CALPHAD: Computer Coupling of Phase Diagrams and Thermochemistry 53 (2016) 97–104

Contents lists available at ScienceDirect

CALPHAD: Computer Coupling of Phase Diagrams and

Thermochemistry
journal homepage: www.elsevier.com/locate/calphad

Thermodynamics of the reciprocal NaCl–KCl–NaNO3–KNO3 system

D. Sergeev a,n, E. Yazhenskikh a, N. Talukder a, D. Kobertz a, K. Hack b, M. Müller a
a
Forschungszentrum Jülich GmbH, IEK-2, Jülich D-52425, Germany
b
GTT-Technologies, Kaiserstraße 100, D-52134 Herzogenrath, Germany

art ic l e i nf o a b s t r a c t

Article history: Thermodynamic properties of the reciprocal NaCl–KCl–NaNO3–KNO3 system are of interest for selecting
Received 4 January 2016 compositions, which can be used as phase change materials in thermal energy storage. In the presented
Received in revised form work two mixtures 7.5KCl–92.5NaNO3 and 12.5NaCl–87.5KNO3 on the diagonal sections of the reciprocal
8 April 2016
system with the liquidus temperatures of 549 K and 561 K, respectively were studied by DSC and DROP
Accepted 9 April 2016
calorimetry. The heat capacities and the enthalpy increments of the solid and the liquid phases were
Available online 19 April 2016
obtained for both mixtures (ΔH549 K ¼ 9.5 kJ/mol, ΔH514–549 K ¼ 16.9 kJ/mol, ΔH519–561 K ¼ 17.0 kJ/mol). The
Keywords: combination of these results allows the calculation of the phase transition enthalpies. DROP-calorimeter
Phase change materials was used in a mode of slow heating rate (0.5 K/min) thermal analysis, where the DROP-calorimetric
Potassium
detector was applied for the direct determination of phase transition enthalpies. The thermodynamic
Sodium
properties were calculated using our own dataset of the reciprocal system. The analysis of our experi-
Chloride
Nitrate mental and calculated results of enthalpy increments and heat capacities has confirmed that our dataset
Enthalpy increment can be used for prediction of thermodynamic properties of the full reciprocal system. Based on this
conclusion three additional compositions 50NaNO3–5.5KCl–44.5KNO3, 28NaNO3–12.4KCl–59.6KNO3 and
13.1NaNO3–11.2KCl–75.7KNO3 of the system with liquidus temperatures at 481 K, 510 K, and 549 K were
suggested as potential phase change materials (ΔH481 K ¼9.8 kJ/mol, ΔH496–510 K ¼ 12.7 kJ/mol and Δ
H510–549 K ¼15.8 kJ/mol).
& 2016 Elsevier Ltd. All rights reserved.

1. Introduction and 561 K, respectively, on the diagonal sections of the reciprocal

system have been selected. These mixtures are candidates for
Salt systems are of high interest for energy-saving technologies phase change materials in thermal energy storage. Figs. 1 and 2
as phase change materials for thermal energy storage (TES) [1–6]. show experimental and calculated results of the diagonal sections
The primary principles for TES are based on sensible and latent KCl–NaNO3 and NaCl–KNO3 with scaled-up part in the region of
heat of the material. These properties characterize the enthalpy 0–0.3 mol of KCl and NaCl, respectively. As it follows from these
increment (H°T – H°298.15), which can be determined by differential diagrams, the melting process of these mixtures proceeds in three
scanning (DSC) and drop calorimetry (DROP). Using various salt different transitions (transitions between phases II–III, III–IV, and
mixtures allows the designing of cascaded TES over different IV–V) close to the melting points. The enthalpies of these transi-
temperature ranges [7–9]. Suitable salts for TES are NaCl, KCl, tions can be interesting for designing of cascaded thermal energy
NaNO3, and KNO3 [1–3,10]. These cheap salt mixtures can cover storage. On one side the enthalpy increment of each composition
in the reciprocal system can be calculated by using thermo-
the very large temperature interval from 208 °C to 800 °C, which is
dynamic parameters published in [11], but on the other hand these
useful for concentrating solar power plants and industrial pro-
calculations should be verified with experimental determined
cesses. Hence the thermodynamic properties of mixtures includ-
ones. Therefore, the aim of this work was to study the two mix-
ing these salts have a practical as well as a fundamental
tures 7.5KCl–92.5NaNO3 and 12.5NaCl–87.5KNO3 by methods of
significance.
DSC and DROP.
Phase equilibria in the reciprocal NaCl–KCl–NaNO3–KNO3 sys-
tem have been studied experimentally and were calculated using a
new thermodynamic dataset including liquid and solid phases
2. Experimental
[11]. By means of these results, two mixtures 7.5KCl–92.5NaNO3
and 12.5NaCl–87.5KNO3 with the lowest melting points of 549 K 2.1. Samples

n
Corresponding author. The pure compounds NaCl and KCl (Alfa (AESAR), ultra dry,
E-mail address: [email protected] (D. Sergeev). 99.99%), NaNO3 and KNO3 (Sigma-Aldrich, 99.995%) were used for

http://dx.doi.org/10.1016/j.calphad.2016.04.001
0364-5916/& 2016 Elsevier Ltd. All rights reserved.
98 D. Sergeev et al. / CALPHAD: Computer Coupling of Phase Diagrams and Thermochemistry 53 (2016) 97–104

identical measurement conditions were performed for each sam-

ple to determine the heat capacity. The first run is the baseline to
determine the heat dissipation of both the empty crucible and the
sample holder set-up, the second run is the reference with a
sapphire slice sample of a mass comparable to that of the sample
mass, and the last one is the sample measurement. Each run was
repeated two or three times to achieve reproducibility of DSC
signals. The heat capacity ( C p° , J mol  1 K  1) of a sample was cal-
culated by the following equation:
nr DSCs − DSCb °
C p° (s) = ˙ ˙C p (r ) ,
ns DSCr − DSCb (1)

where subscripts s – sample, r – reference and b – baseline, n –

mole fraction of the substance (mol), DSC – signal of thermopile
(μV). C p° (r ) of sapphire was taken from the Netzsch database de-
livered with the software [12]. Before starting the measurements a
working volume of the DSC was evacuated and filled with He to
Fig. 1. Experimental and calculated data of the diagonal section KCl–NaNO3 of the
obtain a dry and oxygen free atmosphere and to improve the heat
reciprocal NaCl–KCl–NaNO3–KNO3 system from [11].
exchange. During experiments He flow was 15 ml/min. The mass
of the samples was about 45–60 mg. This amount was enough to
cover the bottom of the pans and to fill up about half of the vo-
lume. The standard temperature programme started with a stand-
by temperature of 45 °C, followed by a heating segment of
10 K min  1 up to the end temperature. A scanning rate of
10 K min  1 was chosen according to standard conditions for DSC
measurements. After the measurement, the temperature was
ramped back to stand-by at 45 °C without switching off the fur-
nace to ensure reproducible results. To check the stability of the
baseline after three runs (baseline–reference–sample) we have
used another three runs: sample without baseline–baseline–re-
ference. Phase changes of the pure compounds C6H5COOH (395 K),
RbNO3 (437 K), KClO4 (574 K), Ag2SO4 (699 K) and CsCl (749 K)
were used for temperature calibration. The average temperature
deviation was 72 K. The measurement precision for sapphire was
0.8%. More details are provided in Kobertz et al. [13].

2.2.2. Drop calorimeter (DROP)

A Calvet-type drop calorimeter, model mHTC 96 (Setaram) was
Fig. 2. Experimental and calculated data of the diagonal section NaCl–KNO3 of the
used. The thermopile in the furnace zone contains an alumina
reciprocal NaCl–KCl–NaNO3–KNO3 system from [11].
drop container surrounded by 34 type S thermocouples (Pt/
PtRh10%) in order to record a maximum of the heat flux.
The drop calorimeter was used in two different modes. The first
the preparation of the mixtures. The melting temperatures and one is the determination of the enthalpy increment (H°T – H°298.15)
DTA-curves of these compounds are given in our previous work at various temperatures (T) by the step-wise drop calorimetric
[11]. All handling with the samples was carried out in a glove box technique. In this case a piece of the sample and a piece of the
under purified argon. The 7.5 mol% KCl–92.5 mol% NaNO3 and the reference (α-Al2O3, NIST Standard Reference Material SRM 720,
12.5 mol% NaCl–87.5 mol% KNO3 mixtures were prepared directly purity 99.95%) within a similar mass of about 20–40 mg were
into the crucible according to the molar fraction with the accuracy dropped alternately from room temperature to the hot zone of the
of the weighed portion of 0.1 mg. The prepared mixtures with an DROP detector with a waiting time of about 10 min. It was re-
amount of 3 g were melted and kept for 2 h at 673 K under va- peated two or three times at each temperature. The atmosphere
cuum in sealed glass ampoules to achieve homogeneity. The inside of the chamber was a slow He flow (5 ml/min). The en-
homogeneity was confirmed by differential thermal analysis in thalpy increment (H°T – H°298.15)s of the sample was calculated by
[11] during heating–cooling cycles as well as in this work by DSC the following relationship:
and DROP measurements of the samples in sealed glass ampoules.
All measurements showed good reproducibility. After the heat mr ∆Hs
(HT° − H298.15
°
)s = ⋅ ⋅(HT° − H298.15
°
)r ,
treatment the mixtures were prepared in a glove box for use in ms ∆Hr (2)
DSC and DROP. where subscripts s – sample, r – reference, m – mass of substance
(g), ∆H – measured peak area ( μV˙s). The heat increment (H°T –
2.2. Instruments H°298.15)r of the reference was taken from the polynomial of Iv-
tanThermo database [14]. The samples prepared as described in
2.2.1. Differential scanning calorimeter (DSC) Section 2.1 were transported from the glove box in a desiccator
A differential scanning calorimeter (DSC 404C Pegasus and loaded immediately under a flow of He into the magazine of
(Netzsch)) with Pt–Rh oven (25°C–1600 °C) and type S thermo- the drop calorimeter.
couple (Pt/PtRh10%)) was used. The sample holder has a specific The second mode is the slow heating rate (0.5 K/min) thermal
design for precise heat capacity measurements. Three runs under analysis (SHRTA) with DROP-calorimetric sample holder. This
 D. Sergeev et al. / CALPHAD: Computer Coupling of Phase Diagrams and Thermochemistry 53 (2016) 97–104 99

3. Experimental results and discussion

3.1. Phase transition temperatures

To determine phase transition temperatures of the mixtures in

sealed glass ampoules the DSC was used with heating and cooling
scan rates of 2 and 5 K/min. Fig. 4 with a grey inset shows ex-
perimental data for the 7.5KCl–92.5NaNO3 mixture obtained with
a scan rate of 5 K/min. The grey picture shows the experimental
results by heating and cooling, where two phase transitions at
386 K and 549 K are identified. These temperatures are in a good
agreement with experimental data in [11] (Fig. 1). The first one at
386 K is related to the transition LT-HT of the (K,Na)NO3 solid
solution (Fig. 1, transition I–II) and the second one at 549 K to the
transition IV–liquid (Fig. 1). The calculated eutectic point is shifted
to 6% KCl at 558 K in comparison to the selected value of 7.5%
(Fig. 1). The white background picture in Fig. 4 is the scaled-up
part of the heating DSC respond in the region of 460–550 K. A
detailed analysis of the signal resulted in a continuous phase
transition beginning at 491 K. This transition is related to three
processes (Fig. 1, transitions between II–III–IV–liquid). As was
mentioned in [11] these peaks are superimposed and cannot be
separated by using a lower heating rate. By way of example, in this
study a heating rate of 2 K/min was also used, but in this case the
signal was lower and not representative as with 5 K/min.
Fig. 5 shows heating and cooling DSC-curves for the
12.5NaCl–87.5KNO3 mixture in a sealed glass ampoule at a scan
rate of 2 K/min. Six phase transition temperatures were found:
two at 390 K and 398 K representing the solid–solid transitions of
nitrates (Fig. 2, transition I–II); at 519 K and 540 K – (transitions II–
III, III–IV); at 553 K (continuous transition between III–IV–V); and
at 561 K (transition from phase V to the liquid). All phase transi-
tion temperatures were taken from the heating curve. The cooling
signals showed distinct peaks and the typical undercooling effects
(Fig. 5). All analogous transition peak areas (enthalpies) were the
Fig. 3. DROP-calorimetric detector [16] with sample for SHRTA measurements: 1 –
thermopile of 34 thermocouples, 2 – ceramic vessel, 3 – furnace thermocouple, 4 – same on heating and cooling. It shows that these transitions are
sealed ampoule with a sample, and 5 – reference. reversible and they were reproduced by repetitions.

3.2. Heat capacity

method was applied for the first time for the determination of the The determination of the heat capacity of the solid phase was
fusion enthalpy of the 50NaCl–50KCl mixture [15]. The advantage carried out in a platinum pan with a scan rate of 10 K/min by the
of this method compared to the differential thermal analysis (DTA) use of the DSC (Section 2.2.1). In this case, the ration method with
or the DSC lies in the construction of the DROP-calorimetric sapphire as a reference was applied. Experimental values of the
sample holder (Fig. 3), which consists of 34 thermocouples and heat capacities for the solid phase of the 7.5KCl–92.5NaNO3 and
allows measuring heat flow around a sample. It enables using a 12.5NaCl–87.5KNO3 mixtures are shown on Figs. 6 and 7. Seven
slow heating rate (about 0.5 K/min) and thus, to do measurements and eleven experimental series for each composition were
under conditions near to equilibrium. In this case the thermal ef-
fects of continuous reactions cannot be detected, but the transition
enthalpies at one temperature can be obtained. To increase the
signal of the thermal effects in this study samples with a weight of
about 1500 mg have been used, which were sealed in glass am-
poules under vacuum. To calibrate the temperature and sensitivity
of the instrument Pb (601 K), In (429 K), NaNO3 (576 K), KNO3
(608 K and 402 K), 50NaNO3 50KNO3 (495 K) were applied as
references. The average temperature deviation was 72 K. The
sensitivity coefficient was approximated with the following
equation:

KH = 0.370±0.019 T –18.573±5.259 ( R2 = 0.98962)

Fig. 4. Heating (red/upper) and cooling (blue/lower) DSC-curves of 7.5KCl–
where T – temperature (°C) and KH – sensitivity coefficient (μV/W). 92.5NaNO3 mixture at a scan rate of 5 K/min. Figure in grey inset shows the full
temperature range while the full figure presents the scaled-up part in the region of
Three cycles of heating and cooling were carried out for each 460–550 K. (For interpretation of the references to colour in this the reader is re-
sample. ferred to the web version of this article.)
100 D. Sergeev et al. / CALPHAD: Computer Coupling of Phase Diagrams and Thermochemistry 53 (2016) 97–104

Table 1
Coefficients of the linear equation of heat capacity (J mol  1 K  1) C°p ¼ATþB for the
7.5NaNO3  92.5KCl and 12.5NaCl–87.5KNO3 mixtures taken from DSC
measurements.

Temperature range (K) A 7 ΔAa B 7 ΔBa

7.5KCl–92.5 NaNO3
298–549 2.241  10  1 5.81  10  3 15.504 2.232
549–700 – – 111.67 5.63

12.5NaCl–87.5KNO3
298–398 0.338 2.80  10  3  15.258  0.954
398–561 0.114 4.29  10  3 56.878 1.007
561–670 0.162 2.24  10  3 27.089 1.4
Fig. 5. Heating (red/upper) and cooling (blue/lower) DSC-curves of 12.5NaCl–
87.5KNO3 mixture at a scan rate of 2 K/min. (For interpretation of the references to a
Standard uncertainties are obtained taking into account the statistical scatter
colour in this the reader is referred to the web version of this article.).
of the experimental data and are u(C°p) ¼ ΔAT þ ΔB J mol  1 K  1.

Fig. 6. Heat capacity Cp° of solid and liquid phases of the 7.5KCl–92.5NaNO3 mixture
determined in a platinum pan and a glass container.
Fig. 8. Enthalpy changes deduced by DSC and DROP calorimeter of the 7.5KCl–
92.5NaNO3 mixture in comparison with calculations [11].

Fig. 7. Heat capacity Cp° of solid and liquid phases of the 12.5NaCl–87.5KNO3
mixture determined in a platinum pan and a glass container.

performed, respectively. These results are in a good agreement Fig. 9. Enthalpy changes deduced by DSC and DROP calorimeter of the 12.5NaCl–
87.5KNO3 mixture in comparison with calculations [11].
with data on Figs. 4 and 5. It should be noted that the sequence of
the processes II–III–IV–liquid (Fig. 1) of the 7.5KCl–92.5NaNO3
mixture started at 491 K according to Fig. 4 and the DSC-curve has transition enthalpies.
a maximum at 514 K (Fig. 6, experimental data in platinum pan). The determination of the heat capacities of the liquid phase
This temperature was chosen for further calculations of phase was carried out in a glass container with a scan rate of 10 K/min. In
 D. Sergeev et al. / CALPHAD: Computer Coupling of Phase Diagrams and Thermochemistry 53 (2016) 97–104 101

Table 2 Table 4
Enthalpy increment H°T–H°298.15 of the 7.5KCl–92.5NaNO3 and 12.5NaCl–87.5KNO3 Phase transition temperatures (T) and corresponding transition enthalpies (ΔtrH) of
mixtures taken from DROP measurements. the 7.5KCl–92.5NaNO3 and 12.5NaCl–87.5KNO3 mixtures calculated using the
thermodynamic dataset [11] and determined with different experimental techni-
7.5KCl–92.5NaNO3 12.5NaCl–87.5KNO3 ques DSC þDROP and SHRTA.

T (K) H°T – H°298.15a (kJ mol  1) T (K) H°T – H°298.15a (kJ mol  1) Phase transition temperature T (K) ΔtrHa (kJ mol  1)

388 9.2 394 11.1 Calculations DSCþ DROP SHRTA

402 11.0 409 15.6
430 14.8 420 16.7 7.5KCl–92.5NaNO3
451 17.2 430 16.8 386 0.44 1.0e 0.3
472 20.8 440 20.0 514 – 4.3 –
482 21.9 445 20.2 549 – 12.1 9.5
492 24.7 455 20.7
503 25.7 469 21.9
513 28.7 482 24.4 12.5NaCl–87.5KNO3
523 30.9 492 26.2 391–398 4.6 3.6 3.9
534 32.8 502 26.9 519 2.3d 11.1b 4.1c 12.2 1.7 4.5
544 37.0 514 32.9 540 – 3.8 1.5
552 41.1 518 29.0 553 – 4.3 0.9
564 47.6 522 33.2 561 – 0.4
585 48.5 523 33.5
a
605 52.4 533 36.5 Relative standard uncertainty of the experimental data ur(ΔtrH)¼ 7 10%.
b
625 55.4 543 40.5 Without Cp contribution 6.2 kJ/mol from 508 to 561 K.
c
645 57.7 553 45.0 Taken from DROP data instead of 6.0 kJ/mol (see Fig. 7).
d
665 60.4 574 47.2 Transition temperature is 508 K.
e
665 58.5 584 49.2 Standard uncertainty 7 0.4 kJ/mol.
604 50.6
614 52.5
624 53.5 Figs. 6 and 7 also show the calculation results performed by
645 56.7
FactSage using our new database as published in [11]. These data
a
Standard uncertainty u(H°T – H°298.15) ¼ 71.0 kJ/mol. were calculated according to the following equation from [16]:
⎛ ∂H ⎞ ∂nφ φ ∂Hmφ

this case the study of liquid phase in a platinum pan was not Cp=⎜ ⎟ =∑ Hm+ ∑ nφ ,
⎝ ∂T ⎠ p, n ∂T ∂T (3)
i φ φ
possible because of high diffusion and evaporation rates of the
φ
samples. In sealed glass containers these effects are eliminated, where φ is phase index, and n is phase fraction of each phase. The
but there is another challenge that the glass containers of the calculated values of the heat capacities are in a good agreement
reference and the sample cannot be identical, which results in a with a deviation of 20.83 J mol  1 K  1 for the liquid phase of the
higher error in the calculation of the heat capacity. To avoid this 7.5KCl–92.5NaNO3 composition and about 12 J mol  1 K  1 for the
problem the technique suggested in [15] was used, where the 12.5NaCl–87.5KNO3 mixture in the region of 400–500 K. Differ-
values of the heat capacity obtained in the glass container were ences are observed for the phase transition ranges. The calculated
corrected relative to the results obtained in the platinum pan for heat capacity of the 7.5KCl–92.5NaNO3 mixture shows a con-
the solid phase. The heat capacity of the liquid and solid phases tinuous solid solution transition II–III (Fig. 1) in the region from
determined by the DSC is shown in Figs. 6 and 7, respectively. 430 to 500 K. In the C p° measurement this transition is shifted to
These results are in a good agreement except for the peaks of the higher temperature 491 K. The solid–liquid transition of the cal-
solid phase transition for both mixtures. In the glass container culated C p° is in the region of 525–555 K. This peak is higher cor-
these peaks were higher. responding to the larger enthalpy of fusion. In the case of the
The heat capacities of the 7.5KCl–92.5NaNO3 and 12.5NaCl–87.5KNO3 mixture the phase transition temperatures are
12.5NaCl–87.5KNO3 mixtures were linearly approximated for dif- in a good agreement with experimental data (Fig. 7). Some mis-
ferent temperature ranges given in Figs. 6 and 7. In this case the matching of transition enthalpies (area of peaks) is observed in the
heat capacity is separated from the contribution of the phase region of 390–400 K the calculated data show no peaks and in the
transitions enthalpies. The coefficients of linear equations ob- region of 500–560 K there is a contradiction of values of the phase
tained are given in Table 1. transitions (Fig. 7). This difference between the measured and

Table 3
Coefficients of the equation H°T–H°298.15 ¼A  T þB  T2 þ C (kJ mol  1) for the enthalpy increment of the 7.5KCl–92.5NaNO3 and 12.5NaCl–87.5KNO3 mixtures calculated from
DROP measurements.

Temperature range /K A 7 ΔAa B 7 ΔBa C 7 ΔCa

7.5KCl–92.5NaNO3
388–514  0.20703  0.0971 3.97  10  4 1.08  10  4 29.8938 21.71271
514–549  5.58718  1.37422 5.54  10  3 1.29  10  3 1437.868 365.717
549–665 0.1253 0.0097 – –  23.5773  6.04666

12.5NaCl–87.5KNO3
409–518 0.12566 0.0055 – –  36.15266  2.53974
514–553  6.40652  1.28119 6.31  10  3 1.2  10  3 1658.898 341.7555
553–645 0.12511 0.00546 – –  24.35977  3.27538

a
Standard uncertainties are obtained taking into account the statistical straggling of the experimental data and are u(H°T–H°298.15) ¼ΔA  T þ ΔB  T 2 þ ΔC kJ mol  1.
102 D. Sergeev et al. / CALPHAD: Computer Coupling of Phase Diagrams and Thermochemistry 53 (2016) 97–104

Fig. 10. Heating (red/lower) and cooling (blue/upper) curves of the slow heating rate (0.5 K/min) thermal analysis with DROP-calorimetric sample holder for NaNO3,
50NaNO3–50KNO3, 12.5NaCl–87.5KNO3, and 7.5KCl–92.5NaNO3. (For interpretation of the references to colour in this the reader is referred to the web version of this article.)

Table 5
Liquidus temperatures (T), temperature intervals of phase transition (ΔT), and
corresponding enthalpy increments (ΔH) of potential compositions of the re-
ciprocal NaCl–KCl–NaNO3–KNO3 system for using as phase change materials.

T (K) Composition ΔT ΔH (kJ mol  1)

Calculations with the dataset [11]

481 50NaNO3–5.5KCl–44.5KNO3 1 9.8
510 28NaNO3–12.4KCl–59.6KNO3 14 12.7
549 13.1NaNO3–11.2KCl–75.7KNO3 39 15.8
Experiment
549 92.5NaNO3–7.5KCl 1 9.5
35 16.9
561 87.5KNO3–12.5NaCl 42 17.0

calculated values can be explained by the fact that the heat ca-
pacity is calculated as a differential of the enthalpy (Eq. (3)) under
equilibrium conditions. In case of the experiments however some
kinetic effects occur under quasi-equilibrium. Due to the differ-
ence between the techniques (calculations and experiments)
complete agreement concerning phase transformations is difficult
to achieve. It is better to consider the enthalpy itself as it can be
seen in Figs. 8 and 9 in the following section.
Fig. 11. Calculated phase diagram of the reciprocal NaCl–KCl–NaNO3–KNO3 system
3.3. Enthalpy increments and phase transition enthalpies [11], where red dots show the chosen compositions (Table 5). (For interpretation of
the references to colour in this figure legend, the reader is referred to the web
The enthalpy increment (H°T – H°298.15) at various temperatures version of this article.)
(T) was determined by the step-wise DROP-calorimetric technique
(described in Section 2.2.2). The 7.5KCl–92.5NaNO3 and the fitted coefficients of these values are given in Table 3.
12.5NaCl–87.5KNO3 mixtures were studied in the temperature The combination of the DROP data with the DSC results
ranges 388–665 K and 394–645 K, respectively. The experimental (DSCþ DROP) allows for the calculation of phase transition en-
results are given in Table 2 and presented in Figs. 8 and 9. The thalpies of the studied samples under the assumption that these
 D. Sergeev et al. / CALPHAD: Computer Coupling of Phase Diagrams and Thermochemistry 53 (2016) 97–104 103

allows the conclusion that the larger part of the transition enthalpy
of the 7.5KCl–92.5NaNO3 mixture can be attributed to the transition
temperature at 549 K. The values of the transition enthalpies were
calculated from the integral area of the peaks (Fig. 10, yellow fields)
and they are given in Table 4. The comparison of the transition
enthalpy ΔtrH549(7.5KCl–92.5NaNO3) determined by the SHRTA and
the DSCþDROP value gives a difference of 2.6 kJ/mol. This differ-
ence can be related to the enthalpy of the continuous transitions IV–
V–liquid (Fig. 1) in the temperature region from 514 to 549 K. The
transition enthalpies of the 12.5NaCl–87.5KNO3 mixture were cal-
culated from cooling curves (Fig. 10 and Table 4). In this case the
difference between the DSCþ DROP and the SHRTA values is 7.7 kJ/
mol, which can also be related to the enthalpy of the continuous
transitions III–IV–V–liquid in the temperature range from 519 to
561 K (Fig. 2). The calculated values of ΔtrH are also given in Table 4.

3.4. Potential phase change materials

The thermodynamic properties obtained allow for the conclu-

Fig. 12. Calculated enthalpy increment of the chosen compositions (Fig. 11) with sion that the 7.5KCl–92.5NaNO3 mixture is most suitable for use as
liquidus temperatures at 481 K, 510 K, and 549 K of the reciprocal system phase change material (PCM) at the same conditions, because
NaCl–KCl–NaNO3–KNO3 [11].
of a narrow temperature interval and a large enthalpy increment
(Table 5). The comparison of experimental and calculated data for
transitions are not continuous and proceed at one temperature. both mixtures (7.5KCl–92.5NaNO3 and 12.5NaCl–87.5KNO3) shows
The coefficients of the heat capacity from Table 1 were used for the very good agreement of the enthalpy increment. Based on this
calculation of the heat increment ending at each phase transition result it can be concluded that the thermodynamic dataset [11]
temperature. The phase transition enthalpies were determined as allows for calculating of thermodynamic properties of the re-
a difference between heat increments calculated from the DSC and ciprocal NaCl–KCl–NaNO3–KNO3 system with a low uncertainty.
the DROP approximation (Figs. 8 and 9 and Table 4). In this case Therefore, three more compositions of this system (Fig. 11 and
the contributions of the heat capacity and the phase transition Table 5) with liquidus temperatures at 481 K, 510 K, and 549 K
enthalpy are separated from the heat increment, but it should be have been chosen, which can be used as PCM. Calculated enthalpy
taken into account that the values of the phase transition en- increments are plotted in Fig. 12. It can be seen that the tem-
thalpies are valid for temperature ranges, not only for one tem- perature intervals of the phase transitions (ΔT) of the composition
perature. It should also be noted that for the 12.5NaCl–87.5KNO3 with the melting point 481 K is equal 1 K and for T3 ¼549 K is 39 K.
composition the enthalpy of transition at 553 K and 561 K cannot These values are given in Table 5 with corresponding enthalpy
be separated because of the narrow temperature interval and the increments (ΔH), which allows for selecting suitable compositions
small values of the enthalpies. for cascaded thermal energy storage. It should be noted that the
The comparison of the DSC þDROP with the DROP approx- melting temperature at 481 K1 of the 50NaNO3-5.5KCl-44.5KNO3
imation lines (Figs. 8 and 9) shows very good agreement for the mixture (ΔtrH481 ¼9.8 kJ/mol) is the lowest in the reciprocal sys-
liquid phase of the 12.5NaCl–87.5KNO3 mixture and small devia- tem and this mixture can compete with the most useful “solar salt”
tions for the temperature region of 398–519 K (2.0 kJ/mol at 50NaNO3-50KNO3 with ΔtrH495 ¼9.5 kJ/mol [17].
519 K). There are also small deviations for the liquid phase of the
7.5KCl–92.5KNO3 mixture (1.5 kJ/mol at 660 K). It means that the
heat capacity of these regions should be higher and should cor- 4. Conclusions
respond to the A coefficients from Table 3, that is 125.66 and
125.3 J mol  1 K  1, respectively. Two mixtures 7.5KCl–92.5NaNO3 and 12.5NaCl–87.5KNO3 were
Figs. 8 and 9 also show the calculation results, which go in line studied by the DSC and the DROP calorimetrical methods in the
with the DROP measurements. A deviation of about 2.3 kJ/mol is temperature regions 298–700 K and 298–670 K, respectively. The
observed only for the liquid phase of the 7.5KCl–92.5KNO3 mix- heat capacities of the solid and liquid phases were determined for
ture, which can be related to higher enthalpy of transition. Tran-
both mixtures on the assumption of a linear approximation. A
sition enthalpies, calculated and deduced by the DSC and DROP are
drop calorimeter was used in two modes: (1) the step-wise drop
given in Table 4.
calorimetric technique and (2) the slow heating rate (0.5 K/min)
The slow heating rate (0.5 K/min) thermal analysis (SHRTA) was
thermal analysis (SHRTA) with the DROP-calorimetric sample
applied for direct determination of phase transition enthalpies
holder. The first mode was applied for the determination of the
(Section 2.2.2). Fig. 10 shows some examples of the experimental
enthalpy increment of the solid and liquid phases. These data are
data for NaNO3, 50NaNO3–50KNO3, 12.5NaCl–87.5KNO3, and
in a good agreement with the calculated results. The combination
7.5KCl–92.5NaNO3, where NaNO3 and 50NaNO3–50KNO3 were used
of the heat capacities and the enthalpy increments allows calcu-
as reference measurements. Transition temperatures of the
lating the phase transition enthalpies. The second mode of the
12.5NaCl–87.5KNO3 and the 7.5KCl–92.5NaNO3 mixtures are in good
DROP calorimeter was also used for the determination of ΔtrH, but
agreement with the DSC results (Figs. 4 and 5). The peak shapes of
in this case thermal effects were measured without contributions
the references are quite identical on heating and cooling. The width
of continuous reactions. The thermodynamic properties obtained
of these peaks is about 10 K, which corresponds to a process
allow for the conclusion that the 7.5KCl–92.5NaNO3 mixture is
duration of about 20 min. In the case of the 7.5KCl–92.5NaNO3
most suitable for use as phase change material (PCM), because of a
mixture the cooling peak is sharper with a width of about 30 K
(60 min), but the larger thermal effect was observed in the region of
the first 10 K. Under consideration of kinetics these thermal effects 1
This value was confirmed by the differential thermal analysis: 482 72 K.
104 D. Sergeev et al. / CALPHAD: Computer Coupling of Phase Diagrams and Thermochemistry 53 (2016) 97–104

narrow temperature interval and a large enthalpy increment. The techniques, Renew. Sust. Energy Rev. 27 (2013) 724–737.
analysis of our experimental results and calculated values has [4] B. Zalba, J.M. Marı ́n, L.F. Cabeza, H. Mehling, Review on thermal energy storage
with phase change: materials, heat transfer analysis and applications, Appl.
confirmed that our dataset can be used for prediction of thermo- Therm. Eng. 23 (2003) 251–283.
dynamic properties of the full reciprocal system. Based on this [5] T. Bauer, D. Laing, R. Tamme, Overview of PCMs for concentrated solar power
conclusion three additional compositions of the system with li- in the temperature range 200 to 350 °C, Adv. Sci. Technol. 74 (2010) 272–277.
quidus temperatures at 481 K, 510 K, and 549 K were suggested as [6] R. Tamme, T. Bauer, J. Buschle, D. Laing, H. Müller-Steinhagen, W.-
D. Steinmann, Latent heat storage above 120 °C for applications in the in-
potential phase change materials. dustrial process heat sector and solar power generation, Int. J. Energy Res. 32
(2008) 264–271.
[7] F. Dinter, M. Geyer, R. Tamme, Thermal Energy Storage for Commercial Ap-
Acknowledgements plications, Springer-Verlag, Berlin, Heidelberg, New York, USA, 1991.
[8] H. Michels, R. Pitz-Paal, Cascaded latent heat storage for parabolic trough solar
power plants, Sol. Energy 81 (2007) 829–837.
The authors would like to thank the Alexander von Humboldt [9] R. Muren, D.A. Arias, B. Luptowski, Performance based cost modeling of phase
Foundation for supporting this research. change thermal energy storage for high temperature concentrating solar
power systems, in: Proceedings of the ASME 2009 International Mechanical
Engineering Congress & Exposition, Lake Buena Vista, FL, 2009, pp. 193–202.
[10] Q. Peng, J. Ding, X. Wei, J. Yang, X. Yang, The preparation and properties of
Appendix A. Supplementary material multi-component molten salts, Appl. Energy 87 (2010) 2812–2817.
[11] D. Sergeev, E. Yazhenskikh, D. Kobertz, K. Hack, M. Müller, Phase equilibria in
the reciprocal NaCl–KCl–NaNO3–KNO3 system, Calphad 51 (2015) 111–124.
Supplementary data associated with this article can be found in the [12] S. D.A.I. Ditmars, S.S. Chang, G. Bernstein, E.D. West, Enthalpy and heat-ca-
online version at http://dx.doi.org/10.1016/j.calphad.2016.04.001. pacity standard reference material: synthetic sapphire (α-Al2O3) From 10 to
2250 K, J. Res. Nat. Bur. Stand. 82 (1982) 159–163.
[13] D. Kobertz, M. Müller, A. Molak, Vaporization and caloric studies on lead ti-
tanate, Calphad 46 (2014) 62–79.
References [14] G.V. Belov, V.S. Iorish, V.S. Yungman, Ivtanthermo for windows - database on
thermodynamic properties and related software, Calphad 23 (1999) 173–180.
[1] Q. Peng, X. Yang, J. Ding, X. Wei, J. Yang, Design of new molten salt thermal [15] D. Sergeev, D. Kobertz, M. Müller, Thermodynamics of the NaCl–KCl system,
energy storage material for solar thermal power plant, Appl. Energy 112 (2013) Thermochim. Acta 606 (2015) 25–33.
682–689. [16] K. Hack, M.H.G. Jacobs, II.23-Heat balances and Cp calculations, in: K. Hack
[2] M.M. Kenisarin, High-temperature phase change materials for thermal energy (Ed.), The SGTE Casebook, second edition,Woodhead Publishing, 2008,
storage, Renew. Sust. Energy Rev. 14 (2010) 955–970. pp. 273–281.
[3] B. Cárdenas, N. León, High temperature latent heat thermal energy storage: [17] M. Kamimoto, Thermodynamic properties of 50 mol% NaNO3  50% KNO3
phase change materials, design considerations and performance enhancement (HTS2), Thermochim. Acta 49 (1981) 319–331.