Stable fluorescent NH3 sensor based on MAPbBr3 encapsulated by tetrabutylammonium cations

Article abstract of DOI:10.1016/j.jallcom.2020.155386

Recently, organic-inorganic halide perovskite materials were investigated on gas sensing due to their excellent optical properties and gas sensitivities. Here, we designed a stable fluorescent perovskite-based sensor for NH₃ detection, in which the CH₃NH₃PbBr₃ (MAPbBr₃) film was deposited on the GeO₂ substrate and also capped by tetrabutylammonium (TBA) ligand as a stabilizing agent. This as-fabricated MAPbBr₃-TBA-based sensor exhibited the relatively strong and stable photoluminescence (PL) intensity, thereby expanding the fluorescence response range for NH₃ sensing. Upon exposure to NH₃ gas, the PL intensity quenched rapidly by 62.5% with short response time (61 s) and recovery time (65 s). The sensor also possessed a linear relationship between the PL intensity and concentration of NH₃ in the range of 0–100 ppm, and presented excellent reversibility, high gas selectivity, and humidity stability. Furthermore, the NH₃ sensing mechanism was investigated based on X-ray diffraction (XRD), Fourier-transform infrared spectroscopy (FT-IR), Differential thermal analysis-Thermogravimetry (DTA-TG) and fluorescence lifetime measurements, in which the NH₃ molecules might permeate the capped TBA ligand and then induce structure transformation of inner MAPbBr₃ crystal. This study indicated that the as-prepared MAPbBr₃-TBA-based sensor might provide promising applications on NH₃ gas detection.

Full text of DOI:10.1016/j.jallcom.2020.155386

Journal of Alloys and Compounds 835 (2020) 155386
Contents lists available at ScienceDirect
Journal of Alloys and Compounds
journal homepage: http://www.elsevier.com/locate/jalcom
Stable fluorescent NH3 sensor based on MAPbBr3 encapsulated by
tetrabutylammonium cations
Guishun Li , Wenqian Zhang , Changkun She , Shicheng Jia , Shaohua Liu , Fangyu Yue ,
Chengbin Jing *, Ya Cheng , Junhao Chu
Engineering Research Center for Nanophotonics and Advanced Instrument of Ministry of Education, Key Laboratory of Polar Materials and Devices (Ministry
of Education), The Extreme Optoelectromechanics Laboratory, Department of Materials, School of Physics and Electronic Science, East China Normal
University, 500 Dongchuan Road, Shanghai, 200241, China
a r t i c l e i n f o
a b s t r a c t
Article history:
Recently, organic-inorganic halide perovskite materials were investigated on gas sensing due to their
excellent optical properties and gas sensitivities. Here, we designed a stable fluorescent perovskite-based
sensor for NH3 detection, in which the CH3NH3PbBr3 (MAPbBr3) film was deposited on the GeO2 sub-
strate and also capped by tetrabutylammonium (TBA) ligand as a stabilizing agent. This as-fabricated
MAPbBr3-TBA-based sensor exhibited the relatively strong and stable photoluminescence (PL) in-
tensity, thereby expanding the fluorescence response range for NH3 sensing. Upon exposure to NH3 gas,
the PL intensity quenched rapidly by 62.5% with short response time (61 s) and recovery time (65 s). The
sensor also possessed a linear relationship between the PL intensity and concentration of NH3 in the
range of 0e100 ppm, and presented excellent reversibility, high gas selectivity, and humidity stability.
Furthermore, the NH3 sensing mechanism was investigated based on X-ray diffraction (XRD), Fourier-
transform infrared spectroscopy (FT-IR), Differential thermal analysis-Thermogravimetry (DTA-TG) and
fluorescence lifetime measurements, in which the NH3 molecules might permeate the capped TBA ligand
and then induce structure transformation of inner MAPbBr3 crystal. This study indicated that the as-
prepared MAPbBr3-TBA-based sensor might provide promising applications on NH3 gas detection.
© 2020 Elsevier B.V. All rights reserved.
Received 12 February 2020
Received in revised form
23 April 2020
Accepted 25 April 2020
Available online 1 May 2020
Keywords:
MAPbBr3
Tetrabutylammonium ligand
PL quenching
Structure transformation
NH3 sensor
1. Introduction
and TiO2 [9,10]. For example, Chen et al. reported an NH3 sensor
based on the SnO2-nanorods/reduced graphene oxide composites
As a raw material, ammonia (NH3) is extensively utilized in the
fields of agriculture and chemistry industries. However, NH3 as
toxic gas can cause serious harm to the human body, even at quite
low concentration. Generally, human tolerance levels of NH3 gas
were defined at 15 min exposure limit of 35 ppm and an 8 h
exposure limit of 25 ppm [1,2]. So it is necessary to adopt effective
methods to monitor NH3 gas precisely. In this regard, researchers
have proposed many novel detection techniques based on various
sensing materials, such as metal oxides [3e5], conducting polymers
[6e8] and quantum dots [1]. Most of them can be classified as
electrical-type and optical-type sensors, mainly depending on the
changes in electrical or optical signals induced by NH3 gas.
with fast response time (8 s) and recovery time (13 s) at 200 ppm
NH3 [3]. Recently, Maity et al. designed a novel and wearable NH3
sensor by combining the polyaniline and multiwall carbon nano-
tubes based on strong
P-P interaction between them [11]. Despite
these significant developments, it is still challenge to achieve supe-
rior sensors with high stability and sensitivity.
In the last decades, metal halide perovskites have drawn much
attention due to their excellent optical and electrical properties.
Thus, the perovskites were widely investigated in several fields
such as solar cells [12], light-emitting diodes [13], photo-detectors
[14] and lasers [15]. Besides, the perovskites can be also applied as
the sensors to detecting NO2 [16], O2 [17], O3 [18], NH3 [19e24], SO2
[25], humidity [26], temperature [27], organic vapors [28,29] and
metal ions [30]. For the NH3 gas sensing, it was firstly reported by
Zhu et al. that the NH3 molecules were able to react with the
CH3NH3PbI3 (MAPbI3), followed by the color changing from brown
to colorless [22]. To enhance the sensitivity and stability, Ruan et al.
For the metal oxide-based sensors, many available sensing ma-
terials were developed including ZnO, In2O3, SnO2, SnS2, MoO3, WO3
* Corresponding author.
E-mail address: cbjing@ee.ecnu.edu.cn (C. Jing).
https://doi.org/10.1016/j.jallcom.2020.155386
0925-8388/© 2020 Elsevier B.V. All rights reserved.

2
G. Li et al. / Journal of Alloys and Compounds 835 (2020) 155386
designed and fabricated a fiber probe based on MAPbX3 (X ¼ Cl, Br
or I) single crystals [21]. In their case, the recorded photo-
luminescence (PL) decreased dramatically by 60%, upon exposure
to NH3 vapor. However, the fabrication of this sensor may require a
complex single-crystal growth process, and the accurate NH3 con-
centration was not mentioned. Herein, it is an urgent need to
develop simple and facile perovskite sensors to detect the low
concentration NH3 with long-term stability.
the perovskite precursor solutions with/without TBA were spin-
coated on GeO2 substrates at 2000 rpm for 60 s Meanwhile, to
induce the crystallization quickly, 100 mL toluene was dropped onto
the substrates at 10 s during the spinning process. Finally, the as-
prepared samples were annealed at 60 ꢀC for 10 h.
2.4. Characterizations
Recently, a long alkyl chain salt tetrabutylammonium-bromide
(TBA) was used as a passivation agent to stabilize the perovskite,
owing to the ability to control crystallization [31]. For the sensing
application, we expected that this capping layer on the surface of
perovskite could increase its stability with suitable gas permeability.
In order to explore this possibility, in this paper, we prepared the
TBA-coated MAPbBr3 films on GeO2 substrates to investigate the
influences of TBA on NH3 sensing. The pre-deposited GeO2 sub-
strates possessed the rough surface which facilitated the adhesion
and dispersion of perovskite crystals compared to the smooth
glasses. The sensing performance of the as-prepared MAPbBr3-TBA-
based sensor was determined that the PL intensity quenched rapidly
upon exposure to NH3 gas and then recovered to the initial PL in-
tensity after removing NH3 gas. The NH3 sensing mechanism was
also studied based on XRD, FT-IR, DTA-TG and PL lifetime measure-
ments. Hence, it paves the way for the application of organic-
inorganic halide perovskite on NH3 detection.
The film morphologies and element distributions were observed
by using a scanning electron microscope combined with an energy-
dispersive X-ray spectroscope (SEM/EDS, Gemini 450). X-ray diffrac-
tion patterns were measured by a rotating Cu K
diffractometer (XRD, Rigaku Ultima IV) under 35 kV and 20 mA. The
scanning range was set as 5ꢀ<2 < 50ꢀ with a speed of 10ꢀ minꢁ1. The
a anode X-ray
q
infrared spectrum was recorded on a Bruker 80 V Fourier Transform
Infrared Spectrometer (FT-IR). The PL spectra data of samples were
collected using the fluorescence spectrophotometer with a slit width
of 2 nm (LengGuang, F97pro). The thermal behaviors were measured
by differential thermal analysis (DTA) and thermogravimetry (TG)
using a Mettler Toledo TGA/SDTA 851e system; measurements were
carried out from 25 to 700 ꢀC (heating rate: 10 ꢀC/min) under N2 flow.
Finally, the PL decay lifetimes were measured by an Edinburgh In-
struments FLS920 fluorescence spectrometer equipped with a pico-
second pulsed diode laser (
time-correlated single-photon counting unit.
l
¼ 371.6 nm, pulse width ¼ 66.3 ps) and a
2. Experimental
2.5. Fabrication of NH3 gas sensors
2.1. Preparation of GeO2 substrates
The schematic measurement setup is shown in Fig. 1. The quartz
cuvette (10*10*40 mm3) was used as a gas chamber to hold perov-
skite samples. The PL spectra of samples were excited by a Xe lamp
source and then recorded by a photomultiplier tube (PMT) detector.
Different concentrations of NH3 gas were obtained via mixing
appropriate pure N2 and NH3 gases. The target gas was blown into
the chamber by gas sampling pump with a constant flow rate of
0.5 L/min. The gas response was determined bycalculating the value
of response (I0eI)/I0ꢂ100%, where the I0 was the initial PL intensity
under NH3-free condition, and I was the PL intensity in the present of
NH3 gas. The response time Tres was defined as the time for the
sensor to reach 90% of maximum PL intensity change upon exposure
to NH3 gas, and the recovery time Trec as the time to recover 90% of
maximum PL intensity change after removing NH3 followed by
exposure of pure N2 gas. To measure the effect of ambient moisture
on the response of this gas sensor, we carried out the PL measure-
ment under different relative humidity (RH) conditions (using the
vapors of saturated salt solutions including MgCl2, K2CO3, NaBr,
NaCl, KCl and K2SO4 at room temperature).
The growth of GeO2 layers was taken by a modified liquid phase
deposition method which reported by Jing et al., previously [32]. In
detail, 0.75 mL of aqueous ammonia (26e28 wt%) was dissolved in
deionized water (25 mL) at 65 ꢀC, followed by adding 1.5 g GeO2
powders. After stirring for 30 min, the transparent solution was
obtained. Then, the nitric acid (36 wt%) was added dropwise to the
above solution until the pH was adjusted to the value of 2. After-
ward, the glass sheets (10*10*1 mm3), sonicated successively in
acetone and ethanol for 20 min, were soaked into the above
aqueous solution. The deposition process was carried out for 36 h at
room temperature. Finally, the white GeO2 layers were observed on
the glass sheets, and then these GeO2 substrates were washed by
deionized water and ethanol, followed by drying at 100 ꢀC for 1 h.
2.2. Synthesis of CH3NH3Br
CH3NH3Br (MABr) was synthesized by mixing methylamine
(MA, 30 wt% in ethanol) with equimolar amounts of aqueous HBr
(40 wt% in water). First, the MA in anhydrous ethanol was stirred
and cooled to 0 ꢀC by adding HBr acid slowly. Subsequently, the
reaction solution was stirred continuously for 2 h under N2 atmo-
sphere and then evaporated by rotary evaporation at 50 ꢀC. The as-
obtained MABr precipitate was washed with diethyl ether several
times to remove the residual HBr and then dried at 60 ꢀC in a
vacuum oven overnight.
3. Results and discussion
Fig. 2 presents the SEM images and EDS maps of the as-prepared
perovskites deposited on glass or GeO2 substrates. It can be seen
that the perovskite MAPbBr3-TBA possesses the piece-like shape
adhered to a smooth glass substrate (Fig. 2a), while the morphol-
ogies of MAPbBr3 and MAPbBr3-TBA on GeO2 layers show the
typical perovskite grains (Fig. 2b and c). It is reasonable that the
GeO2 layer has a rough surface (Fig. 2b), which may facilitate the
heterogeneous nucleation and dispersion of perovskite crystals,
compared to the smooth glass substrate. Clearly, it appears the
cubic-structure grains in the MAPbBr3-TBA sample (Fig. 2d), in
which the grain sizes are smaller than that of MAPbBr3 (Fig. 2c),
indicating the TBA ligand may modify the outer surface perovskite
crystal to reduce grain size [31]. In order to verify whether these
cubic grains are MAPbBr3-TBA, we selected some grains (the rect-
angular area in Fig. 2d) and measured the elemental composition
2.3. Preparation of perovskite films
Initially, the MAPbBr3 (0.3 M) precursor solution was prepared
by dissolving equimolar amounts of MABr (67.2 mg) and PbBr2
(220 mg) in N, N-Dimethylformamide (DMF, 2 mL) under vigorous
stirring. Then, the above solution was divided into 2 equal parts.
The TBA raw material (19.3 mg) was added in one part of them with
the ratio of TBA/MA at 20%. All the solutions were stirred at 50 ꢀC
overnight. For the perovskite film deposition, one droplet (40 mL) of

G. Li et al. / Journal of Alloys and Compounds 835 (2020) 155386
3
Fig. 1. Schematic diagram of experimental setup for NH3 sensing.
by EDS, as shown in Fig. 2f-j. The result indicates the presence of C
(43.7 at%), N (11.2 at%), Pb (13.0 at%) and Br (32.2 at%), which is
closed to the experimental compositions. Fig. 2e shows the cross-
sectional image of MAPbBr3-TBA/GeO2 substrate with the thick-
FT-IR spectra were also conducted to verify the presence of TBA
in perovskite, as shown in Fig. 3b. The broadband peak at 3450-
3300 cmꢁ1 is observed in two samples, mainly assigned to the NeH
stretching of amine salts and the OH- stretching of incidental water
vapor. In addition, the vibration peak at 2960 cmꢁ1 can be assigned
to the CH- stretching characteristic of both TBAþ and MAþ cations,
while the peak at 2865 cmꢁ1 corresponds to the eCH2- group of
only TBAþ cations [37]. The above results indicate the presence of
TBAþ and MAþ cations in the MAPbBr3-TBA sample.
To study the effect of the substrate and TBA on optical pros-
perities of perovskite films, Fig. 4 gives the PL spectra of perovskite
samples including the MAPbBr3-TBA on glass, MAPbBr3 on GeO2
and MAPbBr3-TBA on GeO2. In these three samples, MAPbBr3-TBA
on GeO2 presents the maximum PL intensity. It implies that the
GeO2 layer may provide a good crystal growth substrate for
perovskite and the TBA ligand could passivate the surface defects
and grain boundaries of perovskite crystal by interacting with Pb
atoms, thereby inhibiting the non-radiative recombination [35,38].
Moreover, a slight red-shift of the peak appears in the TBA-doped
sample. We infer that the TBA ligand attached to perovskite sur-
face may form several intermediate energy levels, contributing to
the tiny reduction in energy bandgap. This red-shift phenomenon is
similar to that of TBA-doped CsPbBr3 perovskite which reported
ness at about 10 mm. It is worth noting that the layers of both GeO2
and perovskite are difficult to distinguish because the perovskite
film is too thin and several grains are embedded into the gap of the
GeO2 layer (see EDS data in Fig. S1).
Fig. 3a gives the XRD patterns of MAPbBr3 and MAPbBr3-TBA
films. In these two samples, it appears two prominent diffraction
peaks at 15.2ꢀ and 30.4ꢀ, identified with the (100) and (200) planes
of MAPbBr3 crystal structure, respectively [33]. In the MAPbBr3-TBA
sample, the peak at low angle 7.6ꢀ is also observed, assigned to the
dimensional 2D perovskite phase [31,34]. Since the large size cation
TBAþ is difficult to insert into 3D MAPbBr3 crystal, it may grow on
the surface of perovskite as a capping layer [35,36]. Interestingly,
the main diffraction peaks (15.2ꢀ) appear slight right-shift and
become broaden upon the addition of TBAþ cations (see Fig. S2),
indicating the reduction in grain size of the 3D perovskite, which is
consistent with the SEM data. This case also demonstrates that the
TBA ions would form a passivation layer on the surface of perov-
skite to limit the growth of perovskite rather than insert into the
inner of 3D crystal.
Fig. 2. SEM image of (a) MAPbBr3-TBA on smooth glass substrate, (b) GeO2 layer, (c) MAPbBr3 on GeO2 substrate, (d) MAPbBr3-TBA on GeO2 substrate. (e) cross-sectional image of
MAPbBr3-TBA/GeO2 substrate. (f, g, h, i) EDS maps of (d) rectangular area for C, N, Pb, and Br, respectively. (f) Elemental microanalysis.

4
G. Li et al. / Journal of Alloys and Compounds 835 (2020) 155386
Fig. 3. XRD patterns (a) and FT-IR transmission spectra (b) of MAPbBr3 and MAPbBr3-TBA films.
previously by Song and co-authors [35]. Nevertheless, in several
other studies, the blue-shift characteristic was found upon loading
the long-chain amine salts, attributed to the reduced 3D perovskite
structure dimensionality [38,39]. In contrast, the main roles of TBA
dopant in this work may be utilized to modify and passivate surface
grain boundaries and defects.
To investigate the photo-stability of as-prepared samples, we
measured the PL intensity of perovskites as a function of time (see
Fig. S3). There is no significant decrease of the PL intensity in
perovskite MAPbBr3-TBA, while the PL intensity in pristine MAPbBr3
drops by 10% within 300 s. The result indicates the TBA layer can
keep the perovskite more stable to light irradiation. Actually, this is
an important prerequisite for fluorescent sensors.
As previously reported, long-chain alkyl ammonium cations were
reacted with PbBr2 and only fitted the periphery of the PbX6 octa-
hedra with their chains dangling [40]. Thus, at a low doping con-
centration of TBA, the perovskite crystal surface is not completely
coveredbyTBAþ cations, whichgivesNH3 moleculesachancetoenter
the interior. Also, Nazzal et al. reported the surface passivation layer
could be permeable to gases with the help of the photo-activation of
vibration modes of the crystal lattice induced by photon-phonon
coupling [41].
In order to investigate the response of the perovskite sensor
towards NH3 gas, we measured the color changes in fluorescence
before and after the exposure of NH3 under ultraviolet (UV) light, as
shown in Fig. 5. It can be observed that the significant quenching of
fluorescence occurs when blowing NH3.
Fig. 6a presents the response transients of MAPbBr3-based and
MAPbBr3-TBA-based sensors to 100 ppm NH3 gas under 365 nm
illumination. Upon exposure to NH3, the pristine MAPbBr3 shows
the maximal PL quenching of 74.4%, while the MAPbBr3-TBA shows
the value of 62.5% within ~90 s. It seems that the TBA passivation
layer causes a negative impact on NH3 detecting. However, this
negative impact could be negligible, because the PL intensity is able
to recover fully within a short time after the removal of NH3 gas.
The response and recovery times are calculated as 61 s and 65 s,
respectively. In addition, it should be noted that neither of these
sensors reaches the entire PL quenching. It suggests that the NH3
molecules may be limited to interact with the near-surface region
of perovskite inducing the surface defects. Actually, in case of
immersing into high-pressure NH3 gas, the NH3 molecules could
not only contact with the surface of perovskite, but also penetrate
deeply into the bulk [42,43].
To investigate the reproducibilityof theas-prepared sensor, Fig. 6b
gives the reversible response of the MAPbBr3-TBA-based sensor
versusthetimeunderconstantconcentrationNH3 gas(100 ppm). The
PL intensity quenches rapidly upon exposure to NH3 gas, afterward, it
recovers to the initial level when removing NH3. This excellent
reversibility to NH3 paves a way for the practical application of a gas
sensing system. In addition, we find that different response and re-
covery times are illustrated during these cycles, probably resulting
from the changes in gas flow or the microstructure state of perovskite
after reacting with NH3. Further study is underway.
To explore the sensitivity of the perovskite sensor towards NH3
gas, we recorded the PL response under different concentrations of
NH3 gas (see Fig. 7a). As the NH3 concentration increases, the PL
Fig. 4. PL spectra of MAPbBr3 and MAPbBr3-TBA films on glass or GeO2 substrates,
Fig. 5. Photographs of fluorescence changes in MAPbBr3-TBA film before and after NH3
upon excitation of 365 nm.
treatment, under UV light.

G. Li et al. / Journal of Alloys and Compounds 835 (2020) 155386
5
Fig. 6. (a) The response transients of MAPbBr3-based and MAPbBr3-TBA-based sensors to 100 ppm NH3 gas. (b) Reproducibility of MAPbBr3-TBA-based sensor when switching
between N2 and NH3 gas, alternately.
intensity of the sensor decreases proportionally. From Fig. 7b, an
excellent linear relationship is exhibited between the PL intensity
and concentration of NH3 in the range of 0e100 ppm. The slope is
18.7 with R2 of 0.977. The limit of detection (LOD) is calculated by an
that the ethanol and acetone act as vacancy filler, increasing the
electron-hole recombination in perovskite [45].
To explore the NH3 sensing mechanism, we investigated the
structure and composition changes in perovskite before and after
excess NH3 treatment. Beforehand, the MAPbBr3-TBA sample was
treated in a high concentration NH3 atmosphere for several hours.
At this time, the fluorescence became weak and took a long time to
recover after removing NH3 gas.
Fig. 9a depicts the XRD patterns of MAPbBr3-TBA samples before
and after excess NH3 treatment. After NH3 treatment, several
prominent diffraction peaks of MAPbBr3 crystal become weak even
disappeared such as (100) and (200), whereas the low angle at 7.7ꢀ
(2D phase) is almost unchanged. Moreover, the diffraction peaks of
(110) and (200) appear right-shift, indicating the size decrease of
crystal, mainly due to the decomposition of MAPbBr3. These results
imply that the NH3 molecules might tend to react with the inner
MAPbBr3 rather than outside 2D TBA capping layer, leading to the
phase transformation of MAPbBr3 [21].
FT-IR spectra were also used to evaluate the changes in com-
positions of the MAPbBr3-TBA sample, as shown in Fig. 9b. There is
no obvious change in 2960 cmꢁ1 and 2865 cmꢁ1 peaks before and
after NH3 treatment. The result suggests that no significant loss of
MA and TBA after NH3 treatment. Previously, Huang et al. reported
the MA gas could evaporate from perovskite upon exposure to NH3,
owing to that the CH3NH3þ cations would be replaced by the NH4þ
cations (NH3 þ CH3NH3þ / NH4þ þ CH3NH2[) [42]. In the MAPbBr3-
TBA sample treated by NH3, we speculate that the generated MA
gas molecules may still remain in the perovskite. These MA mole-
cules may locate in uncoordinated states to form the specific in-
termediate phase or adhere to the TBA layer by van der Waals force
[21,46].
equation as LOD ¼ 3
s=S, where s is the relative standard deviation
and S is the slope of the calibration curve. On the basis of this
equation, the LOD for this sensor is calculated to be 0.46 ppm. It
worth noting that the PL peak appears a slight blue-shift when
exposure to excess NH3 gas (see Fig. S4), implying the NH3-induced
structure transformation of perovskite [22].
Previously, it was reported that the humidity had a negative ef-
fect on perovskite solar cells, as the phase transformation of perov-
skite can take place upon exposure to water vapor [17,44]. To verify
the effect of humidity on perovskite sensors, we carried out the
humidity measurement of perovskite sensors under different RH
condition (see Fig. 8a). For the MAPbBr3 and MAPbBr3-TBA perov-
skite sensors, the PL quenching appears upon exposure to 98% RH
water vapor, in which the extent of PL decline was 22.1% and 13.9%,
respectively. Expectantly, for the MAPbBr3-TBA-based sensor, the PL
can be restored absolutely after removing water vapor, attributed to
the protective effect of the TBA passivation layer. As can be seen in
the insert of Fig. 8a, the maximum values of PL quenching are
determined under different RH environment, indicating that the
water vapor has little effect on the performance of MAPbBr3-TBA-
based sensor.
The selectivity performance is also one of important parameters
to NH3 sensor. So we measured the response of MAPbBr3-TBA-
based sensor toward various saturated organic vapors, including
ethanol, acetone, cyclohexane, toluene and chlorobenzene vapors
(see Fig. 8b). It can be observed that this sensor displays an excel-
lent specificity toward NH3 (100 ppm) in comparison with other
saturated organic vapors. Interestingly, the enhanced PL response
occurs upon exposure to ethanol or acetone vapor. It may be due to
Furthermore, to elucidate the sensing mechanism from thermal
decomposition behavior, we carried out the DTA-TG measurements
Fig. 7. MAPbBr3-TBA-based gas sensor. (a) The PL quenching toward different concentration of gaseous NH3 (0e100 ppm), (b) the plot of PL intensity versus concentration of NH3.

6
G. Li et al. / Journal of Alloys and Compounds 835 (2020) 155386
Fig. 8. (a) The humidity response transients of the MAPbBr3-based and MAPbBr3-TBA-based sensors under 98% RH condition, the insert shows the PL quenching in MAPbBr3-TBA
sample under different RH condition. (b) The selectivity of MAPbBr3-TBA-based sensor to various saturated organic vapors at room temperature (25oC).
Fig. 9. XRD patterns (a) and FT-IR transmission spectra (b) of MAPbBr3-TBA samples before and after high concentration NH3 treatment.
on MAPbBr3-TBA (see Fig. S5). In the 1st derivative curve of refer-
ence MAPbBr3-TBA sample (see Fig. S5a), three weight loss peaks
are located at 318, 372 and 643 ꢀC, ascribed to the loss of TBAþ,
MAþ, and inorganic PbBr2, respectively [47,48]. For the NH3 treated
MAPbBr3-TBA (see Fig. S5b), a weak loss peak at 147 ꢀC can be
observed, which may be attributed to the uncoordinated MA [21].
Fig. 10. Schematic diagram of NH3 sensing mechanism for MAPbBr3-TBA-based sensor.

G. Li et al. / Journal of Alloys and Compounds 835 (2020) 155386
7
Moreover, the weight loss at 372 ꢀC becomes weak, indicating the
Acknowledgements
reduction of protonated MAþ content in perovskite. It is suggested
that the thermal stability of perovskite decreases upon exposure to
NH3, which could be one of the evidences that the NH3-induced
decomposition of perovskite.
This work was financially supported by National Natural Science
Foundation of China (61775060, 61275100, 61761136006, 61790583
and 61874043).
In addition, we measured the time-resolved PL to further un-
derstand the sensing mechanism (see Fig. S6). The MAPbBr3-TBA
sample presents a long PL lifetime at 57.65 ns, while it decreases to
the 27.52 ns after NH3 treatment, indicating the increase of defect.
It may be caused by perovskite decomposition, leading to the non-
radiative recombination.
Appendix A. Supplementary data
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.jallcom.2020.155386.
To shed light on the NH3 sensing mechanism, Fig. 10 depicts the
schematic diagram of reversible NH3 sensing processes. It can be
considered as the cation exchange mechanism [42]. Under the
excitation of ultraviolet (UV) light, the strong PL spectrum is emitted
from the MAPbBr3-TBA perovskite sample. Initially, the NH3 mole-
cules would permeate the capped TBA ligand due to the small mo-
lecular size of NH3 and photon-phonon coupling [41] and then
induce the decomposition and structure transformation of MAPbBr3
crystal. This process will bring about the formation of NH4PbBr3 and
MA, leading to the PL quenching [21]. Based on the above charac-
terization measurements and analysis, we speculate that the gaseous
MA might interact with NH4PbBr3 to form the intermediate phase
NH4PbBr3$MA rather than sublimating from inside [21,49]. Mean-
while, the outside TBA ligands with large steric hindrance would also
retard the MA sublimation. After the removal of NH3, the chemical
conversion of NH4þ þ CH3NH2 / NH3 þ CH3NHþ3 occurs via a tran-
sition state of H3N$$H$$CH2NH2 [46]. Consequently, the primary
perovskite is restored gradually and then emerging the initial strong
PL signal.
References
[1] T.W. Sung, Y.L. Lo, Ammonia vapor sensor based on CdSe/SiO2 core-shell
nanoparticles embedded in sol-gel matrix, Sensor. Actuator. B Chem. 188
(2013) 702e708.
[2] Y. Zhou, X. Li, Y. Wang, H. Tai, Y. Guo, UV illumination-enhanced molecular
ammonia detection based on a ternary-reduced graphene oxide-titanium
dioxide-Au composite film at room temperature, Anal. Chem. 91 (2019)
3311e3318.
[3] Y. Chen, W. Zhang, Q. Wu, A highly sensitive room-temperature sensing
material for NH3: SnO2-nanorods coupled by rGO, Sensor. Actuator. B Chem.
242 (2017) 1216e1226.
[4] P.G. Su, L.Y. Yang, NH3 gas sensor based on Pd/SnO2/RGO ternary composite
operated at room-temperature, Sensor. Actuator.
202e208.
B Chem. 223 (2016)
[5] N. Du, H. Zhang, B.D. Chen, X.Y. Ma, Z.H. Liu, J.B. Wu, D.R. Yang, Porous Indium
oxide nanotubes: layer-by-layer assembly on carbon-nanotube templates and
application for room-temperature NH3 gas sensors, Adv. Mater. 19 (2007)
1641e1645.
[6] S.C. Hernandez, D. Chaudhuri, W. Chen, N.V. Myung, A. mulchandani, Single
polypyrrole nanowire ammonia gas sensor, Electroanalysis 19 (2007)
2125e2130.
[7] J. Gong, Y. Li, Z. Hu, Z. Zhou, Y. Deng, Ultrasensitive NH3 gas sensor from
polyaniline nanograin enchased TiO2 fibers, J. Phys. Chem. C 114 (2010)
9970e9974.
[8] S. Virji, J. Huang, R.B. Kaner, B.H. Weiller, Polyaniline nanofiber gas sensors:
examination of response mechanisms, Nano Lett. 4 (2004) 491e496.
[9] Z. Li, H. Li, Z. Wu, M. Wang, J. Luo, H. Torun, P. Hu, C. Yang, M. Grundmann,
X. Liu, Y. Fu, Advances in designs and mechanisms of semiconducting metal
oxide nanostructures for high-precision gas sensors operated at room tem-
perature, Mater. Horiz. 6 (2019) 470e506.
[10] X. Wang, Y. Cui, T. Li, M. Lei, J. Li, Z. Wei, Recent advances in the functional 2D
photonic and optoelectronic devices, Adv. Opt. Mater. 7 (2018) 1801274.
[11] D. Maity, R.T.R. Kumar, Polyaniline anchored MWCNTs on fabric for high
performance wearable ammonia sensor, ACS Sens. 3 (2018) 1822e1830.
[12] I. Turkevych, S. Kazaoui, N.A. Belich, A.Y. Grishko, S.A. Fateev, A.A. Petrov,
T. Urano, S. Aramaki, S. Kosar, M. Kondo, E.A. Goodilin, M. Graetzel,
A.B. Tarasov, Strategic advantages of reactive polyiodide melts for scalable
perovskite photovoltaics, Nat. Nanotechnol. 14 (2019) 57e63.
[13] L. Zhao, K.M. Lee, K. Roh, S.U.Z. Khan, B.P. Rand, Improved outcoupling effi-
ciency and stability of perovskite light-emitting diodes using thin emitting
layers, Adv. Mater. 31 (2019), 1805836.
[14] B. Yang, Y. Wang, T. Wei, Y. Pan, E. Zhou, Z. Yuan, Y. Han, M. Li, X. Ling, L. Yin,
X. Xie, L. Huang, Solution-processable near-infrared-responsive composite of
perovskite nanowires and photon-upconversion nanoparticles, Adv. Funct.
Mater. 28 (2018) 1801782.
[15] G. Xing, N. Mathews, S.S. Lim, N. Yantara, X. Liu, D. Sabba, M. Gratzel,
S. Mhaisalkar, T.C. Sum, Low-temperature solution-processed wavelength-
tunable perovskites for lasing, Nat. Mater. 13 (2014) 476e480.
[16] Y. Zhuang, W. Yuan, L. Qian, S. Chen, G. Shi, High-performance gas sensors
based on a thiocyanate ion-doped organometal halide perovskite, Phys. Chem.
Chem. Phys. 19 (2017) 12876e12881.
4. Conclusions
In summary, a novel NH3 sensor has been developed by using the
TBA-capped MAPbBr3 film on the GeO2 substrate. The TBA ligand
could passivate the surface defects and grain boundariesofperovskite
in which leading to the strong PL emission. For the MAPbBr3-TBA-
based gas sensor, upon exposure to 100 ppm NH3 gas, the PL intensity
rapidly decreases by 62.5% with short response time (61 s) and re-
covery time (65 s). Furthermore, this as-prepared sensor possesses an
excellent linear relationship between the PL intensity and concen-
tration of NH3 in the range of 0e100 ppm with the detection limit of
0.46 ppm. According to the XRD, FT-IR, DTA-TG and PL lifetime
measurements, the NH3 sensing mechanism is speculated that the
NH3 molecules would permeate the capped TBA ligand and then
induce the decomposition and structure transformation of MAPbBr3.
On the basis of these considerations, the MAPbBr3-TBA-based
perovskite sensor could pave the way for practical NH3 gas detection.
Declaration of competing interest
The authors declare that they have no known competing
financial interests or personal relationships that could have
appeared to influence the work reported in this paper.
[17] H.F. Fang, S. Adjokatse, H.T. Wei, J. Yang, G.R. Blake, J.S. Huang, J. Even,
M.A. Loi, Ultrahigh sensitivity of methylammonium lead tribromide perov-
skite single crystals to environmental gases, Sci. Adv. 2 (2016), e1600534.
[18] G. Kakavelakis, B. Ghosh, Solution processed CH3NH3PbI3-xClx perovskite
based self-powered ozone sensing element operated at room temperature,
ACS Sens. 3 (2018) 135e142.
[19] A. Maity, B. Ghosh, Fast response paper based visual color change gas sensor
for efficient ammonia detection at room temperature, Sci. Rep. 8 (2018)
16851.
[20] N. Gupta, O. Nanda, R. Grover, K. Saxena, A new inorganic-organic hybrid
halide perovskite thin film based ammonia sensor, Org. Electron. 58 (2018)
202e206.
[21] S. Ruan, J. Lu, N. Pai, H. Ebendorff-Heidepriem, Y.B. Cheng, Y. Ruan,
C.R. McNeill, An optical fibre-based sensor for the detection of gaseous
ammonia with methylammonium lead halide perovskite, J. Mater. Chem. C 6
(2018) 6988e6995.
CRediT authorship contribution statement
Guishun Li: Conceptualization, Investigation, Methodology,
Writing - original draft. Wenqian Zhang: Validation, Data curation.
Changkun She: Formal analysis, Visualization. Shicheng Jia:
Formal analysis, Visualization. Shaohua Liu: Project administra-
tion, Resources. Fangyu Yue: Supervision, Resources. Chengbin
Jing: Funding acquisition, Supervision, Writing - review & editing.
Ya Cheng: Supervision, Resources. Junhao Chu: Resources, Writing
- review & editing.
[22] Y. Zhao, K. Zhu, Optical bleaching of perovskite (CH3NH3)PbI3 through room-
temperature phase transformation induced by ammonia, Chem. Commun. 50

8
G. Li et al. / Journal of Alloys and Compounds 835 (2020) 155386
(2014) 1605e1607.
[36] H. Li, S. Yang, S. Gong, J. Wu, S. Pan, Z. Chen, Q. Zhao, C. Shou, Q. Shen,
Perovskite films with a sacrificial cation for solar cells with enhanced stability
based on carbon electrodes, J. Alloys Compd. 797 (2019) 811e819.
[37] F. Zhang, C. Chen, S.V. Kershaw, C. Xiao, J. Han, B. Zou, H. Zhong, Ligand-
controlled formation and photoluminescence properties of CH3NH3PbBr3
nanocubes and nanowires, ChemNanoMat 3 (2017) 303e310.
[38] P. Kour, M. Chenna Reddy, R. Naphade, S. Ogale, Quaternary alkylammonium
salt incorporated 2D/3D mixed halide perovskite with highly enhanced
photoluminescence and arrested iodide/bromide phase segregation, Apl.
Mater. 6 (2018), 086107.
[23] C. Bao, J. Yang, W. Zhu, X. Zhou, H. Gao, F. Li, G. Fu, T. Yu, Z.A. Zou, A resistance
change effect in perovskite CH3NH3PbI3 films induced by ammonia, Chem.
Commun. 51 (2015) 15426e15429.
[24] S. Sasmal, A. Sinha, B. Donnadieu, R.G.S. Pala, S. Sivakumar, S. Valiyaveettil,
Volatility and chain length interplay of primary amines: mechanistic inves-
tigation on the stability and reversibility of ammonia-responsive hybrid pe-
rovskites, ACS Appl. Mater. Interfaces 10 (2018) 6711e6718.
[25] X. You, J. Wu, Y. Chi, Superhydrophobic silica aerogels encapsulated fluores-
cent perovskite quantum dots for reversible sensing of SO2 in a 3D-printed gas
cell, Anal. Chem. 91 (2019) 5058e5066.
[39] Z. Xiao, R.A. Kerner, L.T. Zhao, N.L. Ran, K.M. Lee, T.W. Koh, G.D. Scholes,
B.P. Rand, Efficient perovskite light-emitting diodes featuring nanometre-
sized crystallites, Nat. Photon. 11 (2017) 108e115.
[26] L. Zhou, J.F. Liao, Z.G. Huang, J.H. Wei, X.D. Wang, W.G. Li, H.Y. Chen,
D.B. Kuang, C.Y. Su, A highly red-emissive lead-free indium-based perovskite
single crystal for sensitive water detection, Angew. Chem. Int. Ed. 58 (2019)
5277e5281.
ꢀ
ꢀ
[40] L.C. Schmidt, A. Pertegas, S. Gonzalez-Carrero, O. Malinkiewicz, S. Agouram,
ꢀ
G.M. Espallargas, H.J. Bolink, R.E. Galian, J. Perez-Prieto, Nontemplate syn-
[27] J.W. Choi, N. Cho, H.C. Woo, B.M. Oh, J. Almutlaq, O.M. Bakr, S.H. Kim, C.L. Lee,
J.H. Kim, Investigation of high contrast and reversible luminescence thermo-
chromism of the quantum confined Cs4PbBr6 perovskite solid, Nanoscale 11
(2019) 5754e5759.
thesis of CH3NH3PbBr3 perovskite nanoparticles, J. Am. Chem. Soc. 136 (2014)
850e853.
[41] A.Y. Nazzal, L. Qu, X. Peng, M. Xiao, Photoactivated CdSe nanocrystals as
nanosensors for gases, Nano Lett. 3 (2003) 819e822.
[28] M.Y. Zhu, L.X. Zhang, J. Yin, J.J. Chen, L.J. Bie, B.D. Fahlman, Physisorption
induced p-xylene gas-sensing performance of (C4H9NH3)2PbI4 layered
perovskite, Sensor. Actuator. B Chem. 282 (2019) 659e664.
[42] W. Huang, J.S. Manser, S. Sadhu, P.V. Kamat, S. Ptasinska, Direct observation of
reversible transformation of CH3NH3PbI3 and NH4PbI3 induced by polar
gaseous molecules, J. Phys. Chem. Lett. 7 (2016) 5068e5073.
[29] J.H. Kim, S.H. Kim, Sub-second pyridine gas detection using a organometal
halide perovskite functional dye, Dyes Pigments 134 (2016) 198e202.
[30] Y. Liu, X. Tang, T. Zhu, M. Deng, I.P. Ikechukwu, W. Huang, G. Yin, Y. Bai, D. Qu,
X. Huang, F. Qiu, All-inorganic CsPbBr3 perovskite quantum dots as a photo-
luminescent probe for ultrasensitive Cu2þ detection, J. Mater. Chem. C 6
(2018) 4793e4799.
[31] I. Poli, S. Eslava, P. Cameron, Tetrabutylammonium cations for moisture-
resistant and semitransparent perovskite solar cells, J. Mater. Chem. A 5
(2017) 22325e22333.
[43] L. Zhang, F. Yu, Q. Li, J. Su, J. Li, M. Li, Understanding interactions between
halide perovskite surfaces and atmospheric/VOC gas molecules: an ab initio
investigation, J. Phys. D Appl. Phys. 51 (2018) 315302.
[44] S.K. Sharma, C. Phadnis, T.K. Das, A. Kumar, B. Kavaipatti, A. Yella, Reversible
dimensionality tuning of hybrid perovskites with humidity visualization and
application to stable solar cells, Chem. Mater. 31 (2019) 3111e3117.
[45] H. Chen, M. Zhang, R. Bo, C. Barugkin, J. Zheng, Q. Ma, S. Huang, A.W.Y. Ho-
Baillie, K.R. Catchpole, A. Tricoli, Superior self-powered room-temperature
chemical sensing with light-activated inorganic halides perovskites, Small 14
(2018) 1702571.
[32] C. Jing, J. Hou, Y. Zhang, Fabrication of thick (>10 mm) GeO2 ceramic films by a
facile liquid phase deposition process, J. Am. Ceram. Soc. 90 (2007)
3646e3650.
[46] Y. Zong, Y. Zhou, M. Ju, H.F. Garces, A.R. Krause, F. Ji, G. Cui, X.C. Zeng,
N.P. Padture, S. Pang, Thin-film transformation of NH4PbI3 to CH3NH3PbI3
[33] Y. Liu, Q. Xu, S. Chang, Z. Lv, S. Huang, F. Jiang, X. Zhang, G. Yang, X. Tong,
S. Hao, Y. Ren, Brightly luminescent and color-tunable green-violet-emitting
halide perovskite CH3NH3PbBr3 colloidal quantum dots: an alternative to
lighting and display technology, Phys. Chem. Chem. Phys. 20 (2018)
19950e19957.
[34] L.N. Quan, M. Yuan, R. Comin, O. Voznyy, E.M. Beauregard, S. Hoogland,
A. Buin, A.R. Kirmani, K. Zhao, A. Amassian, D.H. Kim, E.H. Sargent, Ligand-
stabilized reduced-dimensionality perovskites, J. Am. Chem. Soc. 138 (2016)
2649e2655.
perovskite: a methylamine-induced conversion-healing process, Angew.
Chem. Int. Ed. 55 (2016) 14723e14727.
[47] L. Dimesso, M. Dimamay, M. Hamburger, W. Jaegermann, Properties of
CH3NH3PbX3(X¼I, Br, Cl) powders as precursors for organic/inorganic solar
cells, Chem. Mater. 26 (2014) 6762e6770.
ꢀ
[48] S. Gonzalez-Carrero, R.E. Galian, J. Perez-Prieto, Maximizing the emissive
properties of CH3NH3PbBr3 perovskite nanoparticles, J. Mater. Chem. A 3
(2015) 9187e9193.
[49] D.L. Jacobs, L. Zang, Thermally induced recrystallization of MAPbI3 perovskite
under methylamine atmosphere: an approach to fabricating large uniform
crystalline grains, Chem. Commun. 52 (2016) 10743e10746.
[35] L. Song, X. Guo, Y. Hu, Y. Lv, J. Lin, Y. Fan, N. Zhang, X. Liu, Improved perfor-
mance of CsPbBr3 perovskite light-emitting devices by both boundary and
interface defects passivation, Nanoscale 10 (2018) 18315e18322.