Screening Hofmann compounds as CO2 sorbents: Nontraditional synthetic route to over 40 different pore-functionalized and flexible pillared cyanonickelates

Article abstract of DOI:10.1021/ic301893p

A simple reaction scheme based on the heterogeneous intercalation of pillaring ligands (HIPLs) provides a convenient method for systematically tuning pore size, pore functionality, and network flexibility in an extended series of pillared cyanonickelates (PICNICs), commonly referred to as Hofmann compounds. The versatility of the approach is demonstrated through the preparation of over 40 different PICNICs containing pillar ligands ranging from ～4 to ～15 A in length and modified with a wide range of functional groups, including fluoro, aldehyde, alkylamine, alkyl, aryl, trifluoromethyl, ester, nitro, ether, and nonmetalated 4,4′-bipyrimidine. The HIPL method involves reaction of a suspension of preformed polymeric sheets of powdered anhydrous nickel cyanide with an appropriate pillar ligand in refluxing organic solvent, resulting in the conversion of the planar [Ni₂(CN)₄] _n networks into polycrystalline three-dimensional porous frameworks containing the organic pillar ligand. Preliminary investigations indicate that the HIPL reaction is also amenable to forming Co(L)Ni(CN)₄, Fe(L)Ni(CN)₄, and Fe(L)Pd(CN)₄ networks. The materials show variable adsorption behavior for CO₂ depending on the pillar length and pillar functionalization. Several compounds show structurally flexible behavior during the adsorption and desorption of CO₂. Interestingly, the newly discovered flexible compounds include two flexible Fe(L)Ni(CN)₄ derivatives that are structurally related to previously reported porous spin-crossover compounds. The preparations of 20 pillar ligands based on ring-functionalized 4,4′-dipyridyls, 1,4-bis(4-pyridyl)benzenes, and N-(4-pyridyl)isonicotinamides are also described.

Full text of DOI:10.1021/ic301893p

Article
pubs.acs.org/IC
Screening Hofmann Compounds as CO2 Sorbents: Nontraditional
Synthetic Route to Over 40 Different Pore-Functionalized and
Flexible Pillared Cyanonickelates
Jeffrey T. Culp,*,†,‡ Catherine Madden,† Kristi Kauffman,† Fan Shi,†,‡ and Christopher Matranga†
†National Energy Technology Laboratory, Pittsburgh, Pennsylvania 15236, United States
‡URS Corporation, South Park, Pennsylvania 15129, United States
S
* Supporting Information
ABSTRACT: A simple reaction scheme based on the heterogeneous intercalation of pillaring ligands (HIPLs) provides a
convenient method for systematically tuning pore size, pore functionality, and network flexibility in an extended series of pillared
cyanonickelates (PICNICs), commonly referred to as Hofmann compounds. The versatility of the approach is demonstrated
through the preparation of over 40 different PICNICs containing pillar ligands ranging from ∼4 to ∼15 Å in length and modified
with a wide range of functional groups, including fluoro, aldehyde, alkylamine, alkyl, aryl, trifluoromethyl, ester, nitro, ether, and
nonmetalated 4,4′-bipyrimidine. The HIPL method involves reaction of a suspension of preformed polymeric sheets of powdered
anhydrous nickel cyanide with an appropriate pillar ligand in refluxing organic solvent, resulting in the conversion of the planar
[Ni2(CN)4]n networks into polycrystalline three-dimensional porous frameworks containing the organic pillar ligand. Preliminary
investigations indicate that the HIPL reaction is also amenable to forming Co(L)Ni(CN)4, Fe(L)Ni(CN)4, and Fe(L)Pd(CN)4
networks. The materials show variable adsorption behavior for CO2 depending on the pillar length and pillar functionalization.
Several compounds show structurally flexible behavior during the adsorption and desorption of CO2. Interestingly, the newly
discovered flexible compounds include two flexible Fe(L)Ni(CN)4 derivatives that are structurally related to previously reported
porous spin-crossover compounds. The preparations of 20 pillar ligands based on ring-functionalized 4,4′-dipyridyls, 1,4-bis(4-
pyridyl)benzenes, and N-(4-pyridyl)isonicotinamides are also described.
The exceptional design flexibility of these materials is
exemplified in a few well-documented cases such as the metal
organic frameworks (MOFs),7 zeolite imidazolate frameworks
INTRODUCTION
■
Carbon dioxide separations using solid sorbents will require the
development of porous materials with high affinities and high
selectivities for CO2 adsorption. Driven in part by this goal,
research on porous coordination polymers (PCPs), commonly
referred to as metal organic frameworks,1,2 has become
exceedingly popular over the past 15 years.3−8 The building
block approach to the design and synthesis of these materials
provides a significant level of control over the pore chemistries,
pore sizes, and surface areas of these materials. The
development of synthetic routes toward materials with
functional groups within the pore structure to enhance both
the adsorption potential and selectivity of a particular guest has
become an active area of research, especially in the field of CO2
capture and separation. The approach typically involves either
the use of ligands containing the desired functionality during
the assembly of the PCP9−12 or postsynthetic modification after
the PCP has formed.13−18
́
(ZIFs),19 and Materiaux de l’Institut Lavoisier (MIL) series,10
where a particular synthetic strategy or network topology has
been found that allows multiple substitutions of bridging
ligands and/or metal ions to form a structurally homologous
series of materials from one general synthetic procedure. These
modular approaches to material design offer great benefits for
systematically designing a series of materials where the effects
of pore modifications, including pore size and chemical
functionalization, can be correlated with sorbent performance.
Another commonly employed modular design strategy for
the preparation of PCPs involves the pillared-layer motif
wherein an extended two-dimensional network is built into a
three-dimensional solid through the incorporation of pillar
Received: August 29, 2012
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ic301893p | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
ligands which link one layer to the next.12,20−23 Polymeric
cyano-bridged networks containing d8 tetracyano complexes
(commonly referred to as Hofmann24 compounds) with their
inherent planar M[M′(CN)4]n networks are well-suited for this
pillared-layer design strategy.25−34 Because of their linear
bridging abilities, dipyridyl derivatives such as 4,4′-bipyridine,
1,4-bis(4-pyridyl)benzenes, and N-(4-pyridyl)isonicotinamides
are excellent candidates as potential pillar ligands in these
systems. As cyano-bridged networks, these pillared Hofmann
materials are also closely related to the widely studied family of
Prussian blue based porous sorbents.35−45
structure change and guest adsorption in these materials is
driven by a complicated interplay of energetics describing the
phase transition, stabilizing effects of guest adsorption, relief of
mechanical strain in the crystal, and nucleation effects.
Understanding this interplay and being able to selectively
tune dynamic behavior in a sorbent is a key step in the rational
design of new separation materials for both liquid and gas
applications. The ability to systematically tune this dynamic
behavior in the PICNIC structure makes them an exceptional
platform for advancing our understanding of these materials.
Because of its commercial availability and ability to adopt a
convenient rigid linear bridging arrangement, 4,4′-bipyridine
(4,4′-Bpy) has steadily gained popularity as a versatile ligand in
the field of coordination polymers.46 It is somewhat surprising,
considering the growing interest in pore-functionalized
coordination polymers, that more materials containing
functionalized versions of this common ligand are not reported.
In addition, the 4,4′-bipyridine molecule is well-suited to
perform as a functionalized ligand since substitution can occur
at the 3-position and not interfere with coordination of the ring
nitrogen. These 3-functionalized derivatives of 4,4′-bipyridine,
however, are not readily available through commercial sources,
and few direct literature preparations are available. Recent
advances in Suzuki coupling reactions and the commercial
availability of many of the necessary starting reagents and
palladium catalysts have made these compounds much more
accessible.47 Since these linear dipyridyl compounds are very
attractive as potential pillar ligands, we developed a synthetic
route to several 3-R-4,4′-bipyridines along with a number of
structurally related and ring-functionalized 1,4-bis(4-pyridyl)-
benzenes and N-(4-pyridyl)isonicotinamides.
EXPERIMENTAL SECTION
■
All starting materials were purchased from commercial sources in at
least reagent grade purity and used as received. Tricyclohexylphos-
phine and tris(dibenzylideneacetone)dipalladium(0) are abbreviated as
P(Cy)3 and Pd2(DBA)3, respectively. Anhydrous 1,4-dioxane,
acetonitrile, ethanol, and toluene were purchased from commercial
sources and transferred under N2 using a gastight syringe. Water was
purified using a Barnstead EASYpure LF water purification system and
degassed by refluxing under N2 3−4 h prior to use. All reactions were
performed under N2 using standard Schlenk line techniques. The pillar
ligand 4,4′-azopyridine (azopyr) is commercially available from Sigma-
Aldrich. The pillar ligand 2,6-naphthyridine was purchased from the
Florida Center for Heterocyclic Compounds and used as received. The
pillar ligand 4,4′-bipyrimidine (4,4′-Bpm) was purchased from
Oakwood Products, Inc. and used as received. The pillar ligands 1,2-
bis(4-pyridyl)acetylene (dpac),55 1,4-bis(4-pyridyl)buta-1,3-diyne
(DPBD),56,57 1,4-bis(4′-pyridylethynyl)benzene (DPAC-Bz),58 and
N-(4-pyridyl)isonicotinamide59 were prepared by literature methods.
The CHN analyses were provided by Robertson Microlit Laboratories,
Inc., Ledgewood, NJ. The NMR measurements were done using a
Bruker 300 MHz proton NMR spectrometer. Melting points were
determined using a Mettler STARe TGA/DSC thermogravimetric
analyzer on samples of 4−6 mg loaded in 100 μL aluminum pans with
a temperature ramp of 5 °C/min under N2.
These ligands were found to work well as pillars in the
synthesis of three-dimensional Hofmann compounds which we
refer to in the following discussion as pillared cyanonickelates,
or PICNICs. These materials are conveniently synthesized in
bulk with near-quantitative yields through a heterogeneous
intercalation of pillaring ligand (HIPL) reaction, resulting in the
expansion of a polymeric quasi-two-dimensional nickel−
cyanide network into covalently linked pillared porous three-
dimensional structures. The method is found to be highly
versatile with nearly three dozen pillar ligands successfully
incorporated into the PICNIC structure, including several
newly reported ring-functionalized bipyridyl, dipyridylamide,
and dipyridylbenzene ligands. The ligands can be varied in both
length and functionalization to give a wide range of porous
materials all derived from the same structural motif. In this
sense, PICNICs can be viewed as a novel family of porous
materials with readily tunable pore properties akin to other
well-known tailorable systems such as MOFs, ZIFs, and MILs.
A previous report revealed that the PICNIC compound
Ni(1,2-bis(4-pyridyl)ethylene)Ni(CN)4 shows structurally dy-
namic behavior during the adsorption and desorption of gases,
including CO2 and N2.28 This flexibility appears to be a
common trait in many other PICNIC materials as several new
flexible members in the family, including flexible Co(L)Ni-
(CN)4 and Fe(L)Ni(CN)4 derivatives, have been discovered
and are reported herein. The CO2-induced structural flexibility
in the Fe(L)Ni(CN)4 compounds is especially interesting
considering the spin-crossover behavior reported for several
analogous Fe(L)M(CN)4 compounds.25,26,30,31,48 Flexible
PCPs (or flexible MOFs) are an interesting class of materials
from both a fundamental and a practical standpoint.28,49−54 The
Synthesis of Pillar Ligands. Functionalized 4,4′-Bipyridines
(Bpy-R). The preparation of 3-R-4,4′-bipyridines was carried out using
a Suzuki coupling reaction between an excess of 1 equiv of pyridine-4-
boronic acid and a 3-R-4-halopyridine where the halo group can be
chloro or bromo or by reaction of 4-bromopyridine with a 3-R-
pyridine-4-boronic acid. The reaction procedure was borrowed
extensively from a method reported by Kudo et al. for the coupling
of pyridineboronic acids.47 All compounds were prepared by the same
general method and purified by recrystallization from either hexanes or
ethyl acetate. The detailed preparation of 3-amino-4,4′-bipyridine is
used as an example. Synthetic details for the remaining Bpy-R
compounds are included in the Supporting Information.
3-Amino-4,4′-bipyridine (Bpy-NH2). A three-necked flask was
charged with 1.9 g (15 mmol) of 3-amino-4-chloropyridine, 2.7 g
(22 mmol) of pyridine-4-boronic acid, 225 mg (0.28 mmol) of
tris(dibenzylideneacetone)dipalladium(0) (Pd2(DBA)3), and 180 mg
(0.64 mmol) of tricyclohexylphosphine (P(Cy)3) and purged with N2.
The mixture was suspended in 40 mL of deoxygenated 1,4-dioxane
(commercial anhydrous grade packaged under Ar). A solution of 8 g
(38 mmol) of K3PO4 in 25 mL of deoxygenated water was added by
syringe through a septum. The flask was added to an oil bath at 60 °C
and slowly heated to reflux (100 °C). The mixture was held at reflux
with rapid stirring overnight (∼18 h). Upon being cooled to room
temperature, the mixture was poured into a separatory funnel and the
lower aqueous phase removed and discarded. The dioxane layer was
collected and filtered and the dioxane removed under reduced
pressure. The residue was triturated in 75 mL of ethyl acetate, treated
with activated carbon and anhydrous magnesium sulfate, and then
heated to reflux for 10 min with stirring. The mixture was filtered hot
through a fine glass frit. The filtrate was concentrated and cooled to
yield 2.1 g (12.3 mmol) of the tan crystalline product. Mp: 170 °C
(lit.60 166−168 °C). H NMR (300 MHz, CDCl3): δ 8.75, m, 2H;
1
8.21 s, 1H; 8.11, d, 1H; 7.43, m, 2H; 7.03, d, 1H; 3.85, s, 2H. Anal.
B
dx.doi.org/10.1021/ic301893p | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Chart 1. Structures and Abbreviations for Pillar Ligands (L) Used To Assemble PICNIC Coordination Polymers as
(Ni(L)[Ni(CN)4])n
Calcd for C10H9N3 (171.2): C, 70.16; H, 5.30; N, 24.54. Found: C,
69.87; H, 4.99; N, 24.07.
Functionalized N-(4-Pyridyl)isonicotinamides (PINA-R). The syn-
thesis of the PINA-R compounds was accomplished by using a
modification of a previously reported method for the preparation of N-
(4-pyridyl)isonicotinamide59 wherein isonicotinoyl chloride hydro-
chloride was reacted with 1 equiv of the appropriate 3-R-4-
aminopyridine at 80 °C in dry pyridine overnight. The products
were recrystallized from acetone/water. The synthesis of PINA-F is
given as an example. Synthetic procedures for the remaining DPBz-R
compounds are included in the Supporting Information.
N-(3-Fluoro-4-pyridyl)isonicotinamide (PINA-F). A solution of 1.03
g (0.0092 mol) of 4-amino-3-fluoropyridine in 25 mL of dry pyridine
was added to a mixture of 1.60 g (0.009 mol) of isonicotinoyl chloride
hydrochloride in 25 mL of dry pyridine. The flasks were then
transferred into a hot oil bath and reacted at 80 °C under N2 for 18 h.
The mixture was then concentrated to a few milliliters under reduced
pressure, and 50 mL of a 5% aqueous solution of sodium bicarbonate
was added. The precipitate was collected by filtration, recrystallized
from acetone/water, and dried to yield 1.61g (0.0074 mol) of a light
beige solid. Mp: 180 °C. 1H NMR (300 MHz, CDCl3): δ 8.86, d, 2H;
8.47, m, 3H; 8.37, s, 1H; 7.73, d, 2H. Anal. Calcd for C11H8FN3O
(217.2): C, 60.83; H, 3.71; N, 19.35. Found: C, 60.66; H, 3.38; N,
19.25.
Synthesis of PICNICs. Nickel Compounds. The pillared
cyanonickelate compounds Ni(L)[Ni(CN)4], where L is a pillar
ligand, were prepared by adaptation of a previously reported
method28,61 by reaction of 1.05:1 molar ratios of pillar ligands to
anhydrous Ni2(CN)4 (prepared by drying the hydrate under vacuum
at 130 °C for 2−3 h) in dry toluene or dry acetonitrile/dry toluene
mixtures at reflux under N2. For pillar ligands in the PINA-R series,
reactions were done in dry acetonitrile due to low solubility of the
ligands in toluene (note that addition of the dry acetonitrile to
anhydrous Ni2(CN)4 is mildly exothermic and results in an immediate
color change of the Ni2(CN)4 due to coordination of the acetonitrile
Functionalized 1,4-Bis(4-pyridyl)-3-R-benzenes (DPBz-R). The
preparation of 1,4-bis(4-pyridyl)-3-R-benzenes was carried out using
a Suzuki coupling reaction similar to the method described above for
the preparation of 3-R-4,4′-bipyridines by reaction of an excess of
∼2.5−3.0 equiv of 4-pyridineboronic acid with 1 equiv of a 3-R-1,4-
dihalobenzene, where halo is chloro, bromo, or iodo. The detailed
preparation of 1,4-bis(4-pyridyl)aniline (DPBz-NH2) is given as an
example. Synthetic procedures for the remaining DPBz-R compounds
are included in the Supporting Information.
1,4-Bis(4-pyridyl)aniline (DPBz-NH2). A three-necked flask was
charged with 1.3 g (5.2 mmol) of 1,4-dibromoaniline, 2.5 g (18 mmol)
of pyridine-4-boronic acid, 300 mg (0.33 mmol) of Pd2(DBA)3, and
225 mg (0.8 mmol) of P(Cy)3 and purged with N2. The mixture was
suspended in 40 mL of deoxygenated 1,4-dioxane (commercial
anhydrous grade packaged under Ar). A solution of 4 g (19 mmol)
of K3PO4 in 20 mL of degassed water was added by syringe through a
septum. The reaction was heated at reflux (100 °C) under N2 with
rapid stirring overnight (∼18 h). Upon being cooled to room
temperature, the mixture was poured into a separatory funnel and the
lower aqueous phase removed and discarded. The dioxane layer was
collected and filtered and the dioxane removed under reduced
pressure. The residue was dissolved in chloroform and washed twice
with 5 g of Na2CO3 in 25 mL of water. The chloroform solution was
dried with anhydrous magnesium sulfate and the solvent removed
under reduced pressure. The solid was taken up in 80 mL of hot ethyl
acetate, treated with activated carbon, and filtered hot through a fine
glass frit. The solid was crystallized by concentration and cooling of
the filtrate to yield 0.75 g (3.0 mmol) of beige product. Mp: 197 °C.
1H NMR (300 MHz, CDCl3): δ 8.70, m, 4H; 7.49, m, 4H; 7.27, m,
1H; 7.12, m, 1H; 7.04, d, 1H. Anal. Calcd for C16H13N3 (247.3): C,
77.71; H, 5.30; N, 16.99. Found: C, 76.92; H, 5.34; N, 16.32.
C
dx.doi.org/10.1021/ic301893p | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Scheme 1. Formation of PICNICs via the HIPL Reaction
solvent). The anhydrous Ni2(CN)4 starting materials and Ni(L)[Ni-
(CN)4] products are insoluble in the reaction solvent and as such are
present throughout the reaction as a suspension. The typical reactant
loading is ∼1 mmol/30 mL of solvent with a reaction time of 24 h
sufficient for most pillar ligands, with some sterically hindered pillars
requiring up to 72 h for completion. Reaction progress can be
monitored qualitatively by observing the color change of the initially
orange anhydrous Ni2(CN)4 coordination polymer to what is often a
light violet or dull green color of the Ni(L)[Ni(CN)4] products or
qualitatively by taking small aliquots of the suspended solid and
filtering out the product for subsequent thermogravimetric analyses in
air where the formula weights of the products can be determined by
the residual weight percent of NiO above 500 °C. It is noted that
reaction times can be reduced for more sterically demanding insertion
reactions when performed solvothermally in sealed Teflon-lined high-
pressure Parr reactors at temperatures of 150 °C using dry toluene.
Each of the pillar ligands listed in Chart 1 could be intercalated into a
Ni(L)[Ni(CN)4] product using this basic reaction scheme; however,
PICNIC-61 and PICNIC-65 irreversibly collapse to nonporous phases
after removal of guest solvents. The PICNICs containing Pyz, Bpy,
Bpane, Bpene, and DPAC pillar ligands have previously been
described.28,33 Characterization of PICNIC compounds by TGA and
X-ray powder diffraction is included in the Supporting Information.
Prior to CO2 adsorption measurements, the PICNIC compounds were
activated by extraction in boiling chloroform or acetone until TGA
showed solvent exchange was complete (typically 2−4 h) and then
dried under vacuum at 115 °C for a minimum of 6 h.
PICNICs Containing Other Transition Metals. Preliminary
reactions were performed to test whether PICNIC derivatives could
be formed with additional transition metals. Reactions to form
Co(L)[Ni(CN)4]n, where L = AmMe-Pyr and pXdAm, Fe(L)[Ni-
(CN)4]n, where L = Pyz, Bpy, Bpene, and pXdAm, and Fe(L)Pd-
(CN)4, where L = Bpy, were accomplished using a method similar to
that of the Ni-containing PICNICs with minor modification. For the
Co- and Fe-based PICNICs, the hydrated polymeric Co[Ni(CN)4]n,
Fe[Ni(CN)4]n, and Fe(L)[Pd(CN)4]n were more effective when used
directly without the need to remove coordinated waters prior to the
intercalation reaction when dry ethanol or dry acetonitrile was used as
the reaction solvent. Sample characterizations are included in the
Supporting Information.
ments, samples (5−10 mg) were run in Pt pans at 15 °C/min under a
dry air purge of 100 mL/min to temperatures >550 °C. To determine
the sample purity, the following calculations are performed using the
TGA data. The expected formula weight of the desired PICNIC
compound was calculated on the basis of the Ni(L)Ni(CN)4 structural
unit. The complete oxidation in air will lead to two NiO groups per
formula unit, which have a formula weight of 149.38. The ratio of the
formula weight of nickel oxide to the formula weight of the PICNIC is
149.39/(expected formular weight of PICNIC). Dividing the mass of
NiO obtained in the TGA run in air by the mass of the corresponding
guest-free PICNIC should result in the same ratio. On the TGA curve,
the material was determined to be guest free at the plateau after guest
solvents were lost (see the Supporting Information, Figures S16−S66).
For the FeNi and CoNi PICNICs, oxide formulas of FeNiO1.5 and
CoNi1.5 were used to determine formula weights on the basis of
common oxidation states and powder diffraction analysis (see the
Supporting Information, Figures S67 and S68).
RESULTS AND DISCUSSION
■
Synthesis of Pillar Ligands. Due to their commercial
availability and ability to adopt a convenient linear bridging
arrangement, pyrazine and 4,4′-bipyridine are commonly used
ligands in the preparation of coordination polymers. Several
functionalized versions of pyrazine are available from
commercial suppliers; however, the symmetry of the molecule
requires functional groups at the position α to the ring
nitrogen, which could cause complications in coordination
chemistry due to the possibility of the functional group to
sterically hinder access to the ring nitrogen, resulting in a single
coordination of the ligand at the unhindered nitrogen and the
failure to act as a bridging ligand. For the PICNIC compounds,
the amino- and methylpyrazines successfully resulted in porous
structures, whereas the methoxy derivative gave a very low
porosity solid, and the attempt at the chloro derivative was
unsuccessful as determined from TGA (see the Supporting
Information).
The situation is improved with functionalized 4,4′-bipyridine
derivatives since functionalization can occur at the 3-position
and not interfere with coordination of the ring nitrogen. These
3-functionalized derivatives of 4,4′-bipyridine, however, are not
readily available through commercial sources. We were able to
prepare both the 3-functionalized 4,4-bipyridine derivatives and
related 1,4-bis(4-pyridyl)benzenes by slight modification of a
convenient method reported by Kudo47 in which pyridine-4-
boronic acid is coupled with a 3-functionalized 4-halopyridine
or a 3-functionalized 1,4-dihalobenzene, the palladium catalyst
Pd2(DBZ)3, K3PO4, and tricyclohexylphosphine in a mixture of
dioxane and water. Products were obtained in fair yields of
sufficient purity by recrystallization. Furthermore, the reaction
conditions were amenable to a wide range of functional groups
and were even found to be effective for aryl chlorides. Using
Instrumentation. Isotherms were collected on a pressure−
composition isotherm measurement system (Advanced Materials
Corp.) for pressures up to ∼18 atm for CO2 over a temperature
range of −25 to +30 °C. The instrument is designed on the basis of a
conventional Sievert apparatus. Prior to the measurements, samples
(∼500−750 mg) were degassed under vacuum at 115 °C for >6 h.
Low-pressure N2 isotherms and five-point BET surface areas were
measured on a Quantachrome Autosorb 1C volumetric instrument on
samples (∼100 mg) degassed under dynamic vacuum for >6 h at 115
°C. Powder X-ray diffraction measurements were performed on a
PANalytical X’Pert Pro MPD powder diffractometer having a θ−θ
configuration, a Cu X-ray source operated at 45 kV and 40 mA, and an
X’Celerator detector with a monochromator. Thermogravimetric
analyses were performed using a Mettler STARe TGA/DSC
thermogravimetric analyzer. For PICNIC decomposition measure-
D
dx.doi.org/10.1021/ic301893p | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
this general synthetic procedure, a large number of 3-
functionalized bipyridines (Bpy-R) and dipyridylbenzenes
(DPBz-R) were prepared. The ligand abbreviations and
associated structures are shown in Chart 1.
reaction also has the benefit of forming the product in
essentially quantitative yields using only a few percent excess of
molar equivalents of pillar ligand to Ni2(CN)4. Such high
percentage yields can often be difficult to achieve using
traditional synthetic approaches where the coordination
polymer is precipitated from a coordinating solvent solution
or prepared using solvothermal methods.
Another route to functionalized pillar ligands was found by
simply using the well-known amide coupling reaction between
an amine and an acid chloride as used to form N-(4-
pyridyl)isonicotinamide.59 Modifications of this reaction using
commercially available 3-substituted 4-aminopyridine deriva-
tives yielded three functionalized versions of the pillar ligand as
shown in Scheme 1 (PINA-R). Additional pillar ligands used in
the preparation of PICNICs obtained through commercial
sources or by literature preparations are also included in Chart
1.
Synthesis of PICNICs. The preparation of the PICNIC
materials was accomplished in essentially quantitative yield
using a generalized synthetic method we call HIPL and is
outlined in Scheme 1. The synthetic technique is a modification
to a similar method first reported by Mathey et al. over two
decades ago.61 The brief report by Mathey suggested that
pillared cyanonickelates could be formed by the reaction of
pillaring ligands such as pyrazine or 4,4′-bipyridine with a
suspension of anhydrous polymeric Ni2(CN)4 for a few
minutes in boiling chloroform; however, no detailed inves-
tigations of the materials or further refinement or discussion of
the method appeared in the literature.
We investigated the method previously and found it
convenient for the preparation of polycrystalline PICNICs
with common pillar ligands such as pyrazine, 4,4′-bipyridine,
1,2-dipyridylacetylene, and 1,2-bis(4-pyridyl)ethylene.28,33
While the products of the reaction scheme may lack the high
degree of crystallinity typically obtained through more
conventional solution-based or solvothermal synthetic meth-
ods, the quantitative yields and exceptional convenience of the
method make it very well-suited for preparing a highly diverse
array of structurally analogous samples of sufficient structural
integrity to allow for a rapid screening of material structure/
property relationships, such as CO2 adsorption affinity, for
example. Thus, we decided to take full advantage of the HIPL
method and greatly expanded it in scope. The method was
found to be effective using pillar ligands that coordinate
through pyrazine, pyridine, or amine nitrogen and is amenable
to a wide variety of functional groups.
While the exact mechanism of the HIPL reaction is not
known, it most likely proceeds through a surface adsorption/
intercalation process since the polymeric Ni2(CN)4 is highly
insoluble in common organic solvents. What is somewhat
remarkable is how efficiently relatively long and even
functionalized linkers can insert into the structure. As such,
the HIPL method takes advantage of the stable preformed
polymeric 2-D Ni2(CN)4 networks which direct the final
layered structure and helps to suppress potential complications
that may arise when coordination polymers are assembled using
functionalized ligands with multiple coordination sites and free
metal ions from solution. This structure-directing ability is
perhaps best exemplified in the preparation of PICNIC-27, with
the 4,4′-Bpm pillar ligand producing a rare open pore structure
containing a nonmetalated 2,2′-bipyridyl structural unit.62−65
The reaction is facilitated by the presence of reactive
unsaturated Ni sites produced after dehydration of Ni2(CN)4
hydrate and by the small particle size of the Ni2(CN)4 starting
material, which aids in the suspension of the particles and
provides a high surface area for the intercalation reaction. The
While attempts to use hydrated Ni2(CN)4 directly (without
prior evacuation of coordinated water) were not as effective,
exploratory reactions using Fe(H2O)2[Ni(CN)4] hydrate and
Co(H2O)2[Ni(CN)4] hydrate were found to proceed well in
dry ethanol or dry acetonitrile without first drying the starting
polymeric cyanometalates. It is interesting to compare the
method for reactions involving Fe since derivates of pillared
FeNi(CN)4 compounds have previously been reported using
traditional precipitation reactions from solution. For the
Fe(pyrazine)Ni(CN)4 compound, formation of the product
occurs in near-quantitative yield from the reaction of molar
equivalents of a ferrous salt with pyrazine and [Ni(CN)4]2− in
methanol/water mixtures.32 For solution reactions involving
the more reactive 4,4′-azopyridine pillar ligand, however, the
propensity of the Fe2+ and 4,4′-azopyridine linker to form a
polymeric product complicates the reaction, and formation of
the targeted Fe(4,4′-azopyridine)Ni(CN)4 compound required
a 100-fold excess of Fe2+.25 A report for the solution
precipitation preparation of Fe(4,4′-dipyridylacetylene)M-
(CN)4 showed the potential for the pillar ligand to be adsorbed
into the pore network as a guest.26 Investigations in our
laboratory also indicated that forming Fe-based PICNICs with
bpy or bpene pillar ligands using solution reactions was difficult.
Tests using the HIPL method were more effective with
reactions using molar equivalents of FeNi(CN)4 or FePd(CN)4
and pillar linkers, including pyz, bpy, bpene, and pXdAm,
producing porous products. Considering the exceptional spin-
crossover behavior25,26,32,48,66−68 reported for many Fe(L)M-
(CN)4 compounds, it appears promising that the HIPL
reaction may provide a means of preparing many more
magnetically interesting compounds in this series which may be
otherwise difficult to obtain by precipitation from solution.
General Structure of PICNICs. While the polycrystalline
nature of the materials is well-suited for the kinetic demands of
the reaction scheme, their small particle sizes make structure
determinations for these compounds challenging due to the
broad nature of the diffraction peaks. Model compounds exist
in the literature, however, that do provide a means to assign a
qualitative description of how the PICNICs are assembled. The
crystal structures of several related compounds, including
Fe(bpene)Pt(CN)4,31 Fe(DPAC)M(CN)4,26 Fe(pyrazine)M-
(CN)4,32,48 Fe(4,4′-azopyridine)M(CN)4,25 Zn(N-(4-pyridyl)-
isonicotinamide)Ni(CN)4,27 and Ni(α,ω-diaminoalkane)Ni-
(CN)4,29 all show the same general structural motif, which
involves an inorganic 2-D MA[MB(CN)4] grid network where
the MA metal ion has octahedral coordination and the MB
complex is square planar. Pillaring of the 2-D networks is
achieved through the coordination in the direction perpendic-
ular to the square planar complex of an organic diaza bridging
ligand (pillar ligands) to the octahedral MII complex. This
results in formation of an extended 3-D network. Moreover, the
absence of pillar ligands on the four-coordinated square planar
d8 tetracyanometalate significantly enhances the porosity of
these materials. As such, they are extended versions of the well-
known Hofmann clathrates,24,29,69−73 where coordinated NH3
E
dx.doi.org/10.1021/ic301893p | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
ligands on the octahedral metal sites have been replaced with
pillar ligands.
original solvent guests can be exchanged by simple refluxing for
several hours of the material in acetone or chloroform. This
solvent extraction procedure also has the added benefit of
helping to remove any excess ligand which may have adsorbed
in the pore structure during the synthesis.26,31,61 The solvent
guests can be removed by heating under vacuum to produce a
family of porous materials with interesting CO2 adsorption
behaviors.
CO2 Adsorption. Pore-Functionalized PICNICs. Included
in the current Article is one of the most extensive groups of
pore-functionalized derivatives within a structurally related
family of porous coordination polymers reported to date. Some
of the best known families of structurally homologous porous
coordination polymers containing organic functional groups
within their pore structures include ZIFs,19 isorecticular metal
organic frameworks (IRMOFs),7 and the MIL series.10 Similar
to the widely studied ZIF, IRMOF, and MIL materials, the
PICNIC family of materials have a broad distribution in pore
dimensions and pore functionalization while maintaining an
open pore structure available for CO2 adsorption.
The proposed general structure for the PICNIC materials
involves the same structural motif as reported for other similar
pillar Hofmann compounds. On the basis of this structural
model, the interlayer spacing in each PICNIC can be
determined by the d-spacing of the (001) diffraction peak in
the X-ray powder diffraction pattern. The interlayer spacing for
each compound prepared and the associated diffraction
patterns are included in the Supporting Information. A
representative sample of diffraction patterns for several
PICNICs showing the span of interlayer spacing in the series
is shown in Figure 1. The interlayer spacing for each material is
close to that predicted from typical N−Ni bond lengths and the
molecular dimensions of the respective pillar ligand.
The shortest pillar ligands include those based on the
pyrazine ligand. Powder X-ray diffraction patterns for Ni(L)-
Ni(CN)4, where L = Pyz, Pyz-Am, and Pyz-Me, are indicative
of a structurally homologous series with a pillared motif similar
to that reported for Fe(Pyz)M(CN)4 (see the Supporting
Information, Figure S3). The CO2 adsorption isotherms
measured at 0 °C for four Pyz-PICNICs are shown in Figure
2A. As expected within an isostructural series, the adsorption
capacity decreases in an inverse relationship with the size of the
functional group. The adsorption capacity for PICNIC-01
approaches 1.5 CO2 per pore and decreases to slightly less than
1 CO2 per pore for PICNIC-03, where Pyz-Me is the pillar
ligand, whereas the CO2 uptake for PICNIC-04 with the
−OMe substituent is negligible. The incorporation of the amine
functional group in PICNIC-02 has little effect on the CO2
affinity of the pore, likely as a result of the insufficient basicity
of the aromatic amine to significantly affect the CO2 adsorption
potential. It is interesting to note that the CO2 capacity of
PICNIC-03 with Pyz-Me as the pillar ligand is slightly below 1
CO2 per pore, suggesting that the pore dimensions are on the
order of the molecular dimensions of CO2. As shown in Figure
2B, the pore constriction has a significant effect on the
adsorption of N2, which is decreased to a negligible amount in
Figure 1. Representative series of powder X-ray diffraction patterns for
several PICNICs. The range of interlayer spacings (ca. 7−19 Å) in the
materials is indicated by the position of the (001) interlayer reflection
marked by an asterisk in the figure. See Chart 1 for ligand and PICNIC
abbreviations.
Evidence for the presence of a porous structure in the
PICNIC compounds can be gained by thermal gravimetric
analysis, which shows a clear step that accompanies the loss of
guest solvents prior to the combustion of the compound to the
metal oxide in air (see the Supporting Information). The
Figure 2. Adsorption isotherms for the Ni(Pyz-R)Ni(CN)4 series: (A) CO2 at 0 °C, (B) N2 at 0 °C. Key: PICNIC-01, R = H (blue tilted squares);
PICNIC-02, R = NH2 (red squares); PICNIC-03, R = CH3 (black circles); PICNIC-04, R = OCH3 (green triangles).
F
dx.doi.org/10.1021/ic301893p | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
the PICNIC-03 material. The pure isotherm data suggest that
the pore constriction provided by the methyl functional group
may provide an effective means for increasing CO2 adsorption
selectivity versus other larger gases such as CH4 and N2.
Inclusion of substituted PINA-R pillars offers the potential
for incorporating two functional groups into the pore structure,
with one being the functional group on the 3-position of one of
the pyridines in the molecule and the other being the amide
linker between the pyridine rings. The affinity of the PINA
molecule to generate hydrogen-bonding interactions is clearly
observed in the reported crystal structure of the material.74
Powder X-ray diffraction patterns for the HIPL-prepared PINA-
R derivatives of Ni(L)Ni(CN)4, where L = PINA, PINA-F,
PINA-Me, and PINA-NO2, are indicative of a structurally
homologous series with a pillared motif similar to that reported
for Zn(PINA)Ni(CN)4 (see the Supporting Information,
Figure S2). The CO2 adsorption isotherms for the four
PICNIC compounds are shown in Figure 3. All four
Figure 4. Adsorption isotherms for CO2 at 0 °C on the Ni(Bpy-
R)Ni(CN)4 series. See Chart 1 for ligand and PICNIC abbreviations.
pillared PICNIC is included as well. Again, the CO2 adsorption
capacity shows trends similar to those observed for the PICNIC
compounds with the Pyz-R and PINA-R pillar ligands, where
the functional group in the pore is detrimental to the
adsorption capacity due to the steric influence of the functional
group. The two PICNICs out of order with regard to the trend
are the two compounds with OMe and C(O)H functional
groups which show slightly higher CO2 uptakes than the CH3
and NH2 functionalized materials. The five-point N2 BET
surface areas for the series of materials are listed in Table 1. The
Table 1. N2 Five-Point BET Surface Areas (SAs) of the Bpy-
R Series of PICNICs
pillar
ligand
five-point BET SA (m2/
SA (m2/
mmol)
sample
fw
g)
PICNIC-20 Bpy
377.6
391.7
392.7
395.6
445.6
405.7
622
466
453
600
448
443
441
542
235
183
178
237
200
180
180
206
Figure 3. CO2 adsorption isotherms at 0 °C on the Ni(PINA-
R)Ni(CN)4 series: R = H, PICNIC-30 (black tilted squares); R =
CH3, PICNIC-31 (green triangles); R = F, PICNIC-32 (red squares);
R = NO2, PICNIC-33 (purple circles). See Chart 1 for ligand and
PICNIC abbreviations.
PICNIC-21 Bpy-CH3
PICNIC-22 Bpy-NH2
PICNIC-23 Bpy-F
PICNIC-24 Bpy-CF3
PICNIC-25 Bpy-Ald
PICNIC-26 Bpy-OMe 407.7
PICNIC-27 4,4′-Bpm 379.6
compounds are porous structures, with the nonfunctionalized
PINA pillar in PICNIC-30 providing the material with the
highest adsorption capacity for CO2 in the series. As observed
with the pyrazine series, none of the functional groups on the
pillar ligand show any significant enhancement to the CO2
adsorption potential. The total adsorption capacity for CO2
decreases with increasing functional group size from ∼2 CO2/
mol for PICNIC-30 to ∼1.5 CO2 for PICNIC-31 to ∼1 CO2
for PICNIC-33. The CO2 capacity for PICNIC-32 with a fluoro
functional group is slightly lower than expected on the basis of
this trend due to a slightly smaller interlayer spacing in the
material (see the Supporting Information, Figure S2). All of the
samples show a relatively low saturation pressure of ∼2 atm of
CO2.
A third structurally analogous series of PICNICs was
successfully prepared via substitution of BPY pillars for BPY-
R pillars as verified by the similar diffraction patterns (see the
Supporting Information, Figure S1) observed for the materials.
Normalized CO2 adsorption isotherms measured at 0 °C for
the seven PICNIC samples containing the Bpy-R pillar ligands
are shown in Figure 4. The isotherm for the related 4,4′-Bpm
surface areas follow trends similar to those of the CO2
absorption behaviors in the materials, with the Bpy, Bpy-F,
and 4,4′-Bpm pillared PICNICs having the highest surface areas
and the remaining Bpy-R pillared compounds having lower and
very similar surface areas.
The similarity in isotherm behavior for the PICNIC-20
through PICNIC-26 samples is further evidence that the
materials are isostructural and amenable to inclusion of the
wide range of functional groups included on the BPY-R ligands.
Nevertheless, as observed in the PYZ-R and PINA-R pillared
PICNICs, the functional groups on the BPY-R ligands appear
to have no significant enhancement to the CO2 adsorption
affinity of the pore systems. The presence of the functional
group is in fact even detrimental in terms of gravimetric uptake,
where the less massive nonfunctionalized BPY sample PICNIC-
20 performs best in both the low- and high-pressure regions.
The CO2 adsorption capacities for the Pyz-R, PINA-R, and
Bpy-R series are consistent with a structurally homologous
family of materials where the pore functional groups occupy a
G
dx.doi.org/10.1021/ic301893p | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 5. Adsorption (solid symbols) and desorption (open symbols) isotherms for CO2 at −15 °C on several newly discovered structurally
dynamic PICNICs (the structure of the pillar ligand is shown).
portion of the void space and thereby limit the total CO2
uptake. The reduction in void volume due to the functional
group correlates well with the decrease in CO2 adsorption
capacity within each series of materials. What is perhaps
unexpected is how the functional groups appear to have little or
no effect on the CO2 affinity in the low-pressure regions of the
isotherms, the region most important in postcombustion
capture of CO2 from power plant flue gas. In each series of
materials, the PICNIC containing the nonfunctionalized pillar
ligand functions as well or better than most of the pore-
functionalized materials.
Structurally Dynamic PICNICs. A recent report highlighted
the ability of PICNIC-60 to act as a structurally dynamic
sorbent with the Bpene pillar ligand.28 The behavior was
unusual in the fact that shorter pillar ligands such as Pyz and
Bpy did not show structural flexibly. Perhaps even more
unusual was the lack of structural flexibility with other pillar
ligands of similar molecular geometry such as DPAC and
Bpane.28 As discussed above, none of the PINA-R derivatives
(PICNIC-30, -31, -32, and -33) show structurally dynamic
behavior even though the length and geometry of the PINA-R
pillar ligands are very similar to those of Bpene.
This result may be due to several factors affecting the CO2
adsorption in these materials. One influence may be due to the
electronic properties of the pyridine ring acting to diminish the
effectiveness of the functional group. This is highly likely, for
example, in the case of the −NH2 functional group, where the
basicity of the amine nitrogen is decreased several orders of
magnitude relative to that of the alkylamine analogue. A second
factor to be considered is the relatively narrow pore space
which exists between the pillar ligands. This inter pillar distance
is on the order of the length of a CO2 molecule. Computation
modeling of the CO2 packing arrangement in the structurally
similar Fe(Pyz)Ni(CN)4 material showed a favored orientation
where the CO2 was bridged between two pyrazine pillars via
hydrogen bonds between the ring hydrogens and CO2 oxygens,
with significant CO2−CO2 interactions in the adsorbed state.75
This preferred packing arrangement for CO2 may explain why
the nonfunctionalized pillars provide a highly competitive CO2
adsorption potential compared to the functionalized materials
since the steric influence of the functional group would be a
strong driving force for an alternative, and perhaps less
energetically favorable, CO2 packing arrangement. In addition,
the presence of a functional group in a confined space may limit
the ability of the adsorbed CO2 molecule and functional group
to orient in the proper positions in which to take advantage of
the attractive forces that would be present between the two in a
nonhindered environment.76 Efforts are under way to obtain
single crystals for several samples so that detailed structural
models of the materials can be obtained and the effects of the
functional groups on the packing arrangement of CO2 can be
calculated to better understand this intriguing result.
These results indicate that the structurally dynamic behavior
of PICNIC-60 is not a simple function of interlayer spacing or
molecular geometry. This conclusion is further demonstrated
by several newly discovered flexible PICNICs whose isotherms
are shown in Figure 5. The isotherms are all plotted at the same
temperature of −15 °C. The pillar ligands vary in both
functionality and molecular dimensions and include the
nonlinear linker pXdAm, along with several linear DPBz-R
analogues. The interlayer spacing is shortest for PICNIC-11
(∼10 Å) with the pXdAm pillar ligand and longest in PICNIC-
70 (∼19 Å) with the IsoNicBz pillar ligand.
Several interesting results are noted in the isotherm data
shown in Figure 5. First, the widths of the hysteresis loops and
adsorption threshold pressures (Pth) vary over a relatively wide
pressure range of 3−10 atm. Additionally, each of the flexible
PICNICs with the exception of pXdAm shows an uptake of
approximately 1 equiv of CO2 prior to the Pth. For the pXdAm
PICNIC, the uptake of CO2 prior to Pth is essentially half that
of the others. It is also interesting to note the similarity in Pth
for PICNIC-11 and -70 even though the pXdAm and IsoNicBz
pillar ligands are structurally quite different and the CO2
adsorption capacities of the two materials differ by at least a
factor of 2.
This result is somewhat surprising since one might expect the
PICNIC-70 material with the much longer IsoNicBz ligand to
have a larger energy requirement for structural reorganization.
In contrast, the isotherm results for PICNIC-40, -41, -42, and
-43, which all contain similar DPBz-R analogues, all show
similar Pth values, as might be expected due to the similarity in
pillar ligands. In fact, the Pth values for the DPBz-R PICNICs
H
dx.doi.org/10.1021/ic301893p | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
are only slightly higher than that of PICNIC-60 and show
similar CO2 adsorption capacities as well.28 This is in line with
what one might expect on the basis of the molecular
dimensions of the pillar ligands.
The interpretation, however, gets more complicated when
additional DPBz-R PICNICs are studied. The isotherms
plotted in Figure 6 for three additional DPBz-R-based
Figure 7. Powder X-ray diffraction patterns for PICNIC-11 (B)
toluene loaded and (C) guest free and PICNIC-12 (D) toluene
loaded, (E) guest free, and (F) after readsorption of toluene. The pillar
ligands are pXdAm and pXdAm-F4, respectively. The structure
transition of PICNIC-11 is reversible for CO2 adsorption as shown
in Figure 5, but the structure transition of PICNIC-12 is not reversible
for CO2 adsorption. PICNIC-12 will, however, reversibly desorb and
adsorb toluene as shown by the three lower diffraction patterns. The
calculated XRD pattern of the structurally related Cd(pXdAm)Ni-
(CN)4-o-toluidine (A) as reported by Yuge et al. is included as a
reference.34.
Figure 6. Adsorption isotherms measured at 0 °C for CO2 on three
nonflexible Ni(DPBz-R)Ni(CN)4-based PICNICs (structures of pillar
ligands are shown). The adsorption behavior is in contrast to that of
PICNIC-40, -41, -42, and -43 with similar DPBz-R ligands as shown in
Figure 5.
8. Both materials were prepared by the HIPL method, and the
two pillar ligands involved, Bpene and AzoPyr, respectively,
PICNICs do not show the same dynamic isotherms. The
isotherms for PICNIC-44, -49, and -51 show more type I
behavior, albeit there does exist a slightly distorted low-pressure
region for PICNIC-51. A significant contrast in behaviors is also
observed when comparing the similar DPBz-Est and DPBz-
OMe PICNICs. The DPBz-Est pillared PICNIC-49 CO2
isotherm in Figure 6 shows nonflexible behavior, while the
DPBz-OMe pillared PICNIC-43 CO2 isotherm shown in
Figure 5 is more complex with two separate transitions. The
two materials also show a large difference in their level of
structural stability between their guest-loaded and guest-free
states as evidenced by their respective powder X-ray diffractions
patterns (see Figures S9 and S10 in the Supporting
Information).
Another example of a large contrast in sorbent behavior
resulting from a relatively small change in pillar chemistry is
obtained when the pXdAm pillar in PICNIC-11 is replaced
with a fluorinated ring derivative, pXdAm-F4, to give PICNIC-
12. While PICNIC-11 is structurally dynamic during CO2
adsorption, PICNIC-12 does not reopen to adsorb CO2. As
indicated by the powder diffraction patterns for PICNIC-12
shown in Figure 7, a significant loss of structural order occurs in
PICNIC-12 when the guest toluene molecules are removed
from the pore system after the initial synthesis. However,
readsorption of toluene by the material regenerates the original
structure. Thus, the structural transition between collapsed and
open pore structures is reversible in PICNIC-12, but the energy
required to drive the structural transition cannot be supplied by
CO2 adsorption alone.
Figure 8. Contrasting CO2 adsorption behaviors at 0 °C for PICNIC-
60 and PICNIC-62 even though the pillar ligands involved, Bpene and
AzoPyr, respectively, are nearly identical in molecular dimensions.
differ only in the substitution of a CC bond for a NN
bond between the two pyridine rings. While the normalized
adsorption capacities are nearly identical, PICNIC-60 is
structurally dynamic and PICNIC-62 is not.
The mechanism for the behavior of these materials is still
under investigation in our laboratory. It is clear in some cases,
e.g., PICNIC-60 and PICNIC-11, from X-ray powder
diffraction analysis that the isotherm behavior in the flexible
PICNICs is due to a structural transition that affects the
interlayer spacing in the materials. This structural transition
appears to result from a change in the in-plane bonding
geometry of the Ni−CN−Ni network, which effectively
A final and equally impressive demonstration of this fine
balance between pillar ligand and sorbent property in the
PICNIC compounds is shown by comparison of the CO2
adsorption isotherms for PICNIC-60 and PICNIC-62 in Figure
I
dx.doi.org/10.1021/ic301893p | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
changes the tilt angle of the pillar ligand. This variation in tilt
angle affects the interpillar distance and the available free
volume for guest adsorption. However, for some of the flexible
PICNICs the shifts in interlayer spacings are much more subtle,
suggesting that a mechanism involving ring orientation of the
pillar ligands may be at play. The phase transition has been
shown to be an interplay of energetic terms which ultimately
lower the thermodynamic potential of the system and drive
both the structure change and incorporation of gas guest. As
such, one expects that even subtle changes in the ligand
structure, hydrogen bonding within the material, ligand−ligand
interactions, and the ability of a gas guest to adsorb and
stabilize the open pore network would collectively play a role in
dictating the conditions for this phase change to occur. In this
regard, the PICNIC architecture affords an opportunity to
systematically interrogate these energetics by varying the pillar
ligands.
PICNICs Containing Other Metals. With the success of the
HIPL technique in the preparation of nickel-based PICNICs,
preliminary investigations were conducted to test the wider
applicability of the method in the synthesis of Fe- and Co-
containing PICNICs. Porous materials were formed for
Co(AmMe-Pyr)Ni(CN)4, Co(pXdAm)Ni(CN)4, Fe(Pyz)Ni-
(CN)4, Fe(Bpy)Ni(CN)4, Fe(Bpene)Pd(CN)4, and Fe-
(pXdAm)Ni(CN)4 by a method similar to that used in the
preparation of the nickel-based PICNICs. The CO2 adsorption
isotherms for several of these compounds are shown in Figures
9 and 10. The CO2 adsorption behavior follows trends similar
Figure 10. Adsorption (solid symbols)/desorption (open symbols)
cycles for CO2 measured at −15 °C for two Co(L)Ni(CN)4 PICNICs,
where L = AmMe-Pyr (red circles) and pXdAm (black tilted squares).
See Chart 1 for ligand abbreviations.
interesting spin-crossover behavior. On the basis of the
preliminary results reported herein, the HIPL method may
provide a convenient avenue to the preparation of other
magnetically interesting Fe-based PICNICs that are challenging
to produce by traditional precipitation methods, some of which,
e.g., the Bpene and pXdAm materials, may show unusual CO2
adsorption influences on their structural and magnetic
properties.
CONCLUSION
■
A number of ring-functionalized dipyridyl-based bridging
ligands were synthesized using Suzuki coupling reactions and
traditional amide coupling reactions. The functionalized ligands
were used in the formation of a structurally homologous series
of porous materials built on the pillared cyanonickelate
structure motif. The porous coordination polymers were
prepared in essentially quantitative yield by a convenient
heterogeneous intercalation of pillaring ligand reaction using
insoluble polymeric nickel cyanide as a structural building
block. The use of a preformed 2-D network as a building block
helped to direct the final structure of the material into a 3-D
porous framework with little interference from the functional
groups. The structure-directing property of the prefabricated 2-
D Ni2(CN)4 networks is particularly well-demonstrated by the
formation of PICNIC-27, a rarely observed porous coordina-
tion polymer containing a nonmetalated 4,4′-bipyrimidine pillar
ligand.63 The CO2 adsorption isotherms and powder X-ray
diffraction analyses were consistent with the formation of an
isostructural series of PICNICs when pillar ligands of similar
molecular geometries were used. The CO2 adsorption
behaviors within an isostructural series showed a decrease in
capacity for the functionalized materials due to occupation of
the void volume by the functional groups. In spite of the large
number of functional groups incorporated into the materials,
none were found which showed any significant enhancement of
CO2 affinity in the materials.
Figure 9. Adsorption (solid symbols)/desorption (open symbols)
cycles for CO2 on Fe(L)M(CN)4 PICNIC derivatives, where M = Ni,
L = Bpy (green tilted squares, T = 0 °C), M = Ni, L = Pyz (blue
squares, T = 0 °C), M = Ni, L = Bpene (black circles, T = 0 °C), and
M = Ni, L = pXdAm (red triangles, T = −25 °C). See Chart 1 for
ligand abbreviations.
to those observed with the nickel-based PICNICs, with the
Bpene and pXdAm pillar ligands giving structurally dynamic
materials and the Pyz and Bpy pillars giving nonflexible
samples. The Fe(Pyz)Ni(CN)4 prepared by the HIPL method
reported herein shows the expected color change related to the
spin-crossover transition32 and CO2 adsorption behavior similar
to the results reported by Southon et al. for the Fe(Pyz)Ni-
(CN)4 material prepared by the conventional precipitation
method.48 The results are encouraging since, to date, only a few
pillar ligands have been reported for iron-based PICNICs in the
literature and each of these reported structures shows
Several of the PICNICs showed structurally dynamic
behavior during the adsorption and desorption of guests. The
structural flexibility is likely the result of corrugated Ni2(CN)4
sheets that have the ability to adjust the bond angles within the
cyanide-bridged network, which ultimately affects the tilt angle
J
dx.doi.org/10.1021/ic301893p | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
of the pillar ligands; however, the adsorption/desorption
mechanisms in these materials are still under investigation in
our laboratory. The structurally dynamic materials showed a
range of threshold pressures for CO2 adsorption. As such,
PICNICs are an unusual class of compounds where the
structural flexibility of the material can be tuned by small
adjustments in the properties of the pillaring ligand. With the
interest in structurally dynamic materials as gas-selective
sorbents, PICNICs may prove to be a valuable class of sorbent
for understanding the subtle interplay between thermody-
namics and host−guest behavior. The ability to potentially tune
the Pth and ΔPhys in these materials may also provide benefits in
applications involving high-pressure capture of CO2.
The assembly technique used for the synthesis of nickel-
based PICNICs was also found to be suitable for preparing
other derivatives containing iron and cobalt in place of the six-
coordinate nickel site. The iron and cobalt analogues also
showed the ability to form structurally dynamic materials
similar to the nickel-based compounds. The iron-based
materials are particularly attractive in light of the recent
literature reports of similar compounds having spin-crossover
behavior at room temperature. Thus, the HIPL method may
prove to be an alternative synthetic route or a convenient
method for rapidly screening a large number of Fe(L)M(CN)4
compounds for spin-crossover behavior. The structurally
dynamic behavior that appears to be a common trait for
these pillared cyanonickelates could also lead to several new
Fe(L)M(CN)4 compounds with interesting associations
between host−guest interactions or CO2 adsorption/desorp-
tion and magnetic properties.
ACKNOWLEDGMENTS
■
This technical effort was performed in support of National
Energy Technology’s ongoing research in CO2 capture under
RES Contract DE-FE0004000. Catherine Madden was
supported by the U.S. Department of Energy (DOE) Office
of Fossil Energy Mickey Leland Energy Fellowship (MLEF)
Program. We thank Dr. Bret Howard at the National Energy
Technology Laboratory for assistance with the X-ray powder
diffraction measurements.
REFERENCES
■
(1) Batten, S. R.; Champness, N. R.; Chen, X. M.; Garcia-Martinez,
J.; Kitagawa, S.; Ohrstrom, L.; O’Keeffe, M.; Suh, M. P.; Reedijk, J.
CrystEngComm 2012, 14, 3001.
(2) Biradha, K.; Ramana, A.; Vittal, J. J. Cryst. Growth Des. 2009, 9,
2969.
(3) Li, J. R.; Ma, Y. G.; McCarthy, M. C.; Sculley, J.; Yu, J. M.; Jeong,
H. K.; Balbuena, P. B.; Zhou, H. C. Coord. Chem. Rev. 2011, 255, 1791.
(4) Li, J. R.; Sculley, J.; Zhou, H. C. Chem. Rev. 2012, 112, 869.
(5) Liu, J.; Thallapally, P. K.; McGrail, B. P.; Brown, D. R. Chem. Soc.
Rev. 2012, 41, 2308.
(6) Meek, S. T.; Greathouse, J. A.; Allendorf, M. D. Adv. Mater. 2011,
23, 249.
(7) Yaghi, O. M.; O’Keeffe, M.; Ockwig, N. W.; Chae, H. K.;
Eddaoudi, M.; Kim, J. Nature 2003, 423, 705.
(8) Zhou, H. C.; Long, J. R.; Yaghi, O. M. Chem. Rev. 2012, 112, 673.
(9) Banerjee, R.; Furukawa, H.; Britt, D.; Knobler, C.; O’Keeffe, M.;
Yaghi, O. M. J. Am. Chem. Soc. 2009, 131, 3875.
(10) Devic, T.; Horcajada, P.; Serre, C.; Salles, F.; Maurin, G.;
Moulin, B.; Heurtaux, D.; Clet, G.; Vimont, A.; Greneche, J. M.; Le
Ouay, B.; Moreau, F.; Magnier, E.; Filinchuk, Y.; Marrot, J.; Lavalley, J.
C.; Daturi, M.; Ferey, G. J. Am. Chem. Soc. 2010, 132, 1127.
(11) Eddaoudi, M.; Kim, J.; Rosi, N.; Vodak, D.; Wachter, J.;
O’Keeffe, M.; Yaghi, O. M. Science 2002, 295, 469.
(12) Kitaura, R.; Fujimoto, K.; Noro, S.; Kondo, M.; Kitagawa, S.
Angew. Chem., Int. Ed. 2002, 41, 133.
ASSOCIATED CONTENT
■
S
* Supporting Information
CHN analysis for pillar ligands, NMR analysis of pillar ligands,
calculated formula weights of PICNICs from TGA, powder X-
ray diffraction patterns for PICNICs, TGA of PICNICs, and
powder XRD of residual iron−nickel oxides and cobalt−nickel
oxides after TGA.. This material is available free of charge via
the Internet at http://pubs.acs.org.
(13) Cohen, S. M. Chem. Rev. 2012, 112, 970.
(14) Costa, J. S.; Gamez, P.; Black, C. A.; Roubeau, O.; Teat, S. J.;
Reedijk, J. Eur. J. Inorg. Chem. 2008, 1551.
(15) Demessence, A.; D’Alessandro, D. M.; Foo, M. L.; Long, J. R. J.
Am. Chem. Soc. 2009, 131, 8784.
(16) Hong, D. Y.; Hwang, Y. K.; Serre, C.; Ferey, G.; Chang, J. S.
Adv. Funct. Mater. 2009, 19, 1537.
AUTHOR INFORMATION
(17) Hwang, Y. K.; Hong, D. Y.; Chang, J. S.; Jhung, S. H.; Seo, Y. K.;
Kim, J.; Vimont, A.; Daturi, M.; Serre, C.; Ferey, G. Angew. Chem., Int.
Ed. 2008, 47, 4144.
■
Corresponding Author
*Fax: (412)-386-4542. E-mail: jeffrey.culp@contr.netl.doe.gov.
(18) Morris, W.; Doonan, C. J.; Furukawa, H.; Banerjee, R.; Yaghi, O.
M. J. Am. Chem. Soc. 2008, 130, 12626.
Notes
(19) Phan, A.; Doonan, C. J.; Uribe-Romo, F. J.; Knobler, C. B.;
O’Keeffe, M.; Yaghi, O. M. Acc. Chem. Res. 2010, 43, 58.
(20) Gao, C. Y.; Liu, S. X.; Xie, L. H.; Ren, Y. H.; Cao, J. F.; Sun, C.
Y. CrystEngComm 2007, 9, 545.
(21) Maji, T. K.; Uemura, K.; Chang, H. C.; Matsuda, R.; Kitagawa,
S. Angew. Chem., Int. Ed. 2004, 43, 3269.
(22) Song, P.; Liu, B.; Li, Y. Q.; Yang, J. Z.; Wang, Z. M.; Li, X. G.
CrystEngComm 2012, 14, 2296.
(23) Wang, X. F.; Wang, Y.; Zhang, Y. B.; Xue, W.; Zhang, J. P.;
This Article was prepared as an account of work sponsored by
an agency of the U.S. Government. Neither the U.S.
Government nor any agency thereof, nor any of their
employees, makes any warranty, express or implied, or assumes
any legal liability or responsibility for the accuracy, complete-
ness, or usefulness of any information, apparatus, product, or
process disclosed, or represents that its use would not infringe
privately owned rights. Reference herein to any specific
commercial product, process, or service by trade name,
trademark, manufacturer, or otherwise does not necessarily
constitute or imply its endorsement, recommendation, or
favoring by the U.S. Government or any agency thereof. The
views and opinions of authors expressed herein do not
necessarily state or reflect those of the U.S. Government or
any agency thereof.
Chen, X. M. Chem. Commun. 2012, 48, 133.
(24) Hofmann, K. A.; Hochtlen, F. Ber. Dtsch. Chem. Ges. 1903, 36,
̈
1149.
(25) Agusti, G.; Cobo, S.; Gaspar, A. B.; Molnar, G.; Moussa, N. O.;
Szilagyi, P. A.; Palfi, V.; Vieu, C.; Munoz, M. C.; Real, J. A.;
Bousseksou, A. Chem. Mater. 2008, 20, 6721.
(26) Bartual-Murgui, C.; Salmon, L.; Akou, A.; Ortega-Villar, N. A.;
Shepherd, H. J.; Munoz, M. C.; Molnar, G.; Real, J. A.; Bousseksou, A.
Chem.Eur. J. 2012, 18, 507.
The authors declare no competing financial interest.
K
dx.doi.org/10.1021/ic301893p | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(27) Chen, X.; Zhou, H.; Chen, Y. Y.; Yuan, A. H. CrystEngComm
2011, 13, 5666.
(28) Culp, J. T.; Smith, M. R.; Bittner, E.; Bockrath, B. J. Am. Chem.
Soc. 2008, 130, 12427.
(29) Iwamoto, T. J. Inclusion Phenom. Mol. Recognit. Chem. 1996, 24,
61.
(61) Mathey, Y.; Mazieres, C.; Setton, R. Inorg. Nucl. Chem. Lett.
1977, 13, 1.
(62) Bodarhouillon, F.; Humbert, T.; Marsura, A.; Devains, J. B. R.;
Dusausoy, O.; Bouhmaida, N.; Ghermani, N. E.; Dusausoy, Y. Inorg.
Chem. 1995, 34, 5205.
(63) Jacobs, T.; Clowes, R.; Cooper, A. I.; Hardie, M. J. Angew.
Chem., Int. Ed. 2012, 51, 5192.
(30) Lemus-Santana, A. A.; Rodriguez-Hernandez, J.; Gonzalez, M.;
Demeshko, S.; Avila, M.; Knobel, M.; Reguera, E. J. Solid State Chem.
2011, 184, 2124.
(31) Munoz-Lara, F. J.; Gaspar, A. B.; Munoz, M. C.; Arai, M.;
Kitagawa, S.; Ohba, M.; Real, J. A. Chem.Eur. J. 2012, 18, 8013.
(32) Niel, V.; Martinez-Agudo, J. M.; Munoz, M. C.; Gaspar, A. B.;
Real, J. A. Inorg. Chem. 2001, 40, 3838.
(33) Culp, J. T.; Natesakhawat, S.; Smith, M. R.; Bittner, E.;
Matranga, C.; Bockrath, B. J. Phys. Chem. C 2008, 112, 7079.
(34) Yuge, H.; Noda, Y.; Iwamoto, T. Inorg. Chem. 1996, 35, 1842.
(35) Chapman, K. W.; Southon, P. D.; Weeks, C. L.; Kepert, C. J.
Chem. Commun. 2005, 3322.
(64) Janiak, C.; Uehlin, L.; Wu, H. P.; Klufers, P.; Piotrowski, H.;
Scharmann, T. G. J. Chem. Soc., Dalton Trans. 1999, 3121.
(65) Maekawa, M.; Tominaga, T.; Sugimoto, K.; Okubo, T.; Kuroda-
Sowa, T.; Munakata, M.; Kitagawa, S. CrystEngComm 2012, 14, 1345.
(66) Bartual-Murgui, C.; Ortega-Villar, N. A.; Shepherd, H. J.;
Munoz, M. C.; Salmon, L.; Molnar, G.; Bousseksou, A.; Real, J. A. J.
Mater. Chem. 2011, 21, 7217.
(67) Bonhommeau, S.; Molnar, G.; Galet, A.; Zwick, A.; Real, J. A.;
McGarvey, J. J.; Bousseksou, A. Angew. Chem., Int. Ed. 2005, 44, 4069.
(68) Halder, G. J.; Kepert, C. J.; Moubaraki, B.; Murray, K. S.;
Cashion, J. D. Science 2002, 298, 1762.
(69) Buttner, H. G.; Kearley, G. J.; Howard, C. J.; Fillaux, F. Acta
Crystallogr. 1994, B50, 431.
(70) Dunbar, K. R.; Heintz, R. A. In Progress in Inorganic Chemistry;
Karlin, K. D., Ed.; Wiley: New York, 1997; Vol. 45, p 283.
(71) Iwamoto, T. J. Mol. Struct. 1981, 75, 51.
(36) Culp, J. T.; Matranga, C.; Smith, M.; Bittner, E. W.; Bockrath, B.
J. Phys. Chem. B 2006, 110, 8325.
(37) Hu, L.; Zhang, P.; Chen, Q. W.; Yan, N.; Mei, J. Y. Dalton Trans.
2011, 40, 5557.
(38) Kaye, S. S.; Long, J. R. J. Am. Chem. Soc. 2005, 127, 6506.
(39) Kaye, S. S.; Long, J. R. Chem. Commun. 2007, 4486.
(40) Motkuri, R. K.; Thallapally, P. K.; McGrail, B. P.; Ghorishi, S. B.
CrystEngComm 2010, 12, 4003.
(41) Natesakhawat, S.; Culp, J. T.; Matranga, C.; Bockrath, B. J. Phys.
Chem. C 2007, 111, 1055.
(42) Reguera, L.; Balmaseda, J.; Del Castillo, L. F.; Reguera, E. J.
Phys. Chem. C 2008, 112, 5589.
(43) Thallapally, P. K.; Motkuri, R. K.; Fernandez, C. A.; McGrail, B.
(72) Iwamoto, T.; Miyoshi, T.; Sasaki, Y. Acta Crystallogr. 1974, B30,
292.
(73) Rayner, J. H.; Powell, H. M. J. Chem. Soc. 1952, 319.
(74) Kumar, D. K.; Jose, D. A.; Das, A. Langmuir 2004, 20, 10413.
(75) Culp, J. T.; Chen, D.-L.; Liu, J.; Chirdon, D.; Goodman, A.;
Johnson, J. K. Eur. J. Inorg. Chem. 2013, 2013, 511.
(76) Vogiatzis, K. D.; Mavrandonakis, A.; Klopper, W.; Froudakis, G.
E. ChemPhysChem 2009, 10, 374.
P.; Behrooz, G. S. Inorg. Chem. 2010, 49, 4909.
(44) Windisch, C. F.; Thallapally, P. K.; McGrail, B. P. Spectrochim.
Acta, Part A 2009, 74, 629.
(45) Windisch, C. F.; Thallapally, P. K.; McGrail, B. P. Spectrochim.
Acta, Part A 2010, 77, 287.
(46) Biradha, K.; Sarkar, M.; Rajput, L. Chem. Commun. 2006, 4169.
(47) Kudo, N.; Perseghini, M.; Fu, G. C. Angew. Chem., Int. Ed. 2006,
45, 1282.
(48) Southon, P. D.; Liu, L.; Fellows, E. A.; Price, D. J.; Halder, G. J.;
Chapman, K. W.; Moubaraki, B.; Murray, K. S.; Letard, J. F.; Kepert,
C. J. J. Am. Chem. Soc. 2009, 131, 10998.
(49) Bureekaew, S.; Shimomura, S.; Kitagawa, S. Sci. Technol. Adv.
Mater. 2008, 9.
(50) Ferey, G.; Serre, C. Chem. Soc. Rev. 2009, 38, 1380.
(51) Hamon, L.; Llewellyn, P. L.; Devic, T.; Ghoufi, A.; Clet, G.;
Guillerm, V.; Pirngruber, G. D.; Maurin, G.; Serre, C.; Driver, G.; van
Beek, W.; Jolimaitre, E.; Vimont, A.; Daturi, M.; Ferey, G. J. Am. Chem.
Soc. 2009, 131, 17490.
(52) Kauffman, K. L.; Culp, J. T.; Allen, A. J.; Espinal, L.; Wong-Ng,
W.; Brown, T. D.; Goodman, A.; Bernardo, M. P.; Pancoast, R. J.;
Chirdon, D.; Matranga, C. Angew. Chem., Int. Ed. 2011, 50, 10888.
(53) Mu, B.; Li, F.; Huang, Y. G.; Walton, K. S. J. Mater. Chem. 2012,
22, 10172.
(54) Serre, C.; Bourrelly, S.; Vimont, A.; Ramsahye, N. A.; Maurin,
G.; Llewellyn, P. L.; Daturi, M.; Filinchuk, Y.; Leynaud, O.; Barnes, P.;
Ferey, G. Adv. Mater. 2007, 19, 2246.
(55) Champness, N. R.; Khlobystov, A. N.; Majuga, A. G.; Schroder,
M.; Zyk, N. V. Tetrahedron Lett. 1999, 40, 5413.
(56) Dellaciana, L.; Haim, A. J. Heterocycl. Chem. 1984, 21, 607.
(57) Rodriguez, J. G.; MartinVillamil, R.; Cano, F. H.; Fonseca, I. J.
Chem. Soc., Perkin Trans. 1 1997, 709.
(58) Fasina, T. M.; Collings, J. C.; Lydon, D. P.; Albesa-Jove, D.;
Batsanov, A. S.; Howard, J. A. K.; Nguyen, P.; Bruce, M.; Scott, A. J.;
Clegg, W.; Watt, S. W.; Viney, C.; Marder, T. B. J. Mater. Chem. 2004,
14, 2395.
(59) Gardner, T. S.; Wenis, E.; Lee, J. J. Org. Chem. 1954, 19, 753.
(60) Seki, K.; Ohkura, K.; Terashima, M.; Kanaoka, Y. Heterocycles
1986, 24, 799.
L
dx.doi.org/10.1021/ic301893p | Inorg. Chem. XXXX, XXX, XXX−XXX