10.1002/anie.201809028
Angewandte Chemie International Edition
COMMUNICATION
perovskites, yet, they used gamma radiation to induce
polymerization, which makes this technique unpractical for most
applications.[14] Other attempts have been made to incorporate
pre-synthesized conducting polymers into perovskites, however,
they often result in poorly crystalline and ill-characterized
materials.[15,16] Moreover, no attention has been given to the
electronic properties of these materials, nor have there been any
reports on the effect that doping might exert in this kind of
materials, despite being well-known in the field of conducting
polymers that doping is essential to achieve high conductivities.
We therefore sought out to study the effect of polydiacetylene
incorporation and doping in the optical and electronic properties
of 2D hybrid perovskites.
cations in 1–N2, that can be identified by vibrational spectroscopy
or by material’s digestion followed by solution-state 13C nuclear
magnetic resonance (NMR). The presence of unreacted diynes
is consistent with a topochemical polymerization, which would be
governed by a stochastic process that would inexorably leave
some unpolymerized monomers.[19] The 13C NMR of the organic
fraction shows the appearance of new signals between 113 and
128 ppm attributable to C–C double bonds, as well as a new set
of signals for aliphatic and vinylic carbons. Thus, solution NMR
of the isolated organic, is also consistent with polydiacetylene
formation (Figure S15).
To our surprise, when samples of 1-as were heated under air
instead of nitrogen, crystals turned black instead of yellow
(Figure 2e). This material, hereafter called 1-O2, is still highly
crystalline and has very similar PXRD pattern and IR spectrum
as 1-N2, however, the UV-Vis-NIR absorption spectra of 1-O2
revealed that this material now absorbs well into the near-IR
(NIR) region of the spectra with an optical bandgap of 1.4 eV,
that is, 1.6 eV smaller than its parent compound, 1-as (Figures
S11 and S14). Notably, absorption in the NIR part of the
spectrum is highly desirable for photovoltaic applications and it
is indeed where most 2D, and some 3D, perovskites fail to
absorb, which makes this an attractive alternative to increase
light absorption.
We started our studies by synthesizing the conjugated dialkynyl
ammonium
salt
of
deca-3,5-diyn-1-amine:
(H3N+CH2CH2CºCCºCC4H9) or DDA. The hydrochloric salt of
DDA (or DDA•Cl) can be easily obtained from commercially
available precursors in two synthetic steps and with good yields
(See Supporting Information (SI) for details). DDA was
incorporated into
a 2D hybrid perovskite with formula:
[DDA]2PbBr4 (1-as) which was characterized by IR, UV-Vis,
single-crystal (SCXRD) and powder (PXRD) X-ray diffraction; all
of which are consistent with the sought after 2D perovskite
structure with an optical bandgap of 3.0 eV (Figure S11). The
SCXRD structure clearly shows the presence of two neighboring
triple bonds with CºC distances of 1.196 and 1.167 Å and
intermolecular C(sp) to C(sp) distances that are as small as 3.79
and 3.85 Å, clearly in a distance range that could allow for
topochemical polymerization (Figure 2a,b).[17,18]
Despite the striking differences in color, 1-N2 and 1-O2 have
almost identical PXRD pattern and IR spectra. This suggests that
structurally, similar materials were formed in both cases, and
thus we hypothesize that the color differences can be attributed
to the presence of oxygen, which caused partial oxidation of
polydiacetylene moiety. Such oxidation could result in the
generation of free carriers and a concomitant shrink of its optical
bandgap. To probe this hypothesis, we performed the thermal
treatment of 1-as under an iodine (I2) atmosphere. Due its
volatility and lower redox potential than oxygen, molecular iodine
is a commonly used oxidant for conducting polymers, as such, it
could act as an alternative oxidant that could then confirm if
oxygen is indeed responsible for the observed changes.
Treatment of 1-as with iodine yielded 1-I2, which has almost
identical PXRD pattern and light absorption properties than 1-O2
(Figures S16-S18). Alternatively, 1-N2, can also be exposed to
iodine vapors at room temperature to yield 1-N2/I2; which also
presents similar absorption and PXRD pattern to those of 1-O2
and I-I2 (Figure S18). Importantly, neither 1-I2 or 1-N2/I2, show
evidence of iodine addition to triple or double bonds[20] halide
swap,[21] nor does the observed effects can be explained by
intercalation of I2.[22]
Inspired by Tieke’s work,[10-13] we first tried to promote
polymerization by irradiating samples of 1-as with UV light (250
and 354 nm). Such treatments did not result in any measurable
change in PXRD pattern or UV-ViS absorbance. The lack of
reactivity of 1-as towards UV light can be rationalized by
considering the high absorption coefficients of lead bromide
layers, particularly in the UV region, which prevents UV light from
reaching the organic fragments and thus, preventing their
polymerization. We then tried heating a crystalline sample of 1-
as at 160 °C (slightly below its decomposition temperature at
180 °C, Figure S12) for 24 hours under an inert atmosphere (N2).
Such treatment resulted in a highly crystalline yellow material,
hereafter called 1-N2 (Figure 2d). 1-N2 shows an expanded unit
cell when compared to 1-as. Particularly the c axis (which is
directly related to the layer to layer distance) expands from 19.8
to 22.4 Å (Figure 2f). The color change and unit cell expansion
come with the appearance of a C=C stretch at 1618 cm–1 in the
vibrational spectrum (Figure S13). All of these changes are
consistent with polydiacetylene formation (Figure 1a).[14]
Similarly, UV-Vis absorption data shows the appearance of a
broad shoulder at about 520 nm, attributable to π-π* transitions
in highly conjugated polydiacetylene fragments (Figure 2g).[14]
Importantly, regardless of the duration of the thermal treatment
duration, we always observed remaining, non-polymerized DDA
This article is protected by copyright. All rights reserved.
Ortiz-Cervantes, Carmen
