Tailoring adsorption induced phase transitions in the pillared-layer type metal-organic framework DUT-8(Ni)

Article abstract of DOI:10.1039/C7DT00015D

Tailoring the characteristics of gating transitions in the porous network, Ni₂(ndc)₂dabco (ndc = 2,6-naphthalenedicarboxylate, dabco = 1,4-diazabicyclo[2.2.2]octane), also termed DUT-8(Ni) (DUT = Dresden University of Technology), was achieved by systematically adjusting the critical synthesis parameters. The impact of the starting composition and solvent mixtures in the synthesis was found to critically affect the guest-response properties of the obtained materials. A comprehensive set of physical characterization methods, namely thermal analysis,¹H NMR of digested crystals, solid state¹³C NMR, PXRD, SEM, IR and Raman spectroscopy shows that the crystallite size is a crucial factor, determining the differing characteristics such as “gate pressure” and adsorption capacity in the guest-responsive switching behaviour of DUT-8. Crystallites smaller than 500 nm in size retain the open form after removal of the guest molecules resulting in typical “Type Ia” isotherm, whereas crystallites larger than 1 μm transform into the “closed pore” form and therefore can show a characteristic “gate opening” behaviour during gas adsorption. The particle size distribution of DUT-8(Ni) can be tailored by changing the synthesis conditions and consequently the slope of the isotherm at the “gating step” is affected. The in depth analysis of synthesis conditions and switching behaviour is an important step towards a better understanding of the fundamental principles responsible for guest responsive porosity switching in the solid state.

Full text of DOI:10.1039/C7DT00015D

View Article Online
View Journal
Dalton
Transactions
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: N. Kavoosi, V.
Bon, I. Senkovska, S. Krause, C. Atzori, F. Bonino, J. Pallmann, S. Paasch, E. Brunner and S. Kaskel, Dalton
Trans., 2017, DOI: 10.1039/C7DT00015D.
This is an Accepted Manuscript, which has been through the
Volume 45 Number
1 7 January 2016 Pages 1–398
Royal Society of Chemistry peer review process and has been
accepted for publication.
Dalton
Transactions
Accepted Manuscripts are published online shortly after
An international journal of inorganic chemistry
www.rsc.org/dalton
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
author guidelines.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the ethical guidelines, outlined
in our author and reviewer resource centre, still apply. In no
event shall the Royal Society of Chemistry be held responsible
for any errors or omissions in this Accepted Manuscript or any
consequences arising from the use of any information it contains.
ISSN 1477-9226
PAPER
Joseph T. Hupp, Omar K. Farha et al.
Efficient extraction of sulfate from water using
framework
a Zr-metal–organic
rsc.li/dalton

Please do not adjust margins
Page 1 of 12
Dalton Transactions
View Article Online
DOI: 10.1039/C7DT00015D
Dalton Transactions
ARTICLE
Tailoring adsorption induced phase transitions in the pillared-
layer type metal-organic framework DUT-8(Ni)
Received 00th January 20xx,
Accepted 00th January 20xx
Negar Kavoosi,a Volodymyr Bon,a Irena Senkovska,a Simon Krause,a Cesare Atzori,b Francesca
Bonino,b Julia Pallmann,c Silvia Paasch,c Eike Brunnerc and Stefan Kaskela*
DOI: 10.1039/x0xx00000x
www.rsc.org/
Tailoring the characteristics of gating transitions in the porous network, Ni2(ndc)2dabco (ndc
= 2,6-
naphthalenedicarboxylate, dabco = 1,4-diazabicyclo[2.2.2]octane), also termed DUT-8(Ni) (DUT = Dresden University of
Technology), was achieved by systematically adjusting the critical synthesis parameters. The impact of the starting
composition and solvent mixtures in the synthesis was found to critically affect the guest-response properties of the
obtained materials. A comprehensive set of physical characterization methods, namely thermal analysis, 1H NMR of
digested crystals, solid state 13C NMR, PXRD, SEM, IR and Raman spectroscopy shows that the crystallite size is a crucial
factor, determining the differing characteristics such as “gate pressure” and adsorption capacity in the guest-responsive
switching behaviour of DUT-8. Crystallites smaller than 500 nm in size retain the open form after removal of the guest
molecules resulting in typical “Type I” isotherms, whereas crystallites larger than 1 µm transform into the “closed pore”
form and therefore can show a characteristic “gate opening” behaviour during the gas adsorption. The particle size
distribution of DUT-8(Ni) can be tailored by changing the synthesis conditions and consequently the slope of the isotherm
at the “gating step” is affected. The in depth analysis of synthesis conditions and switching behaviour is an important step
towards a better understanding of the fundamental principles responsible for guest responsive porosity switching in the
solid state.
Introduction
However, there are still numerous open questions in the
Structural phase transitions induced by external stimuli such as understanding of “gating” phenomena from both experimental
guest molecules, temperature, electromagnetic irradiation and theoretical point of view.
etc., are unique features of flexible metal-organic frameworks, From the theoretical point of view, various fundamental
also termed soft porous crystals.1-4 Among various kinds of concepts for understanding the flexible behaviour were
flexibility such as breathing, swelling, ligand flip, sub-network developed, for example: (i) the adsorption-induced
displacement, gating stands out as a pronounced and step- deformation can be interpreted as a response of the porous
wise transition with an amplitude of cell volume change often material to an adsorption induced stress; (ii) a rigorous
by far exceeding any volume changes observed in traditional thermodynamic description of the coupled energetics of the
porous solids.5 We therefore address these materials in the crystallographic phase transition(s) and the adsorption
following as “switching MOFs” or “gate pressure MOFs”. After isotherms of the different polymorphs within the osmotic
guest removal, “gate pressure MOFs” usually transform into a ensemble.17, 18
dense structure that can be reopened by an external stimulus.6 The synthesis rules and procedures for the targeted synthesis
Switching between a “porous” and “non-porous” (or dense) of “gate pressure” materials and control over the “gate
state offers tremendous opportunities for the application of pressure” is a crucial issue for any application. In the case of
these MOFs in gas storage, separation, and sensing.7-16 layered MOFs such as ELM-11 (ELM – Elastic Layered Material),
the first “gate pressure MOF” discovered, extensive
investigations focussed on the adsorption induced transitions
a.Department of Inorganic Chemistry, Technische Universität Dresden, Bergstrasse
in the presence of nitrogen, carbon dioxide, n-butane,
66, D-01062 Dresden, Germany
19-21
methane, and alcohol vapours.6,
Tuning of the “gate
b.Department of Chemistry, NIS and INSTM Reference Centre, University of Torino,
Via G. Quarello 15, I-10135 and Via P. Giuria 7, I-10125, Torino, Italy
c. Department of Bioanalytical Chemistry, Technische Universität Dresden,
Bergstrasse 66, D-01062 Dresden, Germany
pressure” was achieved by varying the metal ions and type of
layer-terminating anions.22 The dynamics of the “gating” effect
was investigated in situ using various gases as guest
molecules.23
† Corresponding author: Prof. Stefan Kaskel, E-mail: Stefan.Kaskel@tu-dresden.de
Electronic Supplementary Information (ESI) contains nitrogen and argon
physisorption data, theoretical and experimental pore size distribution, PXRD
patterns, scanning electron microscopy images, 1H NMR, FT-IR and TG data. See
DOI: 10.1039/x0xx00000x
Pillared-layer MOFs can be regarded as a second prototypical
class of switching MOFs as discovered by Dybtsev,24 but
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 1
Please do not adjust margins

Please do not adjust margins
Dalton Transactions
Page 2 of 12
ARTICLE
Journal Name
initially only volume changes below 15% were reported in the Lee et al. who reported the material with the saVmieew AsrttirculecOtnulirnee
DOI: 10.1039/C7DT00015D
absence of solvent. However, a complete closing of the and composition as for DUT-8(Ni) but described it as a rigid
framework was not achieved in this case.
A unique member of this family is Ni2(ndc)2dabco (ndc – 2,6- transformations during the guest removal.32 This material was
naphtalenedicarboxylate, dabco 1,4-diazabicyclo- synthesized using slightly modified synthesis conditions.
MOF with a clear “Type I” isotherm and no structural
-
[2.2.2]octane), also known as DUT-8(Ni) (DUT = Dresden Consequently, DUT-8(Ni) is a unique solid, showing either
University of Technology).25-30 The highly flexible Ni2-paddle switchable or rigid behaviour depending on the way it is
wheel in combination with the ndc ligand allows to achieve the synthesized. So far there are only few reports describing a
complete pore closing after removing the guest molecules. synthesis dependent switchability tuning in porous solids.9, 33,
34, 35
Consequently this compound shows the highest cell volume
change (> 250 %) among all pillared layer MOFs and a In the following we describe a detailed study on synthesis and
pronounced “gate opening” effect, initially reported at p/p0 = pre-treatment parameters affecting “gate pressure” and
0.1 for nitrogen adsorption at 77 K.29 The complex structural hysteresis in DUT-8(Ni). In order to elucidate the fundamentals
diversity of DUT-8(Ni) in the presence of nitrogen, carbon of solid state porosity switching and in the context of the
dioxide, ethane, ethylene3 and n-butane was unravelled using textural characteristics behind, we have focussed on a global
in situ powder X-ray diffraction (PXRD) at controlled pressure analysis using various spectroscopic techniques, scanning
and temperature.26 Characteristic host-guest interactions were electron microscopy, elemental analysis, TG etc. and
detected applying in situ 129Xe NMR.31
However, recent observations
correlated the outcome with gas physisorption results
repeated revealing the degree of switchability.
during
adsorption/desorption cycling using n-butane and InfraSORP
technology revealed significant stress built-up upon repeated
adsorption caused by grain boundary formation and
disintegration of larger crystals.27 Three different type of
crystal interfacial surface areas may in principal affect the
gating phenomena by adding a kinetic and/or thermodynamic
contribution to alter the energetics of switching: a) the outer
surface area (Ao) of a monodomain crystal (surface energy); b)
the interfacial area between two domains (grain boundary or
interfacial domain area, Aid) with different orientation in one
Experimental
Materials
Chemicals and solvents were purchased from Sigma Aldrich
and they were used without further purification. Nickel (II)
nitrate hexahydrate (99.999%), 2,6-naphthalenedicarboxylic
acid (99.0%) and 1,4-diazabicyclo[2.2.2]octane (ReagentPlus®,
≥99%) were handled under inert atmosphere. N,N-
Dimethylformamide (99.8%) and methanol (99.8%) were
purchased in anhydrous form.
crystal; c) the interfacial area of
a monodomain or
multidomain crystal covered by a second phase (which can be
also an amorphous outer layer (Aol), a polymer, or a crystalline
shell). For DUT-8, it was shown that an increasing Aid causes a
shift of the gate opening pressure to higher pressure
Synthesis
Flexible DUT-8(Ni) (1)
The synthesis procedure of DUT-8(Ni) reported in ref. was
29
(equivalent to higher chemical potential of the guest slightly modified insofar as all chemicals were dissolved
molecule)27. As we show in this contribution, the gate opening
separately: Ni(NO3)2·6H2O (0.407 g, 1.40 mmol) in 6 mL DMF,
pressure does not reflect a transition at equilibrium, but is
kinetically controlled. The concerted pore opening of unit cells
can be interpreted as a cooperative lattice movement (soft
phonon) propagating throughout the crystal. If adjacent grains
have different crystallographic orientation, the interfacial
domain area (grain boundary) acts as a phonon propagation
barrier and shifts the transition pressure towards higher
chemical potential of the transition inducing guest molecules.
However, if the interface between two domains creates gas
accessible channels, these defects will act as nucleation
centers for the phase transformation. Thus, the number of
such defects in a crystal will determine the transition
probability and consequently the gate opening pressure. A
monodomain crystal without grain boundaries in the closed
state may be kinetically so severely hindered, that an
adsorption induced transition is not any longer observed and
the crystal remains closed.
H2ndc (0.303 g, 1.40 mmol) in 15 mL DMF and dabco (0.100 g,
0.89 mmol) in 9 mL methanol (MeOH). Subsequently the
mixture was transferred into a Teflon vessel (50 mL) and
heated in an autoclave to 393 K with a heating rate of 4 K/min
and held at that temperature for 48 h. After cooling to room
temperature (RT) during 3 hours, the sample was washed
several times, first with 30 mL DMF and then with 30 mL
ethanol and at the end with 30 mL DCM. Afterwards, the
washing step with 150 mL of DCM was continued for 3 days.
The resulting solid was filtered in argon flow and activated in
dynamic vacuum at 393 K for 4 h. DUT-8(Ni) crystals have poor
heat conductivity, therefore for efficient activation of sample it
is important to prevent aggregation of the crystals. In order to
facilitate the homogeneous heat transfer to all crystallites
instantly without solvothermal stress during the desolvation, a
flask with a wide flat bottom was used. This method allows for
a
homogeneous heating of all the crystals without
The strong influence of domain size on “gate pressure” was an
initial motivation to study the impact of crystal morphology in
DUT-8(Ni) in more detail. Another observation triggering our
intentions for “gate pressure” tuning came from a report by
temperature gradients and simultaneous evacuation. Note: A
special vacuum safe flask is required for this operation.
Evacuation (at 10-3 mbar) continued over night at room
2 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 3 of 12
Dalton Transactions
Please do not adjust margins
Journal Name
temperature. During the activation, the colour of the sample Ni(NO3)2·6H2O and in
ARTICLE
5
´b 0.601 g (2.07 mmol) Ni(NO3)2·6H2O
View Article Online
DOI: 10.1039/C7DT00015D
changes from green (open pore phase) to dark yellow (closed were used and in both cases, the ratio of H2ndc to dabco was
pore phase). Finally, the Schlenk tube was purged with argon kept constant (0.303 g H2ndc, 0.1 g dabco).
and transferred into the Glove Box.
Rigid DUT-8(Ni) (2)
Table1. Synthesis parameters used for tuning DUT-8(Ni) materials1
The “rigid” form of DUT-8(Ni) was synthesized using a similar
procedure as reported by Lee et al.32 In a typical synthesis,
Ni(NO3)2·6H2O (0.434 g, 1.50 mmol), H2ndc (0.294 g, 1.36
mmol) and dabco (0.336 g, 3.00 mmol) were mixed and
dissolved in 30 mL of DMF using an ultrasonic bath for 10 min.
After mixing of all chemicals, a cloudy suspension was
1
2
3a
3b
3c
3d
3’a
3’b
3’c
3’d
4a
4b
4c
5a
5b
5’a
5’b
1.0
1.1
1.1
1.1
1.1
1.1
1.0
1.0
1.0
1.0
1.1
1.1
1.1
1.0
0.5
0.5
1.5
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
0.636
2.206
0.199
0.382
0.735
1.470
0.250
0.636
0.993
2.143
2.206
2.206
2.206
2.206
2.206
0.636
0.636
9.0
-
-
-
-
21.0
30.0
30.0
30.0
30.0
30.0
21.0
21.0
21.0
21.0
29.5
25.0
19.0
30.0
30.0
21.0
21.0
393/10
408/10
408/10
408/10
408/10
408/10
393/10
393/10
393/10
393/10
408/10
408/10
408/10
408/10
408/10
393/10
393/10
obtained,
which
is
in
disagreement
with
the
procedurereported by Lee. A clear solution could not be
obtained even after ultrasonication of the cloudy precipitate
for 2 hours. The suspension was placed in a Teflon-lined
autoclave and heated in an oven to 408 K with a heating rate
of 4 K/min and held at that temperature for 72 hours. The
resulting material was washed with 30 mL DMF two times and
subsequently immersed in 30 mL of dry EtOH for 3 days. The
light green powder was filtered out and activated in dynamic
vacuum for 5 hours at 443 K. After activation, the colour
remained light green. The sample was kept under inert
atmosphere.
-
9.0
9.0
9.0
9.0
0.5
5.0
11.0
-
0.0
9.0
9.0
1For the starting compounds, the molar ratio normalized to H2ndc is
given.
Variation of dabco content: series 3
Samples a-d of series
procedure as described for
3
were synthesized using the same
, but utilising varying amounts of
2
dabco. In all the cases, the ratio of Ni2+:H2ndc 1.1:1 was kept
constant (0.434 g Ni2+, 0.294 g H2ndc). The following amounts
of dabco were utilised: in 3a 0.030 g (0.27 mmol), in 3b 0.058 g
(0.52 mmol), in 3c 0.112 g (1.00 mmol) and in 3d 0.224 g (2.00
mmol).
Powder X-ray diffraction
Powder X-ray diffraction (PXRD) patterns were measured at
room temperature on a STOE STADI P diffractometer using Cu–
Kα1 radiation (λ = 1.5405 Å) and a 2D detector (Mythen,
Dectris). All measurements were performed in transmission
geometry using a rotating flatbed sample holder, 2Θ steps of
0.015° and exposition time of 20 s per step. All activated
samples were sealed using decathlon aptonia survival blanket
(film composition: 4.0% Resin, 92.0% polytetrafluoroethylene
(PTFE), 4.0% aluminium).
Variation of dabco content: series 3’
Samples a-d of series 3´ were synthesized using the procedure
described for
1. For all the samples, a constant ratio of
Ni2+:H2ndc 1:1 was used (0.407 g Ni2+, 0.303 g H2ndc). The
amount of dabco was varied as follows: in 3´a 0.039 g (0.35
mmol), in 3´b 0.100 g (0.89 mmol), in 3´c 0.156 g (1.39 mmol)
and in 3´d 0.336 g (3.00 mmol).
Adsorption
Variation of solvents ratio: series 4
Nitrogen (at 77 K) and argon (at 87 K) physisorption
experiments were performed using a BELSORP-max apparatus.
A liquid nitrogen bath was used to thermostat the sample at
77 K. Argon physisorption measurements were performed
using a special external adsorption cell, mounted on a closed
cycle He-cryostat. Measurements were carried out using 25 -
35 mg of the MOF. For all experiments, nitrogen and argon
gases with purity of 99.999% were used.
All samples were synthesized using procedure described for 2,
but the ratio of solvents (DMF to MeOH) was varied: in 4a 29.5
mL DMF and 0.5 mL MeOH, in 4b 25 mL DMF and 5 mL MeOH,
and in 4c 19 mL DMF and 11 mL MeOH.
Variation of nickel to ligands ratio: series 5
Samples a-b of series
method used for preparation of sample
5
were synthesized with the same
, but the ratio of Ni to
2
other compounds was changed. In both cases, a ratio of
H2ndc:dabco = 1:2.2 was used (0.294 g H2ndc, 0.336 g dabco).
In the case of 5a, 0.394 g (1.35 mmol) of Ni(NO3)2·6H2O and in
5b, 0.197 g (0.68 mmol) of Ni(NO3)2·6H2O was added.
Variation of nickel content: series 5´
Thermogravimetric analyses
Thermogravimetric analyses (TG) were performed in synthetic air
flow in a temperature range from room temperature to 1000 °C
with a heating rate of 5 °C/min using STA 409 PC from NETZSCH
Company.
Samples
procedure as described for
components was changed. In
a
and
b
of series 5´ were synthesized using the same
but the ratio of Ni to other
0.204 g (0.70 mmol)
1
5
´
a
Spectroscopic measurements
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 3
Please do not adjust margins

Please do not adjust margins
Dalton Transactions
Page 4 of 12
ARTICLE
Journal Name
DRIFT spectra were measured in the range of 4000 - 400 cm-1 using
a VERTEX-70 spectrophotometer (Bruker).
As expected, the XRD pattern of sample 1 matches the
theoretical calculated pattern of DUT-8(Ni) for the closed
View Article Online
DOI: 10.1039/C7DT00015D
FTIR ATR spectra were collected at 2 cm-1 resolution in the 4000 -
400 cm-1 range using a Bruker Alpha instrument equipped with an
ATR accessory (diamond crystal) and placed inside the Glove Box.
Raman spectra were recorded on powder samples inserted in a
home-made cell with a suprasil quartz cuvette by using a Renishaw
inVia Raman microscope spectrometer and an Ar+ laser emitting at
514 nm. Photons scattered by the sample were dispersed by a 1800
lines/mm grating monochromator and were simultaneously
collected on a CCD camera. The collection optic was set at 20X. The
spectral collection setup consisted of 50 acquisitions, each of 10 s.
phase (cp) (Fig. 1a) and shows “gate opening” behaviour
during the physisorption (Fig. 1b). In contrast, sample 2 shows
an XRD pattern typical for the open pore phase (op) of DUT-
8(Ni) (Fig. 1a) and a “Type Ia” isotherm (Fig. 1b) with a pore
volume comparable to that of 1 after “gate opening”.
1
According to the elemental analysis as well as to the H NMR
spectra after digestion of the samples, no significant
differences in the composition of
(Table S1, Fig. S6, S7, ESI).
1 and 2 could be found
Scanning electron microscopy
The SEM images were recorded using a Hitachi Microscope SU8020.
Elemental analysis
The CHN analysis was performed on a EuroEA Elemental Analyser.
The amount of Ni in the samples was determined by ICP-OES using
an Optima 7000 DV instrument.
Nuclear magnetic resonance
The solutions for 1H NMR were prepared by dissolving the activated
samples in a mixture of 1 mL (99.9%) dimethyl sulfoxide-d6 and 20
μl of DCl (36% solution in D2O). The spectra were measured on a
Bruker DRX-500 spectrometer. Solid-state 13C NMR experiments
were performed on a Bruker AVANCE 300 spectrometer operating
at a resonance frequency of 75.5 MHz using a double-resonance 4
mm MAS NMR probe and a sample spinning rate of 12 kHz. Ramped
1H-13C cross polarization (CP) and SPINAL proton decoupling during
signal acquisition was applied. The chemical shift was referenced
with adamantane.
Results and Discussion
To identify pivotal parameters responsible for the differences in the
switching characteristics of Ni2(ndc)2(dabco) it was essential to
reproduce the syntheses of “flexible” and “rigid” materials, in the
following denoted as 1 and 2, respectively. For the synthesis of 1
the amount of dabco was slightly increased29 to improve the yield.
When following the reported procedure for the synthesis of “rigid”
DUT-8(Ni) (2) reported by Lee et al.32 the solution became a cloudy
suspension after mixing of all chemicals. This was surprising, since
the published procedure reported: “The suspension was treated by
ultrasonic radiation to dissolve solid chemicals until a clear solution
in green colour is formed.” However, in our hands the precipitate
could not be dissolved even after 3 hours of sonication. A reason
might be slightly different water content of solvents or starting
materials. The suspension was placed into an oven, which was pre-
heated to 408 K beforehand, and kept at that temperature for 3
days. The obtained green precipitate was collected by filtration,
washed and activated at 443 K in dynamic vacuum for 5 hours. After
activation, samples 1 and 2 were analysed by PXRD, NMR, and
nitrogen physisorption.
Figure 1. a) Calculated and experimental PXRD patterns of
activated “flexible” ( ) and “rigid” ( ) DUT-8(Ni); b) Nitrogen
physisorption experiments at 77 K for DUT-8(Ni) “flexible” (
and “rigid” ( ) (adsorption – filled symbols, desorption – open
symbols) Inset: logarithmic plot.
1
2
1
)
2
In addition, we performed TG analysis in synthetic air in order
to estimate the thermal stability and metal/organics ratio.
Sample
1
shows slightly higher thermal stability than sample
2
2
and starts to decompose at 380 °C. The decomposition of
starts at 340 °C (Fig. S8, ESI). Decomposition of both samples
proceeds in one step. In the case of sample the thermal
1
decomposition is accompanied by the formation of
intermediates, which are further oxidized in the temperature
range of 500 – 550 °C, causing increasing of the mass of nearly
4%. The residual mass for samples
1 and 2 is 24.12 and
23.89%, respectively, which is in a good agreement with the
calculated value of 22.72%. The PXRD analyses of the residues
show phase pure NiO (Fig. S9, ESI).
4 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 5 of 12
Dalton Transactions
Please do not adjust margins
Journal Name
ARTICLE
Thus, despite the pronounced differences in flexibility, the Using the same ratio of components described for
2
, if nickel
View Article Online
compositions of
1
and
2
are identical within the limits of errors nitrate is added to the H2ndc-dabco sDuOspI:e1n0.s1i0o3n9/Cin7DDT0M00F1,5Da
of the methods. A detailed analysis of defects in lower green, cloudy precipitate in the following denoted as “Ni-ndc-
concentration level is beyond the scope of this contribution dabco” forms. A DRIFT spectrum of this precipitate contains
but has been addressed recently using EPR in combination both additional absorption bands at 785 and 1630 cm-1 found
with NO adsorption.25
in spectrum of
DRIFT spectra of 1 and 2 are apparently very similar (Fig. 2a). pathway of and
However, the spectrum of rigid compound 2 contains a few nucleation of DUT-8(Ni), while the rigid version
additional absorption bands at 1630, 785, and 510 cm-1 which an intermediate salt. The ATR-IR spectra of
and
are absent in the spectrum of 1. The first one could correspond confirm, even if less evidently, the presence of these two
to the asymmetric vibration of the carboxylic group. vibrational fingerprints of the rigid form.
The Raman spectra of and (Fig. 2b) are significantly
shows: i) in the 1700 –
2
. This indicates differences in the formation
: For flexible one may assume the direct
is formed via
(Fig. S14),
1
2
1
2
1
2
1
2
different. In particular, compound
2
1300 cm-1 range peaks shifted in frequency and with different
relative intensity; ii) in the lower frequency range the intensity
increase of the bands at 1010, 784, and 224 cm-1.
Furthermore, solid state 13C NMR spectra were measured on
“as made” and “activated” forms of materials 1 and 2 (Fig. 3).
Figure 2. a) DRIFT spectra for DUT-8(Ni) “flexible” (
), Ni-ndc-dabco, and (H2dabco)(ndc) salt; b) Raman spectra
of DUT-8(Ni) “flexible” ( ) and “rigid” ( ).
1) and “rigid”
Figure 3. Solid state 13C CP NMR spectra of “as made” and
activated form of and
(2
1
2
1
2.
Therefore, the slight shift of the latter to the higher
frequencies could be a sign of a different coordination mode of
carboxylic groups as it could be expected for defective paddle
wheels, formed because of the rapid crystallization.
The spectra of the “as made” form for
1 and 2 are almost
identical and contain two signals between 30 and 40 ppm
corresponding to the carbon atoms of the dabco ligand and
methyl group of DMF. In the range 135 – 165 ppm, three sharp
and one broad signal are observed. These signals are assigned
to the five independent carbon atoms in the naphthalene core
of the ndc2- linker and C-H group of DMF. The broad signal of
quaternary carbon of the ndc2- located near the metal center is
identified in the low field of NMR spectrum at 260 ppm (Inset
Fig. 3).28 Activation of the materials leads to a significant
change in the 13C CP NMR spectra. As it has been shown
The additional absorption band at 785 cm-1 can be assigned to
the νsN-C3 vibration of dabco molecule.36
On the other hand, it is known that dabco is able to form salts
with naphthoic acid where carboxylate anions are hydrogen
bonded to the protonated dabco via COO-/H–N+ hydrogen
bonds.37 Indeed, after mixing of H2ndc (0.294 g, 1.37 mmol)
and dabco (0.336 g, 3.0 mmol) in DMF (30 mL), a white
crystalline precipitate (H2dabco)(ndc)) is formed. A DRIFT
spectrum of the salt contains the absorption band at 785 cm-1,
exactly the same wavenumber observed for the additional
earlier,28 activation of sample
1 leads to the broadening of all
signals, belonging to the framework and disappearing of the
DMF signals. The broadening of the signals could be explained
either by chemical shift distribution of the carbon atoms in the
cp phase or more restricted motions of the linkers in the cp
band of 2 (Fig. 2).
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 5
Please do not adjust margins

Please do not adjust margins
Dalton Transactions
Page 6 of 12
ARTICLE
Journal Name
compared with the op phase. The aforementioned chemical phase. Higher amounts of dabco lead also to ViaewnAritnictleeOnnsliitnye
DOI: 10.1039/C7DT00015D
shift distribution could be caused by the deformation of the increase of “finger print” absorption bands at 785 and 1630
linker molecules as a consequence of structure closing. In the cm-1 in 3c and 3d, characteristic for the spectrum of “rigid”
case of 2, removal of the guest molecules from the framework DUT-8(Ni) (Fig. S15, ESI).
has no influence on the oscillation of the carbon atoms Thus, the dabco amount has a crucial role in the synthesis
showing no peak broadening or anisotropy. At the same time, triggering the formation of small amounts of “flexible” DUT-8
a shift of all signals to the high field was observed and can be
explained by desolvation of the framework, namely the
absence of the ligand-solvent interactions.
(
1
1
) if utilised in lower concentrations. However, pure “flexible”
could not be obtained and very low dabco amounts utilised
in the synthesis, caused a dense phase formation containing
of series 3’, synthesized in
It is interesting to note that the signal at ca. 260 ppm is no dabco. In contrast, samples
b - c
considerably broader in the rigid material than in flexible DUT- DMF:MeOH mixture, are pure cp phase according to the XRD
8(Ni). This might be due to a larger degree of disorder (or (Fig. 3) and DRIFT (Fig. S16, ESI) analyses. Sample 3’d shows
defects) for rigid DUT-8(Ni) in agreement with EPR only a minor impurity of op phase. Thus methanol has an
spectroscopic studies.25 Thus, solid state 13C NMR spectra essential influence on the phase formation.
show an essential difference between “flexible”
DUT-8(Ni) in their activated form as well as absence of further explored in series
1
and “rigid”
2
Therefore, in the next step the role of solvent composition was
4
. Keeping the Ni2+/H2ndc/dabco
additional organic species.
ratio constant (1.1:1:2.2), the MeOH content in DMF was
varied from 1.6% to 36% (MeOH content in the typical
Variation of synthesis parameters
synthesis of flexible
1 is 30%).
The main difference in the synthesis procedure of “flexible”
1
PXRD patterns of the samples 4a and 4b, synthesized using only
and “rigid”
nature of the solvent (or more specifically solvent ratio) (Table MOF. Only if the methanol concentration is higher than 20%, a
1). For the synthesis of the flexible version of the framework, cp op phase mixture is formed. It is also reflected in the
2 is the molar ratio of starting compounds and minor amounts of methanol, show exclusively op phase of the
/
the ratio of the nickel nitrate, H2ndc and dabco of 1:1:0.64 is decreasing intensity of IR absorption bands at 1630 and 785
used, which is very close to the ideal composition of the cm-1 (Fig. S17, ESI).
resulting MOF. As the solvent a DMF/methanol mixture in Finally, in series
volumetric ratio close to 2:1 is used. The synthesis protocol of Ni2+ does not influence the formation of expected phases, and
the “rigid” differs by the increased molar amount of dabco independent of nickel amount utilized (within the investigated
5 and 5’ we could show, that the amount of
2
up to 2.2 equivalents in the synthesis (Table 1). Moreover, the range) the resulting crystalline phase is mostly solvent directed
synthesis is performed in pure DMF at slightly higher (Fig. S1, ESI). Similar to results obtained via PXRD and SEM
temperature.
techniques, the variation of the nickel amount in the synthesis
To find out the critical parameter, which directs the formation causes no changes in the DRIFT spectra (Fig. S18, ESI).
of “flexible” or “rigid” MOF, we performed a wide variation of
synthesis parameters such as molar ratio of the starting Nitrogen and argon physisorption
materials in the system Ni2+/H2ndc/dabco as well as solvent One of the main differences between the “flexible” and “rigid”
mixture composition, starting from original synthesis DUT-8(Ni) materials is the response during nitrogen adsorption
procedure for “flexible”
In the first series, the amount of dabco was varied from 0.20 to technique to distinguish the switchability behaviour of samples
1.47 equivalents (samples 3a 3d), keeping all other synthetic prepared under various conditions as the isotherms differ
1
and “rigid”
2
.
at 77 K and thus physisorption experiments are an ideal
-
parameters constant according to the synthetic procedure for significantly in slope but show a comparable total pore volume
2. In parallel, the syntheses were carried out under conditions (Fig. 1b). For investigation of the pore size distribution in
used for
equivalents (samples 3’a
If a lower amount of dabco is used than needed for ideal volume (Vp) of 0.97 cm3g-1 and BET surface area (SBET) of 2635
stoichiometric composition (in case of samples 3a
3b and 3’a), m2g-1 derived from Ar adsorption isotherm are in perfect
1
and variation of the dabco from 0.25 to 2.14 MOFs, argon physisorption at 87 K was used as it is
-
3’d).
recommended (Fig. S3, ESI).39 The experimental micropore
,
the formation of a non-porous coordination polymer with agreement with values calculated using Poreblazer from the
composition Ni3(ndc)3(DMF)2(Me2NH)2, reported earlier by Du crystal structure of DUT-8(Ni) (Vp = 0.99 cm3g-1, S = 2528 m2g-
et al.38, was observed (Fig. 4a and 4c). It shows characteristic 1). For calculating the pore size distribution NLDFT was applied
reflections at 2Θ = 9.22° and 9.57° in the XRD pattern, as well (zeolite/silica adsorption branch kernel based on a cylindrical
as an additional band at 1666 cm-1 in FTIR spectra, which is pore model). The obtained pore size distribution plot has a
assigned to the coordination mode of the carboxylic group maximum at 8.7 Å. This value is by 1.1 Å smaller than the value
(Fig. S15, S16, ESI).
All activated samples of series
calculated geometrically using Poreblazer software for DUT-
3
represent a mixture of cp and 8(Ni) op structure (Fig. S4 and S5 ESI).40 The largest included
op phases. Furthermore, a decreasing intensity of the main sphere, calculated for the structure, has a diameter of 10.0 Å.
reflection of the cp phase at 2Θ = 7.5° with increasing dabco Given the limitations of the NLDFT kernels, the pore size
content in the synthesis (samples 3c and 3d) indicates the calculated from the adsorption isotherm and crystal structure
substantial role of dabco to form the targeted DUT-8(Ni) are not in excellent but reasonable agreement. We assume
6 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 7 of 12
Dalton Transactions
Please do not adjust margins
Journal Name
ARTICLE
that the desorption branch of the isotherm represents the are different from typical isotherms usually obtained for “gate
View Article Online
thermodynamic equilibrium of the switching process as for pressure” MOFs. In fact, nitrogen adsorption isotherms of
closing no diffusional activation barrier has to be overcome d are a superposition of the isotherms characteristic for
(Fig. 1b inset). In the case of “flexible” material , this “gate 2. Therefore the pore volume, calculated from nitrogen uptake
DOI: 10.1039/C7DT000315aD
and
-
3
1
1
closing” pressure is observed at p/p0 = 0.01, and the structural plateau before “gate opening” can be attributed to the
transformation from op to cp takes place accompanied by the amount of the “rigid” phase in the sample. The ratio between
sharp drop in nitrogen uptake from 588 cm3g-1 at p/p0 = 0.015 this value and total pore volume represents the ratio between
to 40 cm3g-1 at p/p0 = 0.003.
the amounts of “rigid” and flexible phases and is 35:65 for
In contrast, the desorption branch of
2
matches the adsorption samples 3a and 3b and 56:44 for samples 3c and 3d.
branch without any hysteresis, indicating no structural changes As
in the sample during physisorption.
expected,
sample
3’a
of
the
composition
Ni3(ndc)3(DMF)2(Me2NH)2 has no porosity in the crystal
structure and therefore shows nearly no nitrogen uptake (Fig.
Nitrogen adsorption isotherms for samples from the series
3
and 3’ reflect the findings from the XRPD analysis. All samples 4d). The samples 3’b and 3’c show “gate opening” behaviour,
of series
3
show isotherms characteristic for flexible MOFs, but but slightly different uptake in saturation.
a)
b)
c)
d)
e)
f)
Figure 4. PXRD patterns of activated samples and nitrogen adsorption isotherms at 77 K for samples 3 - 4.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 7
Please do not adjust margins

Please do not adjust margins
Dalton Transactions
Page 8 of 12
ARTICLE
Journal Name
2, show aVipewarAtrtiicclleeOsniliznee
Further increasing of dabco amount in the synthesis of 3’d images obtained for DUT-8(Ni) “rigid”
leads to a significant drop in the total pore volume and magnitude close to the resolution limit oDfOthI:e10S.1E0M39/iCns7DtrTu0m00e1n5Dt.
changes the isotherm slope in the gate region (Fig. 4d). In general, the poor electrical conductivity of MOFs is
Interestingly, the “gate opening” pressure is influenced by the disadvantageous for high-resolution SEM studies.
synthetic conditions and is shifted to higher relative pressure Nevertheless, a rough estimation results in an average particle
from p/p0 = 0.13 for to p/p0 = 0.16 in the 3d and to p/p0 = size of 200 - 500 nm (Fig. 5).
1
0.28 for the samples 3a, 3b, 3c and 3’c. An interesting In the samples 3a and 3b, where an additional non-porous
observation is that significant differences in the “gate opening” phase of Ni3(ndc)3(DMF)2(Me2NH)2 is observed in PXRD, SEM
pressure are observed (for example for 3’b and 3’c) while the images contain beside the typical bricks of DUT-8(Ni) also
“gate closing” pressure is essentially the same (Fig. S2b, ESI). some spherical particles, which may be attributed to
These differences are representative for cooperative switching Ni3(ndc)3(DMF)2(Me2NH)2 (Fig. S10, ESI). Images of 3a – 3d show
in single crystallites (or agglomerates) within the particle a broad particle size distribution for DUT-8(Ni) (Fig. S10, ESI).
ensemble (see also particle size analysis) and indicate their Particles larger than 50 µm as well as very small particles (< 1
strong influence of the activation barrier for opening (kinetics) µm) can be observed in the images. This observation is
while the closing energetics (equilibrium) are not affected (Fig. consistent with the results from PXRDs and isotherms showing
S2a, S2b, ESI). A steep slope at “gate opening” is indicative for the presence of both cp and op phases and 2 steps in the
a particle ensemble with relatively uniform activation barrier, nitrogen adsorption isotherms, if one assumes that only
while
a broadening indicates a broader distribution of crystals exceeding a critical size may lead to switchability. The
activation barriers. Similar phenomena were observed recently broadened “gate pressure” in the isotherms of 3a - 3d, in this
by repeated adsorption/desorption of n-butane on flexible interpretation, can be considered as originating from a broad
DUT-8(Ni) and was attributed to the dislocations and lattice particle size distribution in these samples. As for many
strain within the crystallites.27
cooperative phenomena, it is reasonable to expect that the
According to XRD analysis, samples 4a and 4b (prepared using activation energy for “gate opening” is related to the crystallite
the reported synthetic procedure for “rigid” DUT-8, but using size. In this interpretation, depending on the crystallite size,
various DMF:MeOH ratios) represent the op phase. In sample the crystallites open from cp to op at different pressure while
4c additionally a minor fraction of cp phase is detected. the closing pressure is not affected. However, this
However, the nitrogen adsorption experiments reveal an interpretation is oversimplified, since the first guest removal
unexpected adsorption behaviour: An increasing methanol (activation) process causes larger crystals to transform into an
content in the synthesis leads to a gradual decrease of the intertwined domain structure in which the domain size and the
porosity and total pore volume. Sample 4c shows also a broad grain boundary energy determines the activation energy. As
hysteresis in the whole pressure range (Fig. 4f), indicating the the solvothermal synthesis results in the formation of the op
presence of the “flexible” phase.
phase, the desolvation procedure is decisive in affecting the
In contrast, the variation of the amount of Ni2+ in the synthesis microstructure. Due to the huge crystal contraction during
has only minor influence on both products, and the nitrogen solvent removal, larger crystals disintegrate into a complex
adsorption isotherm, measured on the sample 5´a, synthesized microstructure with grain boundaries of unknown size creating
using the procedure for “flexible” DUT-8(Ni) but only 50% of an interfacial domain area Aid as explained above. A uniform
the Ni2+ in the synthesis, reaches the same total pore volume microstructure with grains of similar size will afford a steep
in saturation, but shows a significantly broadened “gate gate opening characteristic, as the activation energy for
opening” (Fig. S1b, ESI). The variation of the nickel amount in switching the microdomains is identical. We expect op crystals
the procedure for “rigid” DUT-8(Ni) influences only the total differing in size to afford differently sized microdomain
pore volume of resulting materials 5a - 5b (Fig. S1d, ESI), but structures upon desolvation into the cp phase, and hence
very little the shape of the isotherm.
altered activation energies corresponding to altered gate
opening pressures. We hypothesize grain boundaries to act as
the nucleation sites for the gating transition as they may act as
Particle size
Recently the influence of crystallite size on adsorption induced diffusional entries for the gas. Thus, for smaller crystals fewer
phase transitions has been recognized.9, 33, 41 Kitagawa and co- number of grain boundaries and defect statistics could in
workers reported that downsizing of twofold interpenetrated principle even shift the gate opening pressure so far, that the
frameworks of Cu2(dicarboxylate)2(amine) regulates the crystals remain closed in the whole pressure range. This
structural flexibility and may induce shape-memory effects in hypothesis could explain the significantly reduced maximum
coordination frameworks.33 The downsizing of the ZIF-8 uptake in some samples with intermediate particle size. If a
crystallites was shown to suppress ligand flip during crystal approaches a certain lower boundary, we can expect a
adsorption.34 Consequently, an important aspect was to single domain switching mechanism without fracture into
analyze the particle size of the products obtained in the mirodomains similar to other ferroic size phenomena. Below
syntheses under varying conditions.
SEM images of DUT-8(Ni) “flexible”
this critical size only Ao will determine the gate opening
show relatively large pressure. However, crystals varying in size may also differ in
1
crystallites with a size ranging from 20 to 100 µm, which is fully point defect concentration and small variations in defect
consistent with our previous study (Fig. 5).27 In contrast, concentration can also change the activation barrier.
8 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 9 of 12
Dalton Transactions
Please do not adjust margins
Journal Name
ARTICLE
phases and nitrogen adsorption isotherms with broadened
View Article Online
DOI: 10.1039/C7DT00015D
gating transitions. The gradual increase of the methanol
amount in the mixture in the sample series
4 shows a very
similar effect. The sample 4a synthesized with lowest methanol
content consists of the crystallites with maximal size of 500 nm
and therefore shows pure “Type Ia” isotherm and the op
phase in the PXRD. With increasing methanol amount in the
synthesis, the particle size distribution is broadened. In sample
4c the particle size is ranging from 500 nm to 10 µm (Fig. S12,
ESI) also reflecting well the nitrogen adsorption and PXRD
experiments (Fig. 4e-f). As expected from the isotherms and
XRD measurements, the variation of the nickel amount in the
synthesis does not influence the particle size in samples 5´a
and 5´b (Fig. S13, ESI).
Temperature dependent X-ray diffraction study
Kitagawa and co-worker reported, that for interpenetrated pillared-
Figure 5. SEM images: flexible DUT-8(Ni) (1) (top) and rigid DUT-
8(Ni) (2) (bottom).
layer compound Cu2(bdc)2(bpy) (bdc = 1,4-benzenedicarboxylate,
bpy = 4,4′-bipyridine) the pore shape could be switched from the
open to the closed form by heating. Thus, temperature dependent
X-ray diffraction experiments were conducted on 2 (Fig. 6, Fig. S19)
to investigate the behavior of DUT-8 during heating. However, the
rigid DUT-8 remains in the open form until decomposition and
therefore is not temperature responsive.
In our observations, below a critical size, which for DUT-8(Ni)
we estimate based on this study to be approximately 500 nm,
the material does not transform into the closed form by the
guest removal (activation), and an isotherm typical for rigid
microporous materials is observed. The underlying principle
for this observation must rely on
a thermodynamic
stabilization of the op vs. cp phase as this phenomenon is
observed in vacuum without any gases. A possible explanation
relies on the EPR data, and the recently identified high defect
concentrations25 could arise from rapid nucleation and growth.
Especially missing ndc ligands are expected to reduce the
dispersive energetic stabilization of DUT-8(Ni) cp. Another
potential explanation could be differences in surface energy
390 °C
370 °C
350 °C
330 °C
310 °C
290 °C
270 °C
250 °C
230 °C
210 °C
190 °C
170 °C
150 °C
130 °C
110 °C
90 °C
related to Ao. However, as AO is larger for the op phase (
∆Ao =
Ao(op) - Ao(cp) > 0), this explanation would imply the specific
surface energy difference as ∆σ
= σ(op) - σ(cp) < 0 which is
difficult to justify.
Zhang et al. proposed an explanation for the
thermodynamically controlled crystal-size-dependent
70 °C
50 °C
25 °C
4
6
8
10
12
14
16
18
20
22
24
structural transitions in MOFs in the presence of gases. The
calculations showed, that the suppressed adsorption uptake at
the surface of nanoparticles compared to that at the bulk-like
region results in a reduced thermodynamic driving force for
the transition in the nanoparticle over the entire range of bulk
fluid pressures.34
2ϴ/°
Figure 6. Temperature dependent X-ray diffraction patterns of 2
heated in vacuum.
Conclusions
Form an experimental point of view, the overall trend is
consistent for all synthesis parameters. In the sample series 3’
,
Significant differences in the synthesis pathway were found to
be responsible for the formation of “rigid” and “flexible” DUT-
8(Ni) under different synthesis conditions. The high
supersaturation in the synthesis of “rigid” DUT-8(Ni) causes
the formation of intermediate phases and rapid nucleation,
leading to small and defective crystallites showing typical type
I physisorption isotherms. In contrast, the higher solubility
provided in the crystallization of “flexible” DUT-8(Ni) and slow
crystallization causes the formation of large crystals that can
the changes of the dabco amount also strongly affect the
crystallite size. The SEM images sample 3’a show block-like
overgrown crystallites, which according to the XRD should
belong to the Ni3(ndc)3(DMF)2(Me2NH)2 phase (Fig. S11, ESI).
Obviously, the use of methanol in the synthesis creates the
best conditions for growing larger crystals. Increased amount
of dabco in the samples 3’c and 3’d leads to a broadening of
the particle size distribution as it can be directly observed in
the images showing crystallites ranging from several hundred
nanometer to hundred micrometre size. Again, these effects
are indirectly reflected in the PXRD showing mixtures of two
undergo
a transformation into a closed phase upon
desolvation. Samples with crystallites smaller than 500 nm do
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 9
Please do not adjust margins

Please do not adjust margins
Dalton Transactions
Page 10 of 12
ARTICLE
Journal Name
not show the op → cp phase transformation and therefore 4. F.-X. Coudert, Chem. Mater., 2015, 27, 1905-1916.
View Article Online
stay in the open form after the guest molecules removal, 5. A. Schneemann, V. Bon, I. Schwedler,DI.OSI:e1n0k.1o0v3s9k/Ca,7DST. 0K0a0s1k5eDl
whereas particles larger than
transformed into the cp phase.
1
µm can be reversibly and R. A. Fischer, Chem. Soc. Rev., 2014, 43, 6062-6096.
6. A. Kondo, H. Noguchi, S. Ohnishi, H. Kajiro, A. Tohdoh, Y.
For “flexible” DUT-8(Ni), varied synthesis parameters Hattori, W.-C. Xu, H. Tanaka, H. Kanoh and K. Kaneko, Nano
influencing the crystal size can be used to tune the “gate Lett., 2006, , 2581-2584.
6
opening” pressure reflecting the activation barrier of the 7. T. R. C. Van Assche, G. V. Baron and J. F. M. Denayer, Dalton
phase transition, while the “gate closing” pressure remains Trans., 2016, 45, 4416-4430.
constant.
8. T. Song, J. Yu, Y. Cui, Y. Yang and G. Qian, Dalton Trans.,
The reason for a strong effect of synthesis parameters lies in 2016, 45, 4218-4223.
the differences of the crystallization mechanism: An excess 9. O. M. Linder-Patton, W. M. Bloch, C. J. Coghlan, K. Sumida, S.
amount of the dabco in the synthesis of the “rigid” DUT-8(Ni) Kitagawa, S. Furukawa, C. J. Doonan and C. J. Sumby,
is partially consumed by deprotonation of the H2ndc ligand CrystEngComm, 2016, 18, 4172-4179.
resulting in the absence of metal source and the precipitation 10. M. L. Foo, R. Matsuda, Y. Hijikata, R. Krishna, H. Sato, S.
of a (H2dabco)(ndc) salt. In the presence of Ni2+ cations, rapid Horike, A. Hori, J. Duan, Y. Sato, Y. Kubota, M. Takata and S.
nucleation of the defective DUT-8(Ni) structure proceeds in Kitagawa, J. Am. Chem. Soc., 2016, 138, 3022-3030.
the form of
a cloudy microcrystalline precipitate. High 11. J. A. Mason, J. Oktawiec, M. K. Taylor, M. R. Hudson, J.
concentration of defects in this phase was recently confirmed Rodriguez, J. E. Bachman, M. I. Gonzalez, A. Cervellino, A.
by EPR in the presence of NO. Additional bands in the IR- Guagliardi, C. M. Brown, P. L. Llewellyn, N. Masciocchi and J. R.
spectrum support these findings. Thus particle size and defects Long, Nature, 2015, 527, 357-361.
are both decisive factors influencing flexibility. In contrast, the 12. H. Sato, W. Kosaka, R. Matsuda, A. Hori, Y. Hijikata, R. V.
use of stoichiometric ratio of the initial reagents in the Belosludov, S. Sakaki, M. Takata and S. Kitagawa, Science, 2014,
crystallization of “flexible” DUT-8(Ni) in DMF/methanol 343, 167-170.
mixture inhibits complete deprotonation of H2ndc and 13. R. Matsuda, Nature, 2014, 509, 434-435.
therefore slows down the nucleation and crystallization of the 14. P. Serra-Crespo, M. A. van der Veen, E. Gobechiya, K.
target compound. It causes the formation of large crystals (20 Houthoofd, Y. Filinchuk, C. E. A. Kirschhock, J. A. Martens, B. F.
to 100 µm). These large crystals undergo a transformation into Sels, D. E. De Vos, F. Kapteijn and J. Gascon, J. Am. Chem. Soc.,
a closed phase upon desolvation.
2012, 134, 8314-8317.
An important implication of our findings is that switchable 15. N. Yanai, K. Kitayama, Y. Hijikata, H. Sato, R. Matsuda, Y.
MOFs, especially those showing pronounced volume changes, Kubota, M. Takata, M. Mizuno, T. Uemura and S. Kitagawa, Nat.
should always be considered as materials with “history Mater., 2011, 10, 787-793.
dependent adsorption profiles”, because each transition 16. K. Nakagawa, D. Tanaka, S. Horike, S. Shimomura, M. Higuchi
undergone may cause changes in domain size and defect and S. Kitagawa, Chem. Commun., 2010, 46, 4258-4260.
formation which in turn affects the adsorption behaviour. We 17. F.-X. Coudert, A. Boutin, A. H. Fuchs and A. V. Neimark, J.
believe our findings to be an important step towards a Phys. Chem. Lett., 2013,
4, 3198-3205.
fundamental understanding of adsorption induced switching 18. H. Tanaka, S. Ohsaki, S. Hiraide, D. Yamamoto, S. Watanabe
transitions of porous solids.
and M. T. Miyahara, J. Phys. Chem. C, 2014, 118, 8445-8454.
19. D. Li and K. Kaneko, Chem. Phys. Lett., 2001, 335, 50-56.
20. V. Bon, I. Senkovska, D. Wallacher, A. Heerwig, N. Klein, I.
Zizak, R. Feyerherm, E. Dudzik and S. Kaskel, Microporous
Mesoporous Mater., 2014, 188, 190-195.
21. Y. Cheng, H. Kajiro, H. Noguchi, A. Kondo, T. Ohba, Y. Hattori,
K. Kaneko and H. Kanoh, Langmuir, 2011, 27, 6905-6909.
22. H. Kanoh, A. Kondo, H. Noguchi, H. Kajiro, A. Tohdoh, Y.
Hattori, W.-C. Xu, M. Inoue, T. Sugiura, K. Morita, H. Tanaka, T.
Ohba and K. Kaneko, J. Colloid Interface Sci., 2009, 334, 1-7.
23. S. Hiraide, H. Tanaka and M. T. Miyahara, Dalton Trans.,
2016, 45, 4193-4202.
Acknowledgements
Financial support by the “excellence initiative by German federal
and state government” (measure “support the best”) and DFG (FOR
2433) is gratefully acknowledged. V.B. acknowledges German
Federal Ministry of Education and Research (BMBF Project No
05K16OD1) for the financial support. Authors are very grateful to
Prof. S. Bordiga for the valuable discussions of the IR and Raman
spectra.
24. D. N. Dybtsev, H. Chun and K. Kim, Angew. Chem. Int. Ed.,
2004, 43, 5033-5036.
References
25. M. Mendt, F. Gutt, N. Kavoosi, V. Bon, I. Senkovska, S. Kaskel
and A. Pöppl, J. Phys. Chem. C, 2016, 120, 14246-14259.
26. V. Bon, N. Klein, I. Senkovska, A. Heerwig, J. Getzschmann, D.
Wallacher, I. Zizak, M. Brzhezinskaya, U. Mueller and S. Kaskel,
Phys. Chem. Chem. Phys., 2015, 17, 17471-17479.
1. S. Horike, S. Shimomura and S. Kitagawa, Nat. Chem., 2009,
1
, 695-704.
2. G. Ferey and C. Serre, Chem. Soc. Rev., 2009, 38, 1380-1399.
3. Z. Chang, D.-H. Yang, J. Xu, T.-L. Hu and X.-H. Bu, Adv. Mater.,
2015, 27, 5432-5441.
27. V. Bon, N. Kavoosi, I. Senkovska and S. Kaskel, ACS Appl.
Mater. Interfaces, 2015, 7, 22292-22300.
10 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 11 of 12
Dalton Transactions
Please do not adjust margins
Journal Name
ARTICLE
28. N. Klein, H. C. Hoffmann, A. Cadiau, J. Getzschmann, M. R.
Lohe, S. Paasch, T. Heydenreich, K. Adil, I. Senkovska, E. Brunner
and S. Kaskel, J. Mater. Chem., 2012, 22, 10303-10312.
29. N. Klein, C. Herzog, M. Sabo, I. Senkovska, J. Getzschmann, S.
Paasch, M. R. Lohe, E. Brunner and S. Kaskel, Phys. Chem. Chem.
Phys., 2010, 12, 11778-11784.
View Article Online
DOI: 10.1039/C7DT00015D
30. K. Trepte, S. Schwalbe and G. Seifert, Phys. Chem. Chem.
Phys., 2016, 18, 1348.
31. H. C. Hoffmann, B. Assfour, F. Epperlein, N. Klein, S. Paasch, I.
Senkovska, S. Kaskel, G. Seifert and E. Brunner, J. Am. Chem.
Soc., 2011, 133, 8681-8690.
32. J. Y. Lee, L. Pan, X. Huang, T. J. Emge and J. Li, Adv. Funct.
Mater., 2011, 21, 993-998.
33. Y. Sakata, S. Furukawa, M. Kondo, K. Hirai, N. Horike, Y.
Takashima, H. Uehara, N. Louvain, M. Meilikhov, T. Tsuruoka, S.
Isoda, W. Kosaka, O. Sakata and S. Kitagawa, Science, 2013, 339
,
193-196.
34. C. Zhang, J. A. Gee, D. S. Sholl and R. P. Lively, J. Phys. Chem.
C, 2014, 118, 20727-20733.
35. S. Watanabe, S. Ohsaki, T. Hanafusa, K. Takada, H. Tanaka, K.
Mae and M. T. Miyahara, Chem. Eng. J., 2017, 313, 724-733.
36. D. A. Guzonas and D. E. Irish, Can. J. Chem., 1988, 66, 1249-
1257.
37. A. Jacobs, L. R. Nassimbeni, G. Ramon and B. K. Sebogisi,
CrystEngComm, 2010, 12, 3065-3070.
38. Y. Du, A. L. Thompson, N. Russell and D. O'Hare, Dalton
Trans., 2010, 39, 3384-3395.
39. M. Thommes, K. Kaneko, A. V. Neimark, J. P. Olivier, F.
Rodriguez-Reinoso, J. Rouquerol and K. S. W. Sing, Pure Appl.
Chem., 2015, 87, 1051-1069.
40. L. Sarkisov and A. Harrison, Mol. Simul., 2011, 37, 1248-1257.
41. S. Tanaka, K. Fujita, Y. Miyake, M. Miyamoto, Y. Hasegawa, T.
Makino, S. Van der Perre, J. C. Saint Remi, T. Van Assche, G. V.
Baron and J. F. M. Denayer, J. Phys. Chem. C, 2015, 119, 28430-
28439.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 11
Please do not adjust margins

Dalton Transactions
Page 12 of 12
View Article Online
DOI: 10.1039/C7DT00015D
226x120mm (230 x 230 DPI)