(259b) A Review of IR Spectroscopic Studies of Molten Fluoride Salts

Derdeyn, W. B., University of Wisconsin Madison
Scarlat, R., University of Wisconsin Madison


Fluoride salt-cooled high temperature nuclear reactors (FHRs)
and molten salt nuclear reactors (MSRs) have potential advances in terms of
their compact size, high thermal efficiency, modular design, and passive safety
features that project them to have economic advantages in the future
electricity market [1].  These factors are largely enabled by the unique
properties of the proposed molten salt coolants.  Molten fluoride salt
mixtures have high thermal conductivity and heat capacity, low viscosity, low
vapor pressure, high boiling point, and chemical stability under  radiation fields [2].  In particular, the
2LiF:BeF2 mixture, known as flibe, is
being actively investigated as a coolant for FHRs, and
as a coolant and fuel solvent for MSRs.

Infrared (IR) spectroscopic
measurements have been used to investigate molten
fluoride salts to reveal the presence of complexes or covalent
structures.  This data can be useful in answering questions about
speciation of impurities such as corrosion, fission, and neutron activation
products, and in making predictions about transport and reaction mechanisms and
kinetics.  In addition, IR spectrum data can be particularly valuable as
inputs into modeling radiative heat transfer. Radiative heat transfer in
semi-transparent (i.e. participating)  media is
important to understand at the high operating temperature of FHRs and MSRs. The
spectroscopic study of molten salts is quite challenging in practice,
specifically for fluoride salts, considering their high melting points and
degree of corrosive attack of window materials. Such characteristics pose a
constraint for the design of melt holder materials, and window selection.

This article reviews IR studies
of fluoride melts in general; the experimental methods employed and the data
obtained are discussed. Currently, no IR spectrum data
for flibe is reported in
literature, though there have been Raman studies.  The results of recent
molecular dynamics (MD) models of flibe are reported, which will be instrumental in guiding future
IR experimental measurements.  The goal of the paper is to outline the
state of the art of IR spectroscopy in fluoride salts studies and discusses
whether these studies can be used to study covalent
structures and radiative properties of flibe

2.1 Review of Experimental

Wilmshurst and Senderoff developed a method in which they reflected a
monochromatic beam of IR light of varying wavelength from the surface of a
molten salt sample and measured the reflected light’s intensity (Figure 1)[3].  Through the relation of the complex reflectivity
and complex coefficient of refraction, and applying the Kramers-Kronig
(KK) technique, the absorption coefficient and refractive index were then obtained for the melt [4].  The absorption
coefficient was the desired result so that the peak locations of observed
spectra could be correlated to vibrational modes of
hypothesized melt species. 

Fordyce and Baum used an
identical optical arrangement and their measurements relied on the same
principles and analysis as above, but they developed a different optical cell
due to the high vapor pressure of their melts (Figure 2)[5]. 
The melt was contained in a nickel crucible, which was then contained in an
Inconel vessel that had two windows made of KBr or CsI.  The windows were angled
such that the incident IR light would hit the melt and reflect out of the cell
towards the detector. In addition, they employed an Ar
flushing system and Teflon wipers to minimize the vapor deposits on the

Barker developed a windowless cell which was capable of measuring both reflectivity and
absorptivity, depending on the opacity of the sample [6].  The optical
setup consisted of a pair of plane mirrors on either side of a central
spherical mirror, which collected light reflected from the optical cell. 
For opaque liquids, the incident light would hit the melt surface, and just the
reflected component would be detected, enabling
reflectivity measurements.  For semi-transparent liquids
both the transmitted and reflected components could be detected due to a
submerged Pt mirror.  Through separate measurements of the melt
reflectivity and depth, plus calculation of the index of refraction, the
absorption coefficient could then be obtained.

2.2 Review of Experimental

Using the previously mentioned
reflectance setup [3], Wilmshurst collected data on mixtures of LiF, KF, NaF, and ZrF4
[7].  An interesting result was observed when ZrF4 was added to
a NaF:KF mixture (Figure 3).
At high concentration of ZrF4, there was a single large peak at 520
cm-1, but as the concentration decreased, a second peak appeared
around 480 cm-1.  The authors argued that in pure form ZrF4
will form a quasi-lattice structure, but that addition of Na+
and K+ cause a breakdown in the lattice and formation of a complex
ion, such as ZrF62-.


Fordyce used an adaptation of
Wilmshurst’s reflectance cell to obtain IR spectra of Ta(V)
dissolved in binary mixtures of KF, NaF, and LiF [5]. First, spectra of Ta(V)
were obtained at varying concentration in KF:LiF
(Figure 4).  Next, each cation was replaced
with Na+ and the procedure was repeated for the resultant melts
(Figure 5).
The observed spectra were compared with IR spectra of solid, crystalline K2TaF7,
CsTaF6, and TaF5.  It was found
that with K+ present, the melt spectra approximated that of K2TaF7,
and without that of CsTaF6, which are known to consist of TaF72-
and TaF6- complexes, respectively.

Mead measured reflection and
emission IR spectra of LiF at 30˚ C above and
below the melting point (1143˚ C) using the setup of Barker [8]. 
Using classical oscillator (CO) theory, the optical constants of the melt were
calculated and compared to the experimental data.  The CO-calculated
reflection spectra agreed well with the measured. The KK-derived absorption
data varied  compared to the previous study by
Wilmshurst, and this was attributed to extrapolation errors in the latter’s KK

2.3 Review of Modeling Data

Dai et al. investigated the IR
spectrum of flibe using ab initio molecular dynamics
(AIMD) based on density functional theory (DFT)[9]. 
The aim was to compare this method with the results obtained by Salanne et al., who used a polarizable ionic interaction
potential model [10]. Vibrational spectra were obtained
via both AIMD and normal mode analysis (NMA) for the following local structures
in the gas phase at 30 K: BeF­42-, Be2F73-,
and Be3F104-.  The AIMD method was used again to calculate the vibrational spectra of Li2BeF4
at 30 K and 873 K (Figure 6).  The high temperature spectrum displayed a
lot of noise, so a low temperature spectrum was useful to identify distinct
bands.  The spectrum at 30 K showed similar features with the individual
gas phase results, which suggested the existence of these structures in the
liquid phase (Figure 7). 


This review has demonstrated how
different groups of researchers have successfully studied fluoride melts via IR
spectroscopy.  The difficulty of these endeavors has
been emphasized, with regards especially to melt containment and
background thermal radiation.  However, the successful studies have shown
to be quite powerful in their identification of both dissolved melt species and
the bulk species complex structure.  In addition,
with recent modeling advances, there is potential for more
theoretical studies based on MD and DFT to support experimental results. 
Therefore, IR spectroscopy appears to be a powerful and effective tool to
address current research questions regarding the complex melt structure and
radiative heat transfer parameters.


[1]      C. Andreades, R. O. Scarlat, et al., ‘Reheat-Air Brayton
Combined Cycle Power Conversion Design and Performance Under Nominal Ambient
Conditions’, J. Eng. Gas Turbines Power, vol. 136, no. 6, p. 62001,

[2]      P. Sabharwall, M. Ebner, et al.,
‘Molten salts for high temperature reactors: University of Wisconsin molten
salt corrosion and flow loop experiments – Issues identified and path forward’,
no. March, p. 33, (2010).

[3]      J. K. Wilmshurst,
‘Infrared Spectra of Highly Associated Liquids and the Question of Complex Ions
in Fused Salts’, J. Chem. Phys., vol. 39, no. 7, pp. 1779–1788,

[4]      T. S. Robinson
and W. C. Price, ‘The Determination of Infra-Red Absorption Spectra from
Reflection Measurements’, Proc. Phys. Soc. Sect. B, vol. 66, no.
11, pp. 969–974, (2002).

[5]      J. S. Fordyce and
R. L. Baum, ‘Infrared-Reflection Spectra of Molten Fluoride Solutions: Tantalum
(V) in Alkali Fluorides’, J. Chem. Phys., vol. 44, no. 3, pp.
1159–1165, (1966).

[6]      A. J. Barker, ‘A
Compact, Windowless Reflectance Furnace for Infrared Studies of Corrosive
Melts’, J. Phys. E., vol. 6, no. 3, pp. 241–244, (1973).

[7]      J. K. Wilmshurst,
‘Infrared Spectra of Molten Salts’, J. Chem. Phys., vol. 39, no.
10, pp. 2545–2548, (1963).

[8]      D. Mead,
‘Comparison of the Optical and Dielectric Properties of Crystalline and Molten
Lithium Fluoride’, J. Phys. C Solid State Phys., vol. 7, pp.
445–453, (1973).

[9]      J. Dai, H. Han,
et al., ‘First-principle investigation of the structure and vibrational spectra
of the local structures in LiF-BeF2 Molten Salts’, J. Mol. Liq., vol.
, pp. 17–22, (2016).

[10]    M. Salanne,
C. Simon, et al., ‘A first-principles description of liquid BeF2 and its
mixtures with LiF: 2. Network formation in LiF-BeF2’,
J. Phys. Chem. B, vol. 110, no. 23, pp. 11461–11467, (2006).



This paper has an Extended Abstract file available; you must purchase the conference proceedings to access it.


Do you already own this?



AIChE Members $150.00
AIChE Graduate Student Members Free
AIChE Undergraduate Student Members Free
Non-Members $225.00