Therefore, we can study and optimize these conditions all
quantitatively. My thesis work will discuss the effect o f concentration, molar ratio and
microwave temperature as follows. We do not consider the temperature and microwave
power due to the limitations o f the microwave oven we used; these conditions need a
further investigation.
During the optimization o f the microwave synthesis for IRMOF2, we found
metal:ligand ratio can play an important role. The SEM images in Figure 3.7 show the
crystal morphologies at a given concentration and microwave power with different Zn:L
ratio. Although in IRMOF2 structure Zn:L should have a ratio of 4:3, when we design the
ratio of the two precursors (ZnN 0 3 .6 H 20 and 2-bromoterephthalic acid) to be the same as
4:3, we did not obtain a pure crystal phase o f IRMOF2. In fact, cubic IRMOF2 and
another cross shaped crystal were observed co-existing in the final suspensions (Figure
60
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
3.7a). If we increased the Zn:L ratio up to 6:1, no cubes would be found. The square
sandwiched crystal became the major product and shows an XRPD spectrum totally
different from that o f the reported structure. Only when Zn:L ratio o f added reactant is
designed at 5:4, pure cubic powders were observed. Their XRPD spectrums are given in
Figure 3.4b, which confirmed that these cubic micro powders have same structures o f
IRMOF2. These experiments show that a relative lower Zn:L ratio than stoichiometry o f
IRMOF2 can drive the self-recognition process o f reactant molecule towards the cubic
structure.
In IRMOF7 synthesis, we found concentrations o f reactants, however, have a
major impact to the final crystal phase. When the starting solution has concentrated at
120mM (total concentrations o f metal and ligand precursors) (Figure 3.8a), square plate
like crystals were formed. If the solvent is diluted to 16 times by DEF solvent, cube like
crystal and leaf like crystal were found to co-exist in products (Figure 3.8c). Only when
the total concentration is maintained at ~60mM, the cubic crystals became the major
products (Figure 3.8b). A further XRPD comparison in Figure 3.9 shows the products at
this condition still contain another crystal phase, which is hard to be observed in SEM
image. A further optimization will be needed for IRMOF7 synthesis under microwave.
61
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
c)
Figure 3.7 Microwave synthesis o f IRMOF2 in Zn:L molar ratio o f a) 4:3 b) 6:1
and c) 5:4 at concentration o f [L]=0.02mM. Notice the pure product o f
IRMOF2 is only observed when Zn:L=5:4.
62
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
a)
Figure 3.8 Microwave synthesis o f IRMOF7 (molar ratio o f Zn:L=5:4) in
concentration o f a) [L]=53.8 mM, b) [L]=26.9 mM and c) [L]=0.538 mM.
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Figure 3.9
XRPD
patterns
o f IRMOF-7
prepared
by
microwave-assisted
solvothermal synthesis (black), and a simulated XRPD curve based on the
published structure o f Eddaoudi, M. (red).
Compared with IRMOF2, the product o f IRMOF1 is not very sensitive to Zn:L
ratio. We have tried the synthesis o f IRMOF1 in Zn:L ratio = 5:4 and 1:1, both conditions
generated same cubic crystals as those under the ratio = 4:3. This is due to the symmetric
nature of the BDCH 2 ligand in IRMOF1, which can create much easier self-recognition
events than the non symmetric ligand in other MOF motifs.
The concentration factor in microwave synthesis for IRMOF1 also does not affect
the final formation o f the cubic frameworks. In fact, we observed that smaller particles
64
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
c)
h)
Figure 3.10
SEM
images
o f IRMOF-1
produced
in
the
BDCH 2
concentration o f a)50.0 mM, b)25.0 mM, c)12.5 mM, d)6.25 mM, e)3.13
mM, f) 1.56 mM, g)0.78 mM, h)0.39 mM, i)0.20 mM (concentration o f Zn
precursor is reduced correspondently to keep the M:L ratio constant.)
65
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
-3. 5
- 2.5
0.5
001
0.001
Concert rati on of H2BD C ( M)
Figure 3.11
Estimated average size o f IRMOF-1 micro-crystal versus
concentration o f BDCH2 in the reactant solution
(b)
(a)
V
Figure 3.12
Enlarged
SEM
images
o f IRMOF-1
synthesized
concentration o f BDCH2 at (a) 0.05M and (b) 0.0002M.
66
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
from
can be obtained by reducing the reactant concentration in a starting solution. In the
synthesis o f small IRMOF1 crystals, the concentration o f BDCH 2 was diluted from 0.05
to 0.0002 M, and the amount o f zinc precursor was changed correspondingly so the
metal/ligand molar ratio remained 4:3. The microwave treatments were also extended up
to 90 s when the solution was diluted. An ~8 s increase in heating time was needed when
the concentration was diluted in half. Figure 3.11 includes images o f a series o f IRMOF1
synthesized at 8 different concentrations. Each concentration we tried was about half o f
previous concentration. The SEM images were taken on the same scale for an easy
comparison. They showed that particle size shrunk as the reactant concentration reduced.
Figure 3.11
further quantitatively illustrate the relationship between the
concentration o f the reactant in starting solvent and resulting crystal size we obtained.
Submicrometer-sized crystals were observed when the reactant concentration was scaled
down to a few mM. The edge and vertex o f the submicrometer-sized crystal are observed
to be less sharp than those o f micro-size crystals, Figure 3.12.
The effect o f microwave reaction time on crystal formation is also investigated.
No crystal formation is observed when the microwave time is under 20 s. By varying the
reaction time from 25 s to 1 min, the size o f the microcrystals does not change
remarkably.
Proposed Mechanism
Based on the phenomenons we observed in this chapter, we can speculate how the
microwave is able to enhance the crystal growth process. A proposed scheme is
illustrated in Figure 3.13.
67
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
In the conventional solvo-thermal growth, crystals nucleate near the walls or on
dust particles. That results in slow growth because there are very few seeds available. In
the microwave-assisted process, though, we observe crystals throughout the bulk o f the
solution probably because local superheating o f the DEF solvent leads to hot spots that
nucleate crystal growth. More seeds lead to faster growth and higher yields. Once the
seeds start to grow, available reactants are quickly depleted. Therefore the size o f the
crystals can be varied by adjusting the reactant concentration. The ability o f the
microwave technique to control the nucleation process leads to a narrow size distribution,
because all of the crystals are nucleated at once. It also allows new types o f MOFs to be
discovered readily since the growth process is not depending on nucleation on the walls
or dust particle.
One caution with the method: heating a closed bottle containing volatile solvents
and nitrates can produce an explosion. Microwaves create hot spots that can accelerate
the explosion. In particular the pressure in a vessel containing a volatile solvent (e.g.,
ethanol) can be much higher than with conventional synthesis. Microwave leakage is also
dangerous. We perform the experiment in a hood and place a blast shied in front o f the
sample vial. Readers are advised to take appropriate precautions.
3.4 Conclusions
In conclusion, microwave assisted solvothermal synthesis can produce MOF
crystals with well defined shape and uniformed size. The synthesis typically takes under a
minute, but the results are highly repeatable with a high yield. In this chapter we
demonstrate that the reported IRMOF1, IRMOF2, IRMOF3, IRMOF7 and MOF38 can
68
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
be reproduced in the microwave system. Conditions such as metal/ligand ratio,
concentration and microwave power were carefully studied during the synthesis. A MOF
nucleation mechanism was proposed to explain why concentration can vary the size of
the MOF crystal while microwave time can not, and why MOF formations in microwave
w on’t be affected by the particles and wall conditions. In general, MASS is as a rapid,
easy, and broad spectrum method that can be incorporated in most MOF syntheses.
V
M
/
| ■ In
+ Heat
Microwave Assisted
So Ivothermal
Conventional
Method
u
B 88ffi H
Figure 3.13
Proposed crystal nucleation mechanism for a) conventional
solvothermal synthesis and b) microwave assisted solvothermal synthesis.
69
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
3.5 References
1.
Barrer FRS, R. M. in “Hydrothermal synthesis o f zeolites”, Academic Press: New
York, 1982
2.
Laudise, R. A., in “Hydrothermal Synthesis o f Single Crystals.” Cotton, F. A.
(Ed.), Progress in Inorganic Chemistry, Wiley, New York, 1962, 3, 1-47.
3.
Rabenau, A. Angew. Chem., Int. Ed. Engl. 1985, 24, 1026-1040.
4.
Francis, R. J.; O ’Hare, D.; J. Chem. Soc., Dalton Trans. 1998, 3133-3148.
5.
Rosi, N. L.; Eckert, J.; Eddaoudi, M.; Vodak, D. T.; Kim, J.; O ’Keeffe, M.;
Yaghi, O. M. Science 2003, 300, 1127
6.
(a) Eddaoudi, M.; Kim, J.; Rosi, N.; Vodak, D.; Wachter, J.; O ’Keeffe, M.; Yaghi,
O. M. Science 2002, 295, 469. (b) Yaghi, O. M.; Eddaoudi, M.; Li, H.; Kim, J.;
Rosi, N. U.S. patent 20030004364A1, 2003
7.
Lin, W.; Wang, Z.; Ma, L. J. Am. Chem. Soc. 1 9 9 9 ,121, 11249.
8.
Chui, S. S. -Y .; Lo, S. M. -F .; Charmant, J. P. H.; Orpen, A. G.; Williams, L. D.
Science, 1999, 283, 1148
9.
Seo, J. S.; Whang, D.; Lee, H.; Jun, S. I.; Oh, J.; Jeon Y. J.; Kim K. Nature, 2000,
404, 982
10.
Dybtsev, D. N., Nuzhdin, A. L.; Chun, H.; Bryliakov, P. K.; Talsi, E. P.; Fedin, V.
P.; Kim, K. Angew. Chem. Int. Ed., 2006, 45, 916.
70
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
11.
Kosal, M. E.; Chou, J.-H.; Wilson, S. R.; Suslick, K. S. Nature Materials 2 0 0 2 ,1,
118.
12.
Smithenry, D. W.; Wilson, S. R.; Suslick, K. S. Inorg. Chem. 2003, 42, 7719.
13.
(a) Panda, A. B.; Glaspell, G.; El-Shall, M. S. J. Am. Chem. Soc., 2006, 128,
2790. (b) Lu, Q, Gao, F.; Li, D.; Komameni, S. Journal o f Materials 2005, 1, 1.
14.
Tompsett, G. A.; Conner, W. C.; Yngvesson, K. S. ChemPhysChem. 2006, 7, 296.
71
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Chapter 4 New MOFs Discovery and Their Sorption Behavior
Measurements
4.1 Introduction
The rational design of MOFs has been greatly developed by the linkage o f various
molecular building blocks in coordination frameworks .1 Having channels or pores with
different size, shape and functionality, MOFs would show various interesting behaviors
in the process o f sorption, chemical sensing and catalysis.
The objective o f this chapter is to discover a series o f new MOFs materials based
on the rapid microwave-assisted solvothermal synthesis. In our efforts, we tried to
combine several transition metals with bi-carboxylic or tri-carboxylic ligands in different
shapes and bearing different functionalities (shown in Figure 4.1). To date, 14 new
materials have been discovered and characterized. And among them, 3 new MOF
structures have been solved using single X-ray diffraction technique (this part o f work
will be described in Chapter 5). The detailed structure information for most o f our new
materials is still unknown. Therefore we arbitrarily designate an abbreviation name for
each new material that we have synthesized. They are: ZnM O Fl, ZnMOF2, ZnMOF3,
ZnMOF4, ZnMOF5, CuM O Fl, CuMOF2, CuMOF4, TbM OFl, CdM OFl, CdMOF2,
CdMOF3, CoM OFl and CoMOF2. Please note that we designate ‘M OF’ in all these
names to indicate what our project is aiming for, but they may or may not have a
framework structures. An accurate name and structure need further X-ray analysis in
future. We characterized the properties o f these materials by SEM, FTIR, XRPD and
BET. In the end, to explore these new MOFs as potential adsorbent materials for
72
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
preconcentrator, a detailed sorption study for each has been conducted by the TGA
measurements. Four vapors were chosen for the sorption study: toluene and water vapors
were picked as false positive which are common in the air; DMMP vapor was selected as
simulant for the chemical warfare agents (CWA); and nitrobenzene vapor was selected to
simulate TNT explosive.
HO.
H O yj
LI
4 .4 \4 " .4 ” ’-(21 H.23H-poij)fain<}-IO-13-20-tetrayl)t«tMJasO«nioic
BDCH,
tertphthaic aa d
1.2
2-arataio3 -b ro m o U re p h lh a lic a c id
»0
L4
2-mUotfsephttJJlic
4Cld
HC
L3
2-tiifluoromethojy
Ic ie p h U ia lic acid
Oy,OH
A”
U
L3
cis-cydobuUne-
16
2J-ttuophcne(iiC«borylic 2-biom.oUitphihulic
IJ.^-bmztnclncutoxylic
-1 .2 -d ic « fc o iy L c » a d
acid
add
L7
(did
acid
Figure 4.1 List of organic linkers that were applied as building blocks in this thesis
work
73
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
4.2 Experimental
All chemicals used were purchased and used without further purification.
Microwave synthesis o f ZnMOF 1:
Exact amount o f zinc nitrate hexahydrate, Zn(N 0 3 ) 2*6 H 20 , (0.18 g, 0.605 mmol)
and: 4,4’,4” ,4’” -(21H,23H-porphine-5-10-15-20-tetrayl)tetrakis(benzoic acid), (0.199 g,
0.252 mmol), were dissolved in 10 mL diethylformamide. The solution was then sealed
with a Pyrex sample vial and heated with a household microwave oven (800W) for a
reaction time o f 80 seconds.
The crystals are dark purple in cubic shape, with size
ranging from 10 to 30 microns.
Microwave synthesis o f ZnMOF2:
Exact amount o f zinc nitrate hexahydrate, Zn(N 0 3 ) 2*6H 20 , (0.1 g, 0.336 mmol)
and 2-anilino-5-bromoterephthalic acid, (2-anilino-5-BrBDCH2) (0.0847 g, 0.252 mmol),
were dissolved in 10 mL diethylformamide. The solution was then sealed with a Pyrex
sample vial and heated with a household microwave oven (800W) for a reaction time o f
80 seconds. The crystals are light yellow in cubic shape, with size ranging from 2 to 4
microns.
Microwave synthesis o f ZnMOF3:
Exact amount o f zinc nitrate hexahydrate, Zn(N 0 3 ) 2*6 H 20 , (0.15 g, 0.504 mmol)
and 2-trifluoromethoxy terephthalic acid, (2-trifluoromethoxy-BDCH2) (0.0946 g, 0.378
mmol), were dissolved in 10 mL diethylformamide. The solution was then sealed with a
74
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Pyrex sample vial and heated with a household microwave oven (800W) for a reaction
time o f 80 seconds. The crystals are yellow in cubic shape, with size ranging from 4 to 7
microns.
Microwave synthesis o f ZnMOF4:
Exact amount o f zinc nitrate hexahydrate, Zn(N 0 3 ) 2*6H 2 0 , (0.1 g, 0.336 mmol)
and nitroterephthalic acid, (0.0532 g, 0.252 mmol), were dissolved in 10 mL
diethylformamide. The solution was then sealed with a Pyrex sample vial and heated
with a household microwave oven (800W) for a reaction time o f 40 seconds.
The
crystals have irregular shape.
Microwave synthesis o f ZnMOF5
Exact amount o f zinc nitrate hexahydrate, Zn(N 0 3 ) 2*6 H 2 0 , (0.15 g, 0.504 mmol)
and cis-cyclobutane-l,2-dicarboxylic acid (0.0545 g, 0.378 mmol), were dissolved in 10
mL diethylformamide. The solution was then sealed with a Pyrex sample vial and heated
with a household microwave oven (800W) for a reaction time of 125 (85sec stop then 45
sec) seconds. The crystals are in square plate shape.
Microwave synthesis o f CuMOF 1
Exact amount o f cupric nitrate, Cu(N 0 3 ) 2*2 .5 H 20 , (0.15 g, 0.645 mmol) and 2,5Thiophenedicarboxylic acid, (0.0833 g, 0.484 mmol), were dissolved in 10 mL
diethylformamide. The solution was then sealed with a Pyrex sample vial and heated with
a household microwave oven (800W) for a reaction time o f 80 seconds.
75
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Microwave synthesis o f CuMOF2
Exact amount o f cupric nitrate, Cu(N 0 3 ) 2*2 .5 H 2 0 , (0.1 g, 0.430 mmol) and 2(Trifluoromethoxy) terephthalic acid (2-trifluoromethoxy-BDCH2) (0.0807 g, 0.322
mmol), were dissolved in 10 mL diethylformamide. The solution was then sealed with a
Pyrex sample vial and heated with a household microwave oven (800W) for a reaction
time o f 80 seconds.
Microwave synthesis o f CuMOF4
Exact amount o f cupric nitrate, Cu(N 0 3 )2*2 .5 H 2 0 , (0.1 g, 0.430 mmol) and 2bromoterephthalic
acid
(0.079
g,
0.322
mmol),were
dissolved
in
10
mL
diethylformamide.
The solution was then sealed with a Pyrex sample vial and heated
with a household microwave oven (800W) for a reaction time of 60 seconds.
Microwave synthesis o f TbM OFl
Exact amount o f terbium (III) nitrate pentahydrate, Tb(NC>3) 3*5 H 2 0 , (0.1 g, 0.230
mmol), and Terephthalic acid (BDC) (0.0286 g, 0.172 mmol), were dissolved in 10 mL
diethylformamide.
The solution was then sealed with a Pyrex sample vial and heated
with a household microwave oven (800W) for a reaction time o f 30 seconds.
The
crystals have rod shape.
Microwave synthesis o f CdMOF 1
Exact amount o f cadmium nitrate hexahydrate, Cd(NC>3) 2*4 H 2 0 , (0.1 g, 0.324
mmol), and cis-Cyclobutane-l,2-dicarboxylic acid (0.035g, 0.243 mmol), were dissolved
76
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
in 10 mL diethylformamide. The solution was then sealed with a Pyrex sample vial and
heated with a household microwave oven (800W) for a reaction time o f 30 seconds.
Microwave synthesis o f CdMOF2
Exact amount o f cadmium nitrate tetrahydrate, Cd(N 0 3 ) 2*4 H 20 , (0.1 g 0.324
mmol), and Nitroterephthalic acid (0.0456 g, 0.216 mmol), were dissolved in 15 mL
diethylformamide. The solution was then sealed with a Pyrex sample vial and heated
with a household microwave oven (800W) for a reaction time of 30 seconds.
Microwave synthesis o f CdMOF3
Exact amount o f cadmium nitrate tetrahydrate, Cd(N 0 3 ) 2*4 H 20 , (0.1 g, 0.324
mmol), and Terephthalic acid (0.0404 g, 0.243 mmol), were dissolved in 10 mL
diethylformamide. The solution was then sealed with a Pyrex sample vial and heated
with a household microwave oven (800W) for a reaction time o f 30 seconds.
Microwave synthesis o f CoMOFl
Exact amount o f cobalt (II) nitrate hexahydrate, Co(N 0 3 )2*6 H 2 0 , (0.1 g, 0.343
mmol), and Terephthalic acid (0.0428 g, 0.258 mmol), were dissolved in 10 mL
diethylformamide. The solution was then sealed with a Pyrex sample vial and heated
with a household microwave oven (800W) for a reaction time o f 30 seconds.
77
R ep ro d u ced with p erm ission o f the copyright ow ner. Further reproduction prohibited w ithout perm ission.
Microwave synthesis o f CoMOF2
Exact amount o f cobalt (II) nitrate hexahydrate, Co(N 0 3 )2*6 H 2 0 , (0.1 g, 0.343
mmol), and 1,3,5-Benzenetricarboxylic acid (0.0542 g, 0.258 mmol), were dissolved in
10 mL diethylformamide. The solution was then sealed with a Pyrex sample vial and
heated with a household microwave oven (800W) for a reaction time o f 30 seconds.
IR spectra were collected on an Infinity Gold FTIR™ spectrometer. Scanning
Electron Microscopy (SEM) and energy Dispersive X-ray Analysis (EDX) were carried
out on a Hitachi S-4700 Microscope in the Center for Microanalysis o f Materials. Surface
measurements were carried out using a Quantachrome NOVAe instrument from Prof.
Kenneth Suslick’s lab. XRPD spectrums were recorded in a Bruker General Area
Detector Diffraction System in X-ray Lab. TGA experiments were carried out on a
Mettler-Toledo TGA/SDTA 85l e instrument in Jeffery M oore’s lab.
Sample preparation for Thermal Gravimetric Analyses (TGA)
To measure the sorption capacity, each sample was saturated with toluene, water,
DMMP and nitrobenzene vapors respectively. We then loaded the sample powders into
the TGA instrument and measure the weight loss during thermal desorption.
In detail, MOF powders were soaked in CHCI3 for two days to exchange the
mother liquor in framework, and then powders were heated at 150 °C under vacuum to
completely remove the solvate. About 5 mg dried sample was added into a clean test tube.
Tested vapor was carried by 15 seem air flow through a saturator and then passed through
the test tube at 25-70 °C. After 10 //L o f solvent loss was observed from saturator, the
78
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout p erm ission .
saturated samples were collected. TGA analyses were then carried out by heating the
sample from 25 to 300 °C at 15 °C/min.
Sample preparation for BET measurements
The sample prepared for BET measurements followed the same cleaning and
drying procedures as those for TGA measurements.
4.3 Results and Discussions
The SEM images o f our new MOF materials are shown in Figure 4.2. The particle
size ranges from 1-30 pm. ZnMOF4 and CdMOF2 powders have irregular shapes, and
the rest samples all have well defined shapes. The XRPD spectrums further proved that
all these materials have good crystalline qualities. Figure 4.3 displays anhydrate powder
diffraction spectrums o f all compounds. However, exactly how metals and ligands are
coordinated in these materials is difficult to be predicted. We have built several possible
models using Cerius 2 software and compared their simulated powder diffractions with
our experimental spectrums. We found a slight change in coordination geometry or in
bond length will affect the diffraction peak heights and locations considerably. Therefore
the simulation results were not accurate enough to predict the framework structures.
79
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
ZdMOFI
ZnMOF2
ZnMOF3
ZhMOF4
ZnMOF5
CuMOFl
CuMOF2
CuMOF4
CAtfOFl
CdMQF2
CoMOFl
CoMOF2
CdMOF3
TbMOFl
Figure 4.2 SEM images o f 14 new metal-organic materials
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
5
‘
ZdMOFI
! !;
&
ZnMOFS
- -
. A 1, .
CoMQFl
ZdMOF4
CuMOFl
CuMOF2
CuMOF4
CdUQFZ
CdMOF3
TbMOfFl
\
■■
CtMOFl
H
ZdMOF3
Z dM O FZ
i ...
'
CaMQFZ
Figure 4.3 XRPD patterns o f 14 new metal-organic materials
81
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
M aterial
Metal
Organic
BET
Precursor
ligand
surface
IR
();is\ mi C H I
IR
Av
v<Osyin. cm
EDX
Thermal
analysis
stability
area
ZnM OFl
Z n (N 0 3)2
LI
133.53 m2/g
1606.4(s)
1403.9(s)
202.5
Zn, C, O
420 °C
ZnMOF2
Z n (N 0 3)2
L2
1.63 m2/g
1571.7(s)
1410.6(s)
161.1
Zn, C, O, Br
280 °C
ZnMOF3
Z n (N 0 3)2
L3
373.51 nfVg
1623.7(s)
1408.7(s)
215
Zn, C, O, F
350 °C
1413.5(s)
210.2
ZnMOF4
Z n (N 0 3)2
L4
93.50 m2/g
1620.8(s)
1404.8(s)
216
Zn, C, O
320 °C
ZnMOF5
Z n (N 0 3)2
L5
0.0 m '/g
1540.8(s)
1442.5(s)
98.3
Zn, C, O
370 °C
CuM OFl
Cu ( N 0 3)2
L6
0.0 m2/g
1557.2(m)
1382.7(s)
174.5
Cu, C, O, S
250 °C
CuMOF2
Cu ( N 0 3)2
L3
11.69 m2/g
1617.9(s)
1387.5(s)
230.4
Cu, C, O, F
320 °C
CuMOF4
C u( N 0 3)2
L7
9.48 m2/g
1615.l(s)
1395.2(s)
219.9
Cu, C, O, Br
250 °C
TbM OFl
T b (N 0 3)2
bdch
2.29 m2/g
1575.5(s)
1409.7(s)
165.8
Tb, C, O
400 °C
CdM OFl
C d (N 0 3)2
L5
42.81 m2/g
1579.4(s)
1429.0(s)
150.4
Cd, C, O
450 °C
1403.9(m)
175.5
CdMOF2
C d (N 0 3)2
L4
CdMOF3
C d (N 0 3)2
bdch
CoM OFl
C o( N 0 3)2
bdch
CoMOF2
C o( N 0 3)2
L8
IRMOF1
Z n (N 0 3)2
bdch
2
105.54 m2/g
1596.7(s)
1391.3(s)
205.4
Cd, C, O
270 °C
2
3.04 m 2/g
1571.7(s)
1388.5(s)
183.2
Cd, C, O
400 °C
2
12.47 m2/g
1587.1 (s)
1359.5(s)
227.6
Co, C, O
350 °C
108.22 m2/g
1583.2(m)
1377.8(s)
-2 4 9 .8 -
Co, C, O
350 °C
1627.6(s)
1445.3(m)
137.9
1605.4(s)
1391.3(m)
214.1
Zn, C, O
420°C
2
380.17 m2/g
Table 4.1 Physical characterization o f selected MOF compounds
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Table 4.1 summarizes the results o f BET, FTIR, EDX analysis and thermal
stability for each material. Most frameworks have thermal stability in the range o f 300450 °C. All three CuMOFs have relatively low stability due to the weak bonding between
Cu and carboxylic ligand. ZnMOF2 start to decompose at 280°C because anilino group
on the linker ligand is not stable.
Initial EDX analysis shows all materials contain the corresponding metal and
ligand, therefore it eliminates the possibility that the materials maybe one o f their starting
precursors’ crystals phase. A typical EDX spectrum for ZnMOF2 is given in Figure 4.4.
f-ull s c a le - 1 3 6 c o u n ts
Cursor: 2 1 8 3 7 keV
Br
rt Limi n t
0
1
2
3
4
-
5
6
7
i
9
10
Figure 4.4 EDX spectrum o f ZnMOF2.
During the preparation o f the samples for BET measurements, we found that the
pretreatment step o f exchanging the solvate in framework with CHCI3 can improve the
surface measurement results. The large DEF molecule can not be completely removed
from the framework when dried in oven directly; therefore it always exhibits limited
83
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
surface area that can be measured. Our BET measurements show ZnM O Fl, ZnMOF3,
ZnMOF4, CdMOF2 and CoMOF2 have close to or more than 100 m 2/g, indicating well
defined 3D pores or channels existing in these frameworks. The followed TGA tests
shown in Figure 4.4 further proved that large sorption capacities were always observed in
MOFs with high porosity. However, surface areas can not be used as a standard to judge
whether a material contains MOF structure or not. For example, all our CuMOFs have a
square plate shape indicating their frameworks may consist of 2D sheets, they have low
surface area because cavities in their structures could not be well defined. In Chapter 5
we will show that ZnMOF2 contains a 3D cubic framework structure, which have well
defined cavities built inside. However, its porosity is only 1.63 m2/g because the anilino
groups on ligand block most framework openings and therefore dramatically reduce the
accessible volume.
We noticed that in Yaghi’s report most o f the IRMOFs have surface porosity
ranging from 1000 -5600 m 2/g .1_4 While our MOFs in best cases only show 100-300 m 2/g.
We believe that an order o f magnitude between the porosity differences came from the
different models for surface calculations. Our measurements were based on the
Brunauser-Emmett-Teller (BET) model8 which considers the interaction between
different adsorbate layers. Yaghi’s calculation assumes Fangmuir surface adsorption. In
reality, for the porous MOF materials, the adsorption in the nanopore (or angstrom-size
pore) should not be considered as the adsorption o f molecules onto a solid surface, but
rather as the filling o f molecules into a nanospace where a deep potential field is
generated by the overlapping o f all the wall potentials. In this case, the adsorption
isotherm shows a steep rise at very low relative pressure and a plateau after saturation.
84
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
There are six representative adsorption isotherms that reflect the relationship between
porous structure and sorption type .5’6 (Figure 4.5) Type I isotherms are characteristics o f
MOF adsorbents that are microporous. Porous MOFs have a variety o f coordination
architectures with uniform and/or dynamic pore structures. In the conventional porous
materials, such as activated carbons and inorganic zeolites, pore shapes are often slit-like
or cylindrical, respectively. On the other hand, the pore shapes o f coordination polymers
are not necessarily modeled by slit-like and cylindrical pores because they have
crystallographically well-defined shapes, such as squares, rectangles, and triangles.
Unprecedented adsorption profiles have been found in porous coordination polymers,
which are characteristic o f the uniform microporous nature. For example, a square pore
possesses four comer sites where a deeper attractive potential for guests is formed by the
two pore walls than at the midpoint o f the wall . 7 (Figure 4.6)
1
IV
Ml
\\
r
f -
'r 1
v
Relative pressure ----------------------►
Figure 4.5 IUPAC classification o f adsorption isotherms
85
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
WAr= 47
27
!’
7
WAr= 195
1
§
14
........... 1.........
A
A
-
Nfi,r = 504
' "1............ ---------- 1--------- 7............ r............
!
*
&
a?
*•
*
f
1
?
1
I
1
A
X
I'J
i /
%
-jg
”J V
1,
■
■
••
■*'
f
-14
#
-27
*
i
a
i
!
i
$
#
s
-9 ,1 - 4 .6
i
0
i
4 .6
9.1
i
i
4 .6
0
XJA
i
)
'i “
r
f
f
i
i
!
4
i
1 "
*
i
I
1
|
"
■ #
f*
^
*
4
#
JtA
4
c
WA 0
K
$
_
,
i
f ® '4$ ' i
0 &■ A f f
,
i
i
-4 .6 -9 1 -4 .6
0
4 .6
9.1
------------ ►
Figure 4.6 Contours o f constant local density o f adsorbed Ar molecules for
several values of the pore loading (Monte Carlo computer simulations for the
pore size 18.2x54.6x A3). NAr is the number o f argon molecules adsorbed.
These local densities have been averaged along the direction o f the pore axis
and thus show where adsorption is occurring in a cross-sectional view down
the pore axis .7
86
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
We also tried to study the metal-carboxylate coordination by IR spectroscopy .9' 11
A carboxylate ion, RCC>2(-) can coordinate to metals in a number o f ways. Among them,
3 bonding modes are quite common in metal carboxylate coordination compounds. They
are unidentated, chelated and bridging bidentated. Metal carboxylates manifest strong CO stretching frequencies (vc.0,asym 1625-1540 cm "1 and
v c-0,Sym
1450-1360 cm '1).
Unidentated coordination removes the equivalence o f the two O atoms, so it should
increase
v c-0 ,asym,
decrease vc_0iasym. Chelated or symmetrical bridging should not alter the
bond orders. In general, the two frequencies separation (Av=
v c -0,asym - v c-0 ,sym )
can be
indicative o f the bonding modes. Unidented bonding has the separation > 200 cm '1,
chelated bonding has the separation <105 cm '1, in between is the bridging situations.
However, we found that above rules can not be applied to accurately predict some
o f the MOF structures. For example, we now know in IRMOF1 structure the two
carboxylate oxygen atoms are equally bond to the Z 114O cluster. Therefore Av should
have a value less than 200 cm '1. In our experiment record (Figure 4.5), we found the Av is
214 cm '1. In Chapter 5 we will show that ZnM OFl and ZnMOF3 (Figure 4.6) also have
the bridging bidentated bonding between metal and carboxylic ligand. Both o f their Av
are larger than 200 cm ' 1 and therefore do not match the prediction from the theory .9
87
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
120
■
100
■80
- BO
4500
4000
3500
3000
2500
2000
1503
1000
500
Figure 4.7 IR spectrum o f IRMOF1
120
4500
4000
3500
3000
2500
2000
1 500
1000
500
-
80
-
60
-
40
-
20
0
Figure 4.8 IR spectrum o f ZnMOF3
88
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Discussion of Sorption Tests
We explored the sorption capacities o f these materials as potential adsorbents for
preconcentors.(Table 4.2) Figure 4.7 list the sorption capacity of all these MOFs with the
four different vapors respectively. There results were also compared with IRMOF1
powders synthesized by microwave. We found these materials show dramatically
different sorption behaviors. In terms o f average sorption o f all four vapors, MOFs such
as CdMOFl and CuM OFl-4 have smaller or no sorption behavior due to their poor
porosity in frameworks. ZnM OFl, ZnMOF3 and CoMOF2 show relative higher sorption
ability. IRMOF1 gives the highest sorption capability.
110
-DMMP
%
©
Hi
-H 2 0
Tnlimnp
-N itrobenzene
i
60
35
85
135
135
285
Tem perature (Cl
Figure 4.9 Desorption o f toluene, DMMP, nitrobenzene and water vapor from
CoMOF2
89
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Toluene
DMMP
Nitrobenzene
Water
g adsorbate/g MOF
g adsorbate/g MOF
g adsorbate/g MOF
g adsorbate/g MOF
ZnM OFl
0.08
0.37
0.09
0.04
ZnMOF2
0.03
0.04
0.05
0.05
ZnMOF3
0.03
0.18
0.18
0.08
ZnMOF4
0.03
0.11
0.03
0.04
ZnMOF5
0.02
0.01
0.00
0.02
CuM OFl
0.02
0.00
0.00
0.05
CuMOF2
0.00
0.02
0.00
0.00
CuMOF4
0.03
0.04
0.02
0.03
TbM OFl
0.00
0.02
0.01
0.08
CdM OFl
0.00
0.00
0.00
0.00
CdMOF2
0.04
0.02
0.02
0.10
CdMOF3
0.04
0.02
0.04
0.05
C oM OFl
0.01
0.01
0.01
0.03
CoMOF2
0.09
0.08
0.05
0.18
IRMOF1
0.05
0.35
0.39
0.11
n/a
n/a
n/a
0.05
n/a
n/a
n/a
0.00
Material
IRMOF1
adsorb
water at 70°C
ZnMOF3 adsorb
water at 70°C
Table 4.2 The sorption capacity o f four vapors in new metal-organic compounds
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
to l u e n e
DMMP
□ n itr o b e n z e n
□ w a te r
Figure 4.10
Sorption o f toluene, DMMP, nitrobenzene and water in 14
MOFs and IRMOF1
According to the normalized solvent polarities ( E t N) reported by Reichardt,12
water ( E t N= 1 .0 0 ) has the highest polarity, toluene ( E t N= 0 .0 9 9 ) is relatively non-polar
vapor, and nitrobenzene (ETN=0.324) stay in the middle. So far, we don't know the ETN
value for DMMP yet, but based on three methyl groups in its structure we believe DMMP
should have a medium polarity. Therefore, the sequence o f their polarities should be:
water > nitrobenzene, DMMP > toluene.
Based on the vapor interactions with MOFs, we can conclude that ZnM O Fl,
ZnMOF3, ZnMOF4, IRMOF1 should have a framework with medium polarity. ZnMOF2
has a polar framework. We do not observe a clear trend in the rest o f the materials.
91
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Besides the porosity o f adsorbent and polarity o f vapor molecule, the volatility o f
vapor, size and shape o f framework pores will also affect the selectivity and capacity o f
each MOF. The sorption mechanism o f different molecules in MOF cavities needs a
further investigation through both simulated calculations and experimental approaches.
4.4 Conclusions
In this chapter 14 new compounds has been synthesized from various bifuntional
or trifunctional ligands and metal precursors by microwave. They all have well diffracted
XRPD patterns indicating they are microcrystals. EDX and IR spectrums show all the
microcrystals are made o f the corresponding metal ion and organic ligand through the
coordination bonds. The BET measurements showed that ZnM O Fl, ZnMOF3, ZnMOF4,
CdMOF2 and CoMOF2 have porosities around or above 100 m2/g, the rest MOFs are
either non-porous or low porous. Four vapors were selected for the sorption study. They
are DMMP and nitrobenzene as simulants for CWA and explosive molecule, and toluene
and water molecule as interference. The sorption behavior for each compound was
explored by TGA. The measurement results showed these new crystals have various
selectivity to different vapors.
4.5 References
1.
Yaghi, O. M.; O'Keeffe, M.; Ockwig, N. W.; Chae, H. K.; Eddaoudi, M.; Kim, J.
Nature 2003, 423, 705-714.
2.
Eddaoudi, M.; Kim, J.; Rosi, N.; Vodak, D.; Wachter, J.; O'Keeffe, M.; Yaghi, O.
M. Science 2002, 295, 469-472
92
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
3.
Eddaoudi, M.; Kim, J.; Rosi, N.; Vodak, D.; Wachter, J.; O'Keeffe, M.; Yaghi, O.
M. Science 2002, 295, 469-472.
4.
O ’Keefe, M.; Yaghi, O. M. Nature 2004, 427, 523-527.
5.
Brunauer, S.; Deming, L. S.; Deming, W. E.; Teller, E. J. Am. Chem. Soc. 1940,
62, 1723
6.
Gregg, S. J.; Sing, K. S. W. in “Adsorption, Surface Area, and Porosity, Academic
Press, London, 1984.
7.
Bojan, M. J.; Steele, W. A. Carbon 1998, 36, 1417-1423
8.
Sing, K. S. W.; Everett, D. H.; Haul, R. A.; Moscou, L.; Pieroti, R. A.; Rouquerol,
J.; Siemieniewski, T. Pure Appl. Chem., 1985, 57, 603-619.
9.
Mehrotra, R. C.; Bohra, R. in “Metal Carboxylates”, New York, 1983
10.
Deacon, G. B.; Philips, R. J. Coord. Chem. Rev. 1980, 33, 227-250.
11.
Nakamoto, K. in “Infrared and Raman Spectra o f Inorganic and Coordination
Compounds”, 3rd ed., John Wiley & Sons: New York, 1978.
12.
Reichardt, C. Chem. Rev. 1994, 94, 2319-2358.
93
R ep ro d u ced with p erm ission o f the copyright ow ner. Further reproduction prohibited w ithout perm ission.
Chapter 5 Structure Characterization of Some New MOFs
5.1 Introduction
Generally MASS can produce MOF crystals in a short period.1 The products have
good crystalline quality and well defined shapes. But the crystal sizes usually range from
a few hundred nanometers to few microns, which are too small for structure
determination by X-ray. Several attempts have been made in recent to solve the structure
by NMR and neutron diffraction techniques.2'4 However, due to the complexity o f the
MOF structure, single-crystal X-ray is still a major tool for the structure analysis.
To date, single X-ray technique is the most powerful tools that people can use to
completely resolve the structures o f molecule. The minimum requirement for the crystal
suitable for X-ray analysis is >150 pm; therefore the MASS method can not be applied
for the synthesis o f crystals in this size range. In future we would like to try to growing
sub-millimeter size MOF crystals with low microwave power feed and long processing
time. On the other hand, when we optimized the conditions for a specific new MOF
under microwave, then we kept all conditions same except that the microwave oven was
replaced with a heating oven for conventional solvothermal growth, the possibility for us
to collect big crystal suitable for X-ray analysis was more than 50%. Therefore, for any
unknown MOF synthesis, we can always use MASS as a quick tryout to decide if it is
possible to form the MOF product, or to find a suitable growth conditions. O f the 14 new
compounds we discussed in Chapter 4, we now have harvested 7 new crystals by
solvothermal method (shown as Figure 5.1). 3 o f these MOFs structures have been solved
and the rests are still under investigations.
94
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
In this chapter, the structures o f three new MOFs crystals are described. They are:
ZnM O Fl, ZnMOF2 and ZnMOF3. A comparison o f the water sorption between
ZnMOF3 and IRMOF1 is also included in this chapter. Their sorption behavior
differences are discussed and explained based on their structure infomation.
ZtaMOFl
&MOF2
OiMOFI
OiiMOF4
CoMOF2
TbMQFl
&MOF3
Figure 5.1 Crystals o f some new MOFs grown by traditional solvothermal
synthesis.
95
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
5.2 Experimental
Solvothermal synthesis o f ZnM OFl: Exact amount of zinc nitrate hexahydrate,
Zn(N 0 3 )2*6 H 2 0 , (0.18 g, 0.605 mmol) and: 4,4’,4” ,4” ’-(21H,23H-porphine-5-10-1520-tetrayl)tetrakis(benzoic acid) (TPPH4), (0.199 g, 0.252 mmol), were dissolved in 10
mL diethylformamide. The solution was then sealed with a Pyrex sample vial and heated
at 90 °C. The dark purple cubic crystals were collected after one week.
Solvothermal synthesis o f ZnMOF2: Exact amount o f zinc nitrate hexahydrate,
Zn(N 0 3 )2*6 H 2 0 , (0.1 g, 0.336 mmol) and 2-anilino-5-bromoterephthalic acid, (2-anilino5-BrBDCH2) (0.0847 g, 0.252 mmol), were dissolved in 10 mL diethylformamide. The
solution was then sealed with a Pyrex sample vial and heated 110 °C. Yellow hexagonal
crystals were collected after 10 days.
Solvothermal synthesis o f ZnMOF3: Exact amount of zinc nitrate hexahydrate,
Zn(N 0 3 ) 2*6 H 2 0 , (0.15 g, 0.504 mmol) and 2-trifluoromethoxy terephthalic acid, (2CF 3O-BDCH 2) (0.0946 g, 0.378 mmol), were dissolved in 10 mL diethylformamide. The
solution was then sealed with a Pyrex sample vial and heated at 110 °C. Transparent
cubic crystals were collected after one week.
Solvothermal
synthesis
of
CuMOF2:
Exact
amount
o f cupric
nitrate,
Cu(N 0 3 ) 2*2 .5 H 2 0 , (0.1 g, 0.430 mmol) and 2-(Trifluoromethoxy) terephthalic acid (2trifluoromethoxy-BDCFL) (0.0807 g, 0.322 mmol), were dissolved in
96
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
10 mL
diethylformamide. The solution was then sealed with a Pyrex sample vial and heated at
90 °C. Green rod crystals were collected after one week.
Solvothermal
synthesis
of
CuMOF4:
Exact
amount
of
cupric
nitrate,
Cu(N 0 3 )2*2 .5 H 2 0 , (0.1 g, 0.430 mmol) and bromoBDC (0.079 g, 0.322 mmol), were
dissolved in 10 mL diethylformamide. The solution was then sealed with a Pyrex sample
vial and heated at 90 °C. Green crystals in square plate were collected after one week.
Solvothermal synthesis of TbM OFl: Exact amount o f terbium (III) nitrate
pentahydrate, Tb(N 0 3 ) 3*5 H 2 0 , (0.1 g, 0.230 mmol), and Terephthalic acid (BDC)
(0.0286 g, 0.172 mmol), were dissolved in 10 mL diethylformamide. The solution was
then sealed with a Pyrex sample vial and heated at 110 °C. Transparent crystals were
collected after 15 days.
Solvothermal synthesis o f CoMOF2: Exact amount o f cobalt (II) nitrate
hexahydrate, Co(N 0 3 ) 2*6 H 2 0 , (0.1 g, 0.343 mmol), and 1,3,5-Benzenetricarboxylic acid
(0.0542 g, 0.258 mmol), were dissolved in 10 mL diethylformamide. The solution was
then sealed with a Pyrex sample vial and heated 90 °C. Dark red crystals in triangular
shape were collected after one week.
Single X-ray diffraction samples preparation: In order to obtain a single crystal
structure, different mounting methods have been tried in an attempt to get a crystal to
diffract. Due to the open structure o f MOF, further protection need to be considered to
97
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
prevent mother liquor lost from the framework. In general, when MOF crystals were
mounted in paratone oil, crystals still lose solvate easily and gave poor diffraction
patterns. When MOF crystals were kept in liquid nitrogen and tested at 78K, crystals still
diffracted poorly and we don know the reason yet. When the crystals were transferred via
mother liquor and sealed in capillary tubes, better diffraction results were collected.
Approximately one out o f every 3 crystals we mounted in solvent would have a good
diffraction pattern suitable for structure analysis. All diffraction patterns were collected
on a Bruker Molecular Analysis Research Tool (SMART) at room temperature.
5.3 Results and Discussions
Unit cell data were determined for all 3 MOFs by single crystal X-ray diffraction
techniques and are tabulated in Table 5.1.
The
crystals
of
ZnM O Fl,
[Zri4(TPP)],
(TPP=
5,10,15,20-tetra(p-
carboxyphenylporphyrinate) were obtained by DEF solvothermal methods. A perspective
view o f coordination geometry is shown in Figure 5.2. Selected bond distances and
angles are listed in Table 5.2. Four TPP ligands coordinate to the two Zn(II) atoms and
form a ‘paddle wheel’ structure. The separation between the two Zn atoms is 2.926(2)
A.
Each Zn atom is joined to four oxygen atoms from the carboxylic group on four TPP
ligands and form a square coordination geometry. The Z n -0 bond distances are 2.063 and
1.960
A.
These porphyrin ligand and Zn atom further form an infinite 2D square grid
sheets in the crystallographic ab plane as shown in Figure 5.3. These sheets are stacked
on each other and create infinite large square channels along the c-axis (Figure 5.4).
98
R ep ro d u ced with p erm ission o f the copyright ow ner. Further reproduction prohibited w ithout perm ission.
Using the atomic centers o f Zn 2 core and TPP center as the points o f a square, we
measured the edge o f the “square” as 11.765 x 11.936 A.
Compound code
ZnM O Fl
ZnMOF2
ZnMOF3
Dehydrate Formula
[Zn2(TPP)]
[(Zti40 )( 2 -anilino -5 -
[(Zn40 )( 2 -CF 30 -BD C )3]
BrBDC)3]
M orphology
Square plate
Cubic
Cubic
Color
Dark purple
Colorless
Colorless
Temperature
297K
297(1)K
297(2)K
Crystal System
Tetragonal
Cubic
Cubic
Space group
P 4 (l)
Fm-3m
Fm-3m
a
16.6902
A
25.7933
A
25.7650
A
b
16.6902
A
25.7933
A
25.7650
A
c
36.7953
A
25.7933
A
25.7650
A
a
90.0°
90.0°
90.0°
P
90.0°
90.0°
90.0°
Y
90.0°
90.0°
90.0°
V
10249.80
z
6
8
8
R1 [I>2sigma(I)]
0.2084
0.0901
0.0990
wR2
0.4824
0.2487
0.3129
A3
1716.1
AJ
17103.71
A3
Table 5.1 Crystal and structure refinement data for Z nM O Fl-3
99
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
04
C39
i >C33
C12
\
M 28
C40
Cl 3
C11
05
C15
<~7/|C8
06
C 9 > * - — Zn3
C17
C26 •->
(25
Zn2
02
C5 C4
• C1 C20
C21
C24
C42
C45
C2
08
C48
C23
«Zn
Q
C
# 0
# N
Figure 5.2 An ORTEP diagram o f the repeating unit in ZnM O Fl. H atoms are
not shown.
'/ x
Figure 5.3
p
The 2D sheet in ZnM OFl
100
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Figure 5.4
Bond length
The 2D sheet in ZnM O Fl, a view down from c axis.
(A)
Zn(l)-Zn(2) 2.926(2)
Zn(3)-N (l) 2.191(1)
Z n(l)-0(1) 2.063(1)
Zn(3)-N(2) 1.992(1)
Zn(2)-0(2) 1.960(1)
Zn(3)-N(3) 1.995(1)
Zn(3)-N(4) 2.257(1)
Angle(°)
0(l)-C (27)-0(2) 124.5(9)
0(2)-Zn(2)-Zn( 1) 75.5(3)
0(1)-Zn(l)-Zn(2) 85.1(3)
Symmetry transformations used to generate equivalent atoms: #1 x+1, y, z #2 x+1, y-1, z #3 x, y-1, z #4 x1, y, z #5 x-1, y+1, z #6 x, y+1, z
Table 5.2. Selected bond lengths
(A) and angles (°) o f ZnM OFl
101
R ep ro d u ced with p erm ission o f the copyright ow ner. Further reproduction prohibited w ithout perm ission.
02B
Z niB
02M
C 1A
C1B
02C
Br1A
Figure 5.5
An ORTEP diagram o f the repeating unit in ZnMOF2. H atoms are
not shown.
ZnMOF2 [Zn4 0 (2 -anilino-5 -Br-BDC)3], (BDC=Benzenedicaboxyliate) has cubic
frame stmcture, which is the same isoreticular framework o f the IRMOF1-16 reported by
Yaghi and his coworkers. Selected bond lengths and angles are given in Table 5.2. As
shown in Figure 5.5, six bifunctional ligands coordinate with the edges o f Zn40 core, the
resulting octahedral geometry confines the framework to be cubic porous network in 3D
space. Both bromo group and anilino group are highly disordered on all four possible
102
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
positions on phenyl ring. Using the atomic centers o f Z 114O core as the points o f the cube
shown in the Figure 5.6, we measured the cube has a dimension o f 12.882 A. This anilino
group blocked opening or the cavities and therefore reduced the accessible volume inside
framework.
Figure 5.6
A cube packing diagram o f ZnMOF2
103
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Bond length (A)
Z n(l)-0(2) 1.922(6)
N (l)-C (3) 1.407(10)
Z n(l)-0(1) 1.943(1)
Br(l)-C(3) 1.896(10)
C(l)-C(2) 1.496(9)
Angle(°)
0(2)-Zn( 1)-0(2F) 106.9(2)
Zn( 1)-0 ( 1)-Zn( 1A) 109.5(1)
0 (l)-Z n (l)-0 (2 ) 112.0(2)
C (l)-0 (2 )-Z n (l) 130.1(7)
0(2)-C (l)-0(2A ) 126.3(10)
C(3)-N(l)-C(4) 154.5(5)
Symmetry transformations used to generate equivalent atoms: #1 z,x,y #2 y,z,x #3 -x + l/2 ,y ,-z + l/2
#4 -x + l/2 ,-y + l/2 ,z #5 x ,-y + l/2 ,-z + l/2 #6 z,y,x #7 -z + l/2 ,y ,-x + l/2 #8 x,-y,z
Table 5.3. Selected bond lengths (A) and angles (°) o f ZnMOF2
MIA
F11B
Figure 5.7
An ORTEP diagram o f the repeating unit in ZnMOF3.
104
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Figure 5.8 A cube packing diagram o f ZnMOF3
ZnMOF3 [Zn4 0 (2 -txifluoromethoxy-BDC)3],
(BDC=Benzenedicaboxyliate) is
also an isoreticular structure o f the cubic frameworks. Selected bond distances and angles
are displayed in Table 5.4. Using the atomic centers o f (Z 114O) core as the points o f the
cube shown in the Figure 5.8, we measured the ZnMOF3 cube has a dimension o f 12.897
A. The trifluoromethoxy group on the ligand o f ZnMOF2 is much smaller than the
anilino group in ZnMOF3, therefore more accessible volume in frameworks would be
expected. This structure difference explained why ZnMOF3 has a porosity o f 373.51
105
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
m2/g comparable to IRMOF1, while ZnMOF2 only has a low porosity o f 1.63 m2/g. The
trifluoromethoxy group is also disordered in all the four possible positions on phenyl ring.
Bond length
(A)
Z n(l)-0(2) 1.932(6)
C (4)-0(4) 1.409(2)
C(l)-C(2) 1.461(8)
C (9)-0(4) 1.443(9)
Angle(°)
0(2)-C (l)-0(2A ) 129.2(10)
0(2)-Z n(l)-0(2F ) 106.3(2)
C (l)-0(2)-Z n(l) 128.2(6)
C(4)-0(4)-C(9) 110.0(8)
0 (l)-Z n (l)-0 (2 ) 112.5(2)
Symmetry transformations used to generate equivalent atoms: #1 -x+1/2, -y+1/2, z #2 x, -y+1/2, -z+1/2 #3
z, x, y #4 y, z, x #5 y, x, z #6 x, z, y #7 -x+1/2, -y, -z+1/2 #8 z, -y, x #9 -z+1/2, y, -x+1/2 #10 -z+1/2, -y, x+1/2 #11 z, y, x #19 -x+1/2, y, -z+1/2 #21 x, -y, z
Table 5.4 Selected bond lengths
(A) and angles (°) o f ZnMOF3
5.4 Comparison of water sorption between IRMOF1 and ZnMOF3
Functional groups introduced into frameworks by organic ligand can affect the
M OF’s sorption behavior significantly. IRMOF1 and ZnMOF3 are two iso-reticular
structures. They both have CaB6 topology.5 A detailed comparison is listed below. From
Table 5.5 we know that both structures have same space group, same crystal system and
similar unit cell parameters. The major difference is that ZnMOF3 have trifluoromethoxy
group in framework, therefore its porosity is also reduced. Trifluoromethoxy- is a very
stable non-polar functional group and we expect they would repel H20 molecule from
entering the cavities o f frameworks.
106
R ep ro d u ced with p erm ission o f the copyright ow ner. Further reproduction prohibited w ithout perm ission.
Name
IRMOF1
ZnMOF3
Crystal system
Cubic
Cubic
Space group
Fm-3m
Fm-3m
Unit cell dimensions
a = 25.6690(3)
A
b = 25.6690(3) A
c = 25.6690(3)
A
a = 9 0 d eg .
p = 90 deg.
y = 9 0 d eg .
Functional ligand
a = 25.765(5)
A
a = 90°.
b = 25.765(5)
A
p = 90°.
c = 25.765(5)
A
y = 90°.
H O ^P
T
BET measurements o f
380.17 m2/g
373.51 m2/g
powders synthesized
by M ASS
Table 5.5. A structure comparison o f IRMOF1 and ZnMOF3
In our previous TGA measurements we found the sorption capacities o f water in
two MOF materials are 0.1 lg F^O/g for IRMOF1 and 0.08g F^O/g for ZnMOF3
respectively, which are comparable to each other (Figure 5.9). However a differential
study o f the two curves revealed that water molecule has three binding sites in IRMOF1
pore at -65 °C, ~ 1 10 °C and -200 °C respectively, while in ZnMOF3 water has only one
binding site happened at -6 0 °C.(Figure 5.10) Therefore, we re-saturated the two MOF
powders in water vapors at 70 °C and collect the TGA data again. The purpose o f doing
this experiment is to find out if ZnMOF3 would adsorb less water due to the weak
interaction with water molecule. The results meet our expectation well. Figure 5.9 shows
ZnMOF3 almost adsorbed no H 2O at 70 °C, and IRMOF1 still shows a 0.05 g H20/g
107
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
capacity to H 2O vapor. Based on these phenomenons we can conclude that
trifluoromethoxy groups can significantly reduce the water adsorption in frameworks.
This result points out an important way for us to design the appropriate MOF
adsorbents for preconcentrators. As we know, in real applications o f gas detector, H 2O is
a common false positive molecule that would interfere the preconcentration o f species of
interests. Unlike other common interferences such as toluene and benzene, water is a
small and extremely polar molecule that was often a major trouble for many selective
preconcentrators. Here we demonstrate that ZnMOF3 can be useful preconcentrator
adsorbents that will eliminate the water sorption when working at 70 °C.
110
105 -
ZnMOF 3 adsorb at 25C
—
IR M 0F1 adsorb at 25C
ZnMOF 3 adsorb at 7 0 0
IR M 0F1 adsorb at 70C
85 -
80
35
85
135
185
235
285
Temperature (C>
Figure 5.9
Thermal desorption o f H2Q vapor from IRMOF1 and ZnMOF3.
108
R ep ro d u ced with p erm ission o f the copyright ow ner. Further reproduction prohibited w ithout perm ission.
0 .3 5 ■
0 .2 5 ■
—
ZnlvlOF3 ad sorb at 25C
—
ZnM OF3 ad sorb at 70C
IRMOF1 adsorb at 25C
—
IRMOF 1 adsorb at 70C
0 .0 5
135
185
235
285
-0 .0 5
Temperature (C)
Figure 5.10
The differential curves o f water desorption TGA curves from
IRMOF 1 and ZnMOF3
5.5 Conclusions
Based on the findings from chapter 4, we tried to grow the large crystals for each
new compound by solvothermal method under the similar solvent conditions and
concentrations. So far 7 conventional sized crystals (> 100 pm) have been collected.
Structures for 3 o f them have been solved by single X-ray analysis. They are ZnM O Fl,
ZnMOF2 and ZnMOF3. The X-ray revealed all these compounds are MOFs containing a
porous framework in structures. In details, ZnM OFl is porous 2D porphyrin-Zn sheets
stacking along the c axis, large square channels can be found inside. ZnMOF2 and
109
R ep ro d u ced with p erm ission o f the copyright ow ner. Further reproduction prohibited w ithout perm ission.
ZnM 0F3 are both isoreticular frameworks as IRMOF1-16 reported by Yaghi and co­
workers. Their structures both contain well-defined cubic lattice connected by the
octahedral Z 114O cores and dicarboxylate linkers. These structure features help explain
some physic properties and sorption behaviors measured in chapter 4.
For example,
ZnMOF2 contains large anilino group attached on the linker, therefore blocks the pores
and yields a very low porosity based on the BET measurements. ZnMOF3 contains
trifluoromethoxy group attached on the linker and can repel the water adsorption at 70 °C.
5.6 References
1.
Ni, Z.; Mase, R. I. J. Am. Chem. Soc. 2 0 0 6 ,128, 12394-12395.
2.
Lozano, E.; Nieuwenhuyzen, M.; James, S. L. Chem. Eur. J., 2001, 7, 2644
3.
Devi, R. N.; Edgar, M.; Gonzalez, J.; Slawin, A. M. Z.; Tunstall, D. P.; Grewal,
P.; Cox, P. A.; Wright, P. A. J. Phy. Chem. B, 2 0 0 4 ,108, 535-543.
4.
Calleja, M.; Mason, S. A.; Prince, P. D.; Steed, J. W.; Wilkinson, C. New J.
Chem., 2001, 25, 1475-1578.
5.
O ’Keeffe, M.; Hyde, B. G., in “Crystal Structures I: Patterns and Symmetry”,
Mineralogy Society o f America, Washington, DC, 1996.
110
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Chapter 6 Pack MOFs in a MEMS Device
6.1 Introduction
Our original preconcentrator design developed in Chapter 2 was basically a
powder packed reactor design. It has two drawbacks when applied in real applications:
first, it can cause a high pressure drop when sample gas flows through and therefore
should increase the energy budget for the operation o f each analysis; second, the powder
packed device can not resist strong mechanic impact, which can cause an inconsistent
performance. Our goal is to pack MOF powders efficiently into a MEMS device as the
schematic diagram shown in Figure 6.1, therefore a uniform micro MOF crystals layer
coated on glass or silicon substrate need to be achieved. And they should be easily
incorporated into the followed fabrication process.
Yoon and his coworkers have demonstrated that a serial o f zeolite microcrystals
can be readily tethered in the form o f uniformly oriented monolayer on glass and silica
substrate by silane based SAM linkers between the microrystals and the substrates.1'4 The
resulting monolayers of crystals were closely packed and uniformly aligned. But the
typical process needs continuously stir and flux conditions in dried toluene from few
hours to 1 day. In one o f their recent w o rk ,5 they found that ultrasound-aided strong
agitation can also lead to a uniform crystal coating which takes only about 10 minutes.
The coating rate was increased more than 103 higher and -100% degree o f coverage and
lateral close packing was achieved. (Figure 6.2)
111
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
I
I
W afer
500um
Figure 6.1
One o f the designs for MEMS based preconcentrator with MOF coated
inside
112
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Figure 6.2
SEM images o f silicalite crystal on glass prepared by reacting 3-
chloropropyl tethered glass plates with silicalite microcrystals for 2 min under
the conditions o f sonication. The scale bar is 10 pm
Based on the similar strategy, a layer o f MOFs grown on gold surface by using
the thio-based SAMs has been reported in recent by Fischers and et al .6 In their
procedures, a 16-mercaptohexadecanoic acid functionalized gold substrate was immersed
into a supersaturated mixture of IRMOF 1 precursors in DEF solvent for 24 hours. The
coating image in Figure 6.3 shows a non-uniform coverage, the control over the shape
and size of the crystals were also hard to be achieved. The preparation step o f the mother
113
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
solution take even longer time (72 hour) and need to be treated very carefully to prevent
crystallization happened before the substrate immersion. This pioneer work, however,
demonstrated that MOF partials can also be patterned on substrate through SAM linkages.
In this chapter, several attempts have been made to achieve a uniform coating o f
MOFs on the substrate through various methods. These methods include: epoxy,
ultrasonication, CPTES SAMs and electrophoresis. These research works still need to
developed to meet the desired effect as we expected. Nevertheless, the preliminary results
we presented in this chapter should throw some new lights onto the development o f MOF
coating research.
6.2 Experimental
All micro-sized MOF powders we applied were prepared and dried based on the
same microwave procedure described in previous chapters. Therefore we will focus on
the coating procedures in this session.
MOF coated on silicon grass
The silicon grass substrate was provided by Junghoon Yeom from Prof. Mark
Shannon’s lab. The silicon grass was etched by DRIE, with the grass diameter in 15 pm
and spacing distance in 50 pm.
10 mg o f dried IRMOF 1 powders was thoroughly suspended in the 15 mL toluene
solvent with the aid o f ultrasonic bath. A small drop o f the suspension was then
transferred onto the silicon grass by a glass burette. The substrate was dried in air for 1
hour before checking.
114
R ep ro d u ced with p erm ission o f the copyright ow ner. Further reproduction prohibited w ithout perm ission.
Figure 6.3
The model o f anchoring a typical IRMOF 1 building unit to a
carboxylic acid-terminated SAM and optical microscope and a AFM image of
a selectively grown film o f IRMOF 1 on a patterned SAM o f 16-mercaptohexadecanoic acid and 1H,1H, 2H, 2H-perfluorododecane thiol on A u ( lll)
for the mother solution at 25 °C .6
115
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
MOF coated on glass by epoxy
A thin layer of epoxy was applied on top o f a silicon or glass substrate. The epoxy
was left dry in room temperature for 12 hours. The dried IRMOF 1 powders (2-5 pm size)
were then spread onto the epoxy layer with the aid o f a spatula. Then the epoxy was
healed at 180 °C for another 24 hours. A nitrogen gun was used to blow off the
unattached powders, leaving a monolayer o f IRMOF 1 powders on top o f the substrate.
MOF coated on substrate by CPTES SAMs
Glass substrates were rinsed thoroughly
with water, then treated with freshly
prepared piranha solution (a 7:3 v/v mixture o f concentrated H 2SO 4 and 30 aqueous H 2O 2)
for 30 min at 90 °C, rinsed thoroughly with distilled, deionized water, and dried in an
oven. Caution: Prianha solution reacts violently with many organic materials and sold
be handled with extreme care.
A clean glass plate was placed in a round-bottomed flask charged with a toluene
solution (50 mL) o f CPTES (3-chloropropyl triethoxysilcane, 0.3 mL), and the solution
was refluxed for 1 hour under argon protection. After cooling to room temperature, the
glass substrate was removed from the flask and washed with copious amounts o f toluene.
Dried IRMOF 1 powders (10 mg) and CPTES tethered glass substrate were
introduced into a round-bottomed flask containing 30ml o f toluene. The mixture
heterogeneous mixture was refluxed for 2 hour under argon. The IRMOF 1 coated glass
substrate were removed from the flask, washed with copious amounts o f toluene, and
dried in the air for one hour.
116
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
MOF coated on substrate by ultrasonication
A glass substrate was cleaned and tethered with CPTES followed the same
procedure described in section 6.2.3.
Dried IRMOF 1 powders (10 mg, 2-5 pm size) were introduced into a roundbottomed
flask
containing
30ml
o f toluene.
The
heterogeneous
mixture
was
untrasonicated in a water bath. In the meantime, an CPTES tethered glass substrate was
held by a tweezer and immerged in the agitated suspension for 10 minutes. The glass
substrate was then removed with a monolayer o f IRMOF 1 on top o f it.
MOF deposit on interdigital electrodes through dielectrophoresis
The interdigital gold electrodes were prepared by Chang-Young Lee from Prof.
Michael Strano’s group. The silicon substrates with 150 nm thermal oxide were
metallized with 5 nm Ti and 300 nm Au, where electrodes with 6 pm gaps were patterned
by photolithography.
CuMOFl powders (2-4 pm size, lOmg) were suspended in 15 mL DEF by
ultrasonication. Then 2-4 drops o f the suspension were applied on top the silicon
substrate to form a puddle that can cover the whole region o f the interdigital electrodes
area. An ac voltage o f 10 Vpp at 10 MHz was applied on electrodes. After 5 minutes, the
CuMOFl powders were concentrated on the electrodes region and covered the whole
gaps space.
6.3 Results and Discussions
MOF coated on silicon grass
117
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
This is our first tryout to incorporate micro-sized MOF crystal inside a MEMS
based preconcentrator made by silicon. The grass like substrate was designed aiming to
keep the adsorbent powders in preconcentrator when sampling flow and carrier flow
passing through. There are three major drawbacks for this method: 1. the physically
attached powder still can not withstand the high flow rate, detached powders can easily
block the exit hole of the preconcentrator chamber; 2. the MOF adsorbents deposit at the
bottom o f the chamber should have a slow mass transfer rate during the sorption process;
3. when drops o f suspension were transferred into the preconcentrator, the top surface of
the silicon was very easy to get contaminated and therefore cause a serious trouble for the
followed bonding and sealing steps.
Figure 6.4
A MOF based preconcentrator with powder packed on posted
reactor design
118
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
MOF coated on glass by epoxy
Figure 6.5
A silicon wafer coated with MOF powders by epoxy.
This method is relative easier and can be generally applied to any type o f micro­
sized MOF powders. More than 85% coverage can be achieved. The binding force
between the powders and the substrate are strong enough to resist against the high flow
rate. However, the epoxy materials need to be inert to the adsorbates and need to be
thermalstable up to 300 °C.
119
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
MOF coated on substrate by CPTES SAMs
Figure 6.6
A silicon wafer coated with MOF powders by CP-TES SAMs.
CPTES linkers can form a strong coordination bond with the MOF
crystals. But so far we only obtain a low coverage below 10% due to the relative
large powder particles (5-15 pm) we used in previous. Using the smaller size
crystals for tethering and optimizing the reflux conditions has not yet been
investigated. We expect a much higher degree o f close packing (DCP) and degree
of coverage (DOC) should be achieved with the above considerations.
120
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
MOF coated on substrate by ultrasonication
Figure 6.7
A glass substrate coated with MOF powders by ultrasonication.
Figure 6.7 shows the effect o f a silicon substrate coated with IRMOF 1 with the
aid o f ultrasonication. Compared with refluxing method, the sonication method reduced
the coating time from a day to just few minutes. And we achieved a crystal coverage
around 50%.
Compared with small molecules and nanoparticles, microcrystals are much
heavier. Therefore, constant stirring and refluxing is needed to keep them dispersed in the
solution. It has been found that the kinetic energy gained by the microcrystal-tethered
121
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
functional groups from the hot-refluxing solutions is usually not high enough for the
functional groups to undergo linkage reaction with the functional groups tethered to
substrates .5 Therefore, the degree o f agitation o f the microcrystals caused by
ultrasonication is a very important factor that sensitively affects the reaction rate and the
degree of close packing of the crystals.
MOF deposit on interdigital electrodes through dielectrophoresis
a)
b)
c)
d)
Figure 6.8
The dielectrophoresis o f CuM OFl on an interdigital Au plate after
a) 0 sec, b) 30 sec, c) lm in and d) 5 min
122
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
This method is borrowed from the way how people introduced the single-walled
carbon nanotubes (SWNTs) onto the micro capacitor .7 We want to know if the electric
fields can be used for on-chip manipulation and assembly o f MOF particles. As shown in
Figure 6.9, CuM OFl microcrystals suspended in DEF solvent readily respond to
alternation current fields. Dielectrophoresis, particle mobility in ac fields allows precise
manipulation o f particles through a range o f parameters including ac voltage and
frequency and electrode geometry.
6.4 References
1.
Choi, S. Y.; Lee, Y.-J.; Park, Y. S.; Ha, K.; Yoon, K. B. J. Am. Chem. Soc. 2000,
122, 5201.
2.
Kulak, A.; Park, Y. S.; Lee, Y.-J.; Chun, Y. S.; Ha, K.; Yoon, K. B. J. Am. Chem.
Soc. 2 0 0 0 ,122, 9308.
3.
Chun, Y. S.; Ha, K.; Lee, Y.-J.; Lee, J. S.; Kim. H. S.; Park, Y. S.; Yoon, K. B.
Chem. Commun. 2002, 1846.
4.
Park, J. S.; Lee, Y.-J.; Yoon, K. B. J. Am. Chem. Soc. 2 0 0 4 ,126, 1934.
5.
Lee, J. S.; Ha, K.; Lee, Y.-J. Yoon, K. B. Adv. Mater. 2 0 0 5 ,17, 837.
6.
Hermes, S.; Schroder, F.; Chelmowski, R.; Woll, C.; Fischer, R. J. Am. Chem.
Soc. 2 0 0 5 ,127, 13744-13745.
7.
Snow, E. S.; Perkins, F. K.; Houser, E. J.; Badescu, S. C.; Reinecke, T. L. Nature,
2005, 307, 1942-1945.
123
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Appendix A: SEM Operating Instructions for MOF Powder
Observation
The instrument we used for SEM shot is a Hitachi S-4700 Scanning Electro
Microscope set in Frederick Seitz Materials Research Laboratory basement. The
following steps were followed to take SEM pictures for MOF powders and to analysis
them by EDX.
Sample Preparation:
1) The MOF powders need to be dried thoroughly in oven before checking.
2) Choose flat-top SEM sample pan and stick a double side electro-conductive tape on
top of it.
3) Use the sharp tip o f a needle or a tweezer to spread a small amount o f sample
powders on top o f the tape. The attached powder layer needs to be as thin as possible
to maintain a good quality o f conductivity to substrate.
4) If more than two samples are loaded on the tape, stick a small piece o f broken glass
on tape as a marker, and draw a geographic map o f samples relative to the marker.
5) To enhance the SEM image, the samples on tape are further coated with a thin layer
of Au/Pd by a Sputter machine in SEM lab o f MRL. Plasma time is set to be 15-20
seconds.
6) Load the sample pan on the SEM support, and adjust the total height to ~12mm.
Load Sample into SEM Chamber:
1) Fill the liquid nitrogen into the cold trap o f the SEM machine.
124
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
2) Check if the chamber pressure is normal. (I P l= lx l0 8, IP2=1 * 107, IP3=1><106)
3) Open the Monitor Software to observe loading site inside chamber.
4) Switch the buffering chamber knob from ‘A IR’ to ‘S.E.C’ to allow the pressure in
buffering chamber rise back to atmosphere.
5) Open the chamber and install the sample support onto the transfer rod.
6 ) Close the buffering chamber and pump down below 2 psi. Then open the inner
chamber gate to install the sample support at the bottom.
7) Pull out the transfer rod and close the inner chamber gate. Close the Monitor
Software.
Shoot SEM Images:
1) Run the Hitachi S-4700 operation software on PC. Set the power to be ‘Ultra High
Resolution, 10 kv” , then switch on the electron beam.
2) Under the low magnification screen (LOW MAG), use joystick to locate the position
of sample.
3) Switch to high magnification; perform an aperture alignment, XY alignment for the
beam.
4) Carefully adjust the focus, stigma and contrast to acquire a good quality o f SEM view,
(focusing and stigma adjustment operations are coupled and need to be treated
simultaneously.)
5) Press the camera button to capture a SEM picture. To make image comparable, keep
the magnification scale to be constant for each sample shot.
125
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Operations for Element Analysis by EDX:
1) Change the beam power to ‘Analysis, 10kv’
2) Set the sample-detector distance to be 12mm. The image will become blurred.
3) Using the knob on SEM chamber, manually adjust the sample-detector distance to
regain a clear view of SEM image on PC screen.
4) In the Manu bar, check ‘Analysis’ bar to display an analysis tool bar on screen. Using
mouse to define a desired spot or line or square area for EDX analysis.
5) Open the ‘Link ISIS’ software on another PC, set data acquisition step to be 100
second then start recording EDX data.
6 ) After a spectrum is collected, using the build-in element EDX spectrum database to
identify each peak.
Finishing Operations:
1) Save all data and shut down the beam. Open the Monitor software on screen again.
2) Press the ‘Exchange’ button to move the sample back to original loading position
3) Open the chamber inner gate and install the sample support to the transfer rod.
4) Move the sample into the buffering chamber. Close the inner gate. Switch knob to
‘A ir’ to recover the pressure again.
5) Open the buffering chamber and remove the sample.
6 ) Close the buffering chamber and pump down.
126
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Appendix B: MOF Crystal Mounting Procedures for Single X-ray
Structure Analysis
The MOF crystals are highly porous single crystals. Once the crystals are isolated
from the solvent, mother liquor in the framework will lose in one or two minutes. This
feature becomes a major difficulty for the single X-ray analysis.
Several methods to
protect crystals from losing solvent has been tried before, such as wrapping the crystal
with paratone oil or freezing the fresh mounted crystal in liquid nitrogen. These methods
still can not produce a good X-ray diffraction pattern for the structure analysis.
Based on previous experience, when the MOF crystal was mounted and preserved
in mother solvent, we have 1/3 chance to collect a good diffraction data. Therefore, the
basic procedures for mounting the MOF crystals with solvent are given as below.
1) The resulting crystals from solvothermal method usually hang on the inner wall o f the
reacton vials. First, using spatula to transfer all crystals and solvent from the vials to a
glass tray for easy mounting.
2) Target a crystal with no cracks inside and well defined shape under the optical
microscope. Measure the crystal size and decide which size o f capillary to choose for
mounting.
3) Choose a capillary with appropriate inner diameter for the crystal (Capillaries were
purchased from Charles Supper Company). Fill the mother solvent into the capillary
and get rid o f the bubble inside.
4) Use a glass burette to transfer the target crystal with solvent into the capillary. The
crystal will slowly drop to the neck o f the capillary by weight force.
127
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
5) Carefully push the crystal towards the dead-end o f the capillary with the aid o f a glass
fiber. Stop pushing when crystal is stuck in the middle the capillary.
6 ) Break the capillary from the open side and leave the crystal within the dead-end part.
The distance between the crystal and broken end should be less than 5 mm.
7) Seal the broken end with epoxy or wax. Then break the other end o f capillary leaving
~5 mm long distance between the crystal and the broken end and seal the end with
epoxy or wax..
8 ) Preheat a heating plate to 200 °C. Place a small amount o f wax on top o f a brass tube
provided by X-ray lab (25.0 mm in length, 2.4 mm in outer diameter, 1.4 mm in inner
diameter). Heat the brass tube with wax near the heater. Once the wax is melt, insert
the crystal loaded capillary in side the tube quickly and leave 2-3 mm distance from
the crystal to the edge of the tube. (Figure A .l)
Note:
1) The thin glass fiber can be obtained by the following steps: get a melting point glass
tube purchased from Kimble Glass, Inc.; heat the middle section o f the tube with a
propane flame; pull the tube from both side when the glass become soft; less than 100
pm glass fiber can be obtained after several practice.
2) The required distances labeled in Figure A .l were based on the feature size o f the
Single X-ray machine. The updated parameter can be acquired from the X-ray
Diffraction Lab.
128
R ep ro d u ced with p erm ission of th e copyright ow ner. Further reproduction prohibited w ithout p erm ission .
Epoxy or Wax Seal
Mother Liquor
<
5m
m
C
ry
sta
l
2-3m
m
Wax Fixing
Epoxy or Wax Seal
Figure A .l
The final look o f a MOF crystal mounted in solvent
129
R ep ro d u ced with p erm ission o f the copyright ow ner. Further reproduction prohibited w ithout perm ission.
Appendix C: Preconcentrator Operating Instructions
The purge and trap system setup for the preconcentration measurement is shown
in Figure A.2. MOF crystals are loaded in one o f the four grooves (0.2-0.5 pL) on a
Valeo rotor. The loaded rotor is sealed inside a VICI injection valve. The valve is further
installed inside a GC oven, with the sampling pipe and carrier pipe (1/16 inch in diameter)
connected to outside. The vent is connected with a silicon capillary with inner diameter
of 100-200 pm. The outlet o f injection port is connected to the FID detector by a similar
silicon capillary.
P2
Figure A.2
The schematic diagram o f the purge-and-trap setup for the
preconcentration measurement.
In the schematic diagram, we named the groove loaded with MOF as groove# 1. The
rest of three empty grooves are numbered counterclockwise. The rotor is driven by a
VICI multiposition actuator installed outside the oven. In each rotation step the rotor will
rotate 90° in counterclockwise direction, allowing two grooves to be aligned with the
130
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
sampling port and injection port o f VICI valve respectively. The following steps describe
a typical preconcentration gain measurement for 107 ppb DMMP vapor.
1)
Place the DMMP bubbler in the water bath o f Isotemp equipment,and set the
constant temperature at 0 ° C.
2) Open the helium gas tank and adjust the total pressure PI at 20 psi. Set carrier gas
pressure P2 at 5 psi. Set the flow controller #1 at 20 seem as sampling gas flow. Set
the flow controller #2 at 100 seem as dilution gas flow, (so the DMMP vapor at 0 0 C
can be diluted 6 times before entering the sampling port, reducing the DMMP
concentration from 642 ppb by mol to 107 ppb by mol.)
3) Aligned the groove #1 (MOF loaded) with the injection port as the starting position.
Use a small beaker o f water to check if flow comes out from the vent.
4) Increase the GC oven temperature to 250 ° C. Watch the FID signal drop to the
baseline. The MOF trap is then confirmed to be clean for the followed analysis.
5) Reduce the oven temperature to 25 ° C as sampling temperature for preconcentrator.
Press the actuator control button once so the groove #1 is rotate 90
0
counterclockwise to be aligned with sampling port. (Figure A.3a)
6 ) Control the sampling time to be 4 sec - lmin, then rotate another 90 0 (Figure A.3b).
Perform the same sampling conditions and sampling time for the empty groove #2 as
reference.
7) Use actuator to rotate another 90 ° so both groove#l and #2 are in a sealed positions.
(Figure A.3c) Increase the oven temperature to 250 ° C for a thermal desorption.
131
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Injection Valve
Injection Valve
V ent
C arrier
Gas
Carrier
Gas
FID
Injection
Injection
Gas
Gas
(a)
(b)
Injection Valve
Injection Valve
RefI
V ent
Vent
Carrier
Gas
C arrier
Gas
HD
HD
Injection
Gas
Injection
Gas
(c)
Figure A.3
(d)
Schematic diagram o f the experimental procedure used to
measure the gain o f the preconcentrator. A) Firstly, a helium gas containing
ppb levels o f DMMP are fed into a groove in a Valeo gas sampling valve
containing MOF crystals. B) Secondly, gas is fed into an empty tube. C)
Then both grooves are rotated against blank openings and the valve is heated
to desorb the vapors.
D) Finally, the DMMP desorbed from the MOF is
injected into the column.
132
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
8) Maintain the oven temperature at 250 °C. Rotate another 90 ° (Figure A.3d), collect
the signal injected from MOFs in groove# 1 by the FID detector.
9) Wait until the signal drop to the base line again, then rotate another 90 ° (Figure
A.3a). Collect the signal injected from the empty groove#2 by the FID detector.
10) Steps 3-9 can be repeated for another measurement.
11)
133
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Author’s Biography
Zheng Richard Ni was bom in 1976 in Xian, China. His family moved to Shantou
in South China while he was in high school. In 1994, he was admitted into Department o f
Chemistry, University o f Science and Technology o f China. He pursued undergraduate
research developing inorganic nonlinear optical materials. He received his B.Sc. in
Chemistry in July 1999. After that, he continued his graduate study in National
University o f Singapore and explored the synthesis and characterization o f coordination
polymers guided by Prof. Jagadese J. Vittal. In July 2001, he received his Master degree
in Chemistry. He then came to University o f Illinois at Urbana Champaign in January
2002 and joined the research group led by Prof. Richard I. Masel few months later. His
first research project is to design and develop microreactors for hydrogen fuel cell. Two
years later he switch to the microGC project and specialized in Metal-Organic
Framework synthesis and its applications in micro preconcentrators. During his time in
Illinois, he has contributed 2 inventions and 4 papers as first author. He was once chosen
as one o f finalists for Lemelson-Illinois 3 OK Student Award in 2007.
Richard will join the high-tech start-up company, Cbana Laboratory, following
his Ph.D. graduation.
PATENTS
"Rapid production o f metal organic crystals via solvothermal synthesis method" Ni, Z.;
Masel, R. I., U. S. Patent, TF06049
"High gain selective preconcentrator for gas sampling" Ni, Z.; Masel, R. I.; Shannon, M.
A., U. S. Patent, TF05139
134
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
PUBLICATIONS
“Metal-Organic Framework as Novel Adsorbents for Trapping and Preconcentration o f
Organic Phosphonates” Ni, Z.; Jerrell, J. P.; Cadwallader, K. R.; Masel, R. I. Analytical
Chemistry, 2006, in press.
“Rapid Production o f Metal-Organic Frameworks Via Microwave Assisted Solvothermal
Synthesis” Ni, Z.; Masel, R. I. Journal o f American Chemical Society, 2006, 128, 1239412395.
“Synthesis o f High-Temperature Titania-Alumina Supports” Subramanian V.; Ni, Z;
Seebauer, E.; Masel, R. I. Industrial & Engineering Chemistry Research, 2006, 45, 38153820.
"A metal-organic framework based preconcentrator for gas sampling in a micro-gas
chromatograph" Ni, Z; Shannon, M.; Cadwallader K.; Jerrell J.; Masel, R. I. Micro Total
Analysis Systems (juTAS), 2 0 0 5 ,1, 262-264.
“Microreactors and microreaction Engineering” Masel, R. I.; Gold, S.; Ni, Z.
Encyclopedia o f Chemical Processing, Ed. Lee, S., Taylor & Francis, Inc, New York,
2005, pp. 1643-1661.
“Effects o f microreactor geometry on performance: differences between posted reactors
and channel reactors” Ni, Z.; Seebauer E. G.; Masel, R. I. Industrial & Engineering
Chemistry Research, 2005, 44(12), 4267-4271.
135
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Документ
Категория
Без категории
Просмотров
0
Размер файла
2 950 Кб
Теги
sdewsdweddes
1/--страниц
Пожаловаться на содержимое документа