Synthesis, physico-chemical characterization and structure of the elusive hydroxylammonium lead iodide perovskite NH3OHPbI3

Article abstract of DOI:10.1039/c9dt00690g

The synthesis of hydroxylammonium lead iodide NH₃OHPbI₃ was accomplished by means of the reaction between water solutions of HI and NH₂OH with PbI₂ in sulfolane in conjunction with either crystallization by CH₂Cl₂ vapor diffusion or sulfolane extraction with toluene. The appropriate choice of the solvent was found to be crucial in order to attain the desired material. The synthesized compound was extensively characterized by single crystal and powder X-ray diffraction, UV-Vis diffuse reflectance spectroscopy, FT-IR spectroscopy, ¹H-NMR spectroscopy, TG-DTA-QMS EGA (Evolved Gas Analysis), ESI-MS, and CHNS analysis. NH₃OHPbI₃ is an extremely reactive, deliquescent solid that easily oxidizes in air releasing iodine. Furthermore, it is the first reported perovskite to melt (m.p. around 80 °C) before decomposing exothermally at 103 °C. Such a chemical behavior, together with its optical absorption properties (i.e. yellow-colored perovskite), renders this material totally unsuitable for photovoltaic applications. The deliquescence of the material is to be ascribed to the strong hydrophilicity of hydroxylammonium ion. On the other hand, the relatively high Br?nsted acidity of hydroxylammonium (pK_a = 5.97) compared to other ammonium cations, promotes the reduction of atmospheric oxygen to water and the NH₃OHPbI₃ oxidation. The crystal structure, determined by single crystal X-ray diffraction with synchrotron radiation, is orthorhombic, but differs from the prototypal perovskite structure. Indeed it comprises infinite chains of face-sharing PbI₆ octahedra along the c-axis direction with hydroxylammonium cations positioned between the columns, forming layers on the ac plane. The solvent intercalates easily between the layers. The crystal structure is apparently anomalous considering that the expected Goldschmidt's tolerance factor for the system (0.909) lies in the range of a stable prototypal perovskite structure. Therefore, the strong hydrogen bond forming tendency of hydroxylamine is likely to account for the apparent structural anomaly.

Full text of DOI:10.1039/c9dt00690g

View Article Online
View Journal
Dalton
Transactions
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: A. D'Annibale, R.
Panetta, O. Tarquini, M. Colapietro, S. Quaranta, A. Cassetta, L. Barba, G. Chita and A. Latini, Dalton
Trans., 2019, DOI: 10.1039/C9DT00690G.
Volume 45 Number
1
7
January 2016 Pages 1–398
This is an Accepted Manuscript, which has been through the
Royal Society of Chemistry peer review process and has been
accepted for publication.
Dalton
Transactions
An international journal of inorganic chemistry
www.rsc.org/dalton
Accepted Manuscripts are published online shortly after
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
author guidelines.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the ethical guidelines, outlined
in our author and reviewer resource centre, still apply. In no
event shall the Royal Society of Chemistry be held responsible
for any errors or omissions in this Accepted Manuscript or any
consequences arising from the use of any information it contains.
ISSN 1477-9226
PAPER
Joseph T. Hupp, Omar K. Farha et al.
Efficient extraction of sulfate from water using
framework
a Zr-metal–organic
rsc.li/dalton

Page 1 of 27
Dalton Transactions
View Article Online
Synthesis, Physico-Chemical Characterization and Struct
D
u
OI
r
: 1
e
0.10
o
39
f
/C9D
t
T
h
00690G
e
Elusive Hydroxylammonium Lead Iodide Perovskite NH OHPbI3
3
1
1
2
2
Andrea D’Annibale , Riccardo Panetta , Ombretta Tarquini , Marcello Colapietro ,
3
4
4
4
Simone Quaranta , Alberto Cassetta , Luisa Barba , Giuseppe Chita , Alessandro
Latini1*
1
Dipartimento di Chimica, Università degli Studi di Roma “La Sapienza”, Piazzale Aldo Moro 5,
0
0185 Roma, Italy
2
Consiglio Nazionale delle Ricerche - Istituto di Cristallografia, Via Salaria km 29,300, 00015
Monterotondo Scalo, Roma, Italy
3
Dipartimento di Ingegneria dell’Informazione, Elettronica e Telecomunicazioni, Università degli
Studi di Roma “La Sapienza”, Via Eudossiana, 18, 00184 Roma, Italy
4
Consiglio Nazionale delle Ricerche - Istituto di Cristallografia, Sede Secondaria di Trieste, Area
Science Park – Basovizza, Strada Statale 14, km 163.5, 34149 Trieste, Italy
*
corresponding author. Email: alessandro.latini@uniroma1.it
ABSTRACT: The synthesis of hydroxylammonium lead iodide NH OHPbI was accomplished by
3
3
means of the reaction between water solutions of HI and NH OH with PbI in sulfolane in conjunction
2
2
with either crystallization by CH Cl vapor diffusion or sulfolane extraction with toluene. The
2
2
appropriate choice of the solvent was found to be crucial in order to attain the desired material. The
synthesized compound was extensively characterized by single crystal and powder X-ray diffraction,
1
UV-Vis diffuse reflectance spectroscopy, FT-IR spectroscopy, H-NMR spectroscopy, TG-DTA-
QMS EGA (Evolved Gas Analysis), ESI-MS, and CHNS analysis. NH OHPbI is an extremely
3
3
reactive, deliquescent solid that easily oxidizes in air releasing iodine. Furthermore, it is the first
reported perovskite to melt (m.p. around 80 °C) before decomposing exothermally at 103 °C. Such
a chemical behavior, together with its optical absorption properties (i.e. yellow-colored perovskite),
renders this material totally unsuitable for photovoltaic applications. The deliquescence of the
material is to be ascribed to the strong hydrophilicity of hydroxylammonium ion. On the other hand,
the relatively high Brønsted acidity of hydroxylammonium (pKa=5.97) compared to other ammonium
cations, promotes the reduction of atmospheric oxygen to water and the NH OHPbI oxidation. The
3
3
crystal structure, determined by single crystal X-ray diffraction with synchrotron radiation, is
orthorhombic, but differs from the prototypal perovskite structure. Indeed it comprises infinite chains
of face-sharing PbI octahedra along the c-axis direction with hydroxylammonium cations positioned
6

Dalton Transactions
Page 2 of 27
View Article Online
between the columns, forming layers on the ac plane. The solvent intercalates easily between the
DOI: 10.1039/C9DT00690G
layers. The crystal structure is apparently anomalous considering that the expected Goldschmidt’s
tolerance factor for the system (0.909) lies in the range of a stable prototypal perovskite structure.
Therefore, the strong hydrogen bond forming tendency of hydroxylamine is likely to account for the
apparent structural anomaly.

Page 3 of 27
Dalton Transactions
View Article Online
DOI: 10.1039/C9DT00690G
1
. Introduction
Methylammonium lead iodide (CH NH PbI ) has been emerging as one of the most promising
3
3
3
photovoltaic materials. Thereby, an intense research activity has been promoted worldwide over the
last ten years in order to maximize the performances of perovskite-based solar cells and improve their
stability. Indeed, CH NH PbI has already proven to be susceptible to thermal and chemical
3
3
3
1
degradation . Among the countless (organo)ammonium cations forming APbI (A=ammonium
3
cation) lead iodides, only a few possessing proper dimensions are able to form perovskitic structures.
Conversely, too large or too small cations lead to the so-called “yellow” perovskites. Such compounds
are characterized by crystal structures unrelated to the prototypal one and, consequently, do not show
2
optical properties amenable to solar cells applications . Hence, plenty of research, mostly predicated
on the Goldschmidt’s tolerance factor rule, has been devoted to the synthesis of stable perovskitic
structures for many ammonium cation/metal cation/anion combinations3. Since the
hydroxylammonium ion possesses an estimated ionic radius almost identical to the methylammonium
one, it is an attractive ammonium cation for perovskite systems based on the divalent lead
3
4
(
cation)/iodine (anion) pair . Theoretical calculations also suggest that NH OHPbI may exist in a
3 3
stable perovskitic phase. Thus, hydroxylammonium lead iodide could possibly manifest optical and
electronic properties similar to the CH NH PbI ones. Therefore, investigations on the synthesis and
3
3
3
+
characterization of lead-iodine-ammonium perovskites relying on the NH OH cation are to be
3
considered highly desirable.
A paper reporting the synthesis of hydrazinium lead iodides together with the unsuccessful synthesis
5
+
of hydroxylammonium lead iodide has been recently published . Needless to say, even N H has a
2
5
calculated ionic radius nearly identical to that of methylammonium. Hence, hydrazinium lead iodides
with two different stoichiometries (i.e. (N H ) Pb I and N H PbI ) have been prepared. Although
2
5 15
3 21
2
5
3
the latter compound matches the stoichiometric composition of a perovskite, neither of its two
experimentally found polymorphs has turned out to own the correct structure. In fact, both
polymorphs do not comply with the tolerance factor rule probably because of the substantial
contribution to the lattice energy brought about by the hydrogen bond. Thus, the formation of crystal
structures other than the perovskite prototype is most likely favored by the presence of hydrogen
bonds. In addition, none of the authors’ attempts to synthesize the hydroxylammonium lead iodide
by reacting hydroxylammonium iodide with lead iodide in 2-propanol and in dimethylformamide
(DMF) has succeed. The ineffectiveness of their synthetic approaches stems from multiple causes.
First, PbI is not soluble in 2-propanol. Second, hydroxylamine reacts with DMF producing
2
dimethylamine and formohydroxamic acid. Indeed, amides (as other carboxylic and carbonyl
6
,7
8
compounds) are reactive towards hydroxylamine . Nonetheless, a 2017 paper reports the successful

Dalton Transactions
Page 4 of 27
View Article Online
DOI: 10.1039/C9DT00690G
synthesis of the NH OHPbI Cl mixed halide hydroxylammonium perovskite. However, the results
3
2
presented are rather doubtful for various reasons. The most important issues are the reaction solvent
used (DMF) and the questionable results concerning the thermogravimetric analysis, in which the
exothermic decomposition of the hydroxylammonium moiety is not visible. Furthermore, the mass
loss is not compatible with the claimed composition.
A simple approach to the synthesis of hydroxylammonium lead iodide, based on the method adopted
9
by Poglitsch and Weber for the preparation of methylammonium lead iodide, was initially used in
the present research. Such a procedure had been already successfully employed by the authors for the
synthesis of different perovskites10,11,12. Briefly, a 50% aqueous solution of hydroxylamine was added
to a warm (50-60 °C) solution of lead (II) acetate in an excess of 57% aqueous HI. No precipitate
formed also when a large excess of NH OH was added to the solution. Conversely, a yellow solid of
2
very complex and still unidentified composition did precipitate by cooling the solution with an ice-
water bath. Dissolution of this solid in DMF (1g of solid in 2 ml of solvent) and subsequent exposition
of the solution to CH Cl vapor in a closed container resulted in the formation of two distinct solid
2
2
phases. Specifically, a white, unstable powder and yellow crystals were collected at the bottom of the
container. The former might probably be a mixture of lead formohydroxamate and other salts, while
the latter was found out to be dimethylammonium lead iodide NH (CH ) PbI by X-ray powder
2
3 2
3
1
3
diffraction and CHNS analysis. Such results verified the DMF decomposition by hydroxylamine.
On the other hand, being the dimethylammonium lead iodide very soluble in DMF, only the white
powder precipitation occurred when the CH Cl exposition step was skipped. After ruling out both
2
2
the Poglitsch and Weber method and the use of DMF, a selection of one or more solvents suitable for
a direct reaction between NH OHI and PbI (i.e. in order to avoid the production of interfering
3
2
species) was undertaken. The appropriate solvent was to be chosen among the highly dipolar aprotic
ones for ensuring dissolution of the ionic compounds formed during the reaction. Besides, other two
requirements needed to be fulfilled by the solvent:


being chemically inert with respect to hydroxylamine;
showing little or no tendency to form thermally resistant adducts with PbI Indeed, any heat
2
.
°
treatment exceeding 100 C would jeopardize the hydroxylamine chemical stability (i.e.
°
NH OH decomposes at about 100 C).
2
Such restrictions excluded all the common solvents such as the amides, gamma-butyrolactone, N-
methylpyrrolidone (all these solvents are attacked by hydroxylamine), and dimethylsulfoxide, that
o
forms stable, and therefore non-removable (i.e. for temperatures lower than 100 C) adducts with
1
4
PbI . The only stable and commercially available solvents satisfying the above requirements are
2
either sulfolane or N-alkylated cyclic ureas like N,N′-dimethylpropyleneurea (DMPU) and 1,3-

Page 5 of 27
Dalton Transactions
View Article Online
dimethyl-2-imidazolidinone (DMI). Sulfolane and DMI were consequently picked out as solvents for
DOI: 10.1039/C9DT00690G
1
5,16
the NH OHPbI synthesis. DMI was chosen because of its lower donor number
(27.7) compared
3
3
to DMPU (33). Judging by DMF and DMSO donor numbers (27 and 29.8, respectively) and chemical
structures, DMI was considered to be less prone to form stable adducts with PbI than DMPU. Pure
2
PbI is insoluble in sulfolane (even hot) due to the low donor number of the solvent (14.8). Both
2
sulfolane and DMI are highly polar (i.e. dipole moments equal to 4.69 and 4.09 debye, respectively),
chemically and thermally very stable, and possess high values of the relative dielectric constant (43.4
and 37.6 respectively)17,18
.
2
. Experimental Section
2
.1 Materials preparation. Lead (II) nitrate Pb(NO ) (99%) was purchased from Alfa Aesar.
3 2
Hydriodic acid HI 57% aqueous solution (<1.5% hypophosphorous acid as stabilizer), hydroxylamine
NH OH (50% solution in water), 1,3-dimethyl-2-imidazolidinone (≥99.5%, over molecular sieves)
2
and sulfolane (99%) were purchased from Sigma Aldrich. Toluene (99.8%) was purchased from
VWR. Absolute ethanol and dichloromethane (both ≥ 99.9%) were purchased from Carlo Erba
Reagents. The synthesis of NH OHPbI was performed in both sulfolane and DMI by reacting freshly
3
3
prepared NH OHI with an equimolar amount of PbI powder. Lead iodide had been previously
3
2
prepared by reacting aqueous lead nitrate with aqueous hydriodic acid. In a typical synthesis, 286 μl
of 57% aqueous HI and 133 μl of 50% aqueous NH OH were added under stirring to 2 ml of absolute
2
ethanol in a 50 ml two necked round bottom flask at room temperature and under nitrogen
atmosphere. Immediately the solution became bright yellow, then colorless. After stirring for 2-3
minutes the ethanol was removed under vacuum at 55 °C until the liquid started turning from colorless
to light yellow. Such a step was carried out in order to prevent NH OHI from decomposing evolving
3
iodine. Then, nitrogen was let into the flask and 4 ml of solvent were introduced with flowing nitrogen
under stirring, followed by the addition of 1 g of PbI . The suspension was stirred under nitrogen until
2
PbI completely dissolved. The solution was filtered through a 0.45 µm PTFE syringe filter in a new
2
1
2 ml glass vial. The vial was inserted in a 100 ml Pyrex bottle that had been half-filled with
dichloromethane and closed in an argon atmosphere. The solution was let nucleate overnight and
bright yellow crystals begun appearing the following morning. Crystals stemming from the synthesis
in sulfolane were needle-shaped (i.e. acicular), while the ones obtained through the DMI route were
polyhedral. Crystal formation and growth took about a week to complete. Then, the vial was removed
from the bottle and the crystals were filtered. According to the results attained by CHNS and DTA-
TG analyses the crystals for the DMI case turned out to be made of an adduct between DMI and PbI2.
Specifically, the C:H:N ratio and the thermal decomposition pattern (i.e. endothermal decomposition)
agree perfectly with the DMI-PbI adduct composition. In fact, the decomposition of the
2

Dalton Transactions
Page 6 of 27
View Article Online
hydroxylamine moiety is associated to an exothermal DTA signal. Hence, DMI was proved to be
DOI: 10.1039/C9DT00690G
unsuitable for the hydroxylammonium lead iodide synthesis. Therefore, the results presented in the
paper concern only the synthesis in sulfolane, which actually resulted into NH OHPbI crystals.
3
3
Moreover, NH OHPbI powder could be collected from the solution by extracting the sulfolane with
3
3
toluene. After the dissolution of PbI in the NH OHI/sulfolane solution, the liquid is transferred to a
2
3
1
00 ml glass centrifuge tube. Then, 50 ml of toluene are added, the mixture is stirred and then
centrifuged at 2000 rpm for 2 minutes, the supernatant liquid was discarded and the whole process is
repeated. After the second washing, a yellow solid forms and it is washed and centrifuged three times
with 50 ml of clean toluene each time as above.
2
.2 Single crystal X-ray diffraction. Single crystal X-ray diffraction measurements were performed
at the XRD1 beamline of the Elettra synchrotron radiation facility (Trieste, Italy)19,20. Crystals were
immersed in paraffin oil in order to avoid their reaction with air. A suitable crystal 0.07 × 0.05 × 0.05
3
mm was selected with a loop and mounted on a goniometer head. A X-ray wavelength of 0.7 Å and
a silicon diode array area detector (Pilatus 2 M; Dectris) were used for diffraction measurements. The
crystal was kept at a steady T = 100 K during data collection to prevent its degradation. Unit cell
2
1
assignment, integration and data reduction were performed with the XDS program . The structure
2
2
was solved with the Superflip structure solution software and by using Olex2 as the graphical
2
3
24
interface . The model was refined with version 2018/3 of ShelXL using least squares minimisation .
Sulfolane disorder was modeled in order to locate the hydroxylammonium ions. The electron density
2
5
corresponding to their presence was located using PLATON/SQUEEZE .
.3 Powder X-ray diffraction. X-ray diffraction patterns were acquired using a Panalytical X'Pert
2
Pro MPD diffractometer (Cu Kα radiation, λ=1.54184 Å). The diffractometer is fitted with a
X'Celerator RTMS detector. The angular resolution in 2θ is 0.001°; 0.04 rad soller slit, 1/4°
divergence slit, 15 mm mask on the incident beam path, 5.0 mm anti-scatter slit, Ni Kβ filter, 0.04
rad collimator on the diffracted beam path were used for the measurements. The scans were performed
in the 7.5-90° in 2θ angular range. Rietveld refinement of the pattern of the powder sample was
2
6
performed by means of the MAUD software package .
.4 Thermogravimetry-Differential Thermal Analysis (TG-DTA)-Quadrupole Mass
2
Spectrometry Evolved Gas Analysis (QMS-EGA). TG-DTA analysis was performed with a
Netzsch STA 409 PC Luxx thermal analyzer. The evolved gas analysis was performed by coupling
the thermal analyzer with a Balzers-Pfeiffer QMG 421 quadrupole mass spectrometer (m/z range 0-
1
28). The coupling was achieved by connecting the inlet of the mass spectrometer to the outlet of the
TG-DTA apparatus through a stainless steel capillary heated at 150 °C to prevent moisture
condensation. The measurements were performed in sintered alumina crucibles under flowing Ar

Page 7 of 27
Dalton Transactions
3
View Article Online
DOI: 10.1039/C9DT00690G
atmosphere (85 cm /min @ STP, purity ≥ 99.9995%) with a scan rate of 10 K/min in the range RT-
5
00 °C.
2
.5 UV-Vis spectroscopy. Reflectance UV-Vis spectrum was acquired with a Shimadzu (Japan)
UV2600 UV-Vis spectrophotometer equipped with a ISR-2600 Plus integrating sphere. BaSO4
powder was the reflectance reference. The type of the band gap (i.e. direct or indirect, allowed or
forbidden) was determined by comparing the absorption edge attained from the pseudoabsorbance
2
7
spectrum with the values obtained with Tauc's plots constructed for different exponents.
.6 FT-IR spectroscopy. A Bruker Alpha spectrophotometer was used to acquire the FT-IR spectra
2
-1
-1
in the 400-4000 cm wavenumber range (resolution 1 cm ) in DRIFT mode on samples dispersed in
dry KBr powder.
2
.7 ESI-MS. A Finnigan-Mat LCQ mass spectrometer was used for the ESI-MS analysis of the
powder sample. Since DMI is liquid at room temperature (while sulfolane melts at 27.5 °C) the
reaction product was dissolved in DMI (1 mg/ml) in order to perform the analysis. Only positive ions
were analyzed. The operating conditions were optimized for the analysis, i.e.: spray voltage 5.00 kV,
capillary temperature 275 °C, capillary voltage 6 V, tube lens offset -5.00 V. The spray voltage value
2
8
was chosen according to literature data .
.8 CHN analysis. The C, H, N and S contents of the single crystals sample was determined with a
EA 1110 CHNS-O instrument.
2
1
1
2
.9. H NMR. H NMR spectral analysis of the product synthesized in sulfolane was performed to
verify the presence of the hydroxylammonium cation. The spectrum was acquired on a sample
prepared dissolving single crystals of NH OHPbI in DMSO-d using a 400 MHz Bruker Avance III
3
3
6
spectrometer.
3
. Results and Discussion.
3
.1 General considerations. NH OHPbI as single crystals appears in the form of pale yellow
3 3
needles, while when prepared as powder it is a bright yellow solid. Both forms of NH OHPbI are
3
3
reactive materials, quickly darkening due to formation of elemental iodine by atmospheric oxygen
2
9
oxidation favored by the relatively high acidity of the hydroxylammonium ion (pKa=5.97) . The
presence of free iodine after the exposure to air was confirmed by washing the darkened sample with
hexane, which immediately became purple. It is also deliquescent as demonstrated by fig. 1, in which
the freshly prepared powder is shown on the left and on the right the same sample is shown after 3
days at ambient conditions, appearing as a dark sludge. Thus, samples could be preserved for
prolonged time only in sealed glass ampules under inert atmosphere. The reactivity and instability of
the material impeded any kind of laser techniques (Raman, PL) investigation. In fact, even exposure
of the sample to the lowest possible power density decomposed the material.

Dalton Transactions
Page 8 of 27
View Article Online
DOI: 10.1039/C9DT00690G
3
.2 Structure determination.
Crystal Data. C H I NO PbS, Mr = 774.18, orthorhombic, Pmma (No. 51), a = 7.920(6) Å, b =
6
20 3
3
3
9
1
.030(4) Å, c = 10.770(2) Å, α = β = γ = 90°, V = 770.25(1) Å , T = 100 K, Z = 2, Z' = 0.25, μ =
7.091, 6502 reflections measured, 1360 unique (R = 0.0392) which were used in all calculations.
int
The final wR was 0.1259 (all data) and R was 0.0394 (I > 2(I)).
2
1
The structure of NH OHPbI is reported in fig. 2 and the data are summarized in table 1. The unit cell
3
3
is orthorhombic (space group Pmma, no. 51) with lattice constants (at 100 K) a=7.920(6) Å,
b=9.030(4) Å and c=10.770(2) Å, containing 2 formula units. The density is low (pycnometric:
-3
-3
2
.9±0.5 g cm , calculated: 3.338 g cm ), due to the layered assembly of the hydroxylammonium-
lead iodide moieties on the ac planes. The distance between the layers is ~4.7 Å along the b axis.
Lead and iodine atoms are arranged in infinite chains of face sharing octahedra along the a direction.
As in other ammonium lead iodides the distances between the lead atom and the two iodine atoms of
2
12
the asymmetric unit are different because of the effect of the lead 6s lone pair , so the PbI octahedra
6
are distorted (bond lengths and angles are reported in the Supporting Information.
Hydroxylammonium ions are positioned between the Pb-I chains on the ac plane. The structural motif
of the hydroxylammonium ions is comprised of chains along the a direction (parallel to Pb-I chains)
formed by 8 positions per unit cell in which the 2 hydroxylammonium ions are randomly distributed.
The chain motif and its sinusoidal trend (fig. 3) are clearly indicative of strong and geometrically
optimized interactions between the individual units.
Sulfolane molecules are able to be accommodated between the layers because of the hydrogen bonds
formed with the hydroxylammonium ions. The maximum quantity of sulfolane molecules found in
the single crystals (1 per formula unit NH OHPbI ), corresponds to a sulfolane content of about 16
3
3
wt. %. Nonetheless, sulfolane can be removed from the structure quite easily. Indeed, after multiple
washings of the powder samples with toluene (which is unreactive towards NH OHPbI but functions
3
3
as a solvent for sulfolane), the quantity of sulfolane decreased to about 10 wt. %, as demonstrated by
thermal and CHNS analyses. Hence, a solvent with properties very close to those of sulfolane but
more easily removable, such as piperylene sulfone30 would probably allow to obtain pure
NH OHPbI . Unfortunately, piperylene sulfone is not commercially available yet and, therefore,
3
3
could not be tested.
Despite of evidence provided by recent and revised approaches, the NH OHPbI crystal structure did
3
3
3
not correspond to the perovskite-type one expected from the tolerance factor calculations . Using the
available data, in fact, NH OHPbI showed a calculated tolerance factor of 0.909, nearly identical to
3
3
that of CH NH PbI (0.912). Perovskite structures are expected for tolerance factors between 0.8 and
3
3
3
1
. As mentioned in the introduction, also hydrazinium lead iodide (tolerance factor 0.912) does not

Page 9 of 27
Dalton Transactions
View Article Online
possess a perovskite structure although it fulfills the requirement on the tolerance factor. The authors
DOI: 10.1039/C9DT00690G
reasonably assumed that the most probable reason is the tendency of hydrazinium ions to form
hydrogen bonds. Provided that such explanation stands valid for the hydrazinium, it should be even
more valid for the hydroxylammonium that contains the OH group. Hydroxylammonium tendency to
form hydrogen bonds is already known for modifying the crystal structures of compound compared
3
1
to its congeners .
.3 Powder X-ray diffraction. The diffraction pattern of the powder sample and of single crystals
3
grinded in agate mortar are shown in fig. 4, panels A and B, respectively. No relevant differences
were detected in the positions of the reflections, while the trend of the intensities of the peaks is
different. This phenomenon is evidently due to preferred orientation effects present in the grinded
single crystals. In order to minimize preferred orientation artifacts the Rietveld refinement procedure
was performed on the pattern obtained from the powder sample. The cif file generated from the single
crystal data was used for the refinement. The procedure was carried out to attain the unit cell
parameters at room temperature (298 K) by taking into account that the single crystal data had been
acquired at 100 K. The Le Bail fit (figure S1 of the Supporting Information) is in excellent agreement
with the experimental pattern. The peaks that were not fitted correspond to the most intense
reflections of PbI , which is evidently present as impurity. A comparison between the cell parameters
2
achieved at room temperature and at 100 K is reported in table 2. The b axis length is more affected
by the temperature increase (4% elongation) than the a and c axes (0.5% for both). Such a result can
be justified by the weakness of the interlayer interactions along the b axis. This speculation is verified
by the easy removal of the solvent intercalated between the layers (see 3.2).
3
.4 TG-DTA/QMS-EGA. TG-DTA patterns of the powder product after multiple washes with
toluene (i.e. to remove as much sulfolane as possible) and with hexane are displayed in fig. 5 panel
A. An endothermal event occurred at 80 °C. In order to assess the nature of the event (i.e. either
polymorphic transition or melting), another portion of the same sample was heated gently in a beaker
and began melting at about 80 °C. This result was quite surprising. Indeed, to the best of authors’
knowledge, no ammonium lead halide is reported to melt before decomposing. At 103 °C the material
started to decompose. The decomposition occurred in two stages. In fact, two exothermal events, with
two distinct weight loss trends, could be detected. This is more evident in the panels B and C of fig.
5
, in which the TG and DTA signals are compared with the derivative of the TG, respectively. The
exothermal decomposition ends at 227 °C and at 277 °C another weight loss process occurs, but this
is an endothermal one. By considering the composition of the sample, the exothermal events can be
safely attributed to the decomposition of the hydroxylammonium iodide (hydroxylammonium
compounds are well known to decompose exothermally32,33), whereas the endothermal event can be

Dalton Transactions
Page 10 of 27
View Article Online
ascribed to the loss of sulfolane. The latter attribution is further supported by the fact that the
DOI: 10.1039/C9DT00690G
corresponding weight loss started at 277 °C and the boiling point of sulfolane is 285 °C. Another
endothermal event with no weight loss occurred at 410 °C and it corresponds to the melting point of
lead iodide. The sulfolane loss accounts for 10.55 wt. %, while the hydroxylammonium iodide loss
is 22.75 wt. %. The ratio between the hydroxylammonium iodide content and the total mass minus
the sulfolane content is 25.4%, in excellent agreement with the NH OHPbI composition (25.9%
3
3
theoretical). During the loss of hydroxylammonium iodide, the composition of the evolved gases
transported by the argon purge flow was monitored by means of quadrupole mass spectrometry. The
maximum intensity of the ion currents produced by the evolved gases was recorded when the sample
temperature was around 150 °C. The difference between the mass spectrum of the bare carrier gas
and during the evolution of the gases with maximum ion current intensities is shown in fig. 6. Peaks
were found for m/z values 44, 32, 30, 28 and 14 and were attributed to N O (44), O (32), NO (30)
2
2
and N (28 and 14). N O, NO and N are the expected gaseous products of the thermal decomposition
2
2
2
3
4
of hydroxylamine while O was very probably produced in the ion source by the decomposition of
2
N O. Thus, TG-DTA results in conjunction with the QMS analysis of the evolved gases confirm both
2
the presence of hydroxylamine in the sample and the expected stoichiometry of NH OHPbI .
3
3
3
.5 UV-Vis spectroscopy. The pseudoabsorbance spectrum of the powder product is shown in fig. 7,
and the corresponding Tauc’s plot in fig. 8. Three absorption maxima are present, the first at 500 nm,
the second at 386 nm and the third at 293 nm in order of increasing intensity. The Tauc’s plot gives
the best agreement to the pseudoabsorbance spectrum for a direct allowed band gap transition
corresponding to the cutoff wavelength of the second peak, i.e. 442 nm that in turn corresponds to a
band gap of 2.8 eV. Specifically, the band gap energy obtained through the Tauc’s plot is 2.88±0.01
3
5
eV. Such a value is associated to a Shockley-Queisser energy conversion efficiency limit around 8-
9
%. These findings, combined with the high chemical reactivity, render NH OHPbI totally
3 3
unsuitable for photovoltaic applications.
3
.6 FT-IR spectroscopy. Because of the relatively high sulfolane content in the samples, a
comparison between the spectra of powder sample and sulfolane both dispersed in KBr was
mandatory in order to isolate the sulfolane absorptions. Both spectra are displayed in fig. 9. The
spectrum of the sample is clearly dominated by sulfolane absorptions, and in the portion extending
-1
from 400 to 1400 cm only sulfolane absorptions are visible. Overall, only three absorptions can be
ascribed to the hydroxylammonium ion. Needless to say, the vibrations related to the Pb-I part of the
structure are out of the explored energy range. The wavenumbers corresponding to these absorptions
-1
+
are 1510, 2700 and 3090 cm . They first can be assigned to deformation mode of -NH , the second
3

Page 11 of 27
Dalton Transactions
+
View Article O+nline
to combination between -NH rocking and deformation modes, and the third to the -OH and -NH
3
3
DOI: 10.1039/C9DT00690G
3
6
stretching vibrations .
.7 ESI-MS. The chemical identity of material was further verified by the positive ions ESI mass
3
spectrum (fig. 10). All the peaks belong to hydroxylamine adducts and decomposition products, with
the exception of the one with the highest value of m/z, which can be attributed to a mono-
demethylated fragment of the solvent.
3
.10 CHNS analysis. The analysis was performed on a sample of single crystals, giving the following
wt. % values: C=5.89%, H=1.61%, N=1.65% S=3.59%. The C/S weight ratio is in good agreement
with the expected value for sulfolane (calculated: 1.5, found 1.6). By using these values, an estimated
sulfolane content of 14 wt. % for the compound was found. Such a result corroborates the 16 wt. %
calculated from the crystal structure. The N content is also in good agreement with the NH OHPbI3
3
stoichiometry provided that the sulfolane content is taken into account (calculated: 2.2%, found:
+
1
.9%). Similar considerations apply to the expected H/N weight ratio for NH OH (calculated: 0.3,
3
found: 0.4).
1
3
.11 H NMR. The presence of hydroxylammonium moiety in the compound was once again
1
confirmed by the H-NMR spectrum collected from a sample of single crystals dissolved in deuterated
dimethylsulfoxide. Two broad signals at d 9.85 (1H) and d 10.05 (3H) were assigned to hydroxyl and
ammonium protons, respectively. The chemical shift and the shape of ammonium hydrogen signal
are agreement with literature spectral data of organic hydroxylammonium compounds in the same
3
7
solvent .
4
. Conclusion
Hydroxylammonium lead iodide NH OHPbI synthesis could be accomplished only after a careful
3
3
choice of the solvent and experimental conditions. The conventional approaches used for the
syntheses of many other ammonium lead halide perovskites (i.e. Poglitsch method, crystallization
from DMF solutions) turned out not to be effective in the case of the hydroxylammonium lead iodide
perovskite. On the other hand, the use of sulfolane as solvent for the synthesis allowed to achieve the
desired product. XRD single crystal structural analysis showed a layered, low density structure in
+
which the layers, lying on the ac plane, are comprised of NH OH chains interposed between chains
3
of face sharing PbI octahedra . Sulfolane molecules easily intercalate between the layers due to the
6
favorable interactions with hydroxylammonium ions. The existence of weak interlayer interactions
was confirmed by powder XRD that demonstrated the high sensitivity of the b axis length to
temperature variation compared to the a and c axes. The chemical identity of the compound was
checked and verified by different physico-chemical techniques. The compound clearly violates the

Dalton Transactions
Page 12 of 27
View Article Online
DOI: 10.1039/C9DT00690G
Goldschmidt's tolerance factor. Indeed, NH OHPbI does not possess the expected perovskite
3
3
structure. Furthermore, its instability, reactivity and inadequate band gap value preclude any possible
use in photovoltaic devices.
Acknowledgements
The authors wish to thank the Università degli Studi di Roma “La Sapienza” for financial support and
the Elettra-Sincrotrone Trieste S.C.p.A for making available beamtime at the XRD1 beamline.
The authors wish to thank the crystallography Reviewer for his/her truly valuable contribution to the
crystal structure refinement.
Associated Content
The CIF (crystallographic information file) of the determined crystal structure is available.
Supporting Information file is available.

Page 13 of 27
Dalton Transactions
View Article Online
DOI: 10.1039/C9DT00690G
References
1
A. Ciccioli; A. Latini. Thermodynamics and the Intrinsic Stability of Lead Halide Perovskites
CH NH PbX . J. Phys. Chem. Lett. 2018, 9, 3756−3765.
3
3
3
2
M Grätzel. The Rise of Highly Efficient and Stable Perovskite Solar Cells. Acc. Chem. Res. 2017,
0, 487−491.
5
3
G. Kieslich; S. Sun; A.K. Cheetham. An extended Tolerance Factor approach for organic–inorganic
perovskites. Chem. Sci. 2015, 6, 3430-3433.
4
M, Becker; M. Wark. Organic Cation Substitution in Hybrid Perovskite CH NH PbI with
3
3
3
+
Hydroxylammonium (NH OH ): A First-Principles Study. J. Phys. Chem. C 2018, 122, 3548-3557.
3
5
E.V. Campbell; B. Dick; A.L. Rheingold; C. Zhang; X. Liu; Z.V. Vardeny; J.S. Miller. Structures
5
-
4-
-
of a Complex Hydrazinium Lead Iodide, (N H ) Pb I , Possessing [Pb I ] [PbI ] , and I Ions and
2
5 15
3 21
2 9
6
α- and β-(N H )PbI . Chem. Eur. J. 2018, 24, 222-229.
2
5
3
6
W.P. Jencks; M. Gilchrist. The Reaction of Hydroxylamine with Amides. Kinetic Evidence for the
Existence of a Tetrahedral Addition Intermediate. J. Am. Chem. Soc. 1964, 86, 5616-5620.
7
D. Davidson. Hydroxamic acids in qualitative organic analysis. J. Chem. Educ. 1940, 17, 81-84.
8
M. Aamira; Z.H. Shahb; M. Shera; A. Iqbalb; N. Revaprasaduc; M.A. Malikd; J. Akhtare. Enhanced
photocatalytic activity of water stable hydroxyl ammonium lead halide perovskites. Mater. Sci.
Semicond. Process. 2017, 63, 6-11.
9
A. Poglitsch; D. Weber. Dynamic disorder in methylammoniumtrihalogenoplumbates (II) observed
by millimeter‐wave spectroscopy. J. Chem. Phys. 1987, 87, 6373-6378.
1
0
B. Brunetti; C. Cavallo; A. Ciccioli; G. Gigli; A. Latini. On the Thermal and Thermodynamic
In)Stability of Methylammonium Lead Halide Perovskites. Sci. Rep. 2016, 6, 31896.
(
1
1
A. Latini; G. Gigli; A. Ciccioli. A study on the nature of the thermal decomposition of
methylammonium lead iodide perovskite, CH NH PbI : an attempt to rationalise contradictory
3
3
3
experimental results. Sustainable Energy Fuels 2017, 1, 1351-1357.
1
2
R. Panetta; G. Righini; M. Colapietro; L. Barba; D. Tedeschi; A. Polimeni; A. Ciccioli; A. Latini.
Azetidinium Lead Iodide: Synthesis, Structural and Physico-Chemical Characterization. J. Mater.
Chem. A, 2018, 6, 10135-10148.
1
3
A. Mancini; P. Quadrelli; G. Amoroso; C. Milanese; M. Boiocchi; A. Sironi; M. Patrini; G.
Guizzetti; L. Malavasi. Synthesis, structural and optical characterization of APbX3
A=methylammonium, dimethylammonium, trimethylammonium; X=I, Br, Cl) hybrid organic-
inorganic materials. J. Solid State Chem. 2016, 240, 55-60.
(

Dalton Transactions
Page 14 of 27
1
4
View Article Online
DOI: 10.1039/C9DT00690G
N. Ahn; D. Son; I. Jang; S.M. Kang; M. Choi; N. Park. Highly Reproducible Perovskite Solar Cells
with Average Efficiency of 18.3% and Best Efficiency of 19.7% Fabricated via Lewis Base Adduct
of Lead(II) Iodide. J. Am. Chem. Soc. 2015, 137, 8696−8699.
1
5
C. Laurence; J.F. Gal. Lewis Basicity and Affinity Scales Data and Measurement, Wiley, 2010, p.
6.
7
1
6
F. Cataldo. A revision of the Gutmann donor numbers of a series of phosphoramides including
TEPA. Eur. Chem. Bull. 2015, 4, 92-97.
1
7
U. Tilstam. Sulfolane: A Versatile Dipolar Aprotic Solvent. Org. Process Res. Dev. 2012, 16, 1273-
1
278.
1
8
B. J. Barker; J. Rosenfarb; J.A. Caruso. Ureas as Solvents for Chemical Investigations. Angew.
Chem. Int. Ed. Engl. 1979, 18, 503-507.
1
9
A. Lausi; M. Polentarutti; S. Onesti; J.R. Plaisier; E. Busetto; G. Bais; L. Barba; A. Cassetta; G.
Campi; D. Lamba; A. Pifferi; S.C. Mande; D.D. Sarma; S.M. Sharma; G. Paolucci. Status of the
crystallography beamlines at Elettra. Eur. Phys. J. Plus 2015, 130, 43.
2
0
S. Bernstorff; E. Busetto; C. Gramaccioni; A. Lausi; L. Olivi; F. Zanini; A. Savoia; M. Colapietro;
G. Portalone; M. Camalli; A. Pifferi; R. Spagna; L. Barba; A. Cassetta. The macromolecular
crystallography beamline at ELETTRA. Rev. Sci. Instrum. 1995, 66, 1661-1664.
2
1
W. Kabsch. XDS. Acta Cryst. 2010, D66, 125-132.
2
2
L. Palatinus; G. Chapuis. SUPERFLIP - a computer program for the solution of crystal structures
by charge flipping in arbitrary dimensions. J. Appl. Cryst. 2007, 40, 786-790.
2
3
O.V. Dolomanov; L.J. Bourhis; R.J. Gildea; J.A.K. Howard; H. Puschmann. Olex2: A complete
structure solution, refinement and analysis program. J. Appl. Cryst. 2009, 42, 339-341.
2
2
2
4
5
6
G.M. Sheldrick. Crystal structure refinement with ShelXL, Acta Cryst. 2015, C71, 3-8.
A.L. Spek. Structure validation in chemical crystallography. Acta Cryst. 2009, D65, 148-155.
L. Lutterotti; D. Chateigner; S. Ferrari; J. Ricote. Texture, Residual Stress and Structural Analysis
of Thin Films using a Combined X-Ray Analysis. Thin Solid Films 2004, 450, 34-41.
2
7
J. Tauc; R. Grigorovici; A. Vancu. Optical properties and electronic structure of amorphous
germanium. Phys. Status Solidi 1966, 15, 627-637.
2
8
T. Kumar; M. Ramya; V. Srinivasan; N. Xavier. A Simple and Direct LC–MS Method for
Determination of Genotoxic Impurity Hydroxylamine in Pharmaceutical compounds. J. Chromatogr.
Sci. 2017, 1-7.
2
9
T. C. Bissot; R.W. Parry; D.H. Campbell. The Physical and Chemical Properties of the
Methylhydroxylamines. J. Am. Chem. Soc. 1957, 79, 796-800.

Page 15 of 27
Dalton Transactions
3
0
View Article Online
DOI: 10.1039/C9DT00690G
D. Vinci; M. Donaldson; J.P. Hallett; E.A. John; P. Pollet; C.A. Thomas; J.D. Grilly; P.G. Jessop;
C.L. Liotta; C.A. Eckert. Piperylene sulfone: a labile and recyclable DMSO substitute. Chem.
Commun. 2007, 1427-1429.
3
1
L. Meng; Z. Lu; Y. Ma; X. Xue; F. Nie; C. Zhang. Enhanced Intermolecular Hydrogen Bonds
Facilitating the Highly Dense Packing of Energetic Hydroxylammonium Salts. Cryst. Growth Des.
2
016, 16, 7231-7239.
3
2
T. Grewer. The influence of Chemical structure on exothermic decomposition. Thermochim. Acta
991, 187, 133-149.
1
3
3
L.O. Cisneros; W.J. Rogers; M.S. Mannan. Comparison of the thermal decomposition behavior for
members of the hydroxylamine family. Thermochim. Acta 2004, 414, 177-183.
3
4
L.O. Cisneros; X. Wu; W.J. Rogers; M.S. Mannan; J. Park; S.W. North. Decomposition products
of 50% mass hydroxylamine/water under runaway reaction conditions. Process Saf. Environ. 2003,
8
1, 121-124.
3
5
W. Shockley; H. J. Queisser. Detailed Balance Limit of Efficiency of p‐n Junction Solar Cells. J.
Appl. Phys. 1961, 32, 510-519.
3
6
D.L. Frasco; E.L. Wagner. Interpretation of the Infrared Spectra of the Solid Hydroxylammonium
Halides. J. Chem. Phys. 1959, 30, 1124-1130.
3
7
N. Fischer; T.M. Klapötke; D.G. Piercey; J. Stierstorfer. Hydroxylammonium 5-
Nitriminotetrazolates. Z. Anorg. Allg. Chem. 2012, 638, 302-310.

Dalton Transactions
Page 16 of 27
View Article Online
DOI: 10.1039/C9DT00690G
Fig. 1. Panel A: NH OHPbI powder freshly prepared by sulfolane extraction with toluene. Panel B:
3
3
the same sample after 3 days under ambient conditions.

Page 17 of 27
Dalton Transactions
View Article Online
DOI: 10.1039/C9DT00690G
Fig. 2. Representation of the unit cell of NH OHPbI with sulfolane molecules layers between the
3
3
Pb-I-NH OH layers. Red ellipsoids: oxygen atoms; yellow ellipsoids: sulfur atoms; black ellipsoids:
3
carbon atoms; creamy-white ellipsoids: hydrogen atoms; blue ellipsoids: nitrogen atoms; dark grey
ellipsoids: lead atoms; purple ellipsoids: iodine atoms. The probability used to depict the thermal
ellipsoids is 50%. Apparent multiple positions of each carbon and hydrogen atom of sulfolane
molecules are due to the modeling of their disorder.

Dalton Transactions
Page 18 of 27
View Article Online
DOI: 10.1039/C9DT00690G
Fig. 3. The crystal structure of NH OHPbI seen along the b axis. The sinusoidal arrangement of
3
3
+
NH OH ions along the a direction is clearly visible.
3

Page 19 of 27
Dalton Transactions
View Article Online
DOI: 10.1039/C9DT00690G
Fig. 4. Panel A: powder diffraction pattern of NH OHPbI powder prepared by sulfolane extraction
3
3
with toluene. Panel B: powder diffraction pattern of NH OHPbI single crystals grinded in agate
3
3
mortar. The assignment of the most important reflections is reported.

Dalton Transactions
Page 20 of 27
View Article Online
DOI: 10.1039/C9DT00690G
Fig. 5. Panel A: TG and DTA profiles of NH OHPbI powder prepared by sulfolane extraction with
3
3
toluene and washed multiple times with toluene. Panel B: TG profile and its derivative. Panel C: DTA
profile and the derivative of TG profile.

Page 21 of 27
Dalton Transactions
View Article Online
DOI: 10.1039/C9DT00690G
Fig. 6. Difference between the mass spectrum of the carrier gas (argon) used in the TG-DTA
measurement before and during the maximum gas evolution from the sample. Peaks are visible for
m/z values 44, 32, 30, 28 and 14.

Dalton Transactions
Page 22 of 27
View Article Online
DOI: 10.1039/C9DT00690G
Fig. 7. Diffuse reflectance UV-Vis spectrum of NH OHPbI powder.
3
3

Page 23 of 27
Dalton Transactions
View Article Online
DOI: 10.1039/C9DT00690G
Fig. 8. Tauc’s plot of the UV-Vis reflectance spectrum assuming an allowed direct band gap
transition.

Dalton Transactions
Page 24 of 27
View Article Online
DOI: 10.1039/C9DT00690G
Fig. 9. FT-IR diffuse reflectance spectra of NH OHPbI powder (yellow orange line) and sulfolane
3
3
(blue line) both dispersed in KBr.

Page 25 of 27
Dalton Transactions
View Article Online
DOI: 10.1039/C9DT00690G
Fig. 10. ESI-MS spectrum (positive ions) of NH OHPbI dissolved in DMI.
3
3

Dalton Transactions
Page 26 of 27
View Article Online
DOI: 10.1039/C9DT00690G
Formula
6 20 3 3
C H I NO PbS
3.338
17.091
774.18
light yellow
polyhedral
-3
D
μ/mm
calc./ g cm
-1
Formula Weight
Colour
Shape
Size/mm
T/K
3
0.07×0.05×0.05
100
Crystal System
Space Group
orthorhombic
Pmma
7.920(6)
9.030(4)
10.770(2)
90
90
90
770.25(1)
2
a/Å
b/Å
c/Å
°
α/
β/
γ/
°
°
3
V/Å
Z
Z'
0.25
Wavelength/Å
0.700
Radiation type
Synchrotron
2.221
30.974
6502
°
ϑ
ϑ
min
/
max/°
Measured Refl.
Independent Refl.
1360
Reflections with I >1271
(I)
2
R
int
0.0392
56
20
1.860
-2.248
1.151
0.1259
0.1239
0.0413
0.0394
Parameters
Restraints
Largest Peak
Deepest Hole
GooF
wR
wR
2
(all data)
2
R
R
1
(all data)
1
Table 1. Crystallographic data of NH OHPbI .
3
3

Page 27 of 27
Dalton Transactions
View Article Online
Cell parameters (Å)
Single crystal (100 K)
7.920(6)
Powder (298 K)
DOI: 10.1039/C9DT00690G
a
b
c
7.9624(5)
9.030(4)
9.3919(6)
10.770(2)
10.8248(8)
Table 2. Comparison of unit cell parameters obtained at 100 K (single crystal data) and at 298 K
powder data).
(