Synthesis of a two-dimensional organic-inorganic bismuth iodide metalate through: In situ formation of iminium cations

Article abstract of DOI:10.1039/c9cc06625j

(Me₂CNMe₂)Bi₂I₇ represents a new layered organic-inorganic iodido bismuthate. It displays an unprecedented anion topology, a low band gap and good stability. Advanced electronic structure analysis finds the I?I interactions to be decisive for the compound's structural and electronic properties.

Full text of DOI:10.1039/c9cc06625j

Showcasing research from Ralf Tonner’s and Johanna Heine’s
group, Department of Chemistry and Collaborative Research
Center 1083, Philipps-Universität, Marburg, Germany.
Image designed and illustrated by Leander Aurel Taubner.
As featured in:
Volume 55 Number 98 21 December 2019 Pages 14703–14858
ChemComm
Chemical Communications
rsc.li/chemcomm
Synthesis of a two-dimensional organic–inorganic bismuth
iodide metalate through in situ formation of iminium cations
(Me2C==NMe2)Bi2I7 represents a new layered organic–inorganic
iodido bismuthate. It displays an unprecedented anion topology,
a low band gap and good stability. Advanced electronic

structure analysis finds the I I interactions to be decisive
ISSN 1359-7345
for the compound’s structural and electronic properties.
COMMUNICATION
Timothy D. W. Claridge, Christopher J. Schofield et al.
19F NMR studies on γ-butyrobetaine hydroxylase provide
mechanistic insights and suggest a dual inhibition mode
See Ralf Tonner,
Johanna Heine et al.,
Chem. Commun., 2019, 55, 14725.
rsc.li/chemcomm
Registered charity number: 207890

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
Synthesis of a two-dimensional organic–inorganic
bismuth iodide metalate through in situ formation
of iminium cations†
Cite this: Chem. Commun., 2019,
55, 14725
Received 26th August 2019,
Accepted 26th October 2019
Natalie Dehnhardt,
and Johanna Heine
Jan-Niclas Luy, Marvin Szabo, Mirco Wende, Ralf Tonner
*
*
DOI: 10.1039/c9cc06625j
rsc.li/chemcomm
Q
(Me2C NMe2)Bi2I7 represents a new layered organic–inorganic organic–inorganic, layered iodido bismuthates should be avail-
iodido bismuthate. It displays an unprecedented anion topology, a able if the right type of counterion can be found.
low band gap and good stability. Advanced electronic structure
A family of organic cations that has been largely overlooked
analysis finds the Iꢀ ꢀ ꢀI interactions to be decisive for the compound’s in the context of metal halide materials are iminium ions
structural and electronic properties.
(R2CQNR2+; R = H, organic group). They are well known as
reactive intermediates and reagents in organic chemistry,16 but
Lead halide perovskites like (CH3NH3)PbI3 are prominent materials they are also known for their tendency to hydrolyze.17 This
due to their excellent semiconductor properties1 and successful reactivity suggests that they are difficult to work with in the
application in solar cells,2 photocatalysis3 and LEDs.4 Layered preparation of robust materials, yet in early studies of iminium
perovskites such as (C4H9NH3)PbI45,6 offer the benefit of greater cations, halogenido metalate anions were used to isolate these
stability and allow the use of larger, functional organic cations species.18 Later work demonstrated that compounds containing
while maintaining the good semiconductor properties of their iminium cations are readily available with numerous substitution
parent perovskites.7 However, the toxicity of lead compounds patterns19 and easy to isolate as single crystals.20 More recently,
has prompted researchers to look to related halogenido meta- cuprates and argentates featuring (Me2CQNMe2)+ were prepared by
lates for alternatives. Iodido bismuthates appear promising, as in situ condensation of dimethylamine and acetone.21
bismuth compounds are typically non-toxic.8 Compounds like
Here, we present (Me2CQNMe2)Bi2I7 (1), a layered organic–
(CH3NH3)3Bi2I9, which features dinuclear Bi2I9 anions,9 have inorganic iodido bismuthate, the first example of a layer with
been tested in photovoltaic devices, but only show moderate fully occupied metal positions in this class of compounds. We
efficiencies up to 3.17% after significant efforts at optimization.10 investigate the synthesis and reactivity, crystal structure and
In contrast, lead halide perovskite solar cells currently reach well optical properties of 1, revealing its unique topology, high
above 20%.11 One reason for this difference in properties lies in stability and low band gap. We use quantum chemical methods
the structural chemistry of the iodido bismuthates: while iodido to analyze the compound’s electronic properties and demon-
plumbates readily feature layer or network anions, bismuthates are strate a new method to characterize and quantify iodine–iodine
mostly limited to molecular or chain-like anions.12 Only two contacts in extended solids. We also discuss why our findings
types of iodido bismuthates with a layered anion have been are relevant to lead halide perovskites and how the facile
reported to date: the all-inorganic family A3Bi2I9 (A = K, Rb)13 preparation of iminium cations opens up new opportunities
featuring (111) perovskite layers, and (H2AEQT)Bi2/3I4 (AEQT = for metal halide materials in general.
3ꢁ
5,5000-bis-(aminoethyl)-2,20:50,200:500,2000-quaterthiophene), featuring
a metal-deficient (001) perovskite layer.14 However, the synthesis amounts of BiI3 and (NMe2H2)I in acetone (Scheme 1).
(Me2CQNMe2)Bi2I7 (1) is obtained by heating stoichiometric
of the layered mixed halide bismuthate (TMP)1.5[Bi2I7Cl2],
1 crystallizes in P2/n (No. 13) as black blocks. Fig. 1 shows an
(TMP = N,N,N0,N0-tetramethyl-piperazine)15 suggests that more overview of the crystal structure. Bond lengths in both the
cation and anion are in good agreement with literature reports
¨
Department of Chemistry and Material Sciences Center, Philipps-Universitat
Marburg, Hans-Meerwein-Straße, 35043 Marburg, Germany.
E-mail: tonner@chemie.uni-marburg.de, johanna.heine@chemie.uni-marburg.de
† Electronic supplementary information (ESI) available: Synthesis, crystallographic
details, thermal analysis, powder patterns, IR and Raman spectra, absorption spectra,
details of computational studies. CCDC 1939881. For ESI and crystallographic data in
CIF or other electronic format see DOI: 10.1039/c9cc06625j
Scheme 1 Synthesis of 1.
This journal is ©The Royal Society of Chemistry 2019
Chem. Commun., 2019, 55, 14725--14728 | 14725

View Article Online
Communication
ChemComm
Fig. 1 Asymmetric unit of 1, ellipsoids at 50% probability, only majority position shown (left). View along the c axis, cations omitted (middle). Excerpt with
I–I interactions below 4 Å marked with dashed lines (right).
(Tables S3 and S4, ESI†). The iodido bismuthate layers feature
the well-known motif of {BiI4} chains of edge-sharing {BiI6}
octahedra.12 These chains are connected into layers by additional
edge-sharing with adjacent strands. The resulting topology is a
63-honeycomb net. To the best of our knowledge, the anion motif
found in 1 is unique among metalates, although related metal
halide layers with six-membered rings are known, e.g. in BiI3,22
(pipzH2)Mn2F8 (pipz = piperazine)23 or the triple-layered anion in
(H2mpz)2Pb3Br10 (mpz = 1-methylpiperazine).24 The layers in 1
are further connected into a pseudo-network via iodine–iodine
interactions.
1 starts to decompose at 290 1C (Fig. S2, ESI†), similar to
layered organic–inorganic iodido plumbates.5 We also tested
the stability of 1 against humidity, as the tendency of iminium
cations to hydrolyze might be expected to translate to the
metalate as well. However, when a powdered sample of 1 is
Fig. 2 Absorption spectra of BiI3, 1 and (HPy)BiI4, measured in diffuse
stored in humid air for 24 h, no change in powder pattern is
observed (Fig. S4, ESI†). Additionally, long term stability was
investigated on a sample that was aged in air for 5 months.
reflection.
Here, no significant decomposition was observed in the powder full periodicity of the system (see ESI†). The calculated band gap
diffractogram as well (Fig. S4, ESI†). The absorption spectrum of 1.87 eV (TB09) is in excellent agreement with the experimental
of 1, measured in diffuse reflectance, shows an onset of value of 1.8 eV (Table S6, ESI†). An analysis of the projected
absorption at 1.8 eV (Fig. S7, ESI†). A comparison with the density of states (pDOS) shows that the valence band edge is
spectra of BiI3 and (Hpy)BiI425 (Hpy = pyridinium) highlights the mostly composed of orbitals located at the iodine atoms, while
effect of metalate formation and changing anion dimensionality the conduction band consists of equal contributions of Bi- and
(Fig. 2): the onset of absorption is blue-shifted going from BiI3 to I-based orbitals (Fig. 3). Plotting the orbitals at the G-point
1 to (Hpy)BiI4, which features chain-like anions, as expected from confirms these assignments in a real-space picture and demon-
general principles of dimensionality reduction in metalates.26 strates that the highest occupied crystal orbital (HOCO) consists
Weller has shown that iodido bismuthate band gaps correlate not of non-bonding electron pairs at the iodine atoms while the
only with the anion dimensionality, but also with the presence of lowest unoccupied crystal orbital (LUCO) additionally features
iodine–iodine interactions between anions, which typically lower contributions from antibonding Bi–I orbitals. These orbitals are
the band gap.27 We suggest that it is these interactions, also delocalized over the whole unit cell, indicating good carrier
found in 1, that cause a red-shifted onset of absorption in mobility. This picture is similar to results derived for Rb3Bi2I9.13
comparison with Rb3Bi2I9, composed of (111) perovskite layers, In contrast, layered lead-based perovskites typically show a higher
where a band gap of 2.1 eV13 has been reported and no interlayer contribution of the metal-based orbitals to the valence and
iodine–iodine interactions below 4 Å occur. Revealing the nature conduction bands.6 The interlayer interaction between the iodine
of these interlayer interactions and their impact on the band gap atoms was previously found to be an important element for the
motivated our computational investigations for 1.
crystal packing and electronic structure of iodido metalates.27
But the exact nature of these interactions has not been
Following our previous work,28 we chose density functional
theory (DFT) approaches based on the PBE functional with revealed up to now. Thus, we performed an in-depth analysis
dispersion correction by the DFT-D3 method and consideration of the underlying principles governing the Iꢀ ꢀ ꢀI interaction by a
of scalar and spin–orbit relativistic effects taking into account the combination of several advanced tools for bonding analysis. In
14726 | Chem. Commun., 2019, 55, 14725--14728
This journal is ©The Royal Society of Chemistry 2019

View Article Online
ChemComm
Communication
Fig. 3 Negative of the pCOHP (PBE-D3) values of the interlayer Iꢀ ꢀ ꢀI interaction (left); total and partial density of states (pDOS, TB09) with insets of the
highest occupied (HOCO) and lowest unoccupied crystal orbital (LUCO) at the G-point (right).
Fig. 3, we show the projected Crystal Orbital Hamilton Population look into the attractive pEDA terms shows that the electrostatic
(pCOHP29) of the Iꢀ ꢀ ꢀI interaction. In agreement with the results interaction is dominating (56%) which can be understood by the
from the pDOS analysis, an antibonding contribution stemming ionic nature of the organic and inorganic layers, respectively.
from non-bonding electron pair repulsion is found close to the Nevertheless, a significant orbital interaction (44%) is found,
Fermi energy. The integrated pCOHP (IpCOHP) amounts to indicating covalent bonding contributions. These orbital interactions
ꢁ0.131 eV indicating an overall slightly attractive interaction. were further analyzed with the Natural Orbitals for Chemical Valence
The topological analysis by the atoms in molecules (AIM) (NOCV) approach,33,34 identifying iodine–iodine charge shift as the
method30 shows an increased but small electron density at the major contribution (Fig. S8 and S9, ESI†).
bond critical points for the interlayer Iꢀ ꢀ ꢀI interactions compared
But how does this I–I interaction effect the band edges? To
to the intralayer contacts and a positive value for the second this end we reduced this interaction by increasing the interlayer
2
derivative of the electron density (Laplacian, r r) supporting the I–I distance from 3.692 to 4.047 Å (9.6%) while keeping
notion of a non-covalent interaction (Table S7, ESI†).
the other structural parameters constant. A pCOHP analysis
A quantitative analysis of this interaction was carried out (Fig. S10, ESI†) reveals that the antibonding contribution close
with our recently developed energy decomposition analysis for to the Fermi energy decreases more than its bonding counter-
extended systems (pEDA). This analysis decomposes the bonding parts leading to a decreased IpCOHP of ꢁ0.051 eV. In addition,
energy between two fragments in terms of well-defined energy the HOCO is shifted to slightly lower energies (ꢁ0.1 eV). This
contributions that help to gain a qualitative understanding of the shows that the interlayer Iꢀ ꢀ ꢀI interaction leads to a smaller band
bonding interaction from quantitative results. It is thus possible gap, implying the possibility of band gap tuning by controlling
to distinguish effects from quasiclassical electrostatic interac- the iodine–iodine distances.
tions (DEelstat), Pauli repulsion (DEPauli) and orbital interactions
We want to highlight another aspect of our findings: the
(DEorb). Dispersion contributions to the bonding are derived from facile synthesis and inclusion of the iminium cations in 1, and
the DFT-D3 scheme (DEint(disp)) and added to the sum of the its relevance to the synthesis of metal halide perovskites.
electronic interaction terms (DEint(elec)) to give the total inter- Dimethylformamide, a common solvent in metal halide perovskite
action energy (DEint).31–33 Notably, the layered structure of 1 synthesis, can form dimethylamine as a decomposition product35
enables an analysis of a 3-dimensional solid not composed of and subsequently (NMe2H2)+ cations in the presence of acids,
molecular units with an EDA method for the first time, since a phenomenon that has already been observed in CsPbI3
commonly a reasonable fragmentation is not possible. Here, the perovskites.36 Our results indicate that a further reaction
fragmentation is done by including one layer of organic and towards iminium cations may occur when ketones are added
inorganic component each as sketched in Fig. S8a (ESI†) to form to the reaction mixture, for example during the synthesis of
a double layer as one fragment. The pEDA results for 1 are nanoparticles.37 The partial inclusion of larger organic cations
summarized in Table S8 (ESI†). Surprisingly, the interaction in lead halide perovskites does not necessarily change the
energy between these double layers stems nearly equally from crystal structure, but can have significant effects on optical
dispersion (53%) and electronic (47%) contributions indicating a properties and stability.38 While this may occur accidentally, it
more complex situation than pure van-der-Waals binding. A further can also be used deliberately to tune the properties of mixed
This journal is ©The Royal Society of Chemistry 2019
Chem. Commun., 2019, 55, 14725--14728 | 14727

View Article Online
Communication
ChemComm
cation perovskite materials.39 On a broader scope, the wide
range of possible starting materials for iminium ions – amines
and ketones or aldehydes – allows for targeting specific cation
sizes and shapes that were previously unavailable and may yield
unique new anion topologies and material properties, not only
for halogenido bismuthates, but metalates in general.
9 K. Eckhardt, V. Bon, J. Getzschmann, J. Grothe, F. M. Wisser and
S. Kaskel, Chem. Commun., 2016, 52, 3058–3060.
10 S. M. Jain, D. Phuyal, M. L. Davies, M. Li, B. Philippe, C. De Castro,
Z. Qiu, J. Kim, T. Watson, W. C. Tsoi, O. Karis, H. Rensmo, G. Boschloo,
T. Edvinssonand and J. R. Durrant, Nano Energy, 2018, 49, 614–624.
11 M. A. Green, Y. Hishikawa, E. D. Dunlop, D. H. Levi, J. Hohl-Ebinger,
M. Yoshita and A. W. Y. Ho-Baillie, Prog. Photovolt. Res. Appl., 2019,
27, 3–12.
12 N. Mercier, N. Louvain and W. Bi, CrystEngComm, 2009, 11, 720–734.
In summary, we show that layered organic–inorganic iodido
bismuthates are available given the right cation: (Me2CQNMe2)Bi2I7
combines a simple synthetic access with a layered anion topology,
low band gap and good stability and thus holds great promise for
material applications. Our computational investigations show a
band gap in good agreement with experiment and characterize it
as a I - s*(Bi–I) transition. Using energy decomposition analysis
for the first time in an inorganic 3D system, we quantitatively and
qualitatively elucidate the IꢀꢀꢀI interactions found as structural
elements with dispersion interactions and polarization of the iodine
atoms as the major bonding contributions. Our results highlight
iminium cations as prime counterion templating candidates for
new metalate materials with unprecedented anion topologies.
This work is funded by the DFG via the SFB 1083. J. H.
thanks Prof. Stefanie Dehnen for her constant support. N. D.
thanks the Fonds der Chemischen Industrie and the Studien-
stiftung des Deutschen Volkes for their support. We thank the
HLR Stuttgart, HRZ Marburg, and Goethe-HLR Frankfurt for
´
13 A. J. Lehner, D. H. Fabini, H. A. Evans, C.-A. Hebert, S. R. Smock,
J. Hu, H. Wang, J. W. Zwanziger, M. L. Chabinyc and R. Seshadri,
Chem. Mater., 2015, 27, 7137–7148.
14 D. B. Mitzi, Inorg. Chem., 2000, 39, 6107–6113.
15 M.-Q. Li, Y.-Q. Hu, L.-Y. Bi, H.-L. Zhang, Y. Wang and Y.-Z. Zheng,
Chem. Mater., 2017, 29, 5463–5467.
¨
16 A. Erkkila, I. Majander and P. M. Pihko, Chem. Rev., 2007, 107,
5416–5470.
17 H. G. Reiber and T. D. Stewart, J. Am. Chem. Soc., 1940, 62, 3026–3030.
18 M. Lamchen, W. Pugh and A. M. Stephen, J. Chem. Soc., 1954, 4418–4425.
19 N. J. Leonard and J. V. Paukstelis, J. Org. Chem., 1963, 28, 3021–3024.
20 L. M. Trefonas, R. L. Flurry Jr., R. Majeste, E. A. Meyers and
R. F. Copeland, J. Am. Chem. Soc., 1966, 88, 2145–2149.
21 R. Hehl and G. Thiele, Z. Anorg. Allg. Chem., 2000, 626, 2167–2172.
22 M. Ruck, Z. Kristallogr., 1995, 210, 650–655.
23 R. Stief and W. Massa, Z. Anorg. Allg. Chem., 2006, 632, 797–800.
24 L. Mao, P. Guo, M. Kepenekian, I. Hadar, C. Katan, J. Even,
R. D. Schaller, C. C. Stoumpos and M. G. Kanatzidis, J. Am. Chem.
Soc., 2018, 140, 13078–13088.
25 T. Li, Y. Hu, C. A. Morrison, W. Wu, H. Han and N. Robertson,
Sustainable Energy Fuels, 2017, 1, 308–316.
26 E. G. Tulsky and J. R. Long, Chem. Mater., 2001, 13, 1149–1166.
27 A. J. Dennington and M. T. Weller, Dalton Trans., 2018, 47, 3469–3484.
28 N. Dehnhardt, H. Borkowski, J. Schepp, R. Tonner and J. Heine,
Inorg. Chem., 2018, 57, 633–640.
¨
computational resources, Jakob Mobs for preparing reference
compounds, Dr Istemi Kuzu for recording the Raman spectrum
and Dr Robert Wilson for helpful discussions.
¨
29 R. Dronskowski and P. E. Blochl, J. Phys. Chem., 1993, 97, 8617–8624.
30 R. F. W. Bader, Chem. Rev., 1991, 91, 893–928.
31 Dispersion interactions are of course electronic in origin as well.
Distinguishing them from the electrostatic and overlap-based terms
in the pEDA nevertheless gives us additional insight into the nature
of the chemical bond.
Conflicts of interest
32 M. Raupach and R. Tonner, J. Chem. Phys., 2015, 142, 194105.
33 L. Pecher and R. Tonner, Wiley Interdiscip. Rev.: Comput. Mol. Sci.,
2018, e1401.
There are no conflicts to declare.
34 M. P. Mitoraj, A. Michalak and T. Ziegler, J. Chem. Theory Comput.,
2009, 5, 962–975.
References
1 C. C. Stoumpos and M. G. Kanatzidis, Adv. Mater., 2016, 28, 5778–5793. 35 S. Ding and N. Jiao, Angew. Chem., Int. Ed., 2012, 51, 9226–9237.
2 A. K. Jena, A. Kulkarni and T. Miyasaka, Chem. Rev., 2019, 119, 3036–3103. 36 W. Ke, I. Spanopoulos, C. C. Stoumpos and M. G. Kanatzidis, Nat.
3 X. Zhu, Y. Lin, Y. Sun, M. C. Beard and Y. Yan, J. Am. Chem. Soc.,
2019, 141, 733–738.
4 L. N. Quan, B. P. Rand, R. H. Friend, S. G. Mhaisalkar, T.-W. Lee and
E. H. Sargent, Chem. Rev., 2019, 119, 7444–7477.
5 D. B. Mitzi, Chem. Mater., 1996, 8, 791–800.
Commun., 2018, 9, 4785.
37 Q. A. Akkerman, S. G. Motti, A. R. S. Kandada, E. Mosconi,
V. D’Innocenzo, G. Bertoni, S. Marras, B. A. Kamino, L. Miranda,
F. De Angelis, A. Petrozza, M. Prato and L. Manna, J. Am. Chem. Soc.,
2016, 138, 1010–1016.
6 C. C. Stoumpos, D. H. Cao, D. J. Clark, J. Young, J. M. Rondinelli, 38 I. Spanopoulos, W. Ke, C. C. Stoumpos, E. C. Schueller, O. Y.
J. I. Jang, J. T. Hupp and M. G. Kanatzidis, Chem. Mater., 2016, 28,
2852–2867.
Kontsevoi, R. Seshadri and M. G. Kanatzidis, J. Am. Chem. Soc.,
2018, 140, 5728–5742.
7 L. Mao, C. C. Stoumpos and M. G. Kanatzidis, J. Am. Chem. Soc., 39 W. Ke, I. Spanopoulos, Q. Tu, I. Hadar, X. Li, G. S. Shekhawat,
2019, 141, 1171–1190.
8 H. Li, R. Wang and H. Sun, Acc. Chem. Res., 2019, 52, 216–227.
V. P. Dravid and M. G. Kanatzidis, J. Am. Chem. Soc., 2019, 141,
8627–8637.
14728 | Chem. Commun., 2019, 55, 14725--14728
This journal is ©The Royal Society of Chemistry 2019