Redox-Active Guanidines with One or Two Guanidino Groups and Their Integration in Low-Dimensional Perovskite Structures

Article abstract of DOI:10.1002/ejic.201900975

Redox-active guanidines are versatile reagents in redox and proton-coupled electron-transfer (PCET) reactions. Recently, they were integrated as sacrificial electron donors for the prohibition of metal oxidation in perovskite materials. Herein we report t

Full text of DOI:10.1002/ejic.201900975

DOI: 10.1002/ejic.201900975
Full Paper
Electron-Donor Molecules
Redox-Active Guanidines with One or Two Guanidino Groups
and Their Integration in Low-Dimensional Perovskite Structures
Petra Walter,[a] Elisabeth Kaifer,[a] Hendrik Herrmann,[a] Hubert Wadepohl,[a] Olaf Hübner,[a]
and Hans-Jörg Himmel*[a]
Abstract: Redox-active guanidines are versatile reagents in new redox-active guanidines with aromatic (benzene, naphthal-
redox and proton-coupled electron-transfer (PCET) reactions. ene or anthraquinone) cores. Two of these compounds are then
Recently, they were integrated as sacrificial electron donors for integrated in lead iodide materials. The structure, thermal sta-
the prohibition of metal oxidation in perovskite materials. bility and optical band-gap of the resulting materials are evalu-
Herein we report the synthesis and characterization of several ated.
Introduction
the integration of (protonated) redox-active guanidine 1,4-
bis(guanidino)-benzene (2, Figure 2)[17] in perovskite materials
is a possibility to protect the device from oxidative damage.[17]
As sacrificial electron donor, twofold protonated 2 fulfills the
task of capturing dioxygen diffusing into the layer by a proton-
coupled electron transfer (PCET) reaction (2e–,2H+ transfer) in
which the dioxygen is converted to water while the redox-
active organic cation is oxidized.
Redox-active organic molecules are of interest for several appli-
cations, ranging from synthetic chemistry to materials sci-
ence.[1–4] Guanidino-functionalized aromatic compounds (GFAs)
constitute a relatively new class of compounds developed by
our group.[5,6] The redox-properties of these compounds can be
tuned by the number and position of the guanidino groups
attached to an aromatic core. Figure 1a shows as examples the
three compounds 1,4-bis(tetramethylguanidino)-benzene (1),[7]
1,2,4,5-tetrakis(tetra-metylguanidino)-benzene,[8] and a hexakis-
(guanidino)-benzene derivative.[9] The redox potential becomes
more negative with increasing number of guanidino groups. In
their oxidized state, some GFA compounds are versatile proton-
coupled electron-transfer (PCET) reagents (2e–,2H+ transfer as
shown exemplary for compound 1 in Figure 1b).[5,10–12] Reduc-
tion of the oxidized (dicationic) form is favored by the distinct
[5,13,14]
Brønsted basicity
of the reduced state (e.g. protonated
tetramethylguanidine exhibits a pKa value of 15.2 in water and
23.3 in CH3CN, and protonated tetramethylguanidino-benz-
ene [15] a pKa value of 12.2 in water and 20.6 in CH3CN). Further
studies demonstrate a massive increase of the power to oxidize
substrates even with relatively high redox potentials in the pres-
ence of strong acids.[16] Protonated redox-active guanidines
could be re-oxidized by dioxygen, a reaction that is accelerated
by a catalyst.[11] Hence, redox-active guanidines could be used
as redox-catalyst for the oxidation of several organic substrates,
with dioxygen as terminal oxidant.[11] Recently we found that
Figure 1. (a) Lewis structure of redox-active guanidines with guanidino
groups attached to a benzene core. The redox potentials vs. Fc+/Fc (in CH3CN
solution) decrease with increasing number of guanidino groups. (b) Reaction
equations showing the PCET reactivity of compound 1.
[a] Anorganisch-Chemisches Institut, Ruprecht-Karls-Universität Heidelberg,
Im Neuenheimer Feld 270, 69120 Heidelberg, Germany
In the following, we briefly summarize some aspects of the
research on perovskites to motivate the advantages of integrat-
ing redox-active organic cations. Perovskite materials are cur-
rently intensively studied as absorber and hole-transporting
components in solar cell devices.[18–22] Research initially con-
centrated on lead perovskites of the general formula APbX3,
where A denotes a monocation and X a heavier halogen. As
demonstrated already 40 years ago, the band-gap of these
E-Mail: hans-jorg.himmel@aci.uni-heidelberg.de
http://www.uni-heidelberg.de/fakultaeten/chemgeo/aci/himmel/
Supporting information and ORCID(s) from the author(s) for this article are
available on the WWW under https://doi.org/10.1002/ejic.201900975.
© 2019 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA. ·
This is an open access article under the terms of the Creative Commons
Attribution License, which permits use, distribution and reproduction in any
medium, provided the original work is properly cited.
Eur. J. Inorg. Chem. 2019, 4147–4160
4147
© 2019 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
ces.[31] Organic dications such as α,ω-diammonium-alkanes
were also applied.[34] If larger cations are mixed with smaller
ones of comparable solubility, materials with lead iodide layers
of various thickness and consequently of different material
properties might result.[35,36] Some of these “2D perovskites”
have been shown to exhibit an enhanced resistance to humid-
ity.[37,38] It was also possible to modify the optical properties
by the integration of cations with optically active π-systems.[39]
Recently the search for materials in which the environmentally
problematic lead is substituted by less problematic tin was in-
tensified,[40–46] starting with the application of CsSnI3 as hole-
transporting layer in solid-state dye-sensitized solar cells by
Kanatzidis et al.[47] However, the application of tin-containing
perovskite materials is still hampered by the inherent liability
to oxidation of SnII. Several strategies were developed to cir-
cumvent this oxidation liability. Hence the addition of SnF2 was
suggested, being one of the few SnII compounds that are stable
towards oxidation.[48] However, the integration of fluorides is
disadvantageous for the film morphology and the device prop-
erties. Moreover, the stabilizing effect of a thin coating of SnCl2
on perovskite crystallites in a CsSnI3 layer was reported.[49] In
an alternative approach the tin perovskites are synthesized in a
reducing atmosphere of hydrazine vapour.[50] This approach
could unambiguously minimize the formation of defects during
the material synthesis, but does not protect the material in the
long-term from being oxidized in the device.
The integration of sacrificial organic electron donors in pe-
rovskite materials was suggested by our group as a new ap-
proach to prohibit oxidative degradation of the solar-cell devi-
ces.[17] A previous work dealt with compounds 1 and 2. Since
the hydrogen-bond donor propensities of guanidino [NC(NH2)2]
and amino (NH2) groups (with N–H functions that could form
hydrogen bonds to the anions and are at the same time steri-
cally not demanding) are essential for the integration of the
redox-active guanidines in lead iodides, only compound 2 but
not 1 is suitable for integration in lead iodide materials with
extended lead iodide structures. On the other hand, these
groups exhibit reactive N–H sites that lead to oxidative elimina-
tion and aggregation processes of the free molecules. There-
fore, compounds with these groups were isolated in their proto-
nated forms (compounds 2–4 and 8 in Figure 2), since protona-
tion reduces their liability to oxidative degradation to an extent
that makes them easy to handle. Derivatives with peralkylated
guanidino and amino groups, that are less prone to oxidative
degradation, were employed to study the properties (especially
the redox potentials) of the un-protonated new compounds.
Figure 2. The redox-active bisguanidines 1 [p-bis(tetramethyl)guanidino-
benzene] and 2 (p-bisguanidino-benzene) studied previously, and the new
redox-active guanidines 3–10 reported herein.
compounds could be modified through the halogen (or mix-
tures of different halogens).[23] For a “3D perovskite” structure
the choice of possible cations (A) is very much limited, and
include methylammonium (CH3NH3+, usually denoted MA), as
+
well as formamidinium [HC(NH2)2+, FA], guanidinium [C(NH2)3
GUA], and heavy alkali metals, e.g. caesium.
,
Guanidinium ions are particularly attractive due to the ab-
sence of a dipole moment that is made responsible for the large
hysteresis in the current-voltage curves.[24–26] The mechano-
chemical synthesis of hybrid organic-inorganic guanidinium
lead iodides was recently evaluated.[27] Obviously, mixtures of
two of these cations were also studied. For example, the use of
a perovskite APbI3, where A is a mixture of formamidinium and
methylammonium cations, leads to an enhanced short-circuit
current compared with (MA)PbI3.[28] It was shown that solar
cells with Cs0.2(FA)0.8PbI3 exhibit a particularly high ambient
stability (> 1000 h in the dark).[29] Very recently, dications such
as ethylenediammonium ions were introduced as a third cation
component together with formamidinium and methylammo-
nium monocations in tin-lead iodides to achieve high-efficient
solar cells.[30] The spectrum of materials can be further ex-
Results and Discussion
1. Synthesis and Characterization
Compounds 3–5
tended to low-dimensional materials, commonly denoted “2D The synthesis of protonated 3 (p-tetramethylguanidino-aniline)
[31–33]
perovskites”, “1D perovskites” or even “0D perovskites”
started with p-nitroaniline (Scheme 1a). The reaction with
by integration of larger cations such as alkyl- and phenyl-substi- chloro-formamidinium-chloride (obtained from tetramethyl-
tuted ammonium cations, forming e.g. layers of organic mol- urea and oxalyl chloride) gave 1-nitro-4-tetramethylguanidino-
ecules interacting with each other through van-der-Waals for- benzene in 75 % yield (see solid-state structure in the SI). Re-
Eur. J. Inorg. Chem. 2019, 4147–4160
www.eurjic.org
4148
© 2019 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
duction of this compound with dihydrogen in the presence of
3 mol-% PtO2, followed by addition of hydrochloride solution in
diethyl ether, led in 79 % yield to the salt (3+2H)Cl2. Protonated
compound 4 (p-guanidino-dimethylamino-benzene) was ob-
tained by reaction of p-dimethylamino-aniline with cyanamide
(carbodiimide) and subsequent workup with hydrochloride acid
(Scheme 1b).
Figure 3. Illustration of the solid-state structures of (a) (4+2H)Cl2 (left) and
(3+2H)Cl2 (right) and (b) compound 5b. Displacement ellipsoids drawn at the
50 % probability level. Hydrogen atoms bound to carbon omitted. Selected
bond lengths (in Å) for (4+2H)Cl2: N1–C1 1.432(2), N1–C7 1.337(2), N2–C7
1.331(2), N3–C7 1.330(2), N4–C4 1.479(2). Selected bond lengths (in Å) for
(3+2H)Cl2: N1–C1 1.427(3), N1–C7 1.356(3), N2–C7 1.334(3), N3–C7 1.335(3),
N4–C4 1.459(3). Selected bond lengths (in Å) for 5b: N1–C1 2.282(1), N1–C6
1.406(1), N2–C1 1.386(1), N3–C1 1.390(1).
Scheme 1. Synthesis of the new redox-active guanidines 3 and 4 in their
protonated forms, that protect them from oxidative degradation reactions.
Both compounds were crystallized as dihydrochlorides (Fig-
ure 3a). As expected, the two protons are attached to the imino
N atom of the guanidino group and to the amino group. Short
N–H···Cl interactions demonstrate the ability to form hydrogen-
bonded networks, that is essential for the integration in stable
lead iodide materials.
ation equilibrium decreasing the amount of 5b·+ in the diffu-
sion layer. For comparison, compound 1 shows a two-electron
redox process (redox pair 1/12+) at a very similar potential of
E1/2 = –0.22 V in CH3CN.[7] For 2-DMAP, one-electron oxidation
to the radical monocation (“Wurster's blue”) occurs at a poten-
tial (E1/2) of –0.28 V vs. Fc+/Fc, and further one-electron
oxidation to the colourless dication at E1/2 = +0.31 V vs. Fc+/
Fc.[51–53] Hence the potentials for oxidation to the radical mono-
cation are very similar for 5a and 2-DMAP. On the other hand,
the presence of the guanidino group in 5a in place for the
amino group in 2-DMAP decreases significantly the potential
needed for two-electron oxidation.
The redox-activity of compounds 3 and 4 is comparable to
the well-known compound 1,4-bis(dimethylamino)-benzene (2-
DMAP), that could be oxidized to “Wurster's radical cation” and
further to the dication,[51,52] and also similar to compounds 1
and 2.[7] For compounds 3 and 4, adiabatic first ionization ener-
gies of 5.82 and 5.88 eV were calculated (B3LYP/def2-TZVP, see
SI for details). These values are close to the values of 5.36 eV
calculated for 1 and 5.96 eV calculated for 2,[7] indicating that
the PCET reactivity of these compounds is comparable. Since
oxidation of 3 or 4 is accompanied by further reactions due to
the N–H functions, compounds 5a and 5b were synthesized
to study the redox-activity in solution electrochemically in CV
experiments. Compound 5a is an oil, but compound 5b is a
solid at standard conditions, that was crystallized by layering a
diethyl ether solution with n-hexane (Figure 3b). The CV curve
(see Figure 4) recorded for 5a shows two reversible oxidation
waves. The first one at Eox = –0.20 V (E1/2 = –0.24 V) is assigned
to the redox-couple 5a/5a·+ and the second one at Eox
=
+0.06 V (E1/2 = 0.03 V) to 5a·+/5a2+. In addition, smaller and
broader oxidation and reduction waves occur at higher poten-
tials, presumably arising from protonated species readily
formed in the presence of traces of water during the CV meas-
urements. In the case of compound 5b, the two oxidation proc-
esses (Eox = –0.24 V for 5b/5b·+ and Eox = +0.08 V for 5a·+/5a2+
)
are not really reversible (see Figure 4) The low height of the
reduction wave corresponding to the oxidation wave at –0.24 V
(redox-coupled 5b·+/5b) might result from a fast disproportion-
Figure 4. CV curves [CH3CN, Ag/AgCl reference electrode, 0.1
M N(nBu)4(PF6)
as supporting electrolyte, scan speed 100 mV s–1] for compounds 5a and 5b
(measured in direction of oxidation). Potential values given vs. Fc/Fc+.
Eur. J. Inorg. Chem. 2019, 4147–4160
www.eurjic.org
4149
© 2019 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Compounds 6 and 7
The synthesis of bisguanidines with anthraquinone core was
accomplished in one step starting with commercially available
1,4-diamino-anthraquinone. Red-purple crystals of compound 6
were grown by diffusion of n-hexane into a CH3CN solution.
Figure 5 illustrates the solid-state structure of 6. Unfortunately,
a 1,4-bisguanidino-anthraquinone with two NC(NH2)2 groups is
not accessible, at least not by following a similar protocol as for
the synthesis of compound 4. In the UV/Vis spectrum of CH3CN
solutions, compound 6 shows a strong, sharp band at 254 nm
and a broader one around 360 nm in the UV region. Moreover,
a broad band centered at 544 nm was measured in the Vis
region. TD-DFT calculations (B3LYP/def2-TZVP/COSMO, see SI
for details) on 6 found the lowest-energy transition at 558 nm,
in pleasing agreement with the experiments. Similarly, the UV/
Vis spectra of solutions of compound 7 in CH3CN exhibit bands
due to molecular electronic transitions at 252, 349 and 536 nm.
Figure 6 displays the CV curves of compounds 6 and 7. Com-
pound 6 is oxidized at Eox = +0.07 V and reduced back to the
neutral molecule at Ered = +0.01 V vs. Fc/Fc+ (E1/2 = +0.04 V).
Figure 6. CV curves [CH3CN, Ag/AgCl reference electrode, 0.1
M
N(nBu)4(PF6)
as supporting electrolyte, scan speed 100 mV s–1] for compounds 6 and 7
(measured in direction of oxidation). Potential values given vs. Fc/Fc+.
These redox-processes are assigned to the redox-couple 6/62+
.
On the other hand, for 7 two one-electron oxidation waves (at
Eox = –0.01 and +0.12 V) were measured (Figure 6), as well as
two one-electron reduction waves (at Ered = –0.08 and +0.05 V
vs. Fc/Fc+), resulting in E1/2 values of –0.06 V for the redox cou-
ple 7·+/7 and +0.09 V for 72+/7·+. Both redox processes are re-
versible.
Scheme 2. Protonation of red-purple 6 leading to orange (6+2H)(PF6)2.
Figure 5. Illustration of the solid-state structure of compound 6. Displacement
ellipsoids drawn at the 50 % probability level. Hydrogen atoms omitted. Se-
lected bond lengths (in Å): N1–C8 1.2947(17), N2–C8 1.3789(17), N3–C8
1.3817(17), O1–C4 1.2236(17).
Protonation of red-purple 6 with two equivalents of NH4PF6
led to the orange salt (6+2H)(PF6)2 (Scheme 2). In the UV/Vis
spectra, the lowest-energy band of 6 centered at 544 nm
shifted to 456 nm (3548 cm–1 shift) upon protonation (Figure 7).
This large hypsochromic shift could be rationalized by the inter-
actions between the N–H protons and the two carbonyl oxygen
atoms. In this context it is worth mentioning that hypsochromic
shifts upon protonation of not-redox-active mono-guanidine
compounds, including guanidino-benzene, and the influence of
intermolecular hydrogen-bonding to anions were recently ana-
lysed.[54] A similar hypsochromic shift was observed upon pro-
Figure 7. Electronic absorption spectra (in CH3CN) for 6 and (6+2H)(PF6)2. The
inlet shows photos of NMR tubes containing CH3CN solutions of the two
compounds.
Eur. J. Inorg. Chem. 2019, 4147–4160
www.eurjic.org
4150
© 2019 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
tonation of 7 [from λmax = 536 nm for 7 to λmax = 462 nm for blue-shift of 3548 cm–1 for the dication in solution. Further-
(7+2H)(PF6)2]. Figure 8 illustrates the solid-state structures of more, the calculations show that the lowest-energy transition
the protonated compounds.
results from an excitation of an electron from the HOMO into
the LUMO. Figure 9 illustrates the shapes of the HOMO and
LUMO for 6 and (6+2H)2+, from which one can see that the
excitation has charge-transfer character from the electron-rich
ring with the two guanidino groups to the electron-poor quin-
one ring. For the second energetically disfavored tautomer T2,
the calculations found a relatively intense lowest-energy transi-
tion at 541 nm, which does not fit to the experimental results.
Hence both experiment and theory indicate that (6+2H)2+ exists
exclusively in its lowest-energy tautomeric form, T1. Further ex-
periments showed that it was not possible to photo-tautomer-
ize the compound.
Figure 8. Illustration of the solid-state structures of (6+2H)2+ and (7+2H)2+
.
Displacement ellipsoids drawn at the 50 % probability level. C–H hydrogen
atoms omitted. Selected bond lengths (in Å) for (6+2H)2+/(7+2H)2+: N1–C15
1.361(4)/1.364(4), N2–C15 1.324(4)/1.321(4), N3–C15 1.337(4)/1.328(3), N4–
C20 1.352(4)/1.363(3), N5–C20 1.332(4)/1.329(4), N6–C20 1.336(4)/1.324(4),
O1–C6 1.231(3)/1.235(3), O2–C13 1.247(3) /1.243(3).
Quantum-chemical calculations (B3LYP/def2-TZVP, see SI for
details) found two tautomeric structures for (6+2H)2+. A struc-
ture close to the experimental solid-state structure defines the
global-energy minimum structure (T1). It exhibits two asym-
metric N–H···O bridges. In the second tautomer (T2, Scheme 3)
the protons are attached to the oxygen atoms. Consequently,
tautomerization from T1 to T2 is accompanied by reduction of
the quinone to the dihydroquinone and oxidation of the outer
ring with the two guanidino groups. The aromaticity in the C6
ring with the two guanidino groups is removed and instead the
central ring is now participating in the aromaticity (see Lewis
representation in Scheme 3). Isomer T2 exhibits a significantly
higher energy [ΔE = 160 kJ mol–1 (1.66 eV) according to B3LYP/
def2-TZVP, see SI for details], in line with the experimental re-
sults.
Figure 9. Illustration of the isodensity surfaces for the HOMO and LUMO in 6
and in (6+2H)2+ (from B3LYP/def2-TZVP/COSMO calculations).
The CV curves indicate that reversible two-electron reduction
of the protonated compounds (6+2H)(PF6)2 and (7+2H)(PF6)2 is
possible (see SI for more information). For (6+2H)(PF6)2, one
obtains Ered = –0.88 V and Eox = –0.84 V. For (7+2H)(PF6)2, values
of Ered = –0.78 V and Eox = –0.73 V vs. Fc+/Fc were measured.
As shown in Scheme 4, two-electron reduction should be ac-
companied by transfer of the protons from the guanidino
groups to the oxygen atoms and formation of an electron-rich
Scheme 3. Lewis representations of the two calculated tautomers of (6+2H)2+
and their relative energies (according to B3LYP/def2-TZVP, see SI for details).
TD-DFT calculations (B3LYP/def2-TZVP/COSMO) were carried
out to gain further information about the electronic transitions.
In pleasing agreement with the experimental results (λmax.
=
456 nm), the lowest-energy transition was predicted at λ =
476 nm. The blue-shift upon protonation of 3063 cm–1 calcu-
lated for the isolated dication compares with the experimental
Scheme 4. Two-electron reduction of (6+2H)2+ to give the electron-rich di-
hydro-anthraquinone derivative 6+2H.
Eur. J. Inorg. Chem. 2019, 4147–4160
www.eurjic.org
4151
© 2019 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
dihydro-anthraquinone. The results of quantum-chemical calcu- not fully clear, but the positions 4 and 8 of the naphthalene
lations on the reduced form 6+2H support such an electronic core para to the guanidino groups are activated and prone to
structure and also confirm the formation of intramolecular C–C self-coupling reactions to give tetrakisguanidino-binaphth-
O–H···N interactions (see SI).
yl derivatives. Since such reactions are outside the scope of this
work, they were not further persued.
Compounds 8–10
Finally, the salt (8+2H)Cl2 was obtained directly by reaction
of 1,5-diaminonaphthalene-dihydrochloride with cyanamide
(carbodiimide). Its solid-state structure is illustrated in Figure 12.
As expected, the two imino-N atoms of each guanidino group
are protonated, and the structure shows N–H···Cl hydrogen
bonding. The compound (9+2H)(PF6)2 was prepared from 9 and
two equivalents of NH4PF6 (see also Figure 12). Again, protona-
tion occurred exclusively at the imino N atoms of the guanidino
groups and hydrogen-bond interactions to the anions were es-
tablished.
The synthesis started for all three compounds with 1,5-diamino-
naphthalene-dihydrochloride. Reaction with chloroformamidin-
ium-chloride yielded 9, and reaction with 2-chloro-1,3-di-
methyl-4,5-dihydro-1H-imidazolium chloride yielded 10. Com-
pounds 9 and 10 were crystallized from n-hexane and aceto-
nitrile solutions, respectively, and Figure 10 illustrates their
solid-state structures.
Figure 12. Illustration of the solid-state structures of (8+2H)Cl2 (left) and
(9+2H)(PF6)2 (right). Displacement ellipsoids drawn at the 50 % probability
level. Hydrogen atoms bound to carbon omitted. Selected bond lengths (in
Å) for (8+2H)Cl2: N1–C1 1.4241(18), N1–C6 1.3370(19), N2–C6 1.329(2), N3–C6
1.327(2). Selected bond lengths (in Å) for (9+2H)(PF6)2: N1–C1 1.427(3), N1–
C6 1.354(3), N2–C6 1.336(3), N3–C6 1.328(3).
Figure 10. Illustration of the solid-state structures of 9 (left) and 10 (right).
Displacement ellipsoids drawn at the 50 % probability level. Hydrogen atoms
omitted. Selected bond lengths (in Å) for 9: N1–C1 1.407(3), N1–C6 1.298(2),
N2–C6 1.377(3), N3–C6 1.385(3). Selected bond lengths (in Å) for 10: N1–C1
1.392(5), N1–C6 1.293(5), N2–C6 1.386(5), N3–C6 1.383(5).
The CV curves of 9 and 10 show non-reversible oxidation
waves at Eox = +0.43 V vs. Fc/Fc+ for 9 and Eox = +0.07 V vs.
Fc/Fc+ for 10 (Figure 11). The reason for the non-reversibility is
2. Integration of Redox-Active Guanidines in Lead Iodides
Reactions were carried out with compounds 3, 4 and 8 (applied
in their twofold-protonated form). While no clean product could
be isolated in the reactions with compound 8, new organic-
inorganic hybrid materials were obtained by reacting the proto-
nated organic electron-donors 3 and 4 with lead iodide in
aqueous HI solutions. In the reactions, the corresponding
hydrochloride (3+2H)Cl2 resp. (4+2H)Cl2 was first dissolved in a
small amount of concentrated HI (57 % in water) to replace the
counterions (chloride → iodide). Then a second solution con-
taining PbI2 in concentrated HI, heated to a temperature of
75 °C, was added. During re-cooling of the mixture to room-
temperature a microcrystalline precipitate of the raw-product
formed, that was isolated, washed twice with MeOH and dried
under vacuum. Finally, the raw-product was re-dissolved in con-
centrated HI at a temperature of 60 °C, some EtOH added and
the solution cooled back to room temperature. In the case of
the reactions with (3+2H)Cl2, crystals of the product suitable
for structural characterization with single-crystal X-ray diffrac-
tion were obtained during re-cooling to room-temperature, and
the product, (3+2H)Pb2I6, was isolated in 90 % yield. In the case
of (4+2H)Cl2, crystals of the product were obtained at low tem-
Figure 11. CV curves [CH3CN, Ag/AgCl reference electrode, 0.1
M N(nBu)4(PF6)
as supporting electrolyte, scan speed 100 mV s–1] for compounds 9 and 10
showing irreversible oxidation waves. Potential values given vs. Fc/Fc+.
Eur. J. Inorg. Chem. 2019, 4147–4160
www.eurjic.org
4152
© 2019 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
peratures (5 °C). These crystals were washed with pentane and shown in Figure 14b. The guanidine substituents are aligned
hexane, and subsequently dried under vacuum to give the along the chains. In (3+2H)Pb2I6, each (3+2H)2+ unit establishes
product, (4+2H)PbI4, in 70 % yield. Figure 13 and Figure 14 dis- four short N–H···I interactions with the lead iodide chains (Fig-
play the structures of the two new materials. In both cases lead ure 13c). In (4+2H)PbI4, three short N–H···I interactions between
iodide chains are formed, composed of PbI6 octahedra that each (4+2H)2+ unit and the lead iodide chains were found (Fig-
share some of their edges with neighboring octahedra. In ure 14c).
(3+2H)Pb2I6, each octahedron is connected to five neighboring
The volume of the unit cells also differ. For (3+2H)Pb2I6
octahedra. Hence, it shares five out of twelve edges with (monoclinic, space group C2/c), a volume of V = 5222.9(3) Å3
adjacent octahedra, leading to a double-chain type structure was derived. In the case of (4+2H)PbI4 (monoclinic, space group
highlighted in Figure 13b. By contrast, each octahedron in P21/c), the volume is considerably smaller [V = 1906.73(2) Å3].
(4+2H)PbI4 is connected only to two neighboring octahedra, as In this context it should be added that recently 2D layer struc-
Figure 13. (a) Illustration of the structure of (3+2H)Pb2I6 in the crystalline compound from two perspectives (view in the direction of the chains). (b) Perpendicu-
lar view on a segment of one chain highlighting the connectivity of the octahedral units though shared edges. (c) View on one redox-active dication
highlighting its interactions via N–H···I contacts. C–H hydrogen atoms omitted. Vibrational ellipsoids drawn at the 50 % probability level.
Figure 14. (a) Illustration of the structure of (4+2H)PbI4 in the crystalline compound from two perspectives (view in the direction of the chains). (b) Perpendicu-
lar view on a segment of one chain highlighting the connectivity of the octahedral units though shared edges. (c) View on one redox-active dication
highlighting its interactions via N–H···I contacts. C–H hydrogen atoms omitted. Vibrational ellipsoids drawn at the 50 % probability level.
Eur. J. Inorg. Chem. 2019, 4147–4160
www.eurjic.org
4153
© 2019 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
tures were obtained with twofold protonated p-dimethylamino- endothermic process is 33.1 J g–1. At higher temperatures de-
aniline.[55] By contrast, 1D chain structures were obtained with composition commences, with a sharp exothermic decomposi-
twofold protonated p-amino-aniline.
tion peak at ca. 311 °C (exothermic process with an enthalpy
change of 15.4 J g–1). These results clearly show that the
thermal stability of the material is sufficiently high.[58] For
(4+2H)PbI4, a broader endothermic signal around 221 °C can
be assigned to the melting process. The decomposition starts
at temperatures above 270 °C. The quite large difference be-
tween the melting and decomposition temperature might be
interesting for applications. Hence, perovskites with relatively
low melting point, lower than the decomposition temperature,
are attractive for melt processing.[59,60]
Electronic excitation spectra of the materials (3+2H)Pb2I6 and
(4+2H)PbI4 were collected as diffuse reflection measurements
for the materials embedded in a BaSO4 matrix (Figure 15a). The
spectrum recorded for (3+2H)Pb2I6 shows an edge-type struc-
ture with an onset of absorption at ca. 450 nm, and a maximum
of absorption at around 400 nm. For (4+2H)PbI4, a similar onset
of the absorption, but a slight blue-shift of the maximum of
the sharp absorption to ca. 390 nm was measured (Figure 15a).
By using the direct band gap Tauc plot (Figure 15b), optical
band gaps of ca. 2.75 eV were deduced.[56] For comparison, in
a series of lead iodide, bromide and chloride 2D perovskites
with acene alkylammonium cations optical band gaps (esti-
mated with the direct band gap Tauc formalism) between 2.7–
4.5 eV were determined.[57] For (2+2H)Pb2I6 and (3+2H)Sn2I6,
optical band gaps of 2.81 eV and 2.85 eV were derived.[17]
Figure 16. Differential scanning calorimetry (DSC) curves for (3+2H)Pb2I6 and
(4+2H)PbI4, measured under N2 atmosphere with a heating rate of 15 K min–1
.
Finally, high-temperature powder XRD measurements were
carried out for (3+2H)Pb2I6 in the presence of air (diffracto-
grams are included in the SI). These measurements showed no
changes of the material up to a temperature of 240 °C. At
242 °C the peaks decreased and at 250 °C all diffraction peaks
vanished. At much higher temperatures of 420 °C and more,
peaks due to mixed lead oxides appeared (see diffractograms
in the SI). Hence, these measurements confirm a high thermal
stability of the material, even upon exposure to air. This stability
in part arises from the ability of the redox-active guanidine
cations to capture the dioxygen diffusing into the material via
a proton-coupled electron-transfer (PCET) reaction with di-
oxygen. Thereby metal oxidation (PbII → PbIV) is prohibited, as
shown previously for materials with compound 2 [17] exhibiting
a similar redox potential and pKa value and therefore similar
PCET reactivity.
Conclusions
Several new redox-active guanidines with benzene, anthra-
quinone and binaphthyl cores were synthesized and character-
ized. Their redox-properties were analyzed by CV measure-
ments. These new compounds enrich the class of redox-active
Figure 15. (a) Electron spectra (diffuse reflectance of the material in a BaSO4
matrix) of (3+2H)Pb2I6 and (4+2H)PbI4. (b) Tauc plot for (3+2H)Pb2I6 and lin-
ear fit (red line) for estimation of the optical band-gap.
Dynamic difference calorimetry (DSC) measurements for guanidines, that exhibits manifold applications, e.g. as electron
(3+2H)Pb2I6 under inert-gas atmosphere found evidence for a donors and reagents in proton-coupled electron-transfer (PCET)
sharp endothermic peak around 286 °C (Figure 16) that is as- reactions.[5,10] Two of the new compounds were then integrated
signed to the melting process. The enthalpy change for this in lead iodide materials. In addition to their role in establishing
Eur. J. Inorg. Chem. 2019, 4147–4160
www.eurjic.org
4154
© 2019 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
(20 mL). A solution of p-nitroaniline (2.66 g, 19.25 mmol) with tri-
ethylamine (8.4 mL, 60 mmol) in acetonitrile (30 mL) is prepared in
another flask and cooled to –5 °C. At this temperature the solution
of the chloroformamidinium chloride is added in portions to the
nitroaniline solution. Then the solution is stirred for 5 h at 0 °C and
further 16 h at room temperature. The solvent is removed in vacuo
and the residue re-dissolved in aqueous HCl (5 %). After the addi-
tion of a solution of sodium hydroxide in water (40 %), the solution
is extracted with diethyl ether and the organic phases are dried
with K2CO3. The solvent is removed in vacuo, the solid residue dis-
a hydrogen-bond network that stabilizes the material, the
redox-active guanidines could act as sacrificial electron donors
to protect the metal ions from being oxidized by dioxygen (by
PCET with dioxygen), explaining their high thermal stability
even in air. The materials were structurally characterized, and
their thermal stability and optical band-gaps evaluated. Both
materials exhibit chains of lead iodides composed of PbI6 octa-
hedra that share some of their edges with adjacent octahedra,
but nevertheless differ in their structure. The optical band gaps
of 2.75 eV derived for the materials are in a typical region for solved in a small amount of acetonitrile and this solution extracted
with cyclohexane. A yellow solid (3.43 g, 75 %) is obtained after
solvent removal in vacuo. Crystals suitable for structural analysis by
single-crystal X-ray diffraction are grown from a toluene solution.
1H NMR (199.87 MHz, CD3CN): δ = 8.04–7.96 (m, 2 H, CHarom.), 6.67–
6.59 (m, 2 H, CHarom.), 2.72 (s, 12 H, CH3) ppm; elemental analysis
(%) for C11H16N4O2 (236.27 g mol–1): calcd. C 55.92, H 6.83, N 23.71;
found C 56.22, H 6.89, N 23.63.
low-dimensional perovskite materials. The compounds are ther-
mally stable up to more than 200 °C under air. In ongoing work
we are studying the performance of the new redox-active guan-
idines in proton-coupled electron-transfer (PCET) reactions.[4]
Current work on low-dimensional perovskite structures with
integrated redox-active guanidines aims at the synthesis of
2D-materials with lower optical band gaps.
2. Step: p-Tetramethyl-guanidinoaniline Hydrochloride: In a 2-L
flask, p-nitro-tetramethylguanidinobenzene (1.18 g, 5 mmol) and
PtO2 (34 mg, 0.16 mmol) are suspended in 100 mL of ethyl acetate.
The reaction mixture is degassed by three pump-thaw-freeze cycles.
Then the mixture is freezed again and the flask evacuated. The flask
is filled with an atmosphere of dihydrogen. The mixture is warmed
to room temperature and stirred for a period of 20 h. During this
period, the pressure decreases arising from the consumption of di-
hydrogen is compensated from time-to-time. The catalyst is re-
moved by filtration through celite and to the colorless filtrate HCl
Experimental Section
All synthetic work was carried out under argon atmosphere using
the Schlenk technique. The starting reagents PbI2 (99 %), cyan-
amide (99 %), oxalyl chloride (98 %), 1,5-diamino-naphthalene
(97 %) and N,N-dimethyl-p-phenylene-diamine-dihydrochloride
(99 %) were obtained from Sigma Aldrich, HI (57 % in H2O) and
p-nitroaniline (99 %) from Acros Organics and 1,4-diamino-anthra-
quinone (90 %) from abcr GmbH. All solvents were dried with an
MBRAUN MB-SPS-800 solvent purification system, degassed by three
freeze-pump-thaw cycles and stored over molecular sieves. IR spec-
tra were collected on a BIORAD Excalibur FTS 3000. The acquisition
of diffuse reflectance spectra was done on an Agilent Cary 5000
spectrophotometer with BaSO4 pellets of the powder materials. Ele-
mental analysis was carried out at the Microanalytical laboratory of
the chemical institutes at the Heidelberg University. Powder X-ray
diffraction data were recorded with a Stoe Stadi P [Cu-Kα radiation,
λ = 1.5406 Å, Ge(111)-monochromated] in sealed glass capillaries
as sample containers (Ø = 0.5 mm). High-temperature powder
X-ray diffraction was carried out with a graphite heater element in
quartz capillaries (Ø = 0.5 mm). A Cary 5000 spectrophotometer was
used for diffuse reflectance measurements on BaSO4 pellets of the
in diethyl ether (15 mL, 2
M) is added. The precipitate is washed
with ethyl acetate (2 × 30 mL) and dried under vacuum. One ob-
tains p-tetramethylguanidinoaniline-dihydrochloride as a colorless
solid in a yield of 1.26 g (4 mmol, 79 %). Crystals suitable for an
XRD analysis were obtained from glacial acetic acid solutions.
1H NMR (199.87 MHz, CD3OD): δ = 7.06–6.95 (m, 2 H, CH), 2.97 (s,
12 H, CH3) ppm; elemental analysis (%) for C11H18N4·2HCl
(279.1 g mol–1): calcd. C 47.32, H 7.22, N 20.07; found C 46.73,
H 7.28, N 19.59; ESI+: m/z = 207.1758 [MH + -2HCl].
N,N-Dimethyl-p-guanidinoaniline Hydrochloride (4·2HCl): N,N-
Dimethyl-p-phenylenediamine dihydrochloride (4.176 g, 20 mmol)
and carbodiimide (1.264 g, 30 mmol) are heated to reflux in abso-
lute ethanol (50mL) for a period of 30 h. The solution is cooled to
room temperature and then HCl in diethyl ether (10 mL, 2M) is
added. The solvent is removed in vacuo and the residual is washed
with ethanol (20 mL) and diethyl ether (2 × 20 mL). A colorless solid
is obtained (2.8 g, 10 mmol) in a yield of 50 %. Crystals suitable for
an XRD analysis were obtained from a solution in glacial acetic acid.
1H NMR (199.87 MHz, CD3OD): δ = 7.56–7.39 (m, 4 H, CHarom.), 3.23
(s, 6 H, CH3) ppm; elemental analysis (%) for C9H14N4·2HCl
(251.12 g mol–1): calcd. C 43.04, H 6.42, N 22.31; found C 42.01,
H 6.35, N 21.34; ESI+: m/z = 179.1325 [MH+-2HCl].
1
lead iodata materials. H NMR spectra were recorded on a BRUKER
Avance DPX 200 machine, and 13C NMR spectra on a BRUKER
Avance II 400 maschine. Solvent resonances were taken as references
for all 1H NMR spectra. A Discovery DSC 250 from TA Instruments
was used for the DSC measurements. The samples were sealed in
aluminum pans with lids from TA Instruments. The heating rate was
set to 15 K/min. Cyclic voltammetry (CV) measurements relied on
a Metrohm Autolab potentiostate PGSTAT204. The compounds were
dissolved in CH3CN (c = 10–3 mol L–1), with 0.1
M N(nBu)4(PF6) as
(3+2H)Pb2I6: A solution of p-tetramethylguanidinoaniline hydro-
chloride (0.110 g, 0.40 mmol) in aqueous HI (57 %, 5 mL) is prepared
and added to a solution of lead(II)iodide (0.18 g, 0.40 mmol) in
aqueous HI (57 %, 5 mL) at a temperature of 75 °C and stirred for
15 min. Then the clear yellow solution is cooled to room tempera-
supporting electrolyte. An Ag/AgCl electrode was used as reference
electrode. The voltammograms are denoted CV curves in the text.
p-Tetramethylguanidinoaniline-dihydrochloride, 3·2HCl in Two
Steps
1. Step: p-Nitro-tetramethylguanidinobenzene: N,N,N′,N′-Tetra- ture. During this time a fine crystalline precipitate forms. The sol-
methylurea (3.6 mL, 30 mmol) is dissolved in chloroform (30 mL). vent is removed by filtration and the precipitate is washed twice
Then 13.0 mL (155 mmol) of oxalyl chloride is added at room tem- with methanol and dried under vacuum. One obtains (3+2H)Pb2I6
perature and the mixture heated to reflux for a period of 17 h. The
solvent is removed in vacuo and the resulting residue washed twice
with 20 mL portions of diethyl ether, yielding the chloroform-
amidinum chloride salt. This product is dissolved in acetonitrile
as a pale yellow solid (0.25 g, 0.18 mmol) with in a yield of 90 %
with respect to lead(II)iodide. Crystals suitable for a structural char-
acterization by single-crystal X-ray diffraction are obtained by dis-
solving the product in hot aqueous HI (57 %, 4 mL) after addition
Eur. J. Inorg. Chem. 2019, 4147–4160
www.eurjic.org
4155
© 2019 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
of 1 mL of ethanol. Elemental analysis (%) for C11H20N4Pb2I6
(1382.11 g mol–1): calcd. C 9.54, H 1.46, N 4.05; found C 9.57, H 1.53,
N 3.99.
(100.56 MHz, 296 K, CD3CN): δ = 35.3 (CH3), 41.4 (CH3), 48.9 (CH2),
114.5 (CHarom.), 123.1 (CHarom.), 141.7 (Cq), 146.0 (Cq), 155.8 (Cq) ppm;
elemental analysis (%) for C13H20N4 (232.33 g mol–1): calcd. C 67.21,
H 8.68, N 24.11; found C 67.30, H 8.27, N 24.66.
(4+2H)PbI4: A solution of N,N-dimethyl-p-guanidinoaniline hydro-
chloride (0.105 g, 0.40 mmol) in aqueous HI (57 %, 5 mL) is heated
to 75 °C. At this temperature a solution of lead(II)iodide (0.18 g,
0.40 mmol) in aqueous HI (57 %, 5 mL) is added and the reaction
mixture stirred for 1 h. The solution is cooled back to room temper-
ature, and the clear solution is layered with ethanol and stored at
5 °C. The solution is decanted from the precipitated crystals (suit-
able for X-ray diffraction) which are washed with n-pentanol and
hexane and dried subsequently under vacuum. The product
(4+2H)PbI4 is obtained as a pale yellow solid (0.26 g, 0.29 mmol) in
1,4-Bis(tetramethylguanidino)anthraquinone (6): N,N,N′,N′-Tetra-
methyl urea (3.0 mL, 25 mmol) is dissolved in CHCl3 (30 mL) and
oxalyl chloride (10.3 mL, 120 mmol) is added. Subsequently the
reaction mixture is heated to reflux for 17 h. Then the solvent is
removed in vacuo and the remaining solid powder washed twice
with 20 mL portions of diethyl ether. The resulting chloroform-
amidinium-chloride is re-dissolved in CH3CN (25 mL). This solution
is added to a solution of 1,4-diaminoanthraquinone (2.40 g,
10 mmol) and trimethylamine (7.0 mL, 50 mmol) in CH3CN (30 mL)
at –5 °C. The reaction mixture is stirred for 2 h at 0 °C and additional
42 h at room temperature. The solvent is removed in vacuo and the
remaining solid is re-dissolved in aqueous HCl (5 %). Then aqueous
KOH (50 %) is added, and the resulting mixture extracted with di-
ethyl ether. The combined organic phases are washed with brine
and dried with K2CO3 and the solvent is removed in vacuo. The
solid residue is re-dissolved in CH3CN and precipitated by addition
of n-hexane. 1,4-Bis(tetramethylguanidino)anthraquinone (6) is ob-
tained as a red-purple solid (2.7 g, 2.6 mmol) in 62 % yield. Crystals
suitable for structural characterization by single-crystal X-ray diffrac-
tion are obtained by diffusion of n-hexane into a CH3CN solution.
1H NMR (199.87 MHz, CD3CN, 25 °C): δ = 8.01–8.06 (m, 2 H, CHarom.),
7.65–7.70 (m, 2 H, CHarom.), 6.86 (s, 2 H, CHarom.), 2.66 ppm (s, 24 H,
CH3); 13C NMR (100.56 MHz, 296 K, CD3CN): δ = 39.3 (CH3), 122.3
(Cq), 126.2 (CHarom.), 133.0 (CHarom.), 133.3 (CHarom.), 135.8 (Cq), 149.5
a
yield of 72 %. Elemental analysis (%) for C9H16N4PbI4
(899.10 g mol–1): calcd. C 12.02, H 2.24, N 6.23; found C 11.93,
H 1.79, N 6.45.
N,N′-Tetramethylguanidino-N′′,N′′-dimethyl-p-phenylenedi-
amine (5a): N,N,N′,N′-Tetramethylurea (3.0 mL, 25 mmol) is dis-
solved in chloroform (35 mL). Then oxalyl chloride (11.0 mL,
130 mmol) is added and the reaction mixture heated to reflux for
a period of 18 h. Subsequently the solvent is removed in vacuo and
the remaining solid is washed twice with 20 mL portions of diethyl
ether. The resulting chloro-formamidinium chloride is dissolved in
acetonitrile (30 mL). This solution is added to a suspension of N,N-
dimethyl-p-phenylenendiamine-dihydrochloride (4.38 g, 21 mmol)
and triethylamine (13 mL, 93 mmol) in acetonitrile (45 mL) at –5 °C.
Then the reaction mixture is stirred for 3 h at 0 °C and additional
2 h at room temperature. The solvent is removed in vacuo and
the solid residue dissolved in aqueous HCl (5 %). After addition of
aqueous NaOH (40 %) the resulting mixture is extracted with diethyl
ether. The organic phases are dried with K2CO3 and the solvent is
removed in vacuo. The residual oil is purified by distillation under
vacuum. Compound 5a is obtained as a pale-yellow oil (4.3 g,
18 mmol) in a yield of 87 %. 1H NMR (199.87 MHz, CD3CN, 25 °C):
δ = 6.69–6.64 (m, 2 H, CHarom.), 6.53–6.49 (m, 2 H, CHarom.), 2.80 (s,
6 H, CH3), 2.61 ppm (s, 24 H, CH3); 13C NMR (100.56 MHz, 296 K,
CD3CN): δ = 41.4 (CH3), 114.7 (CHarom.), 122.5 (CHarom.), 143.6 (Cq),
145.8 (Cq), 159.0 (Cq) ppm; ESI+: m/z = 234.9722 [MH+]; elemental
analysis (%) for C13H22N4 (234.34 g mol–1): calcd. C 66.63, H 9.46,
N 23.91; found C 66.11, H 9.88, N 24.24.
(Cq), 160.3 (Cq), 184.2 (Cq) ppm; IR (KBr): ν = 3298 (w), 3084 (w),
˜
3065 (w), 3044 (w), 2994 (w), 2936 (w), 2884 (w), 2803 (w), 1892 (w),
1659 (w), 1611 (w), 1503 (w), 1466 (w), 1443 (w), 1429 (w), 1418 (w),
1404 (w), 1375 (w), 1352 (w), 1325 (m) 1273 (m), 1240 (m), 1188
(m), 1165 (m), 1140 (m), 1111 (w), 1064 (s), 1028 (s), 1015 (m), 966
(w), 943 (s), 924 (s), 897 (m), 866 (s), 841 (s), 831 (w), 799 (s), 775
(w), 752 (m), 741 (m), 729 (vs), 719 (s), 691 (s), 664 (w), 654 (w), 637
(m), 610 (w), 571 (m), 548 (w), 534 (m), 500 (w), 480 (w), 457 (w),
438 (m), 419 (w) cm–1; UV/Vis (CH3CN, c = 2.0 × 10–5 mol L–1): λmax
(ε in L mol–1 cm–1) = 544 (4279), 360 (5456), 254 nm (34721) nm;
elemental analysis (%) for C24H30N6O2 (434.52 g mol–1): calcd.
C 66.34, H 6.96, N 19.34; found C 66.05, H 6.93, N 19.49.
(6+2H)(PF6)2: A solution of ammonium hexafluorophosphate
(0.164 g, 0.89 mmol) in CH3CN (10 mL) is prepared. This solution is
added to another solution of 1,4-bis[(tetramethyl)guanidino]anthra-
quinone (0.195 g, 0.448 mmol) in CH3CN at room temperature. The
reaction mixture is heated to 50 °C and stirred at this temperature
for 1 h. Then the mixture is cooled back to room temperature and
the solvent removed in vacuo. The solid residue is re-disssolved in
CH3CN (5 mL) and some charcoal added. Then the mixture is filtered
through celite. The obtained solution is layered with diethyl ether.
N,N′-Dimethylethyleneguanidino-N′′,N′′-dimethyl-p-phenylene-
diamine (5b): Oxalyl chloride (6.5 mL, 82 mmol) is added to a solu-
tion of 1,3-imidazole-2-imidazolidinone (1.75 mL, 15 mmol) in
chloroform (25 mL). Then the reaction mixture is heated to reflux
for 18 h. Subsequently the solvent is removed in vacuo and the
solid residue washed twice with 20 mL portions of diethyl ether. The
resulting chloroformamidinium-chloride is dissolved in acetonitrile
(20 mL). This solution is added to a suspension of N,N-dimethyl-p-
phenylenediamine-dihydrochloride (2.53 g, 12 mmol) and triethyl- One obtains (6+2H)(PF6)2 as orange crystals (0.326 g, 0.448 mmol)
amine (7.5 mL, 54 mmol) in acetonitrile (35 mL) at –5 °C, and the
reaction mixture is stirred for 3 h at 0 °C and additional 2 h at room
temperature. The solvent is removed in vacuo and the solid residue
dissolved in aqueous HCl (5 %). Then aqueous NaOH (40 %) is
added and the resulting mixture extracted with diethyl ether. The
organic phases are dried with K2CO3 and the solvent is removed in
vacuo. For further purification, the solid residue is re-dissolved in
diethyl ether and precipitated with n-hexane. Compound 5b is ob-
tained as a beige solid (1.52 g, 6.54 mmol) in 55 % yield. Crystals
suitable for structural characterization by single-crystal X-ray diffrac-
tion were obtained by layering a diethyl ether solution with n-hex-
ane. 1H NMR (199.87 MHz, CD3CN, 25 °C): δ = 6.64 (m, 4 H, CHarom.),
in quantitative yield. The crystals are suitable for structural charac-
terization by single-crystal X-ray diffraction. 1H NMR (199.87 MHz,
CD3CN, 25 °C): δ = 10.96 (s, 2 H, NH), 8.23–8.28 (m, 2 H, CHarom.),
7.89–7.94 (m, 2 H, CHarom.), 7.32 (s, 2 H, CHarom.), 3.05 ppm (s, 24 H,
CH3); 13C NMR (100.56 MHz, 296 K, CD3CN): δ = 40.8 (CH3), 119.0
(Cq), 127.6 (CHarom.), 127.8 (CHarom.), 133.7 (Cq), 135.9 (CHarom.), 137.6
(C ), 158.1 (C ), 186.9 (C ); IR (KBr): ν = 3464 (w), 2963 (w), 2926 (w),
˜
q
q
q
2818 (w), 1632 (m), 1595 (w), 1587 (w), 1543 (m), 1501 (w), 1476
(w), 1431 (m), 1415 (m), 1350 (m), 1298 (w), 1263 (s), 1225 (w), 1180
(m), 1172 (m), 1115 (w), 1074 (s), 1047 (w), 1038 (w), 1018 (w), 986
(w), 947 (w), 833 (vs), 781 (w), 735 (s), 704 (w), 685 (w), 669 (w), 664
(w), 625 (w), 557 (vs), 503 (w), 494 (w), 444 (w), 419 (w) cm–1; UV/
3.19 (s, 4 H, CH2), 2.81 (s, 6 H, CH3), 2.53 ppm (s, 6 H, CH3); 13C NMR Vis (CH3CN, c = 3.28 × 10–5 mol L–1): λmax (ε in L mol–1 cm–1) = 456
Eur. J. Inorg. Chem. 2019, 4147–4160
www.eurjic.org
4156
© 2019 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
(5040), 321 (9855), 257 (37747) nm; elemental analysis (%) for
C24H32N6O2P2F12 (726.47g mol–1): calcd. C 39.68, H 4.44, N 11.57;
found C 39.72, H 5.23, N 11.43.
1,5-Bisguanidinonaphthalene Dihydrochloride (8·2HCl): A 2
M
solution of HCl in diethyl ether (6.5 mL, 13 mmol) is added to a
suspension of 1,5-diamino-naphthalene (1.00 g, 6.35 mmol) in
50 mL of diethyl ether. The reaction mixture is stirred overnight at
room temperature. After solvent removal under vacuum the crude
product is re-dissolved in 100 mL of ethanol. Cyanamide (1.556 g,
44.0 mmol) is added and the solution is heated to reflux overnight.
The volume of the resulting dark purple solution is reduced under
vacuo to a minimum. Diethyl ether is added and the precipitate
filtered. The residue is washed with diethyl ether and dried under
vacuo to give 1,5-bisguanidinonaphthalene-dihydrochloride,
8·2HCl, as a light-grey solid (1.85 g, 5.9 mmol) in 93 % yield. Crystals
suitable for structural characterization by single-crystal X-ray diffrac-
tion are obtained by layering a methanol solution with diethyl
ether. Elemental analysis (%) for C12H16N6Cl2 (315.20 g mol–1): calcd.
C 45.73, H 5.12, N 26.66; found C 45.24, H 5.43, N 26.07.
1,4-Bis(N,N′-dimethylethylenguanidino)anthraquinone (7): A so-
lution of 1,3-imidazol-2-imidazolidinone (2.7 mL, 25 mmol) in CHCl3
(30 mL) is prepared and oxalyl chloride (10.6 mL, 125 mmol) is
added. The reaction mixture is heated for 20 h under reflux. Subse-
quently the solvent is removed in vacuo and the residual is washed
with diethyl ether and dried under vacuum. The resulting solid
product (2-chloro-1,3-dimethyl-4,5-dihydro-1H-imidazolium chlor-
ide) is dissolved in CH3CN and added to a solution of 1,4-diamino-
anthraquinone (2.4 g, 10 mmol) and triethylamine (7.0 mL,
50 mmol) in CH3CN at –5 °C. After completing the addition, the
reaction mixture is stirred for 2 h at 0 °C and additional 18 h at
room temperature. Then the solvent is removed under reduced
pressure und the remaining solid is re-dissolved in aqueous HCl
(5 %). After addition of aqueous NaOH (25 %), the resulting mixture
is extracted with ethyl acetate. The combined organic phases are
washed with brine, dried with K2CO3 and the solvent is removed in
vacuo. The residue is dissolved in toluene and precipitated by the
addition of n-hexane. 1,4-Bis(N,N′-dimethylethyleneguanidino)-
anthraquinone (7) is obtained as a red-purple solid (2.88 g,
1,5-Bis(tetramethylguanidino)naphthalene (9): Oxalyl chloride
(10.3 mL, 120 mmol) is added to a solution of N,N,N′,N′-tetramethyl-
urea (3.0 mL, 25 mmol) in chloroform (30 mL). Then the reaction
mixture is heated to reflux for 18 h. The solvent is removed in vacuo
and the remaining solid washed twice with 20 mL portions of di-
ethyl ether. The resulting chloroformamidinium chloride is dissolved
in acetonitrile (25 mL), and this solution added to a suspension of
1,5-diaminonaphthalene (1.58 g, 10 mmol) and triethylamine
(7.0 mL, 50 mmol) in acetonitrile (30 mL) at a temperature of –5 °C.
Subsequently the reaction mixture is stirred for 3 h at 0 °C and
additional 2 h at room temperature. The solvent is removed in
vacuo and the solid residue dissolved in aqueous HCl (5 %). Aque-
ous NaOH (40 %) is added, and the resulting mixture extracted with
dichloromethane. The organic phases are dried with K2CO3 and the
solvent is removed in vacuo. The solid residue is washed with tolu-
ene and dried under vacuo to give 1,5-bis(tetramethylguanidino)-
naphthalene (9) is obtained as a pale-rosy solid (2.7 g, 7.6 mmol)
in a yield of 76 %. Crystals suitable for single-crystal X-ray diffraction
could be obtained from a solution in n-hexane. 1 H NMR
(200.13 MHz, CD3CN, 25 °C): δ = 7.52–7.57 (m, 2 H, CHarom.), 7.16–
7.23 (m, 2 H, CHarom.), 6.44–6.48 (m, 2 H, CHarom.), 2.64 ppm (s, 24
H, CH3); 13C NMR (100.56 MHz, 296 K, CD3CN): δ = 39.4 (CH3) ppm.
The poor solubility prohibited the detection of more signals; UV/Vis
(CH3CN, c = 5.53 × 10–5 mol L–1): λmax (ε in L mol–1 cm–1) = 345
(17463), 239 nm (40240); elemental analysis (%) for C20H30N6
(354.49 g mol–1): calcd. C 67.76, H 8.53, N 23.71; found C 66.87, H
8.32, N 23.34.
1
6.69 mmol) in 67 % yield. H NMR (199.87 MHz, CD3CN, 25 °C): δ =
8.01–8.09 (m, 2 H, CHarom.), 7.66–7.71 (m, 2 H, CHarom.), 6.98 (s, 2 H,
CHarom.), 3.31 (s, 8 H, CH2), 2.58 ppm (s, 12 H, CH3); 13C NMR
(100.56 MHz, 296 K, CD3CN): δ = 34.7 (CH3), 48.6 (CH2), 122.1 (Cq),
126.2 (CHarom.), 132.9 (CHarom.), 133.3 (CHarom.), 135.6 (Cq), 147.7 (Cq),
156.5 (Cq), 183.4 (Cq) ppm; IR (KBr): ν = 2948 (w), 2852 (w), 1992
˜
(w), 1995 (w), 1657 (m), 1599 (m), 1537 (w), 1493 (m), 1457 (w),
1441 (w), 1420 (m), 1396 (m), 1368 (w), 1352 (s), 1287 (s), 1243 (m),
1218 (m), 1197 (w), 1167 (m), 1160 (s), 1147 (s), 1117 (w), 1093 (w),
1077 (m), 1059 (vs), 1039 (vs), 1032 (vs), 1024 (s), 987 (s), 980 (s),
968 (m), 943 (s), 904 (m), 892 (w) 878 (s), 853 (s), 800 (s), 766 (m),
760 (m), 729 (vs), 702 (w), 691 (w), 674 (m), 660 (m), 648 (m), 636
(m), 610 (m), 602 (m), 577 (m), 543 (s), 507 (w), 490 (w), 475 (s), 443
(w), 416 (w) cm–1; UV/Vis (CH3CN, c = 3.26 × 10–5 mol L–1): λmax (ε in
L mol–1 cm–1) = 536 (4933), 349 (5494), 252 (42250) nm; elemental
analysis (%) for C24H26N6O2 (430.45 g mol–1): calcd. C 66.95, H 6.06,
N 19.52; found C 66.81, H 5.94, N 19.32.
(7+2H)(PF6)2: A solution of ammonium hexafluorophosphate
(63 mg, 0.39 mmol) in CH3CN (8 mL) is added to a solution of 1,4-
bis(N,N′-dimethylethyleneguanidino)anthraquinone (7, 90 mg,
0.21 mmol) in CH3CN (8 mL). The reaction mixture is stirred for 2 h
at 50 °C and additional 18 h at room temperature. The solvent is
removed in vacuo and the solid residue is re-dissolved in CH3CN
(5 mL). After addition of charcoal the solution is filtered through
celite. Half of the solvent is removed and the solution layered with
diethyl ether. One obtains (7+2H)(PF6)2 (0.144 g, 0.199 mmol) as
(9+2H)(PF6)2:
A solution of ammonium hexafluorophosphate
(33 mg, 0.2 mmol) in CH3CN (2 mL) is added to a solution of 1,5-
bis(tetramethylguanidino)naphthalene (9, 35 mg, 0.10 mmol) in
CH3CN (7 mL). The reaction mixture is stirred for 1 h at 50 °C and
additional 18 h at room temperature. The solvent is removed in
vacuo and the solid residue re-dissolved in CH3CN (3 mL). After
addition of charcoal the solution is filtered through celite. Half of
the solvent is removed and the solution layered with diethyl ether.
From this solution, one obtains (6+2H)(PF6)2 as colorless crystals,
that are structurally characterized by single-crystal X-ray diffraction.
Further analysis is not made.
1
orange crystals in 97 % yield. H NMR (199.87 MHz, CD3CN, 25 °C):
δ = 11.32 (s, 2 H, NH), 8.29–8.344 (m, 2 H, CHarom.), 7.94–7.99 (m,
2 H, CHarom.), 7.47 (s, 2 H, CHarom.), 3.88 (s, 8 H, CH2), 2.95 ppm (s,
12 H, CH3); 13C NMR (100.56 MHz, 296 K CD3CN): δ = 34.8 (CH3),
49.3 (CH2), 119.0 (Cq), 127.6 (CHarom.), 127.8 (CHarom.), 133.7 (Cq),
136.0 (CHarom.), 137.0 (Cq), 156.9 (Cq), 187.0 (Cq) ppm; IR (KBr): ν =
˜
3463 (w), 3167 (w), 3114 (w), 2952 (w), 2903 (w), 1617 (s), 1589 (s),
1578 (s), 1499 (s), 1458 (w), 1429 (m), 1399 (m), 1378 (w), 1343
(m),1311 (m), 1264 (m), 1233(w), 1191 (w), 1161 (w), 1072 (s), 1035
(m), 1021 (w), 983 (w), 946 (w), 901 (m), 878 (vs), 832 (vs), 804 (m),
734 (vs), 714 (w), 668 (w), 648 (m), 590 (w), 557 (vs), 503 (w), 441
1,5-Bis(N,N′-dimethylethylenguanidino)naphthalene (10): Oxalyl
chloride (5.3 mL, 62.5 mmol) is added to a solution of 1,3-imidazol-
2-imidazolidinone (1.35 mL, 12.5 mmol) in chloroform (30 mL). The
reaction mixture is heated for 20 h under reflux. Subsequently the
(m), 417 (w) cm–1; UV/Vis (CH3CN, c = 2.66 × 10–5 mol L–1): λmax (ε in solvent is removed in vacuo and the residue washed with diethyl
L mol–1 cm–1) = 462 (6488), 313 (9503), 255 (51516) nm; elemental
analysis (%) for C24H28N6O2P2F12 (722.44 g mol–1): calcd. C 39.90, H
3.91, N 11.63; found C 40.80, H 4.25, N 13.04.
ether and dried under vacuum. The obtained solid is dissolved in
acetonitrile (12 mL) and added to a solution of 1,5-diamino-naphth-
alene (0.774 g, 4.9 mmol) and triethylamine (3.5 mL, 25 mmol) in
Eur. J. Inorg. Chem. 2019, 4147–4160
www.eurjic.org
4157
© 2019 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
acetonitrile at –5 °C. The reaction mixture is stirred for 2 h at 0 °C
and another 18 h at room temperature. Then the solvent is removed
under reduced pressure und the remaining solid is dissolved in
aqueous HCl (5 %). Aqueous NaOH (40 %) is added and the result-
ing mixture extracted with toluene. The combined organic phases
are dried with K2CO3 and the solvent is removed in vacuo. Finally,
the residue is washed with toluene and dried under vacuum. The
product 1,5-bis(N,N′-dimethylethyleneguanidino)-naphthalene (10)
is obtained as a pale-pink solid (1.50 g, 4.28 mmol) in 87 % yield.
Crystals suitable for structural characterization by single-crystal
CCDC 1938180 (for 3·HCl), 1938184 {for [(3+2H)Pb2I4]}, 1938181 (for
4·2HCl), 1938185 {for [(4+2H)PbI4]}, 1938183 (for 5b), 1938175 (for
6), 1938186 {for [for (6+2H)(PF6)2]}, 1938178 {for [(7+2H)(PF6)2]},
1938176 {for [(8+2H)Cl2]}, 1938174 (for 9), 1938182 {for
[(9+2H)(PF6)2]}, 1938177 (for 10), and 1938179 (for p-nitro-tetra-
methylguanidinobenzene) contain the supplementary crystallo-
graphic data for this paper. These data can be obtained free of
charge from The Cambridge Crystallographic Data Centre.
Details of the Quantum Chemical Calculations
1
Density functional calculations are performed with the program
X-ray diffraction are grown from hot acetonitrile solutions. H NMR
[77,78]
TURBOMOLE.[74,75,76] The B3LYP functional
is used with the
(199.87 MHz, CD3CN, 25 °C): δ = 7.49–7.53 (m, 2 H, CHarom), 7.16–
7.23 (m, 2 H, CHarom.), 6.74–6.70 (m, 2 H, CHarom.), 2.50 (s, 8 H, CH2),
3.25 ppm (s, 12 H, CH3); 13C NMR (100.56 MHz, 296 K, CD3CN):
δ = 34.7 (CH2), 48.8 (CH3) ppm. The poor solubility prohibited the
detection of more signals; UV/Vis (CH3CN, c = 5.31 × 10–5 mol L–1):
λmax (ε in L mol–1 cm–1) = 342 (17933), 238 nm (40807); elemental
analysis (%) for C20H26N6 (350.46 g mol–1): calcd. C 68.54, H 7.48, N
23.98; found C 68.55, H 7.21, N 23.48.
def2-TZVP or def2-SV(P) basis sets.[79] For the calculation of the two-
electron integrals, the resolution-of-the-identity (RI) approxima-
tion[80] is used with the appropriate def2-TZVP or def2-SV(P) auxil-
iary basis sets.[81] To confirm that the optimized structures are min-
ima of the potential energy surfaces and to obtain zero point vibra-
tional energy contributions, vibrational frequencies are calculated.
All electronic energies are obtained with the def2-TZVP basis set.
For the anthraquinone derivatives, the calculations of the vibra-
tional frequencies relied on the smaller def2-SV(P) basis set. Elec-
tronic excitation energies are obtained by time-dependent density
functional calculations.[82,83] The time-dependent density functional
X-ray Crystal Structure Determinations
Crystal data and details of the structure determinations are com-
piled in a table in the SI. Full shells of intensity data were collected
at low temperature with an Enraf Nonius Kappa CCD diffractometer
[Mo-Kα radiation, sealed X-ray tube, graphite monochromator, com-
pounds 3·2HCl, 4·2HCl, 6, (7+2H)(PF6)2, 8·2HCl, 9, (9+2H)(PF6)2, 10],
a Bruker D8 Venture, dual source (Mo- or Cu-Kα radiation, micro-
focus X-ray tube, Photon III Detector, compound 5b) or with an
Agilent Technologies Supernova-E CCD diffractometer [Mo- or
Cu-Ka radiation, microfocus X-ray tube, multilayer mirror optics,
compounds (3+2H)Pb2I6, (4+2H)PbI4, (6+2H)(PF6)2]. Detector frames
(typically w-, occasionally j-scans, scan width 0.4–0.1°) were inte-
calculations considers the influence of the environment by the
[84]
conductor-like screeing model (COSMO)
38.8 (corresponding to CH3CN).
with a value for εr of
Acknowledgments
The authors gratefully acknowledge financial support by the
BMBF (Perosol project) and the Deutsche Forschungsgemein-
schaft (DFG).
grated by profile fitting.[61,62] Data were corrected for air and de-
[62]
tector absorption, Lorentz and polarization effects
and scaled
Keywords: Guanidine ligands · Perovskites · Lead iodide ·
Electron donors · Redox chemistry
essentially by application of appropriate spherical harmonic func-
tions.[62–64] Absorption by the crystal was treated with a semiempiri-
cal multiscan method (as part of the scaling process) and aug-
mented by a spherical correction,[62,63,64] or numerically (Gaussian
grid).[62,64,65] For datasets collected with the microfocus tubes an
illumination correction was performed as part of the numerical ab-
sorption correction.[64]
[1] T. Nishinaga (Ed.), Organic Redox Systems, Synthesis, Properties and Appli-
cations, John Wiley & Sons, Inc. 2016.
[2] J. Broggi, T. Terme, P. Vanelle, Angew. Chem. Int. Ed. 2014, 53, 384–413;
Angew. Chem. 2014, 126, 392–423.
[3] a) S. Zhou, H. Farwaha, J. A. Murphy, Chimia 2012, 66, 418–424; b) J. A.
Murphy, J. Org. Chem. 2014, 79, 3731–3746; c) E. Doni, J. A. Murphy,
Chem. Commun. 2014, 50, 6073–6087.
[4] J. Hankache, O. S. Wenger, Chem. Rev. 2011, 111, 5138–5178.
[5] H.-J. Himmel, Synlett 2018, 29, 1957–1977.
[6] a) B. Eberle, O. Hübner, A. Ziesak, E. Kaifer, H.-J. Himmel, Chem. Eur. J.
2015, 21, 8578–8590; b) H.-J. Himmel, Z. Anorg. Allg. Chem. 2013, 639,
1940–1952.
[7] J. Hornung, O. Hübner, E. Kaifer, H.-J. Himmel, RSC Adv. 2016, 6, 39323–
39329.
[8] A. Peters, E. Kaifer, H.-J. Himmel, Eur. J. Org. Chem. 2008, 5907–5914.
[9] B. Eberle, E. Kaifer, H.-J. Himmel, Angew. Chem. Int. Ed. 2017, 56, 3360–
3363; Angew. Chem. 2017, 129, 3408–3412.
[10] U. Wild, S. Federle, A. Wagner, E. Kaifer, H.-J. Himmel, Chem. Eur. J. 2016,
22, 11971–11976.
The structures were solved by ab initio dual space methods involv-
ing difference Fourier syntheses (VLD procedure),[66] by the heavy
atom method combined with structure expansion by direct meth-
[67]
ods applied to difference structure factors
procedure
or by the charge flip
[68]
and refined by full-matrix least-squares methods
based on F2 against all unique reflections.[69] All non-hydrogen
atoms were given anisotropic displacement parameters. Hydrogen
atoms bound to carbon were input at calculated positions and re-
fined with a riding model. Hydrogen atoms bound to nitrogen were
located in difference Fourier syntheses and refined, either fully or
with appropriate distance and/or symmetry restraints (variable ge-
ometry rigid groups with local C3v symmetry for [NH3]+). P–F and
F···F distances within the [PF6]– anions were restrained to sensible
values during refinement.[70] Suitable adp restraints were ap-
plied.[71]
[11] U. Wild, F. Schön, H.-J. Himmel, Angew. Chem. Int. Ed. 2017, 56, 16410–
16413; Angew. Chem. 2017, 129, 16630–16633.
[12] For some review articles about proton-coupled electron transfer reac-
tions, see: a) S. J. Slattery, J. K. Blaho, J. Lehnes, K. A. Goldsby, Coord.
Chem. Rev. 1998, 174, 391–416; b) C. Costentin, M. Robert, J.-M. Savéant,
Chem. Rev. 2008, 108, 2145–2179; c) J. J. Consepcion, J. W. Jurss, M. K.
Brennaman, P. G. Hoertz, A. O. T. Patrocinio, N. Y. M. Iha, J. L. Templeton,
T. J. Meyer, Acc. Chem. Res. 2009, 42, 1954–1965; d) S. Y. Reece, D. G.
Nocera, Annu. Rev. Biochem. 2009, 78, 673–699; e) J. J. Warren, T. A.
Tronic, J. M. Mayer, Chem. Rev. 2010, 110, 6961–7001; f) S. Hammes-
Due to severe disorder and fractional occupancy, electron density
attributed to solvent of crystallization (acetonitrile and n-hexane)
was removed from the structure of (6+2H)(PF6)2 with the BYPASS
procedure,[72] as implemented in PLATON (squeeze/hybrid).[73] Par-
tial structure factors from the solvent masks were included in the
refinement as separate contributions to Fcalc
.
Eur. J. Inorg. Chem. 2019, 4147–4160
www.eurjic.org
4158
© 2019 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Schiffer, A. A. Stuchebrukhov, Chem. Rev. 2010, 110, 6939–6960; g) S.
Hammes-Schiffer, J. Am. Chem. Soc. 2015, 137, 8860–8871; h) I. Siewert,
Chem. Eur. J. 2015, 21, 15078–15091; i) D. C. Miller, K. T. Tarantino, R. R.
Knowles, Top. Curr. Chem. (Z) 2016, 374:30; j) E. C. Gentry, R. R. Knowles,
Acc. Chem. Res. 2016, 49, 1546–1556; k) J. C. Lennox, D. A. Kurtz, T. Hu-
ang, J. L. Dempsey, ACS Energy Lett. 2017, 2, 1246–1256; l) A. Pannwitz,
O. S. Wenger, Dalton Trans. 2019, 48, 5861–5868.
[41]
[42]
[43]
[44]
C. C. Stoumpos, C. D. Malliakas, M. G. Kanatzidis, Inorg. Chem. 2013, 52,
9019–9038.
M.-G. Ju, J. Dai, L. Ma, X. C. Zeng, J. Am. Chem. Soc. 2017, 139, 8038–
8043.
F. Hao, C. C. Stoumpos, D. H. Cao, R. P. H. Chang, M. G. Kanatzidis, Nat.
Photonics 2014, 8, 489–494.
N. K. Noel, S. D. Stranks, A. Abate, C. Wehrenfennig, S. Guarnera, A.-A.
Haghighirad, A. Sadhanala, G. E. Eperon, S. K. Pathak, M. B. Johnston, A.
Petrozza, L. M. Herz, H. J. Snaith, Energy Environ. Sci. 2014, 7, 3061–3068.
Z. Xiao, W. Meng, J. Wang, D. B. Mitzi, Y. Yan, Mater. Horiz. 2017, 4, 206–
216.
W. Ke, C. C. Stoumpos, M. G. Kanatzidis, Adv. Mater. 2018, 30, 1803230
(1–31).
L. Chung, B. Lee, J. He, R. P. H. Chang, M. G. Kanatzidis, Nature 2012, 485,
486–489.
a) M. H. Kumar, S. Dharani, W. L. Leong, P. P. Boix, R. R. Prabhakar, T.
Baikie, C. Shi, H. Ding, R. Ramesh, M. Asta, M. Gratezel, S. G. Mhaisalkar,
N. Mathews, Adv. Mater. 2014, 26, 7122–7127; b) S. J. Lee, S. S. Shin, Y. C.
Kim, D. Kim, T. K. Ahn, J. H. Noh, J. Seo, S. I. Seok, J. Am. Chem. Soc.
2016, 138, 3974–3977; c) D. Sabba, H. K. Mulmudi, R. R. Prabhakar, T.
Krishnamoorthy, T. Baikie, P. P. Boix, S. Mhaisalkar, N. Mathews, J. Phys.
Chem. C 2015, 119, 1763–1767.
K. P. Marshall, M. Walker, R. I. Walton, R. A. Hatton, Nat. Energy 2016, 1,
16178.
T.-B. Song, T. Yokoyama, C. C. Stoumpos, J. Logsdon, D. H. Cao, M. R.
Wasielewski, S. Aramaki, M. G. Kanatzidis, J. Am. Chem. Soc. 2017, 139,
836–842.
a) C. Wurster, E. Schobig, Ber. Dtsch. Chem. Ges. 1879, 12, 1807–1813;
b) C. Wurster, Ber. Dtsch. Chem. Ges. 1879, 12, 2071–2072; c) C. Wurster,
Ber. Dtsch. Chem. Ges. 1887, 20, 808–810.
a) L. Michaelis, E. S. Hill, J. Am. Chem. Soc. 1933, 55, 1481–1494; b) L.
Michaelis, M. P. Schubig, S. Granick, J. Am. Chem. Soc. 1939, 61, 1981–
1992.
S. V. Rosokha, J. K. Kochi, J. Am. Chem. Soc. 2007, 129, 3683–3697.
I. Antol, Z. Glasovac, D. Margetić, R. Crespo-Otero, M. Barbatti, J. Phys.
Chem. A 2016, 120, 7088–7100.
M. P. Hautzinger, J. Dai, Y. Ji, Y. Fu, J. Chen, I. A. Guzei, J. Wright, Y. Li, S.
Jin, Inorg. Chem. 2017, 56, 14991–14998.
[13] See, for example: I. Kaljurand, A. Kütt, L. Sooväli, T. Rodima, V. Mäemets,
I. Leito, I. A. Koppel, J. Org. Chem. 2005, 70, 1019–1028.
[14] T. Ishikawa (Ed.), Superbases for Organic Synthesis, John Wiley & Sons, Ltd.
2009.
[15] H. Bredereck, K. Bredereck, Chem. Ber. 1961, 94, 2278–2295.
[16] U. Wild, O. Hübner, L. Greb, M. Enders, E. Kaifer, H.-J. Himmel, Eur. J. Org.
Chem. 2018, 5910–5915.
[45]
[46]
[47]
[48]
[17] H. Herrmann, P. Walter, E. Kaifer, H.-J. Himmel, Eur. J. Inorg. Chem. 2017,
5539–5544.
[18] a) B. V. Lotsch, Angew. Chem. Int. Ed. 2014, 53, 635–637; Angew. Chem.
2014, 126, 647–649; b) P. Gao, M. Grätzel, M. K. Nazeeruddin, Energy
Environ. Sci. 2014, 7, 2448–2463; c) C. C. Stoumpos, M. G. Kanatzidis, Acc.
Chem. Res. 2015, 48, 2791–2802; d) A. Senocrate, J. Maier, J. Am. Chem.
Soc. 2019, 141, 8382–8396.
[19] Z. Wang, Z. Shi, T. Li, Y. Chen, W. Huang, Angew. Chem. Int. Ed. 2017, 56,
1190–1212; Angew. Chem. 2017, 129, 1210–1233.
[20] a) M. M. Lee, J. Teuscher, T. Miyasaka, T. N. Murakami, H. J. Snaith, Science
2012, 338, 643–647; b) M. Liu, M. B. Johnston, H. J. Snaith, Nature 2013,
501, 395–398.
[21] Z. Zhao, F. Gu, H. Rao, S. Ye, Z. Liu, Z. Bian, C. Huang, Adv. Energy Mater.
2019, 9, 1802671 (1–28).
[22] X. Zhang, L. Li, Z. Sun, J. Luo, Chem. Soc. Rev. 2019, 48, 517–539.
[23] D. Weber, Z. Naturforsch. B 1978, 33, 1443–1445.
[24] a) M. Szafrański, Thermochim. Acta 1997, 307, 177–183; b) N. De Marco,
H. Zhou, Q. Chen, P. Sun, Z. Liu, L. Meng, E.-P. Yao, Y. Liu, A. Schiffer, Y.
Yang, Nano Lett. 2016, 16, 1009–1016; c) L. Dimesso, A. Quintilla, Y.-M.
Kim, U. Lemmer, W. Jaegermann, Mater. Sci. Eng. B 2016, 204, 27–33.
[25] G. Giorgi, J.-I. Fujisawa, H. Segawa, K. Yamashita, J. Phys. Chem. C 2015,
119, 4694–4701.
[26] A. D. Jodlowski, A. Yépez, R. Luque, L. Camacho, G. de Miguel, Angew.
Chem. Int. Ed. 2016, 55, 14972–14977; Angew. Chem. 2016, 128, 15196–
15201.
[27] M. Wilke, N. Casati, Chem. Eur. J. 2018, 24, 17701–17711.
[28] N. Pellet, P. Gao, G. Gregori, T.-Y. Yang, M. K. Nazeeruddin, J. Maier, M.
Grätzel, Angew. Chem. Int. Ed. 2014, 53, 3151–3157; Angew. Chem. 2014,
126, 3215–3221.
[29] C. Yi, J. Luo, S. Meloni, A. Boziki, N. Ashari-Astani, C. Grätzel, Energy Envi-
ron. Sci. 2016, 9, 656–662.
[30] W. Ke, I. Spanopoulos, Q. Tu, I. Hadar, X. Li, G. S. Shekhawat, V. P. Dravid,
M. G. Kanatzidis, J. Am. Chem. Soc. 2019, 141, 8627–8637.
[31] D. B. Mitzi, J. Chem. Soc., Dalton Trans. 2001, 1–12.
[32] Y. Zhang, J. Liu, Z. Wang, Y. Xue, Q. Ou, L. Polavarapu, J. Zheng, X. Qi, Q.
Bao, Chem. Commun. 2016, 52, 13637–13655.
[49]
[50]
[51]
[52]
[53]
[54]
[55]
[56]
[57]
[58]
[59]
[60]
[61]
For a similar analysis, see for example: A. E. Maughan, J. A. Kurzman, J. R.
Neilson, Inorg. Chem. 2015, 54, 370–378.
K.-z. Du, Q. Tu, X. Zhang, Q. Han, J. Liu, S. Zauscher, D. B. Mitzi, Inorg.
Chem. 2017, 56, 9291–9302.
For TG/DSC analysis of CH3NH3MI3 perovskites (M = Sn or Pb), see the
examples given in ref.[41]
.
T. Li, W. A. Dunlap-Shohl, Q. Han, D. B. Mitzi, Chem. Mater. 2017, 29,
6200–6204.
T. Li, W. A. Dunlap-Shohl, E. W. Reinheimer, P. Le Magueres, D. B. Mitzi,
Chem. Sci. 2019, 10, 1168–1175.
K. Kabsch, in: M. G. Rossmann, E. Arnold (Eds.), International Tables for
Crystallography Vol. F, Ch. 11.3, Kluwer Academic Publishers, Dordrecht,
The Netherlands, 2001.
[33] L. Mao, C. C. Stoumpos, M. G. Kanatzidis, J. Am. Chem. Soc. 2019, 141,
1171–1190.
[62]
a) DENZO-SMN, Z. Otwinowski & W. Minor, Processing of X-ray Diffraction
Data Collected in Oscillation Mode, Methods Enzymol. (1997), p. 276
(Eds.: C. W. Carter, R. M. Sweet), Academic Press; b) SAINT, Bruker AXS
GmbH, Karlsruhe, Germany 2017; c) CrysAlisPro, Agilent Technologies UK
Ltd., Oxford, UK 2011–2014 and Rigaku Oxford Diffraction, Rigaku Polska
Sp.z o.o., Wrocław, Poland 2015–2019.
R. H. Blessing, Acta Crystallogr., Sect. A 1995, 51, 33–38.
a) SCALE3 ABSPACK, CrysAlisPro, Agilent Technologies UK Ltd., Oxford, UK
2011–2014 and Rigaku Oxford Diffraction, Rigaku Polska Sp.z o.o., Wro-
cław, Poland 2015–2019; b) G. M. Sheldrick, SADABS, Bruker AXS GmbH,
Karlsruhe, Germany 2004–2014; c) L. Krause, R. Herbst-Irmer, G. M. Shel-
drick, D. Stalke, J. Appl. Crystallogr. 2015, 48, 3–10.
[34] M. Safdari, P. H. Svensson, M. T. Hoang, I. Oh, L. Kloo, J. M. Gardner, J.
Mater. Chem. A 2016, 4, 15638–15646.
[35] a) D. B. Mitzi, C. A. Feild, W. T. A. Harrison, A. M. Guloy, Nature 1994, 369,
467–469; b) C. C. Stoumpos, D. H. Cao, D. J. Clark, J. Young, J. M. Rondin-
elli, J. I. Jang, J. T. Hupp, M. G. Kanatzidis, Chem. Mater. 2016, 28, 2852–
2867; c) D. H. Cao, C. C. Stoumpos, O. K. Farha, J. T. Hupp, M. G. Kanatzidis,
J. Am. Chem. Soc. 2015, 137, 7843–7850.
[63]
[64]
[36] A. Mei, X. Li, L. Liu, Z. Ku, T. Liu, Y. Rong, M. Xu, M. Hu, J. Chen, Y. Yang,
M. Grätzel, H. Han, Science 2014, 345, 295–298.
[37]
I. C. Smith, E. T. Hoke, D. Solis-Ibarra, M. D. McGehee, H. I. Karunadasa,
Angew. Chem. Int. Ed. 2014, 53, 11232–11235; Angew. Chem. 2014, 126,
11414–11417.
[65]
[66]
[38]
[39]
[40]
C. C. Boyd, R. Cheacharoen, T. Leijtens, M. D. McGehee, Chem. Rev. 2019,
119, 3418–3451.
N. I. Selivanov, A. A. Murashkina, R. Kevorkyants, A. V. Emeline, D. W.
Bahnemann, Dalton Trans. 2018, 47, 16313–16319.
M. Konstantakou, T. Stergiopoulos, J. Mater. Chem. A 2017, 5, 11518–
11549.
W. R. Busing, H. A. Levy, Acta Crystallogr. 1957, 10, 180–182.
a) M. C. Burla, R. Caliandro, B. Carrozzini, G. L. Cascarano, C. Cuocci, C.
Giacovazzo, M. Mallamo, A. Mazzone, G. Polidori, SIR2014, CNR IC, Bari,
Italy, 2014; b) M. C. Burla, R. Caliandro, B. Carrozzini, G. L. Cascarano, C.
Cuocci, C. Giacovazzo, M. Mallamo, A. Mazzone, G. Polidori, J. Appl. Crys-
tallogr. 2015, 48, 306–309.
Eur. J. Inorg. Chem. 2019, 4147–4160
www.eurjic.org
4159
© 2019 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[67] a) P. T. Beurskens, G. Beurskens, R. de Gelder, J. M. M. Smits, S. Garcia-
Granda, R. O. Gould, DIRDIF-2008, Radboud University Nijmegen, The
Netherlands, 2008; b) P. T. Beurskens, in: G. M. Sheldrick, C. Krüger, R.
Goddard (Eds.), Crystallographic Computing 3, Clarendon Press, Oxford,
UK, 1985, p. 216.
[68] a) L. Palatinus, SUPERFLIP, EPF Lausanne, Switzerland and Fyzikální ústav
AV ČR, v. v. i., Prague, Czech Republic, 2007–2014; b) L. Palatinus, G.
Chapuis, J. Appl. Crystallogr. 2007, 40, 786–790.
[69] a) G. M. Sheldrick, SHELXL-20xx, University of Göttingen and Bruker AXS
GmbH, Karlsruhe, Germany 2012–2018; b) W. Robinson, G. M. Sheldrick
in: N. W. Isaaks, M. R. Taylor (Eds.), Crystallographic Computing 4, Ch. 22,
IUCr and Oxford University Press, Oxford, UK, 1988; c) G. M. Sheldrick,
Acta Crystallogr., Sect. A 2008, 64, 112–122; d) G. M. Sheldrick, Acta Crys-
tallogr., Sect. C 2015, 71, 3–8.
[70] a) J. S. Rollett in: F. R. Ahmed, S. R. Hall, C. P. Huber (Eds.), Crystallographic
Computing p. 167, Munksgaard, Copenhagen, Denmark, 1970; b) D. Wat-
kin in: N. W. Isaaks, M. R. Taylor (Eds.), Crystallographic Computing 4, Ch.
8, IUCr and Oxford University Press, Oxford, UK, 1988; c) P. Müller, R.
Herbst-Irmer, A. L. Spek, T. R. Schneider, M. R. Sawaya in: P. Müller (Eds.),
Crystal Structure Refinement, Ch. 5, Oxford University Press, Oxford, UK,
2006; d) D. Watkin, J. Appl. Crystallogr. 2008, 41, 491–522.
[72]
[73]
[74]
a) P. van der Sluis, A. L. Spek, Acta Crystallogr., Sect. A 1990, 46, 194–201;
b) A. L. Spek, Acta Crystallogr., Sect. C 2015, 71, 9–18.
a) A. L. Spek, PLATON, Utrecht University, The Netherlands; b) A. L. Spek,
J. Appl. Crystallogr. 2003, 36, 7–13.
TURBOMOLE V7.2 2017, a development of University of Karlsruhe and
Forschungszentrum Karlsruhe GmbH, 1989–2007, TURBOMOLE GmbH,
since 2007; available from http://www.turbomole.com.
R. Ahlrichs, M. Bär, M. Häser, H. Horn, C. Kölmel, Chem. Phys. Lett. 1989,
162, 165–169.
O. Treutler, R. Ahlrichs, J. Chem. Phys. 1995, 102, 346–354.
A. D. Becke, J. Chem. Phys. 1993, 98, 5648–5652.
P. J. Stephens, F. J. Devlin, C. F. Chabalowski, M. J. Frisch, J. Phys. Chem.
1994, 98, 11623–11627.
F. Weigend, R. Ahlrichs, Phys. Chem. Chem. Phys. 2005, 7, 3297–3305.
K. Eichkorn, O. Treutler, H. Öhm, M. Häser, R. Ahlrichs, Chem. Phys. Lett.
1995, 242, 652–660.
F. Weigend, Phys. Chem. Chem. Phys. 2006, 8, 1057–1065.
R. Bauernschmitt, R. Ahlrichs, Chem. Phys. Lett. 1996, 256, 454–464.
R. Bauernschmitt, M. Häser, O. Treuler, R. Ahlrichs, Chem. Phys. Lett. 1997,
264, 573–578.
[75]
[76]
[77]
[78]
[79]
[80]
[81]
[82]
[83]
[84]
A. Klamt, G. Schürmann, J. Chem. Soc., Perkin Trans. 2 1993, 799–805.
[71] a) J. S. Rollett in: F. R. Ahmed, S. R. Hall, C. P. Huber (Eds.), Crystallographic
Computing p. 167, Munksgaard, Copenhagen, Denmark, 1970; b) A.
Thorn, B. Dittrich, G. M. Sheldrick, Acta Crystallogr., Sect. A 2012, 68, 448–
451.
Received: September 12, 2019
Eur. J. Inorg. Chem. 2019, 4147–4160
www.eurjic.org
4160
© 2019 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim