Synthesis and nanoscale characterization of (NH4) 4ThF8 and ThNF

Article abstract of DOI:10.1021/ic900632g

Synthesis of (NH₄)₄ThF₈ by a solid state reaction of ThO₂ and NH₄HF₂ and the formation of ThNF by ammonolysis of (NH₄)₄ThF₈ and ThF₄ under differ

Full text of DOI:10.1021/ic900632g

5736 Inorg. Chem. 2009, 48, 5736–5746
DOI: 10.1021/ic900632g
Synthesis and Nanoscale Characterization of (NH4)4ThF8 and ThNF
G. W. Chinthaka Silva,† Charles B. Yeamans,‡ Gary S. Cerefice,† Alfred P. Sattelberger,§ and
Kenneth R. Czerwinski*,†
†Harry Reid Center for Environmental Studies, University of Nevada, Las Vegas, Box 454009, 4505 Maryland
Parkway, Las Vegas, Nevada 89154, ‡Department of Nuclear Engineering, University of California, Berkeley,
1140 Etcheverry Hall, M.C. 1730 Berkeley, California 94720-1730, and §Argonne National Laboratory, 9700
Cass Avenue, Bldg 221, Argonne, Illinois 60517
Received November 28, 2008
Synthesis of (NH4)4ThF8 by a solid state reaction of ThO2 and NH4HF2 and the formation of ThNF by ammonolysis of
(NH4)4ThF8 and ThF4 under different experimental conditions were investigated. The solid state reaction of ThO2 with
NH4HF2 led to the terminal product (NH4)4ThF8 through a known intermediate (NH4)3ThF7 and most likely two other
unknown chemical phases as determined by X-ray powder diffraction. Conversion of (NH4)4ThF8 into ThNF occurs
through a ThF4 intermediate phase. Studies on the ammonolysis of ThF4 revealed it converted into ThNF through a
continuous formation of low-stoichiometric thorium-nitride-fluorides such as ThN0.79F1.63 and ThN0.9F1.3. Thermal
behavior of ThNF was also examined under different atmospheres and temperatures, with evaluation of formation
kinetics. The ThNF decomposed to low-stoichiometric thorium nitride fluorides (ThNx/3F4-x) under different
environments up to 1100 °C. Significant morphological changes in the products compared to that of the precursors
confirmed the reaction steps involved. Microstructural characterization of (NH4)4ThF8 and ThNF were performed by
HRTEM and are presented in this work for the first time. The (NH4)4ThF8 product was shown to contain polycrystalline
characteristics in the majority of its nanostructure. On the other hand, ThNF has a high order of nanostructure, which
explains the high thermal stability of the compound up to 1100 °C and the difficulty of making ThNx, in initial target
product, from the described experimental conditions.
1. Introduction
characterizations on most of the thoria-based compounds
such as oxides,4 carbonates,5 oxalates,6 and hydrides7 have
been reported in literature. However, no such data can be
found for thorium fluorides, nitrides, or nitride-fluorides.
Lack of such substantial chemical data of these materials
motivates their synthesis and characterization.
On the basis of previous studies performed on thorium
nitrides (ThNx), three common compounds can be identified.
Those are ThN, Th3N4, 8,9 and Th2N3.10,11 Synthesis of ThNx
involves heating of the thorium metal 12 or thorium hydride
under ammonia or nitrogen atmosphere.10,13 Reacting ThCl4
Actinide mononitrides and their mixed systems such as
(U-Pu)N have been considered for advanced nuclear fuel
applications.1,2 They have excellent thermal, chemical, phy-
sical, and nuclear properties, with well-behaved mutual
solubility properties, and are considered as a fuel material
for the transmutation of minor actinides.3 Thoria-based
nitride fuels can also be considered in these different applica-
tions because of the analogous crystallography of thorium
mononitride (ThN) to other nitride fuels such as UN. How-
ever, there is a very small amount of data reported on the
thoria-based nitrides and related compounds. Microscopic
(6) Ananthasivan, K.; Anthonysamy, S.; Singh, S.; Rao, P. R. V. J. Nucl.
Mater. 2002, 306, 1–9.
(7) Terrani, K. A.; Silva, G. W. C.; Yeamans, C. B.; Balooch, M.;
Olander, D. R. J. Nucl. Mater., accepted for publication.
(8) Benz, R.; Hoffman, C. G.; Rupert, G. N. Phys. Inorg. Chem. 1967, 89,
191–197.
(9) Auskern, A. B.; Aronson, S. J. Phys. Chem. Solids 1967, 28, 1069–
1071.
(10) Olson, W. M.; Mulford, R. N. R. J. Phys. Chem. 1965, 69, 1223–
1226.
(11) Aronson, S.; Auskern, A. B. J. Phys. Chem. 1966, 70, 3937–3941.
(12) Chiotti, P. J. Am. Ceram. Soc. 1952, 35, 123–130.
(13) Uno, M.; Katsura, M.; Miyake, M. J. Less-Common Met. 1986, 121,
615–619.
*To whom correspondence should be addressed. E-mail czerwin2@unlv.
nevada.edu. Phone: (702) 895 0501. Fax: (702) 895 3094.
(1) Matzke, Hj. Advanced LMFBR Fuels; Elsevier: Amsterdam, 1986; p 2.
(2) Takano, H.; Nishihara, K. Prog. Nucl. Energy 2002, 40, 473.
(3) Minato, K.; Akabori, M.; Takano, M.; Arai, Y.; Nakajima, K.; Itoth,
A.; Ogawa, T. J. Nucl. Mater. 2003, 320, 18–24.
(4) Curran, G.; Sevestre, Y.; Rattray, W.; Allen, P.; Czerwinski, K. R. J.
Nucl. Mater. 2003, 323, 41–48.
(5) Dash, S.; Singh, A.; Ajikumar, P. K.; Subramanian, H.; Rajalakshmi,
M.; Tyagi, A. K.; Arora, A. K.; Narasimhan, S. V.; Raj, B. J. Nucl. Mater.
2002, 303, 156–168.
r
2009 American Chemical Society
pubs.acs.org/IC
Published on Web 06/11/2009

Article
Inorganic Chemistry, Vol. 48, No. 13, 2009 5737
Figure 1. XRD patterns indicating the reaction progress of the 100 mg batch of ThO2 mixed with excess NH4HF2. Only a part of the XRD patterns are
shown for clarity of the comparison.
with LiNH2 in liquid ammonia and then heating under
nitrogen gas is another way to synthesize the thorium
nitrides.15 In most of these methods, thorium mononitride
is made by thermally decomposing higher nitrides such as
Th3N4 at 1500 °C under inert atmosphere.8,10,16 However,
synthesis of ThNx with high phase purity is difficult because
of secondary ThO2 formation. In particular, the ThN pro-
duced by these methods shows the presence of ThO2 chemical
phase as it is highly susceptible to oxidation.13 This fact
complicates the synthesis of ThNx using above-mentioned
high-temperature routes.
Therefore, a low-temperature fluoride route based on a
recent research work on synthesizing UN 14 was investigated
to make thorium nitrides and related compounds in this
work. In this method, ThO2(s) was reacted with excess
(4.2 molar) NH4HF2(s) to make ammonium thorium fluor-
ide {(NH4)xThFx+4} which was then used as the precursor
for making the desired nitrides. Heat treatment of
(NH4)xThFx+4 under NH3(g)/ammonolysis and further
heating of any higher nitrides formed under Ar(g) was
envisaged in the subsequent steps. Nevertheless, the expected
ThNx compounds were not observed, and instead ThNF
was obtained as the terminal product up to a temperature of
1100 °C. Similar reactions and product formations have also
been reported by different groups.16
by reacting ThF4 with an aqueous ammonium fluoride
solution. In the current work, the same (NH4)4ThF8 is
reported as the terminal product of the ThO2(s) reaction with
NH4HF2(s) but with an intermediate (NH4)3ThF7(s) phase.
Two more possible intermediate chemical phases with differ-
ent stabilities with respect to reaction kinetics are also dis-
cussed. Ammonolysis of (NH4)4ThF8 produced ThNF. In
this process ThF4 was also identified as an intermediate phase
before forming ThNF. In the literature, Juza et al.20 reported
ThNF formation by heating ThF4 at 850 °C under flowing
ammonia gas confirming what was found in the current work.
However, a systematic study on these systems cannot be
found in literature. Therefore, the products observed when
using different temperatures in the ammonolysis step of
(NH4)4ThF8 and ThF4 and a complete study on the thermal
behavior of ThNF up to 1100 °C under high-purity atmo-
spheres of Ar(g) and N2(g) are presented here. On the basis of
the results of these experiments, a possible reaction mechan-
ism for the formation of ThNF from (NH4)4ThF8/ThF4
under ammonia gas is also proposed. Microscopy was used
for the first time to characterize the morphology and nanos-
tructures of these thorium compounds. These observations
were further used in explaining the formation of different
products at separate reaction steps.
2. Experimental Details
The product of the reaction between ThO2(s) and ammo-
nium hydrogen fluoride (NH4HF2(s)) was reported as
(NH4)4ThF8.17 Penneman et al.18 as well as Glavic et al.19
synthesized the same ammonium thorium fluoride compound
2.1. Synthesis of Ammonium Thorium Fluoride. A solid state
reaction was utilized in synthesizing ammonium thorium fluor-
ides. Two batches of ThO2 (100 mg and 1000 mg) were used in
this experiment. The ThO2 (STREM Chemicals, 99.99%) pow-
der was added into a polyethylene vial with ground NH4HF2
(Fisher Scientific, 99.99%) in a 1:4 molar ratio with 10 mol. %
excess NH4HF2. The sample was then mixed with a spatula for
10 min. The reaction was allowed to progress in the closed vial at
room temperature until completion. Reaction progress was
evaluated by periodically removing a small sample for X-ray
diffraction analysis.
(14) Yeamans, C. B.; Silva, G. W. C.; Cerefice, G. S.; Czerwinski, K. R.;
Hartmann, T.; Burrell, A. K.; Sattelberger, A. P. J. Nucl. Mater. 2008, 374,
75–78.
(15) Brese, N. E.; Disalvo, F. J. J. Solid State Chem. 1995, 120, 378–
380.
(16) Aronson, S.; Cisney, E.; Gingerich, K. A. J. Am. Ceram. Soc. 1967,
50, 248–252.
(17) Wani, B. N.; Patwe, S. J.; Rao, U. R. K.; Venkateswarlu, K. S. J.
Fluorine Chem. 1989, 44, 177–185.
(18) Penneman, R. A.; Ryan, R. R.; Rosenzweig, A. Chem. Commun.
2.2. Ammonolysis of (NH4)4ThF8 and Heating of ThNF. The
resulting (NH4)4ThF8 product was placed on a platinum sheet
1968, 16, 990–991.
(19) Glavic, P.; Slivnik, J.; Bole, A. J. Inorg. Nucl. Chem. 1973, 35, 427–
432.
(20) Juza, R.; Sievers, R. Z. Anorg. Allg. Chem. 1968, 363, 258–272.

5738 Inorganic Chemistry, Vol. 48, No. 13, 2009
Silva et al.
Figure 2. Observed and the calculated XRD powder patterns of (a) (NH4)4ThF8 and (b) (NH4)3ThF7 samples. Calculated and the observed patterns are
shown in continuous and scattered lines, respectively. Calculated patterns of both (NH4)4ThF8 and (NH4)3ThF7 are highlighted in the corresponding
figures. X-axes are in 10-1 nmunits. Square rootofcountswas usedas the scaleofY-axisfor properdisplay ofthe calculatedpatterns. Allowedlinepositions
for the separate phases are marked as vertical lines at the bottom with arrows for the highlighted pattern.
and inserted into a 25 mm diameter quartz-glass tube. The tube
was closed on either end with a 25 mm quartz Solv-Seal
(Andrews Glass Co., Inc.), and sealed using Pyrex Solv-Seal
caps fitted with 15 mm high vacuum Teflon stopcocks. Ammo-
nia gas (high-purity 99.999%, Praxair) was flushed through the
tube for 5 min to pre-equilibrate and purge the system. The
system was heated to 800 °C under a constant flow of ammonia
gas at 1 atm. The gas flow was constant throughout the experi-
ment. Total heating time was varied up to 5 h and then cooled to
room temperature. Once the samples cooled, the ammonia gas
flow was terminated and the product was removed for analysis.
The reaction was also performed at 1000 °C for 45 min and at
1100 °C for 15 min. All heating of the products were performed
with high-purity (99.999%) Ar, Ar/H2 (5%), N2, and N2/H2
(5%). A zirconium metal sponge was used as an oxygen getter in
experiments at 1100 °C.
2.3. Characterization Methods. 2.3.1. X-ray Powder Dif-
fraction (XRD). XRD was used to identify phase distributions
of the synthesized samples using a Philips PANalytical X’Pert
Pro instrument with a Cu KR radiation and a Ni filter. Lattice
parameter refinements and phase quantifications were done
using TOPAS software.21 TOPAS was designed to analyze
powder data and uses the Rietveld method to fit the calculated
patterns to observed patterns. With TOPAS, structure determi-
nation does not include any methods that require integrated
intensity extraction as it is performed in direct space using step
intensity data. Structure determination is made in TOPAS by
initially estimating the cell parameters of the structure, using a
rough knowledge about the cell content, and speculating the
space group of the structure. Initial deduction of the structure
was done using ICSD (Inorganic Crystal Structure Database)
reference patterns as the models for the refinement. The ICSD
XRD pattern numbers 9889 and 14128 were used for
(NH4)4ThF8 and (NH4)3ThF7, respectively. For ThO2 and
ThNF, the ICSD XRD pattern numbers 61586 and 35745 were
used, respectively. LaB6 SRM (Standard reference Material)
660a was used as the internal standard for accurate lattice
parameter refinements.
Electron density map of ThNF was also used to confirm the
crystallography and to support the HRTEM findings. Le Bail
decomposition22 was used to extract the individual observed struc-
ture factor amplitudes (Fobs) of the XRD powder pattern
using Jana2000.23 These observed structure factors were then used
to calculate the electron density maps of the compound by the
charge-flipping algorithm using Superflip program.24 The calculated
electron density maps were visualized using UCSF Chimera.25
(22) Bail, A.; Duroy, H.; Fourquet, J. L. Mater. Res. Bull. 1998, 23, 447–
452.
::
::
::
^
~
(23) DusIek, M.; PetrIoAcIek, V.; Wunschel, M.; Dinnebier, R. E.;
Smaalen, S. J. Appl. Crystallogr. 2001, 34, 398–404.
(24) Palatinus, L.; Chapuis, G. J. Appl. Crystallogr. 2007, 40, 786–790.
(25) Pettersen, E. F.; Goddard, T. D.; Huang, C. C.; Couch, G. S.;
Greenblatt, D. M.; Meng, E. C.; Ferrin, T. E. J. Comput. Chem. 2004, 25(13),
1605–1612.
(21) Coelho, A. A. J. Appl. Crystallogr. 2000, 33, 899–908.

Article
Inorganic Chemistry, Vol. 48, No. 13, 2009 5739
Figure 3. Secondary-electron SEM image (a) and the corresponding EDS spectrum (b) of the synthesized (NH4)4ThF8 and the SEM image of ThO2 (c).
2.3.2. Microscopy. Morphological studies of the samples
were done using scanning electron microscopy (SEM) and
transmission electron microscopy (TEM). SEM imaging was
performed using a JEOL scanning electron microscope model
JSM-5610 equipped with secondary electron (SE) and back-
scattered electron (BE) detectors and an Oxford ISIS EDS
(energy dispersive spectroscopy) system. The acceleration vol-
tage used in SEM was 15 kV. TEM was performed using
TECNAI-G2-F30 with a 300 keV field emission gun. All TEM
images were recorded using a slow scan CCD camera attached
with a Gatan GIF 2000 image filter. The conventional bright
field (BF) diffraction contrast mode and the high resolution
(HR) TEM modes coupled with selected area electron diffrac-
tion (SAED) patterns and Fast Fourier Transformation (FFT)
micrographs were utilized to characterize the samples. X-ray
energy dispersive spectrometry (XEDS) was used to determine
the elemental distributions. Further description of the sample
preparation techniques used for this work can be found in Silva
et al. 26. The thorium samples for TEM analysis were prepared
by the solution-drop method and microtome cutting. In the
solution drop method, the samples were ground to microsize in a
mortar and mixed with ethanol. One drop of the solution was
added onto a 3 mm diameter carbon film supported on a copper
grid that was used as the sample holder. A Leica EM UC6rt
microtome was used for sample cutting of 25 to 75 nm thick-
nesses. More details of these sample preparations are discussed
in Silva et al.26
3. Results
3.1. Formation and Characterization of (NH4)4ThF8.
The XRD patterns of the samples showed that the
terminal product of the reaction of NH4HF2 with ThO2
is (NH4)4ThF8 with a triclinic unit cell and P1 space
group. Further analysis confirmed the formation of an
intermediate (NH4)3ThF7 chemical phase of an orthor-
hombic unit cell and Pnma space group in the reaction.
Figure 1 is a graphical view of the kinetics of ThO2 (100
mg batch) solid state reaction with NH4HF2 at room
temperature. After 1 day of mixing the two reactants, a
new chemical phase was formed according to the XRD
pattern in Figure 1. This XRD pattern has some simila-
rities to the pattern of the terminal (NH4)4ThF8 chemical
phase even though refinement assuming similar crystal-
lography was not successful. The intermediate chemical
phase which was identified as (NH4)3ThF7 was formed
after 2 days. The XRD pattern of the sample after 5 days
displays some of the peaks that correspond to
(NH4)3ThF7 as indicated by arrows. These peaks start
to diminish with the formation of new set of peaks at 2θ
values of 12.6, 15.1, and 15.5°. The XRD pattern formed
with these peaks was prominent and was believed to be
another intermediate phase that was not identifiable as
the known ammonium fluorides of thorium or any other
actinides such as uranium and neptunium. This new and
unknown chemical phase remained until it completely
(26) Silva, G. W. C.; Yeamans, C. B.; Ma, L.; Cerefice, G. S.; Czerwinski,
K. R.; Sattelberger, A. P. Chem. Mater. 2008, 20, 3076–3084.

5740 Inorganic Chemistry, Vol. 48, No. 13, 2009
Silva et al.
b= 0.83667 (7), and c = 0.73372 (6) nm. Lattice para-
meters of orthorhombic (NH4)3ThF7 with a space group
of Pnma were refined to be a=0.3952(9), b=0.7932(5),
and c = 0.70462(5) nm.
The secondary electron SEM image of the (NH4)4ThF8
sample (Figure 3a) indicates well-crystallized sample
particles with an acicular shape (Figure 3a) compared
to the plate-like shape of the reactant, ThO2, (Figure 3c).
Furthermore, the acicular crystals that Penneman et al.18
described can also be clearly seen in this figure. The
elemental distribution of the (NH4)4ThF8 sample ob-
tained using the EDS (Figure 3b) also shows the presence
of the expected components in the sample.
TEM BF image of the (NH4)4ThF8 particles from a
ground sample is shown (Figure 4a). The SAED pattern
in Figure 4b was obtained by focusing on an area of the
large particle seen in the bottom half of Figure 4a, and this
pattern is in [111] zone axis. Figure 4c is a HRTEM image
of the same particle displaying lattice fringes in several
different orientations as indicated in the image. Some of
these domains have high point resolution in two direc-
tions while others exhibit either one or no directional
fringe properties. This observation suggests that the
(NH4)4ThF8 obtained is a polycrystalline material.
On the other hand, Figure 5b shows a well-oriented set
of lattice fringes of (NH4)4ThF8 corresponding to the
(1-11) lattice plane distributed in a larger area than in the
nanocrystalline areas found in Figure 4c. This set of
lattice fringes are also disturbed by another similar-sized
nanoparticle domain, as indicated by the arrow, which
has being grown from the other side of the particle. This
feature further implies a polycrystalline characteristic of
the sample even though this area consists of compara-
tively large nanoparticles.
Figure 4. (a) TEM BF image, (b) [111] electron diffraction pattern, and
(c) HRTEM image of (NH4)4ThF8 ground particle.
3.2. Synthesis and Characterization of ThNF. Ammo-
nolysis of (NH4)4ThF8 or the heat-treatment of the
sample under ammonia atmosphere at three different
temperatures (Table 1) showed that the final product of
this second reaction step is ThNF. At 800 °C, 300 min
were required to synthesize this terminal product. In-
creasing the reaction temperature resulted in a significant
decrease in reaction time, with sample heating at 1100 °C
for 15 min optimal for synthesis of ThNF. The elemental
analysis of this sample determined by EDS is shown in
Table 2. The Th/N/F molar ratio of the compound is
approximately 1:1:0.6. On the basis of the error asso-
ciated with the EDS on quantification and as EDS is
considered a semiquantitative method, the stoichiometry
of the compound can be considered close to 1:1:1 (ThNF).
In the sample heated for 300 min at 800 °C, a secondary
ThO2 phase was identified because of prolonged exposure
to minute oxygen impurities in the system.
The XRD pattern of the ThNF sample synthesized by
ammonolysis of (NH4)4ThF8 at 1100 °C for 15 min
(Figure 6) confirms the sample is single-phased. The
refined lattice parameter of the rhombohedral unit cell
was 0.71312 (5) nm which agrees well with the literature
value of 0.7130 (3) nm.20 The phase density of ThNF was
calculated to be 9.369 (2) g/cm3.
The secondary SEM image of a ThNF sample
(Figure 7a) shows plate-like particle shapes. The 15,000
times magnified SEM image (Figure 7b) further displays
pitted and porous characteristics of the ThNF particle
changed into the terminal (NH4)4ThF8 chemical phase
after 22 days.
Wani et al.13 describe the reaction of solid ThO2 with
solid NH4HF2 at room temperature to be quick (2 h) in
forming (NH4)4ThF8. However, they reported an absence
of the reflection at d=0.815 nm in the XRD pattern of
this product. Furthermore, they did not identify or
matched their XRD pattern to (NH4)3ThF7 which has
some similar reflections to that of (NH4)4ThF8 with the
absence of the peak at d = 0.815 nm (Figure 2). The
chemical phase they were describing can also be due to the
first unknown XRD pattern that was described earlier in
Figure 1 (XRD pattern after 1 day). Further comparison
with this is difficult as the experimental parameters such
as reactant mass or XRD patterns of the products are not
reported.
The observed and the calculated XRD patterns of
(NH4)4ThF8 and a sample containing (NH4)3ThF7 are
shown in Figure 2, panels a and b, respectively. The
patterns of both (NH4)4ThF8 and (NH4)3ThF7 matched
well with the reference patterns in the ICSD of
numbers 9889 and 14128, respectively. Few impurity
peaks in both samples were identified, especially near
0.68 nm d-spacing value in the (NH4)4ThF8 sample. These
impurity peaks together with the amorphous character-
istics of the samples can be accounted for by the excess
NH4F/NH4HF2 salts. The refined triclinic unit cell para-
meters of (NH4)4ThF8 compound are a = 0.85041 (7),

Article
Inorganic Chemistry, Vol. 48, No. 13, 2009 5741
Figure 5. (a) TEM BF image of a ground (NH4)4ThF8 particle. (b) The HRTEM image of the circled area of (a) and (c) the corresponding FFT.
Table 1. Temperature Effect on the Ammonolysis of (NH4)4ThF8
product
temperature/ °C
time/ min
primary phase
weight %
secondary phase
ThO2
weight %
21.0(6)
800
1000
1100
300
45
15
ThNF
ThNF
ThNF
79.0(6)
100
100
Table 2. Elemental Analysis of the As-Synthesized ThNF by EDS
From C to D, the intensities start to change abruptly and
approach a minimum at the end of the line. This observa-
tion suggests a thickness variance across the scanned area
forming a thinner region of about 6 nm, given the length
of the line scanned to be 36 nm. The void spaces of
the ThNF particles detected by SEM imaging were 50
to 200 nm in width. This implies a continuation of void
spaces down to nanoscale of the ThNF compound.
HRTEM of another area of the same particle in
Figure 8a is displayed in Figure 9. Two rectangle areas
of Figure 9 were enlarged using a smooth spatial filter to
obtain HRTEM images in Figure 9b and Figure 10. The
beam direction of these HRTEM images is [110]. Lattice
fringes in Figure 9b provide a width of 0.265 nm for a
single layer of fringes corresponding to a reflection plane
of (221). Intensity difference between areas A and B
reveals more intuitive structural detail of ThNF as de-
picted in Figure 10. Thickness variance as identified
previously is the probable reason for such a difference
in intensities of two areas within 30 nm. Lattice fringes in
area B as shown in Figure 10 provide more detail than
area A. Spacing between two layers of lattice fringes in
Figure 10 is 0.225 nm, which is close to the d-spacing of
(333) lattice plane of ThNF. Furthermore, this spacing is
less than the width of a single layer of lattice fringes in
area A (0.265 nm) demonstrating a shift in the atomic
planes that create the lattice fringes in the HRTEM
images. The electron density map of ThNF calculated
from XRD pattern utilizing charge flipping, which is the
inset to the upper right of Figure 10, also confirms the
structural details observed with HRTEM imaging. As the
electron density map indicates a high electron density at
element
averaged atom percentages (%)
N K ((1.25)
F K ((1.26)
39.0
21.6
Th K ((1.47)
Th/N/F molar ratio
39.4
1:0.99:0.60
surface. Furthermore, the corresponding EDS spectrum
(Figure 7c) of this sample confirms the presence of
thorium, fluorine, and nitrogen. The nitrogen peak par-
tially overlaps with the background peak from carbon
tape that was used to mount the sample. SEM image of
ThF4 is also shown in Figure 7d for the comparison
purposes.
A 50 nm thick sample of ThNF was microtome cut for
TEM imaging. Figure 8a shows a low-resolution TEM
BF image of few ThNF particles thus prepared. As
microtome cutting was applied for the specimen prepara-
tion, the cross-sectional view of the sample at HRTEM is
supposed to show a uniform intensity across a focused
area. However, most of the areas investigated showed
intensity variance of the images at high-resolution.
HRTEM of one such area is shown in Figure 8b. The
area in this image shows a continuous set of ThNF lattice
fringes corresponding to (0-11) planes under [100] beam.
Most of the area displays lattice fringes only in one
direction except for some lattice fringes in the area
depicted by B-C. A line scan across the image on a
randomly selected lattice fringe from A through D
(Figure 8d) shows that the intensity of the image changes
considerably. The intensities from A to B do not vary
much, but it is at highest in the region denoted by B-C.

5742 Inorganic Chemistry, Vol. 48, No. 13, 2009
Silva et al.
Figure 6. Observed (top continuous line) and calculated (highlighted continuous line for ThNF) patterns of ThNF synthesized by heating (NH4)4ThF8 at
1100 °C for 15 min under NH3. Square root of counts was used as the scale of Y-axis for proper display of the calculated patterns.
Figure 7. Secondary electron SEM images of the ThNF sample, (a) showing the morphology at ꢀ5,000 magnification and (b) at ꢀ15,000 magnification,
(c) the corresponding EDS spectrum of ThNF, (d) and the SEM image of ThF4.
locations where Th atoms reside within the crystal struc-
ture of ThNF, the HRTEM imaging will only show the
intensities due to Th atoms. Repetition of such structural
units will give a well-ordered nanostructure (Figure 10).
Atomic resolution of the hexagonal-like order of atoms in
the HRTEM image is different from point-to-point as
indicated by circles in the image. The atomic planes
in (111) direction of the packed rhombohedral unit cells
are in and out of plane slightly such that the depth of
the imaging can change as indicated in the inset of the
Figure 10 (lower to the electron density map). All these
observations confirm the presence of single-phased
ThNF in the sample.
(Figure 11). The lattice parameters, a=1.3049 (3), b=
1.1120 (2), c=0.8538 (2) nm27 for the monoclinic-ThF4
agree fairly well with the refined lattice parameters of the
measured pattern, a=1.30444 (7), b =1.10118 (6), c=
0.85343 (5) nm.
After identifying ThF4 as an intermediate in the con-
version of (NH4)4ThF8 to ThNF, ThF4 was heated under
varying experimental conditions (Table 3) to determine
the reaction mechanism. Removal of fluorine atoms from
the ThF4 and incorporation of nitrogen to fill those
vacant sites can be observed as the low-stoichiometric
chemical phases of thorium nitride fluorides begin to
appear in the samples (Table 3). Figure 12 shows the
XRD powder pattern refinement of the ThN0.79F1.63
sample, which was synthesized by heating ThF4 at 800 °C
for 60 min. Treating ThF4 for 300 min under NH3(g)
3.3. Ammonolysis of ThF4 and the Thermal Behavior of
ThNF. Heating of the (NH4)4ThF8 sample at 800 °C for
60 min formed ThF4 in agreement with Wani et al. The
17
observed XRD powder pattern together with the calcu-
lated pattern of ThF4 chemical phase identified is shown
(27) Benner, G.; Muller, B. G. Z. Anorg. Allg. Chem. 1990, 588, 32–42.

Article
Inorganic Chemistry, Vol. 48, No. 13, 2009 5743
Figure 10. HRTEM of the particle area denoted by letter B in Figure 9.
Inset to the upper right side is the electron density map calculated by
ThNF XRD pattern using charge flipping. A model of the molecule in
(111) direction is also inserted.
was observed after 600 min with no ThO2 impurities
(Figure 14a). However, a 17.6(7) wt % ThO2 secondary
phase was detected on heating ThF4 at 1000 °C for 75 min
(Figure 14b) whereas a 45 min heating did not show any
impurity phase. These three observations suggest that
prolonged heating at elevated temperatures form a ThO2
secondary chemical phase because of the oxygen partial
pressure produced by the quartz furnace tube used in the
experiment.
Figure 8. TEM BF of a microtome cut ThNF sample, (b) HRTEM
image, (c) FFT of HRTEM, and (d) experimental intensity profile along
A-B in HRTEM.
Thermal stability of ThNF under different atmo-
spheres was evaluated and the results summarized
(Table 4). Using high-purity argon (99.999%) at 1100 °C,
ThNF decomposes into ThN0.79F1.63 with a formation
of secondary ThO2 phase probably because of minor
oxygen impurities formed in the experimental setup at
this temperature. Under N2 and N2-H2 (5%) environ-
ments, ThNF decomposed into another, similar low-
stoichiometric thorium-nitride-fluoride (ThN0.90F1.30).
4. Discussion
The terminal product of ThO2(s) reaction with NH4HF2(s)
at ambient conditions was (NH4)ThF8, as previously
reported.17 This work determined the reaction proceeds
through a known intermediate chemical phase (NH4)3ThF7(s)
that was formed after less than 2 days of mixing ThO2 with
NH4HF2. Two other possible phases of unknown chemical
compositions were also intermediates in the reaction. Further
interaction of the remaining ThO2 with NH4HF2/NH4F and
the transformation of (NH4)3ThF7 into the terminal
(NH4)4ThF8 took up to 22 days for a sample mass of
100 mg and 60 days for a sample mass of 1000 mg. Wani
et al. reported 17 that (NH4)ThF8 to be formed after 2 h based
on the XRD pattern obtained for the sample. However, the
sample size was not reported. A sample under 100 mg could
react to completion within a few hours. The XRD pattern in
the previous work did not include the peak near d = 0.815 nm
Figure 9. (a) HRTEM of another area of the particle in Figure 8a.
(b) Enlarged image of the area denoted by letter A with a rectangle. The
area B denoted by a rounded rectangle is used in Figure 10. Onsets of each
image are the corresponding FFT micrographs.
further reduced the fluorine content from the reactant
producing ThNF (79 wt %) along with ThN0.79F1.63
(21 wt %). XRD powder refinement of this sample is
displayed in Figure 13 with highlighting the calculated
pattern for ThN0.79F1.63. A complete formation of ThNF

5744 Inorganic Chemistry, Vol. 48, No. 13, 2009
Silva et al.
Figure 11. XRD powder pattern of the sample synthesized by heating (NH4)4ThF8 at 800 °C for 60 min under NH3(g). The calculated pattern of ThF4 is
highlighted.
Table 3. Ammonolysis of ThF4
product
temp./ °C
time/ min
primary phase
mass %
secondary phase
mass %
800
800
800
800
1000
1000
1100
60
ThN0.79F1.63
ThNF
ThNF
ThNF
ThNF
73.6(2)
79.0 (1)
79.0(6)
100
100
82.4 (6)
100
unreacted ThF4
ThN0.9F1.3
ThN0.9F1.3
26.4(2)
21.0 (1)
21.0(6)
240
300
600
45
75
15
ThNF
ThNF
ThO2
17.6 (7)
which could probably differentiate chemical composition of
the sample to be (NH4)3ThF7 or (NH4)4ThF8. Observation of
another chemical composition could also be possible accord-
ing to the XRD patterns identified for two unknown chemical
phases as indicated in Figure 1. In this current work, it was
found that (NH4)3ThF7 forms within 2 days of mixing ThO2
and (NH4)2HF2 whereas (NH4)4ThF8 takes up to months
depending on the sample size.
Ammonolysis of (NH4)4ThF8 at 800 °C produced ThNF
through ThF4. Therefore, further studies of the thorium
system were performed using ThF4. Heat treatment of
ThF4 under NH3 at different temperatures demonstrated a
ThF4 decomposition through removing some of the fluorine
atoms in the crystal system. Nitrogen incorporates into those
vacant cites of the ThF4-x, resulting in new phases with the
chemical formula of ThNx/3F4-x, occurred because of the
nitrogen pressure in the cover gas. Removal of fluorine and
the incorporation of nitrogen to those vacancies seized after
the Th/N/F molar ratio of ThNx/3F4-x chemical phases
reached 1:1:1, forming ThNF as the terminal product.
Further heating of ThNF up to 1100 °C under ammonia
and other atmospheres examined in this work resulted only in
the formation of low-stoichiometric thorium nitride fluorides
and an impurity phase ThO2 because of minor oxygen
pressure coming from the quartz tubing used in the experi-
ment at elevated temperatures. Experiments performed at
Figure 12. XRD powder refinement of the ThF4 sample after the ammonolysis at 800 °C for 1 h. The calculated pattern of ThN0.79F1.63 is highlighted.
Square root of counts was used as the scale of Y-axis for proper display of the calculated patterns.

Article
Inorganic Chemistry, Vol. 48, No. 13, 2009 5745
Figure 13. Calculated and observed XRD patterns of ThF4 after the ammonolysis at 800 °C 300 min.
Figure 14. XRD powder refinements of ThF4 after the ammonolysis at 800 °C for 600 min (a) and at 1000 °C for 75 min (b).
Table 4. Thermal Behavior of ThNF under Different Atmospheric Conditions
atmosphere
temperature/ °C
time/ min
product
secondary phases
ThO2
Ar
Ar-H2 (5%)
N2
N2-H2 (5%)
N2-H2 (5%)
1100
1100
1100
1100
1100
60
60
60
60
ThN0.79F1.63
ThO2
Th00.90F1.30
ThN0.90F1.30
ThO2
ThNF, ThO2
ThO2
120
different temperatures and heating times using same cover
gas (Figure 14) suggest this observation confirming that the
cover gases were high purity (99.999%) as purchased.
Although variation of the experimental conditions impacted
the stoichiometry of ThNF, no conditions yielded a pure
thorium nitride (ThNx) phase under these experimental

5746 Inorganic Chemistry, Vol. 48, No. 13, 2009
Silva et al.
conditions. Observation of all the thorium-nitrogen-fluor-
ides, however, inferred a possible reaction mechanism for the
formation of ThNF starting from ThO2 as shown below:
nanoscale. Nanostructure of ThNF was easily observed with
the HRTEM, and the crystal structure could also be con-
firmed. ThNF consisted of high-ordered nanostructure.
Therefore, even at elevated temperatures it was difficult to
change the chemical properties of the well-crystallized low-
energy stable chemical phase. This is likely to contribute to
the difficulty of removing further fluorine from ThNF to
produce ThNx. However, the overall thermodynamic favor-
ability of the oxide phase may drive other crystalline chemical
phases such as ThO2.
RT
ThO2 þ 4NH4HF2 ꢁꢁꢁꢁꢁꢁꢁꢁ!ðNH4Þ4ThF8
þ2H2O, viaðNH4Þ3ThF7
800°C,NH3
ðNH4Þ4ThF8 ꢁꢁꢁꢁꢁꢁꢁꢁ! ThF4 þ 4NH4F
5. Conclusion
800°C,NH3
ThF4 ꢁꢁꢁꢁꢁꢁꢁꢁ! ThN0:79F1:63
Solid-state reaction of ThO2 with NH4HF2 at room
temperature formed two products: (NH4)3ThF7 and
(NH4)4ThF8. The (NH4)3ThF7 compound was produced
quickly. The final product of the ammonolysis of
(NH4)4ThF8 at temperatures ranging from 800 to 1100 °C
was found to be ThNF. Attempting to further react the
material under NH3, N2, or N2/H2 (5%) to produce ThNx
chemical phases instead resulted in the formation of low-
stoichiometric ThNF chemical phases with a minor ThO2
phase. Formation of ThO2 was most likely due to the trace
oxygen in the experimental system, often formed at elevated
temperatures over common laboratory equipment materials
like the quartz used in this experiment. Thermal decomposi-
tion of (NH4)4ThF8 to ThNF occurred through a ThF4
intermediate and following a slow nitridization of ThF4
under NH3 through low-stoichiometric thorium-nitro-
gen-fluoride ternary chemical phases. Studies of the thermal
behavior of ThNF under different chemical environments
confirmed that the stoichiometric thorium nitrogen fluoride
(ThNF) is the preferable chemical composition of all thorium
nitride fluorides produced during the (NH4)4ThF8 ammono-
lysis above 800 °C.
Distinct morphologies of the starting materials and final
products of the reactions studied confirmed the chemical
transformations identified by XRD. Polycrystalline character-
istics were identified in the nanostructure of the (NH4)4ThF8
where as in ThNF, long-ranged single crystal areas were
observed over a wide range of the particles. Despite the
appearance of the larger particles shown by SE SEM,
HRTEM demonstrated the final ThNF product to be more
crystalline than its precursors, thus explaining the high stabi-
lity of ThNF toward substitution of fluorine with nitrogen.
800°C
,
NH3
800°C,NH3
ꢁꢁꢁꢁꢁꢁꢁꢁ! ThN0:9F1:3 ꢁꢁꢁꢁꢁꢁꢁꢁ! ThNF
SEM imaging showed that the morphology of the samples
involved in the reactions changed considerably. Plate-like
morphology of the starting ThO2 turned into an acicular
particle shape upon conversion to (NH4)4ThF8. The acicular
(NH4)4ThF8 particles became flat and agglomerated with
partially crystallized facets upon thermal decomposition to
ThF4. Particles of ThNF produced by the ammonolysis of
ThF4 consist of plate-like particles with porous surfaces. The
average particle size and shape of (NH4)4ThF8 shown in the
SEM image (Figure 3) are much different than that of the
starting ThO2, implying a difficulty in ThO2 conversion to
(NH4)4ThF8 at room temperature. Formation of ThF4, on
the other hand, was reasonable based on the high gap in the
crystalline character between (NH4)4ThF8 and ThNF ac-
cording to SEM imaging. Morphology of ThF4 is at an
intermediate position with respect to the change of acicular
shape of (NH4)ThF8 to a thin plate-like particle shape of
ThNF. The porous characteristics of the ThNF particleshave
relatively large surface areas. This increase is likely the
explanation for ThNF reactions with minute oxygen content
in the experimental setup. A longer exposure of ThNF to
minute quantities of oxygen at elevated temperatures (i.e.,
1100 °C) thus produces the ThO2.
TEM also showed some prominent differences in the
nanostrutures of (NH4)4ThF8 and ThNF. Bulk area of the
micro particles of (NH4)4ThF8 exhibited polycrystalline
characteristics with nanoscale crystalline patches. Some of
these patches had fringe details in two directions while others
showed fringe details either in one direction or none, showing
the bulk of the particle to be low in crystallinity. Even in areas
of high crystallinity, continuation of such crystallinity was
disturbed by other polycrystalline characteristics. Therefore,
the (NH4)4ThF8 with its low-ordered crystal structure is
vulnerable for attacks by other chemical agents and is easy
to decompose into other compounds.
Acknowledgment. The authors thank Dr. Anthony
Hechanova for administrating the UNLV Transmutation
Research Program under the financial support of the U.S.
Department of Energy (Grant DE-FG07-01AL67358).
We also thank Dr. Clay Crow from the Department of
Geoscience and Dr. Longzhou Ma at UNLV, for helpful
discussions and support. The authors are indebted to
Tom O’Dou and Trevor Low for laboratory management
and radiation safety.
In the case of ThNF, most of the microparticle areas
imaged showed well-oriented fringes with two-directional
details, implying a high crystallinity. The porous character-
istics identified from SEM imaging were also found at the