We present the influence of positional isomerism on the crystal structure of fluorobenzylammonium copper(II) chloride perovskites A2CuCl4 by incorporating ortho-, meta-, and para-fluorine substitution in the benzylamine structure. Two-dimensional (2D) polar ferromagnet (3-FbaH)2CuCl4 (3-FbaH+ = 3-fluorobenzylammonium) is successfully obtained, which crystallizes in a polar orthorhombic space group Pca21 at room temperature. In contrast, both (2-FbaH)2CuCl4 (2-FbaH+ = 2-fluorobenzylammonium) and (4-FbaH)2CuCl4 (4-FbaH+ = 4-fluorobenzylammonium) crystallize in centrosymmetric space groups P21/c and Pnma at room temperature, respectively, displaying significant differences in crystal structures. These differences indicate that the position of the fluorine atom is a driver for the polar behavior in (3-FbaH)2CuCl4. Preliminary magnetic measurements confirm that these three perovskites possess dominant ferromagnetic interactions within the inorganic [CuCl4]∞ layers. Therefore, (3-FbaH)2CuCl4 is a polar ferromagnet, with potential as a type I multiferroic. This work is expected to promote further development of high-performance 2D copper(II) halide perovskite multiferroic materials.
We present the influence of positional isomerism on the crystal structure of fluorobenzylammonium copper(II) chloride perovskites A2CuCl4 by incorporating ortho-, meta-, and para-fluorine substitution in the benzylamine structure. Two-dimensional (2D) polar ferromagnet (3-FbaH)2CuCl4 (3-FbaH+ = 3-fluorobenzylammonium) is successfully obtained, which crystallizes in a polar orthorhombic space group Pca21 at room temperature. In contrast, both (2-FbaH)2CuCl4 (2-FbaH+ = 2-fluorobenzylammonium) and (4-FbaH)2CuCl4 (4-FbaH+ = 4-fluorobenzylammonium) crystallize in centrosymmetric space groups P21/c and Pnma at room temperature, respectively, displaying significant differences in crystal structures. These differences indicate that the position of the fluorine atom is a driver for the polar behavior in (3-FbaH)2CuCl4. Preliminary magnetic measurements confirm that these three perovskites possess dominant ferromagnetic interactions within the inorganic [CuCl4]∞ layers. Therefore, (3-FbaH)2CuCl4 is a polar ferromagnet, with potential as a type I multiferroic. This work is expected to promote further development of high-performance 2D copper(II) halide perovskite multiferroic materials.
Two-dimensional (2D)
organic–inorganic hybrid perovskites
are evolving into an important class of materials displaying a diverse
range of physical properties such as photovoltaic activity, luminescence,
and ferroelectricity.[1−4] Among the known families of two-dimensional layered perovskites,
two conventional families are the Dion–Jacobson (DJ)[5] and Ruddlesden–Popper (RP)[6] phases, which are commonly defined in terms of their generic
stoichiometries ABX4 and A2BX4, respectively,
for examples with single octahedral layers. The structure of these
compounds can be regarded as derived by slicing the cubic perovskite
aristotype ABX3 along vertices of the BX6 octahedra
and inserting additional moieties between these layers. An alternative
definition of these generic families lies in the relative “staggering”
of adjacent perovskite-like layers. In purely inorganic systems, the
DJ family has “eclipsed” layers (i.e., (0,0) relative
shift in the ab-plane of the idealized tetragonal
unit cell), whereas the RP family exhibits a (1/2,1/2) shift. There
is a third, intermediate case displaying a (1/2,0) shift, which has
been designated DJ2.[7] However, in the case
of hybrid perovskites, the huge variation of interlayer ion sizes
and shapes seriously affects the final structure, resulting in a much
more complex array of possible “shifts” of adjacent
layers, often resulting in adjacent layers being neither “fully
eclipsed” (DJ-like) nor “fully staggered” (RP-like).[8−11] Therefore, recent papers have suggested a more flexible structural
classification of hybrid layered perovskites, beyond the simple “DJ
or RP” types, according to the relative shifts of neighboring
inorganic layers and regardless of the stoichiometries ABX4 and A2BX4.[12−14] Introducing organic
molecules with halogen substitution into the hybrid perovskite structure
has been widely investigated to control the crystal symmetry, which
has been proved to be a powerful approach to modify the physical and
chemical properties.[15−18]Recently copper-based halide perovskites with Jahn–Teller
(J–T) distortion of the 3d9 ion Cu2+ have
been widely explored as a platform for developing new multifunctional
materials due to their interesting thermochromism, ferromagnetism,
and ferroelectricity.[19−22] To date, few copper-based halide perovskites with halogen-substituted
organic molecule spacers have been prepared. 2D layered perovskites
(CEA)2CuCl4 (CEA = 2-chloroethylammonium
and (BEA)2CuCl4 (BEA = 2-bromoethylammonium)
with chlorine and bromine substitution in ethylamine exhibit reversible
and irreversible thermochromism, which can be modulated by incorporating
organic cations.[23] The introduction of
two fluorine substituents in the centrosymmetric (CBA)2CuCl4 (CBA = cyclobutylammonium) structure results
in a polar structure (DF-CBA)2CuCl4 (DF-CBA
= 3,3-difluorocyclobutylammonium) with a ferroelectric phase
transition temperature of 380 K.[24]Despite considerable research, 2D copper layered perovskites with
monofluorine-substituted organic spacers are still relatively unknown.
Herein, for the first time, we present three new 2D hybrid layered
copper(II) chloride perovskites, (2-FbaH)2CuCl4 (2-FbaH+ = 2-fluorobenzylammonium), (3-FbaH)2CuCl4 (3-FbaH+ = 3-fluorobenzylammonium),
and (4-FbaH)2CuCl4 (4-FbaH+ = 4-fluorobenzylammonium),
by incorporating ortho-, meta-,
and para-fluorine substitution into the benzylamine
structure. The monofluorine substitution at various positions in the
benzylamine structure results in significant differences in crystal
symmetry. Here we show that (3-FbaH)2CuCl4 crystallizes
in a polar orthorhombic space group Pca21 at room temperature. In contrast, both isomers (2-FbaH)2CuCl4 and (4-FbaH)2CuCl4 crystallize
in centrosymmetric space groups at room temperature. Magnetic measurements
confirm that all three perovskites exhibit ferromagnetism, indicating
the potential multiferroic behavior of the polar ferromagnet (3-FbaH)2CuCl4.
Experimental Section
Materials
Copper(II) chloride anhydrous (CuCl2, 98%), hydrochloric
acid (HCl, 36%, w/w, aqueous solution), and
ethanol absolute (C2H5OH, 99.99%) were purchased
from Alfa Aesar. 2-Fluorobenzylamine (C7H8NF,
97%), 3-fluorobenzylamine (C7H8NF, 97%), and
4-fluorobenzylamine (C7H8NF, 98%) were purchased
from Fluorochem. All chemicals were directly used without further
purification.
Synthesis
The compounds (2-FbaH)2CuCl4, (3-FbaH)2CuCl4, and
(4-FbaH)2CuCl4 were crystallized by a slow evaporation
method.For (2-FbaH)2CuCl4 (C14H18F2N2CuCl4), CuCl2 (268.9
mg, 2 mmol) was dissolved in concentrated HCl (5 mL) and ethanol (2
mL) with moderate heating. Once fully dissolved, 2-fluorobenzylamine
(0.48 mL, 4 mmol) was added, and the solution was allowed to cool.
By cooling overnight, green plate-shaped crystals were obtained. Elemental
Anal. Calcd (%) for (2-FbaH)2CuCl4: C, 36.74;
H, 3.96; N, 6.12. Found: C, 36.82; H, 3.97; N, 5.99.For (3-FbaH)2CuCl4 (C14H18F2N2CuCl4), CuCl2 (134.45
mg, 1 mmol) was dissolved in concentrated HCl (20 mL) and ethanol
(20 mL) with slow stirring and moderate heating. Once fully dissolved,
3-fluorobenzylamine (0.24 mL, 2 mmol) was added. The produced precipitates
were dissolved by adding excess concentrated HCl to get a clear solution.
By naturally cooling the solvent for a few hours, yellow plate-shaped
crystals were obtained. Elemental Anal. Calcd (%) for (3-FbaH)2CuCl4: C, 36.74; H, 3.96; N, 6.12. Found: C, 36.88;
H, 3.91; N, 5.97.For (4-FbaH)2CuCl4 (C14H18F2N2CuCl4),
CuCl2 (134.45
mg, 1 mmol) was dissolved in concentrated HCl (20 mL) and ethanol
(20 mL) with moderate heating. Once fully dissolved, 4-fluorobenzylamine
(0.24 mL, 2 mmol) was added. The produced precipitates were dissolved
by adding excess concentrated HCl to get a clear solution. By cooling
for a few hours, yellow plate-shaped crystals were obtained. Elemental
Anal. Calcd (%) for (4-FbaH)2CuCl4: C, 36.74;
H, 3.96; N, 6.12. Found: C, 36.88; H, 3.89; N, 5.95.
Characterization
Single
Crystal X-ray Diffraction
Single crystal X-ray
diffraction data were collected on a Rigaku XtaLAB P200 diffractometer
at 93 K and on a Rigaku SCX Mini diffractometer at 173 and 298 K using
Mo Kα radiation (λ = 0.71075 Å). The data were processed
by using Rigaku CrystalClear software.[25] Crystal structures were solved by using structure solution program
SHELXT,[26] and full-matrix least-squares
refinements on F2 were performed by using
SHELXL-2018/3[26] incorporated in the WinGX
program.[27] Absorption corrections were
conducted empirically from equivalent reflections according to multiscans
using CrystalClear.[25] All the hydrogen
atoms were treated as riding atoms, and all non-H atoms were refined
anisotropically.
Powder X-ray Diffraction (PXRD)
Powder X-ray diffraction
data were measured on a PANalytical EMPYREAN diffractometer using
Cu Kα1 (λ = 1.5406 Å) radiation at ambient
temperature. The data were collected in the range of 3°–70°
for 1 h to confirm the purity of each sample. Variable-temperature
PXRD was run on a PANalytical EMPYREAN diffractometer using Mo Kα
radiation, from room temperature to 433 K.
Thermogravimetric Analyses
(TGA)
TGA data were collected
on a STA-780 instrument between 293 and 523 K at a heating rate of
5 K min–1 under flowing N2.
Electrical
Characterization
Dielectric and impedance
spectroscopy measurements were made on pellets ca. 1 mm thick and
10 mm in diameter formed by uniaxially pressing powder under a load
of 2 tons. Silver conductive paste was applied to the opposing pellet
faces and allowed to dry at 373 K. The data were recorded over the
frequency range 100 Hz and 10 MHz at 1 K increments at a heating/cooling
rate of 1 K min–1 between 298 and 473 K by using
a closed cycle cryocooler and furnace.
Magnetic Measurements
The magnetic measurements were
performed on a Quantum Design (MPMS XL) SQUID magnetometer. Data were
collected by cooling a known mass of material within a 100 Oe field
at 10 K intervals between 300 and 2 K.
Results and Discussion
Crystal
Structures
The single crystal X-ray structures
suggest no phase transitions in the regime 93 < T < 298 K, so the crystallographic details will be discussed based
on the structures at 298 K. Details of the structures at 93 and 173
K are provided in the Supporting Information. Crystallographic parameters for all three compounds at 298 K are
given in Table and
selected geometrical parameters in Table .
Table 1
Crystal and Refinement
Data for (2-FbaH)2CuCl4, (3-FbaH)2CuCl4, and
(4-FbaH)2CuCl4 at 298 K
compound
(2-FbaH)2CuCl4
(3-FbaH)2CuCl4
(4-FbaH)2CuCl4
formula
C14H18F2N2CuCl4
C14H18F2N2CuCl4
C14H18F2N2CuCl4
formula
weight
457.64
457.64
457.64
color/habit
green/platelet
yellow/platelet
yellow/platelet
crystal size (mm3)
0.25 × 0.25 × 0.08
0.33 × 0.32 × 0.15
0.17 × 0.10 × 0.06
crystal system
monoclinic
orthorhombic
orthorhombic
space group
P21/c
Pca21
Pnma
a (Å)
5.3813(4)
7.5770(5)
10.6126(7)
b (Å)
5.1660(3)
7.2883(5)
31.3460(2)
c (Å)
31.817(2)
32.525(2)
5.2326(4)
α (deg)
90
90
90
β (deg)
90.975(4)
90
90
γ (deg)
90
90
90
V (Å3)
884.38(10)
1796.1(2)
1740.69(18)
Z
2
4
4
ρcalc (g/cm3)
1.719
1.692
1.746
μ (mm–1)
1.856
1.828
1.886
F(000)
462
924
924
no. of reflns collected
7019
16073
16385
independent reflns
1971
4014
2030
[R(int) = 0.0352]
[R(int) = 0.0421]
[R(int) = 0.0673]
goodness of fit
1.052
1.063
0.998
final R indices (I >
2σ(I))
R1 = 0.0335
R1 = 0.0258
R1 = 0.0304
wR2 = 0.0777
wR2 = 0.0646
wR2 = 0.0711
largest diff peak/hole (e Å–3)
0.409/–0.344
0.254/–0.342
0.293/–0.285
Table 2
Cu–Cl Bond
Lengths and Cu–Cl–Cu
Bond Angles for the Three Structures at 298 K
(2-FbaH)2CuCl4
(3-FbaH)2CuCl4
(4-FbaH)2CuCl4
RS (Å)
2.301(2)
2.292(7)
2.298(2)
2.297(7)
RL (Å)
3.012(2)
3.001(8)
2.978(7)
2.959(8)
RZ (Å)
2.277(2)
2.294(11)
2.280(6)
2.290(11)
(2/√3)[RL – (RS + RZ)/2] (Å)[35]
0.835
0.793
0.796
Cu–Cl–Cu (deg)
162.3(4)–169.0(4)
169.5(3)–171.3(3)
174.3(4)–175.3(6)
The powder X-ray diffraction
(PXRD) patterns of (2-FbaH)2CuCl4, (3-FbaH)2CuCl4, and (4-FbaH)2CuCl4 all exhibit similar features to those calculated
based on their single crystal structures (Figure S1). Rietveld refinements were performed to confirm the purity
of the as-synthesized crystals (Figure S2). Thermogravimetric analysis (TGA) measurements indicate that these
compounds show superior thermal stability, with decomposition temperatures
up to 450, 462, and 464 K for (2-FbaH)2CuCl4, (3-FbaH)2CuCl4, and (4-FbaH)2CuCl4, respectively (Figure S3).The crystal structures of (2-FbaH)2CuCl4,
(3-FbaH)2CuCl4, and (4-FbaH)2CuCl4 can be described as 2D layered perovskite structures, with
single layers of corner-shared CuCl6 octahedra forming
a layer stoichiometry [CuCl4]∞, which
are separated by a double layer of the protonated fluorobenzylammonium
moieties (Figure ),
as found in their non-fluorinated counterpart (BaH)2CuCl4 (BaH+ = benzylammonium).[28] We will first discuss general features and differences of the three
crystal structures from the perspective of the inorganic “framework”,
as these can be discussed systematically based on the known principles
of perovskite crystallography. It is often helpful to rationalize
the unit cell metrics of layered perovskites in relation to the idealized
parent DJ or RP phases, which have tetragonal symmetry as well as
space groups P4/mmm and I4/mmm and incorporate one and two [CuCl4]∞ layers per unit cell, respectively. The
parent benzylammonium compound (BaH)2CuCl4 adopts
the triclinic space group P-1 and has unit cell metrics
at 293 K, a = 10.501(2) Å, b = 10.576(2) Å, c = 16.115(3) Å, α
= 98.00(3), β = 99.33(3)°, and γ = 90.06(3)°.
Thus, this unit cell can be regarded as a 2aDJ × 2aDJ × cDJ supercell of the parent ABX4 DJ structure.
In contrast, the monofluorine substitution on various positions of
the benzylamine molecule results in significant differences in crystal
symmetry, and each of the present fluorinated derivatives crystallizes
with two [CuCl4]∞ layers per unit cell,
rather than one. In that sense, they may be better regarded as derived
from the RP rather than the DJ parent. The data in Table reveal that the corresponding
supercell metrics for the present compounds are aRP × aRP × cRP for (2-FbaH)2CuCl4,
√2aRP × √2aRP × cRP for (3-FbaH)2CuCl4, and 2aRP × cRP × aRP in
the case of (4-FbaH)2CuCl4. Further details
of the underlying distortions are addressed below.
Figure 1
Crystal structures of
(a) (2-FbaH)2CuCl4,
(b) (3-FbaH)2CuCl4, and (c) (4-FbaH)2CuCl4 at 298 K parallel to the layer direction.
Crystal structures of
(a) (2-FbaH)2CuCl4,
(b) (3-FbaH)2CuCl4, and (c) (4-FbaH)2CuCl4 at 298 K parallel to the layer direction.In addition to rationalization of unit cell metrics,
the space
groups of layered perovskites can also often be understood in terms
of the underlying types of distortion of the inorganic framework,
bearing in mind, of course, that it is ultimately the cooperative
relationship of the organic guests and the inorganic framework that
dictates these distortions. In general, for layered hybrid perovskites
there are two main types of distortion mode, viz. tilting of octahedral
MX6 units and relative lateral shifts of adjacent [MX4]∞ layers. In addition, in copper halide
perovskites, there is also the J–T mode, which leads to alternating
long and short in-plane Cu–Cl bonds. In the case of (3-FbaH)2CuCl4, use of the online tool ISODISTORT[29] reveals that the key “standard”
distortion modes[7] are the layer shift mode,
together with two distinct octahedral tilt modes (i.e., both rotation
around the c-axis and out-of-plane tilting, leading
to the common Glazer-like tilt system a–a–c/–(a–a–c)). The additional significant modes are the J–T
mode and a polar mode, acting along the c-axis. We
note that the [CuCl4]∞ framework is only
marginally affected by the polar mode, which is clearly driven by
the positioning of the interlayer moieties; indeed, the space group
would be close to centrosymmetric, either Pbca or Pbcm based on the distortions of the inorganic layer only.
Attempts to derive a plausible model in these centrosymmetric space
groups proved impossible. An example of the Pbca space
group (which permits the two octahedral tilt modes, but not the layer
shift mode) and corresponding unit cell metrics occurs in (NO2C6H4NH3)CuCl4.[30] In the remaining two present examples, matters
are complicated by some structural disorder within the [CuCl4]∞ sheets (Figure S4). After trial refinements, the occupancies of the disordered sites
were fixed at 50:50, in each case; whether this disorder is correlated
to the retention of centrosymmetricity in these cases is unclear.
In the case of (4-FbaH)2CuCl4, there is no octahedral
tilting, but there is a layer shift mode and an additional slight
“rippling” of the [CuCl4]∞ sheets, which leads to the doubled a-axis. Such
a rippling effect is rare in layered perovskites, but this has been
discussed in our recent review[7] and has
been seen in the lead halide (2-fluoroethylammonium)2PbBr4,[31] which has the same
generic cell metrics and same space group, Pnma.
A more dramatic example also occurs in a recently reported polymorph
of the copper halide [C6H5(CH2)4NH3]2[CuCl4].[32] The additional disorder within the [CuCl4]∞ layers is effectively a disorder of the
J–T affected “long” and “short”
in-plane bonds. Such an effect has previously been noted in copper
halides, for example, the chiral ferromagnets (R-MPEA)2CuCl4 and (S-MPEA)2CuCl4.[33] The disorder in (2-FbaH)2CuCl4 is more complex, and it exhibits aspects
of both the J–T disorder and localized, disordered octahedral
tilting, similar to that we recently reported in (3-AbaH)2CuCl4 (3-AbaH+ = protonated 3-aminobenzoic
acid), which crystallizes in the triclinic space group P-1.[34]Figure shows that the “layer shift”
degree of freedom for each of the three structure differs significantly:
these differences are discussed below in the context of the interactions
of the interlayer organic moieties with the [CuCl4]∞ layers.
Figure 2
Crystal structures of (a) (2-FbaH)2CuCl4,
(b) (3-FbaH)2CuCl4, and (c) (4-FbaH)2CuCl4 at 298 K perpendicular to the layer direction.
Crystal structures of (a) (2-FbaH)2CuCl4,
(b) (3-FbaH)2CuCl4, and (c) (4-FbaH)2CuCl4 at 298 K perpendicular to the layer direction.The Cu2+ ions in halide perovskite structures
are strongly
J–T active, giving rise to dramatic differences in the Cu–Cl
bond lengths and hence to structural distortions of the CuCl6 octahedra, with the shorter in-plane bond designated RS, the longer one RL, and
the out-of-plane bond RZ (Table ). We quantitatively measure
the magnitude of the distortion of the octahedra using the equation
(2/√3)[R- (R + R)/2],[35] indicating
that there is no substantial difference in structural distortion for
the three compounds.At each temperature studied, (3-FbaH)2CuCl4 crystallizes in the polar orthorhombic system Pca21 space group, whereas both (2-FbaH)2CuCl4 and (4-FbaH)2CuCl4 crystallize
in the
centrosymmetric monoclinic P21/c and orthorhombic Pnma space groups, respectively.
The protonated fluorobenzylammonium cations in all three structures
are ordered and connected with the [CuCl4]∞ layers through N–H···Cl hydrogen bonds. In
the case of (3-FbaH)2CuCl4 the nature of the
lowering of symmetry can be seen in Figure , which shows an expanded view of two FbaH
moieties and the nature of their H-bonding to a single [CuCl4]∞ layer. In the corresponding centrosymmetric
structure (i.e., that allowing the layer shift mode and the octahedral
rotation, but not octahedral tilt, mode, and space group Pbcm, discussed above) the Cu atoms would sit on an inversion center,
and the two FbaH moieties would be related by a mirror plane through
the CuCl4 plane (Figure ). It can be seen that the positioning of the two FbaH
moieties breaks this symmetry slightly, with a similar, but distinct,
H-bonding network on either side of the plane. In particular, there
is a significantly short N–H---Cl contact (H---Cl ∼
2.38 Å) on one side of the plane, with the corresponding contact
on the opposite side of the plane being 2.55 Å. Although the
underlying reason for this symmetry breaking is not clear, we might
speculate that this is correlated to the ordering of the [CuCl4]∞ layer itself, which is not seen in either
the 2-FbaH or 4-FbaH analogues. Further details of the symmetry-breaking
are given in Figure S5 and Table S13.
Figure 3
Symmetry
breaking in (3-FbaH)2CuCl4. On the
left, the true structure is overlaid with a “dummy”
3-Fba moiety generated by reflection through the Cu–Cl plane.
The resulting two overlapping FbaH molecules are shown on the right,
which clearly shows one of the key features breaking symmetry to be
the positioning of the fluorine atom, in addition to deviations of
the −CH2NH3 group. All hydrogen atoms
are omitted for clarity.
Figure 4
Hydrogen-bonding interactions
for (a) (2-FbaH)2CuCl4, (b) (3-FbaH)2CuCl4, and (c) (4-FbaH)2CuCl4 at
298 K. Note the symmetrical arrangements
of the FbaH moieties in (a) and (c) and the symmetry breaking in (b).
See the Supporting Information for full
details of the H-bonding schemes.
Symmetry
breaking in (3-FbaH)2CuCl4. On the
left, the true structure is overlaid with a “dummy”
3-Fba moiety generated by reflection through the Cu–Cl plane.
The resulting two overlapping FbaH molecules are shown on the right,
which clearly shows one of the key features breaking symmetry to be
the positioning of the fluorine atom, in addition to deviations of
the −CH2NH3 group. All hydrogen atoms
are omitted for clarity.Hydrogen-bonding interactions
for (a) (2-FbaH)2CuCl4, (b) (3-FbaH)2CuCl4, and (c) (4-FbaH)2CuCl4 at
298 K. Note the symmetrical arrangements
of the FbaH moieties in (a) and (c) and the symmetry breaking in (b).
See the Supporting Information for full
details of the H-bonding schemes.As shown in Figure , in (2-FbaH)2CuCl4 and (4-FbaH)2CuCl4 the corresponding interactions of the organic moieties
with the [CuCl4]∞ layer retain their
ideal higher symmetry (specifically, the two moieties are related
by an inversion center in (2-FbaH)2CuCl4 and
a mirror plane in (4-FbaH)2CuCl4). However,
the (4-FbaH+) moieties in the (4-FbaH)2CuCl4 structure only form N–H···Cl hydrogen
bonds with in-plane chlorides, Cl(2) and Cl(3), with no further H-bonding
to the apical chloride Cl(1). In each structure, the two FbaH moieties
per formula unit form molecular “bilayers” in the interlayer
region between the [CuCl4]∞ layers (Figure ). In both the (2-FbaH)2CuCl4 and (4-FbaH)2CuCl4 structures,
the organic amines are arranged with the aromatic groups and F substituents
oriented in the middle of these organic bilayers, forming stronger
hydrogen bonds, N–H···Cl < 2.78 and 2.83
Å for (2-FbaH)2CuCl4 and (4-FbaH)2CuCl4, respectively (Tables S4 and S9). However, the (3-FbaH+) moieties in the noncentrosymmetric
structure (3-FbaH)2CuCl4 form a wider range
of hydrogen bonds, with weaker contacts up to N–H···Cl
= 2.96 Å in addition to the shortest contact of 2.38 Å (Table S6). Thus, the orientation of the cations
within each organic bilayer contributes to the changes in crystal
symmetry.F---F intermolecular interactions are known to play
a minor, complementary
role in determining crystal packing.[36] In
(3-FbaH)2CuCl4 and (4-FbaH)2CuCl4, the intra-bilayer cohesion may be further
enhanced by weak intermolecular F---F interactions (∼3.91and
3.64 Å, respectively) (Figure ). In contrast, in the (2-FbaH)2CuCl4 structure, the amines are unable to orient to produce optimized
F---F interactions, and the F atoms are inevitably directed toward
the [CuCl4]∞ planes rather than the intra-bilayer region. The subtle differences of the overall inter-layer cohesion result in different degrees of layer
shift of neighboring [CuCl4]∞ layers
in all three structures. Figure shows that neighboring [CuCl4]∞ layers in both the (3-FbaH)2CuCl4 and (4-FbaH)2CuCl4 structures are staggered relative to each
other in a style that spans the spectrum from DJ-like to RP-like.
Both structures appear to be near RP, with a greater degree of layer
shift in the (4-FbaH)2CuCl4 structure, consistent
with the enhanced intermolecular F---F interaction (∼3.64 Å).
In contrast, in the (2-FbaH)2CuCl4 structure,
the adjacent octahedral layers are partially eclipsed, in the DJ2
style, despite the stoichiometry.The positional isomerism of
closely related amines clearly has
a significant effect on the structural behavior observed. A similar
effect has been observed in corresponding lead chloride systems.[37] However, the structural details and by implication
the underlying factors driving polar behavior in only one of the isomers
are subtly different. In the present systems, introducing fluorine
at the ortho- and para-positions
retains centrosymmetric symmetry, whereas via meta-substitution the crystal symmetry is reduced to noncentrosymmetric.
In the case of the lead chloride analogues, it is the (2-FbaH)2PbCl4 isomer that is polar (and ferroelectric),
whereas the meta and para isomers
are centrosymmetric. Furthermore, the polar axis in (2-FbaH)2PbCl4 lies along an in-plane direction, whereas in (3-FbaH)2CuCl4 it lies perpendicular to
the [CuCl4]∞ layers (i.e., along the c-axis). The reasons for this are presumably dictated by
the subtle differences in H-bonding opportunities, but identifying
an exact comparison is difficult due to the two unique features of
the Cu-based systems, i.e., the J–T effect and the disordering
within the [CuCl4]∞ layers, neither of
which is present in the Pb-based systems. A potential lone pair effect
from Pb2+ may also contribute toward the observed differences
in structure.
Electrical Properties of (3-FbaH)2CuCl4
The relative permittivity measured of the
polar compound
(3-FbaH)2CuCl4 at 100 kHz appears to show a
peak at ca. 426 K on both heating and cooling (Figure S6). However, no peak was observed in data obtained
at several other frequencies and instead showed a large dispersion
with permittivity values in excess of 10000 at 1 kHz and below; this
indicated that the peak observed at 100 kHz is not intrinsic but is
instead linked to the increasing influence of electrode polarization
and changing contact resistance.[38−40] Collection of dielectric
data at higher temperature was not possible as the sample decomposed.
Variable temperature PXRD was also performed to check for the presence
of a structural transition, possibly associated with a ferroelectric TC. On heating from 298 to 413 K the diffraction
patterns remain unchanged, but the data at 433 K is notably different
with the appearance of new diffraction peaks. The single-crystal structure
of (3-FbaH)2CuCl4 above 400 K has not been determined
because the diffraction in the high-temperature phase is too weak.
Despite the polar nature of (3-FbaH)2CuCl4,
the combination of dielectric and diffraction data does not support
the presence of a ferroelectric-to-paraelectric transition in the
temperature range investigated. It may be that the compound decomposes
before this transition.The electrical properties of (3-FbaH)2CuCl4 were also investigated by using impedance
spectroscopy measurements from room temperature up to 408 K. Above
382 K, the complex impedance, Z*, response indicated
a single semicircular arc which decreased in size with increasing
temperature (Figure a). This indicates a decrease in resistivity with increasing temperature
and semiconducting behavior. The semicircular arc was fitted assuming
a single parallel resistor–capacitor (R–C) circuit and
had an associated capacitance of 1–2 pF at all temperatures
and is consistent with a bulk response.[41] Bulk resistivity (ρ) values were estimated from the intercepts
of the semicircular arc with the x-axis (the real
part of the complex impedance, Z′) and converted
to bulk conductivity, σ = 1/ρ, and plotted in Arrhenius
format (Figure b).
The data show linear behavior consistent with thermally activated
semiconducting properties and with an activation energy of 1.22 ±
0.1 eV. Unfortunately, the level of conductivity means that polarization
switching was not possible to conclusively demonstrate ferroelectricity.
Figure 5
(a) Complex
impedance plane (Z*) plots for (3-FbaH)2CuCl4 as a function of temperature showing a single
semicircular arc associated with a bulk response. (b) Arrhenius plot
of bulk conductivity as a function of temperature (R2 = 0.9993).
(a) Complex
impedance plane (Z*) plots for (3-FbaH)2CuCl4 as a function of temperature showing a single
semicircular arc associated with a bulk response. (b) Arrhenius plot
of bulk conductivity as a function of temperature (R2 = 0.9993).
Magnetic Properties
The magnetic measurements have
been performed in the temperature region 2–300 K with an applied
dc field of 100 Oe. Figures a–c show the magnetic susceptibility (χ) measured
as a function of decreasing temperature T in an applied
magnetic field, where χ increases gradually with decreasing T and then increases very rapidly below around 10 K. No
maximum is observed in the χ vs T plots, consistent
with the spontaneous onset of ferromagnetic order. The dominance of
ferromagnetic interactions within the [CuCl4]∞ layers resembles previously reported 2D layered copper(II) perovskites[42−44] and is further reflected in the high-temperature magnetic response.
The inverse susceptibility (1/χ) data were fitted to the Curie–Weiss
law in the region 50–300 K for (2-FbaH)2CuCl4 and (3-FbaH)2CuCl4 and 50–270
K for (4-FbaH)2CuCl4. This gave the fitted values:
Curie constant, C = 0.53(5), 0.50(2), and 0.46(3)
cm3 mol–1 K, Weiss constant θ =
10.8(2), 8.8(1), and 32.2(3) K, derived effective moment, μeff = 2.059(5), 2.000(2), and 1.918(3) μB,
and Landé factor, g = 2.378, 2.309, and 2.215
for (2-FbaH)2CuCl4, (3-FbaH)2CuCl4, and (4-FbaH)2CuCl4, respectively,
which are within the expected range for one Cu2+ ion with S = 1/2.[42,43] The peaks in χT vs T plots (Figure d–f), occurring at 5 K for (2-FbaH)2CuCl4 and 7 K for (3-FbaH)2CuCl4 and (4-FbaH)2CuCl4, simply reflect the approach
to saturation of the ordered ferromagnetic moment as the temperature
is lowered, which is further reinforced by the field-dependent magnetic
response discussed below. The polar ferromagnet (3-FbaH)2CuCl4 therefore exhibits the potential for multiferroic
behavior.
Figure 6
Magnetic susceptibility (χ) and its inverse 1/χ with
Curie–Weiss fit (red line) for (a) (2-FbaH)2CuCl4 in the region 50–300 K, (b) (3-FbaH)2CuCl4 in the region 50–300 K, and (c) (4-FbaH)2CuCl4 in the region 50–270 K. χT vs T for (d) (2-FbaH)2CuCl4, (e) (3-FbaH)2CuCl4, and (f) (4-FbaH)2CuCl4.
Magnetic susceptibility (χ) and its inverse 1/χ with
Curie–Weiss fit (red line) for (a) (2-FbaH)2CuCl4 in the region 50–300 K, (b) (3-FbaH)2CuCl4 in the region 50–300 K, and (c) (4-FbaH)2CuCl4 in the region 50–270 K. χT vs T for (d) (2-FbaH)2CuCl4, (e) (3-FbaH)2CuCl4, and (f) (4-FbaH)2CuCl4.The magnetization vs magnetic field plots for these three compounds
at 2, 5, 10, and 20 K are shown in Figure . Below the ordering temperatures of around
10 K all three compounds exhibit sharp magnetic switching at low field
with a small amount of hysteresis, in contrast to the linear paramagnetic
response at 20 K. All compounds display hysteresis loops at 2 and
5 K, with a coercive field of approximately 40 Oe for (2-FbaH)2CuCl4, 25 Oe for (3-FbaH)2CuCl4, and (4-FbaH)2CuCl4 at 2 K (Figure S9). At 2 K, the magnetic moment at low field has saturation
values of approximately 1.00, 1.03, and 1.01 μB for
(2-FbaH)2CuCl4, (3-FbaH)2CuCl4, and (4-FbaH)2CuCl4, respectively,
which are comparable to the expected value (1.0 μB) for an S = 1/2 system.[42−44] Furthermore,
the hysteresis effect observed in all compounds is characteristic
of a soft ferromagnet.
Figure 7
Magnetization (M) vs magnetic field (H) at 2, 5, 10, and 20 K for (a) (2-FbaH)2CuCl4, (b) (3-FbaH)2CuCl4, and (c) (4-FbaH)2CuCl4, respectively. Inset: the low-field region
of the hysteresis loops.
Magnetization (M) vs magnetic field (H) at 2, 5, 10, and 20 K for (a) (2-FbaH)2CuCl4, (b) (3-FbaH)2CuCl4, and (c) (4-FbaH)2CuCl4, respectively. Inset: the low-field region
of the hysteresis loops.The most reliable estimate
of the ground state comes from saturation
magnetization below Tc, which is in excellent
agreement with that expected value of Ms = 1 μB for an S = 1/2 spin only
ground state. The values of effective moment derived at higher temperature
in the paramagnetic state are about 10% higher than the expected value
of 1.73Ms under the assumption of no additional
contributions to the moment at higher temperature (i.e., constant
Curie constant). The existence of ferromagnetic correlations in the
[CuCl4]∞ layers can be understood as
arising from the motif of alternating in-plane long and short bonds
at neighboring Cu sites such that the superexchange along the nearly
linear Cu–Cl–Cu path is controlled by the Hund’s
rule correlations at the ligand. As here, the development of long-range
ferromagnetic order is also observed in other 2D layered copper(II)
perovskites,[42−44] which due to the Mermin–Wagner theorem demands
the existence some additional source of anisotropy. Recent DFT calculations
on related [CuCl4]∞-based systems suggest
this may be a single ion anisotropy arising from spin–orbit
interactions due to the large covalency between the Cu and the Cl.[45] The resulting additional moment, estimated as
∼0.1 μB, may also help account for the higher
effective moment observed in the paramagnetic state in our materials.
Conclusions
In summary, we have explored the effect of structural
isomerism
of the interlayer organic cation in the series of layered perovskites
(n-FbaH)2CuCl4 (n = 2, 3, 4). Fluorination of benzylamine at the ortho- and para-positions leads to layered perovskites
(2-FbaH)2CuCl4 and (4-FbaH)2CuCl4 which retain centrosymmetric structures. However, the effects
of resulting interactions with the inorganic framework, and also of
differing weaker interactions within the organic bilayers, lead to
differences in relative shifts of the [CuCl4]∞ layers within each of the compounds and also different octahedral
tilting arrangements, with both these compounds exhibiting disordering
of Cl ligands. More significantly, meta-F-substitution
in (3-FbaH)2CuCl4 leads to a polar material,
with the symmetry being broken by subtle displacements in the positions
of the 3-FbaH moieties, with resulting advantages in H-bonding opportunities
to the [CuCl4]∞ layers. This symmetry-breaking
mechanism seems related to, but distinct from, that recently seen
in some manganese chloride layered perovskites[46] where the rotational degree of freedom of extra-cyclic
chains is suggested to be responsible. In addition, in this case,
we see that the F-substitution of the phenyl ring also plays a key
role. All three compounds display ferromagnetic interactions, and
therefore polar (3-FbaH)2CuCl4 may be a type
I multiferroic, although ferroelectric switching has yet to be demonstrated.
It is notable that there are no phase transitions in any of the three
present compounds in the range 93–293 K. In each case the organic
moieties are well-ordered throughout this regime, but in (2-FbaH)2CuCl4 and (4-FbaH)2CuCl4 the
inorganic layers remain disordered. It might be anticipated that a
more ordered ground state would freeze out on further cooling, and
perhaps this would ultimately lead to a polar phase for these compositions
at temperatures below 93 K. The non-fluorinated parent structure (BaH)2CuCl4 also exhibits disordered inorganic [CuCl4]∞ layers and a centrosymmetric space group.
From this we might speculate that there is a cooperativity between
ordering of the [CuCl4]∞ layers and ordering
of the interlayer organic moieties into a stable noncentrosymmetric
disposition, a phenomenon which is worthy of further study.
Authors: Nicole A Benedek; James M Rondinelli; Hania Djani; Philippe Ghosez; Philip Lightfoot Journal: Dalton Trans Date: 2015-02-17 Impact factor: 4.390
Authors: Ceng Han; Alasdair J Bradford; Jason A McNulty; Weiguo Zhang; P Shiv Halasyamani; Alexandra M Z Slawin; Finlay D Morrison; Stephen L Lee; Philip Lightfoot Journal: Chem Mater Date: 2022-02-21 Impact factor: 9.811