- Email: [email protected]
Nanoparticles as a high-temperature anticorrosion additive to molten nitrate salts for concentrated solar power
Solar Energy Materials and Solar Cells 203 (2019) 110171
Contents lists available at ScienceDirect
Solar Energy Materials and Solar Cells journal homepage: www.elsevier.com/locate/solmat
Nanoparticles as a high-temperature anticorrosion additive to molten nitrate salts for concentrated solar power
T
Udayashankar Nithiyananthama,b, Yaroslav Grosua,∗, Argyrios Anagnostopoulosc, Enrique Carbó-Argibayd, Oleksandr Bondarchukd, Luis González-Fernándeza, Abdelali Zakia, Josu Mirena Igartuab, María Elena Navarroc, Yulong Dingc, Abdessamad Faika,∗∗ a
CIC Energigune, Albert Einstein 48, Miñano, Álava, 01510, Spain Applied Physics II Department, Faculty of Science and Technology, University of the Basque Country (UPV/EHU), PO Box 644, Bilbao, 48080, Spain c BCES Birmingham Centre of Energy Storage, University of Birmingham, United Kingdom d International Iberian Nanotechnology Laboratory, Av. Mestre José Veiga, s/n, 4715-330, Braga, Portugal b
ARTICLE INFO
ABSTRACT
Keywords: Corrosion Molten salt Thermal energy storage Concentrated solar power
Hot corrosion is profoundly detrimental to construction elements in Concentrated Solar Power (CSP) plants affecting their lifetime, running costs and safety. In this work we have studied the anticorrosion effect of TiO2 nanoparticles additives on carbon steel using XRD, XPS with depth profiling, EDX and FIB/SEM techniques. The results revealed that the addition of 1 wt% of TiO2 nanoparticles to molten binary nitrate salt reduces the corrosion rate of carbon steel more than twice and stabilizes the corrosion scale at 390 °C. The anticorrosion effect of TiO2 nanoadditive was attributed to the formation of iron-titanium mixed oxide on the carbon steel surface. It was confirmed by XRD and TGA techniques that addition of TiO2 nanoparticles does not alter the stability of the salt. In view of presented results, the feasibility of molten salts based nanofluids in the CSPs can be reconsidered in terms of improved compatibility with construction materials.
1. Introduction Reaching necessary dispatchability is a major challenge for renewable technologies, which is typically addressed through efficient energy storage. Concentrated solar power (CSP) technology [1] possesses several important advantages in this sense, as it allows for reliable, lowcost and relatively simple thermal energy storage (TES) system [2,3] instead of more expensive and complex chemical or mechanical storage systems [4]. There is a great deal of effort from scientific and engineering community for further improvement of TES technologies at CSP plants by using two-tank or single-tank thermocline configurations [5], where different storage materials and heat transfer fluids (HTF), like molten salts, thermal oils, air, phase change materials, etc., can be used [6]. Corrosion issues [7], particularly with most common storage material NaNO3–KNO3 molten salt [8], are considered important not only in terms of safety and power plant reliability, but also in terms of the electricity cost reduction [9,10]. With these regards, a rapidly growing number of works dedicated to the corrosion of molten salts are evident in the literature [11–47].
∗
Nitrate salts are often the object of investigation due to the high level of maturity of this storage material. HitecXL salt (48 wt% Ca(NO3)2–7wt% NaNO3–45 wt% KNO3) was studied in terms of corrosion with AISI 304, 430 stainless steels, low-Cr alloy steel T22 [11] and with carbon steel A516Gr70, stainless steels AISI 304, AISI 316 [24]. The increase in corrosion rates due to the chlorides in molten nitrate salts was demonstrated for Hitec salt (53 wt% KNO3–40 wt% NaNO2–7wt% NaNO3) [12], Solar salt (60 wt% NaNO3 – 40 wt% KNO3) [13] and KNO3–NaNO2–NaNO3–KCl salt [14]. On the contrary, the positive effect of chromium in steels on the compatibility between different molten nitrate salts and construction materials has been reported [16,21]. The corrosion rates generally increase with temperature as was demonstrated for stainless steels SS-304 and SS-316 at 570 °C and for carbon steel A36 at 316 °C with several nitrate salts mixtures [17]. The same negative effect of temperature stands for Solar salt, which was tested at 550 °C [20], 650 °C [21], 680 °C [25] and for Hitec salt tested at 450, 600 and 680 °C [26]. It should be noted that corrosion rate is often nonlinearly depends on temperature, for example, SS-316 steel was demonstrated to be resistant at 450 °C, while being severely corroded at
Corresponding author. Corresponding author. E-mail addresses: [email protected] (Y. Grosu), [email protected] (A. Faik).
∗∗
https://doi.org/10.1016/j.solmat.2019.110171 Received 17 June 2019; Received in revised form 26 August 2019; Accepted 8 September 2019 0927-0248/ © 2019 Elsevier B.V. All rights reserved.
Solar Energy Materials and Solar Cells 203 (2019) 110171
U. Nithiyanantham, et al.
600 and 680 °C by Hitec salt [26]. It is highly important to carry out the corrosion tests under the conditions close to those of real CSP plants. Ruiz-Cabañas et al. investigated the corrosion of carbon steel (CS) A516.Gr70 at the pilot plant scale by continuously or partially exposing it to solar salt at 390 °C [15]. From the analysis, the continuously exposed CS shows 5.46 μm/ year and partially exposed CS shows 3.57 μm/year corrosion rates [15]. Wang et al. studied compatibility of X80 carbon steel, 304, 316L stainless steels, and 600, 825 nickel alloys with Solar salt and HITEC salt under static and dynamic conditions [18]. The authors demonstrated the corrosion resistance behaviour of the stainless steel and nickel based alloy at 450 °C in the following order 304 < 316L < 600 < 825. They also found that the stainless steel has lower corrosion resistance compared to nickel based alloy [18]. Encinas-Sanchez et al. presented the electrochemical impedance spectroscopy method for monitoring corrosion process at the CSP plants [19]. The corrosion test was performed by using P91 steel at 580 °C for the time period of 1000h and the observed corrosion rate of ∼300 μm/ year. They also found the maximum mass gain of 0.59 mg/cm2 after 72 h corrosion test [19]. Interestingly the mass gain reduced with further exposure time due to the detachments of the corrosion layer from the P91 steel [19]. Walczak et al. described the corrosion problems relevant to TES systems using different combinations of construction materials and molten salts [10]. A recent review on the corrosion aspects of molten nitrate salts is available in Ref. [33]. Corrosion issues for molten salts other than nitrates are also actively discussed. In particular, carbonate [23,41] and chloride [34–42] salts in view of their potential to be used at the third generation CSP plants at temperatures above 600 °C. A review on sensible storage materials at temperatures above 600 °C, including corrosion aspects, can be found in Ref. [48]. Several anticorrosion methods were proposed to cope with compatibility issue in the CSP plants. In particular, Al slurry coating reduces corrosion rates in nitrate salts at 580 °C and in carbonate salts at 650 °C [29]. Recently we proposed a spry-graphitization coating for carbon steel against corrosion attack of molten nitrate salts under cycling conditions up to 500 °C [22,30]. Encinas-Sanchez et al. tested ZrO2–Y2O3 coating for protecting P91 against molten nitrate salt at 500 °C, demonstrating environmental benefits of such combination compared to SS-304 alloy [31]. Porcayo-Calderon et al. compared Ni20Cr coating to SS-304 in molten ZnCl2–KCl salt at 350, 400 and 450 °C [43]. Gomez-Vidal et al. used pre-oxidation for alumina forming alloys to improve the corrosion resistance under isothermal [45] and cycling [35] conditions. MCrAlX coatings were proposed against corrosion of molten chlorides salts [46]. Several strategies, namely, the addition of Mg and pre-oxidation, were suggested by Ding et al. to reduce the corrosion attack of chloride salts at 700 °C [47]. On the other hand, enhancement of thermophysical properties of molten salts due to nanoparticles doping (Al2O3, Graphite, SiO2, TiO2, etc.) are actively discussed in the literature. In particular, the enhancement of thermophysical properties for Al2O3 and SiO2 based nanofluids are well-known and studied by several researchers [48,49]. Furthermore, the in-situ preparation of the TiO2 based nanofluids shows 7.5% enhancement of its specific heat capacity [50]. However, there is no information available regarding the corrosion effect of TiO2 nanoparticles doped molten nitrate salts. With these regards, corrosion aspects of such molten salts based nanofluids become the point of interest [27,28]. At the moment the understanding of the influence of the nanoparticles on the corrosivity of molten salt is very limited as well as the number of corresponding articles [21,28,32,51,52]. Adding 1 wt% of silica nanoparticles into 42.7%Li2CO3–K2CO3 salt was reported to reduce the corrosion rates at 520 °C based on mass variation [51,52]. Fernandez et al. detected the incorporation of alumina and silica nanoparticles into the corrosion scale, which reduced the corrosion rates of SS-347 with Solar salt at 565 °C [28]. Earlier incorporation of Al2O3 and SiO2 nanoparticles into
the corrosion scale was also reported for HitecXL salt explored with carbon steel at 310 °C [27]. Most recently, our research group investigated the corrosion effect of Al2O3 and SiO2 based molten salts nanofluids at 310 and 390 °C with CS [32]. Both Al2O3 and SiO2 doped molten salts demonstrated around 3 times lower corrosion layer thickness compared to pure molten salt at 390 °C [32]. However, the corrosivity was found to be higher when nanoparticles are added to the salt at 310°C due to numerous microbubbles of air entrapped in the interparticle porosity, which increased the rates of oxidation [27,32]. From this limited number of works it is clear that nanoparticles have pronounced effect on the corrosivity of molten salts, however, the mechanism behind this effect is not yet understood. Understanding this mechanism has tremendous potential to increase the feasibility of molten salts based nanofluids at CSP plants, bringing the benefits not only in terms of thermophysical properties enhancement, but also in terms of cost reduction of the construction materials. The aim of this work is to define the dominant phenomena affecting the corrosivity of molten salts based nanofluids. For this purpose, a set of techniques was applied to carbon steel tested with molten binary nitrate salt doped with TiO2 nanoparticles. The corrosion test conditions as well as materials were chosen to be representative for the real CSP application. In particular, 390 °C is the maximum temperature for the cold tank in commercial parabolic through system. While carbon steel is a typical material for such tank. The corrosion rates were found to decrease considerably when nanoparticles were added, which was confirmed by three independent techniques (mass gain, cross-section analysis and XPS depth profiling). It was identified that the formation of mixed iron-titanium oxide is responsible for the stabilization of the corrosion layer and reduction of the corrosion rates. 2. Materials and techniques 2.1. Material The eutectic salt was prepared by mixing pre-dried NaNO3 and KNO3 in 51:49 wt% ratio. Next, the mixture was melted at 360 °C for 6 h and grinded using agate mortar to obtain a homogeneous fine powder of the eutectic salt. Both the NaNO3 and KNO3 were received from SQM with a purity of 99% and the impurities present in the salts are given in our previous work [27]. The TiO2 nanoparticles with a particle size of 21 nm were purchased from Sigma Aldrich. The nanofluids were prepared by using 99 wt% of the eutectic salt with 1 wt% of nanoparticle by physical shaking method. The physical shaking process was carried out through SPEX Sample Prep 9000-series High-Energy shaker mill for the time period of 15 min. This method has been demonstrated to be suitable for molten salts based nanofluids preparation [27]. A516 Gr70 carbon steel (CS) used for the corrosion tests has the following chemical composition: 98.68% Fe, 0.97% Mn, 0.31% C and 0.04% P (weight percentage). 2.2. Corrosion test protocol For the corrosion tests, pieces of CS A516 Gr70 having a size of around 3 × 12 × 14 mm were used. These pieces were cleaned with ultra-sonication for 15 min using acetone followed by ethanol and finally with DI water. It is worth to mention here that, similar to previous work [24,27,30], the samples were used under the conditions close to practical applications without applying any polishing or other external treatment of the CS surface, which may enormously affect the evolution of the corrosion rate. The mass of the CS was controlled before and after the corrosion tests. The corrosion tests were conducted in alumina crucible by immersing CS pieces in 5 g of eutectic salt or TiO2 based nanofluid. The corrosion tests were conducted under air at 390 °C for the periods of 250, 500, 1000 and 1500 h under static immersion condition. After the corrosion tests the samples were carefully washed with hot water. 2
Solar Energy Materials and Solar Cells 203 (2019) 110171
U. Nithiyanantham, et al.
Fig. 1. SEM images of the main steps of cross-section view preparation using FIB technique: a) carbon steel; b) Pt protective layer of ∼500 nm deposited on carbon steel; c) staircase-shaped cavity excavated using focused Ga ion beam; d) 52° tilted view of the cross-section.
2.3. Methods 2.3.1. Scanning electron microscopy - SEM FEI Helios NanoLab 450S DualBeam – Focused ion beam (FIB) with FEG SEM was used to record micrographs of the surfaces of carbon steel samples after the corrosion tests. Also, the FIB technique was deployed to observe cross-sectional view of the corrosion layer. Fig. 1 shows the steps of cross-section preparation procedure. First, Pt protective layer (∼1 μm) was deposited above the spot of interest on the carbon steel surface (Fig. 1a) via electron beam (3 keV) stimulated decomposition of a Pt-containing gaseous precursor (Fig. 1b). Next, a staircase-shaped cavity was excavated using a focused Ga ion beam (30 keV, 9.6 nA) (Fig. 1c). The dimensions of this cavity were around 15 μm wide, 30 μm long and 20 μm deep at the cross-sectional plane. Finally, polishing of the cross-section surface was performed by low-current (30 keV, 100 pA) focused Ga ion beam followed by 52° tilt of the sample for the cross-section analysis (Fig. 1d). Using a low-voltage (3 keV) and lowcurrent polishing combined with Pt protective layer, allowed to preserve the corrosion layer in its initial form and to gain the details of its topology usually inaccessible when mechanical polishing is applied. Additionally, Quanta 200 FEG SEM was used in high vacuum mode
Fig. 2. Raman spectrum of TiO2–nanoparticles. Insert: SEM micrograph.
Fig. 3. SEM micrographs and EDX-analysis of the surface of carbon steel after 1500h corrosion test at 390 °C with nanoparticles-free nitrate salt (a, c, d) and nitrate salt doped with 1 wt% of TiO2 nanoparticles (b, e, f). 3
Solar Energy Materials and Solar Cells 203 (2019) 110171
U. Nithiyanantham, et al.
Fig. 4. SEM micrographs and EDX-mapping of the surface of carbon steel after 1500h corrosion test at 390 °C with nitrate salt doped with 1 wt% of TiO2 nanoparticles.
Fig. 6. XRD diffractograms of carbon steel before and after the 1500h corrosion tests at 390 °C with nitrate salt and nitrate salt doped with 1 wt% of TiO2 nanoparticles.
Fig. 5. XPS spectra of the surfaces of carbon steel (CS) after the 1500h corrosion tests with nitrate salt and nitrate salt doped with 1 wt% of TiO2 nanoparticles (np).
Table 1 Quantitative XPS analysis (at%) of carbon steel after 1500h corrosion tests at 390 °C with nitrate salt (CS) and nitrate salt doped with nanoparticles (CS + TiO2).
with electron beam energies 10 kV, 20 kV and 30 kV with BSED and ETD detectors. Energy-Dispersive X-Ray Spectroscopy (EDX) was performed for elemental mapping. For observing cross-section in SEM, the samples were put in a resin holder made of Aka-Resin and Aka-Cure catalyst (1 mL + 0,135 mL proportion respectively) to maintain the corrosion layer intact during the cutting and polishing processes.
CS + TiO2 CS
Ti2p
Fe2p
O1s
C1s
Mg
5 –
26 32
47 66
22 1
– 1
2.3.3. Thermogravimetric analysis – TGA For the thermogravimetric analysis NETZSCH STA 449 F3 Jupiter was used with a constant airflow of 60 mL/min in the temperature range from 40 °C to 700 °C with the heating rate of 10°C / min.
2.3.2. X-ray diffraction - XRD Bruker D8 Discover X-ray diffractometer was used with a LYNXEYEXE detector using CuKα1 radiation (λ = 1.5418 Å) and Bragg-Brenato θ:2θ geometry. The data collection was carried out at room temperature, between 10° and 80° with a step of 0,02° and dwell time of 1,03s per step. The EVA program was used to determine the phase composition of the material.
2.3.4. X-ray photoelectron spectroscopy characterization – XPS The XPS measurements were performed in an ultra-high vacuum (UHV) system ESCALAB250Xi (Thermo Fisher Scientific). The base pressure in the system was below 5x10−10 mbar. XPS spectra were 4
Solar Energy Materials and Solar Cells 203 (2019) 110171
U. Nithiyanantham, et al.
Figure 7. a) mass gain and b) corrosion layer thickness for carbon steel tested with molten nitrate salt and nitrate salt doped with 1 wt% of TiO2 nanoparticles at 390 °C.
2.3.6. Contact angle measurements A KRUSS HT-2 Contact Angle was used to perform the contact angle (CA) measurements. All the CA measurements were conducted under ambient atmosphere, with a 2 °C/min heating rate. The presented CA values are the average of 3 measurements. CA values dispersion is subject to several factors, among which surface roughness and drop dimension are the most pertinent. To minimize their effect, a sample preparation protocol was followed. The salt eutectic powder was weighed, then milled and afterwards compressed at 40 MPa for 3 min in a 4 mm diameter round compression dye, to form same size pellets of 0.1g. The measurements were conducted on a carbon steel substrate (20 mm × 20 mm x 3 mm), which roughness was measured before and after each measurement using an EPS Interferometer. The surface roughness is found to have a constant value of S = 0.43 ± 0.20 μm. 3. Results
Fig. 8. Cross-section of carbon steel after 1500h corrosion test at 390 °C with a) nitrate salt and b) nitrate salt doped with 1 wt% of TiO2 nanoparticles.
SEM imaging of TiO2 nanoparticles (Fig. 2, insert) confirmed the average size of 21 nm specified by the supplier. Raman spectroscopy is the technique of choice for TiO2 polymorphs characterization. Fig. 2 shows a Raman spectrum from the TiO2 nanoparticles. The main peaks centred at 148 cm −1, ∼400 cm −1, ∼520 cm −1 and ∼640 cm −1 are close to the peaks’ positions in anatase, thus concluding that the TiO2 nanoparticles used in this study are mainly anatase polymorph. Fig. 3 presents SEM images of the carbon steel surface after corrosion test with pure nitrate salt (Fig. 3a) and with nanoparticles doped nitrate salt (Fig. 3b). A pronounced separation of the corrosion layer (peel-off) in the case of pure nitrate salt is clearly seen (Fig. 3a), while homogeneous corrosion appears when TiO2 nanoparticles were added (Fig. 3b). The high-magnification micrographs reveal the difference in the corrosion layer morphology (Fig. 3c–d versus Fig. 3e–f). Specifically, one can see that the nanoparticles are merging with the typical flakes of iron oxide (Fig. 3f) - the nanoparticles no longer possess the well-defined crystal shape as compared to the pristine nanoparticles (Fig. 2, insert). Such observation is in favour of the chemical reaction between the nanoparticles and the corrosion layer. Another argument to support the idea about possible chemical reactions between TiO2 and carbon steel comes from the EDX-analysis. Tables in Fig. 3 list elemental concentrations for the corresponded surfaces. In case of nanoparticles-free nitrate salt apart from iron and oxygen, magnesium was detected. Magnesium is a common impurity in nitrate salts and is often detected in the corrosion layers [11,15,24,27,30]. When salt without nanoparticles is used, Mg reacts with carbon steel during the corrosion test and incorporates into
acquired with a hemispherical analyzer and a monochromated X-ray source (Al Kα radiation, hν=1486.6 eV) operated at 15 keV and power 200 W. The XPS spectra were recorded with pass energies 20 eV and 200 eV for high resolution and survey spectra, respectively. The spectrometer was calibrated by setting the Au 4f7/2 level to 84.0 eV measured on a gold foil. XPS spectra were peak-fitted using Avantage data processing software. For peak fitting, the Shirley-type background subtraction was used. Quantification was done using elemental sensitivity factors provided by the Avantage library. The XPS system features a dual mode Ar ion source MAGCIS (Thermo Fisher Scientific) which can operate in monoatomic and cluster modes. The samples’ surfaces before XPS characterization were cleaned by using Ar cluster ions of 1000 atoms. For Ar cluster cleaning an accelerating voltage of 6 keV was used. For depth profiling the MAGCIS was operated in the monoatomic mode with beam energy 4 keV and beam current 3 μA. 2.3.5. Raman spectroscopy Raman spectra were recorded with Alpha300R spectrometer from WITec using 532 nm laser operated at 9x10−4 W. No radiation induced effects were observed during the measurements. The accumulations of 100 cycles were used for the measurements to achieve a sufficient signal-to-noise ratio. 5
Solar Energy Materials and Solar Cells 203 (2019) 110171
U. Nithiyanantham, et al.
steel. That is why Mg is not found in the corrosion layer after the corrosion tests with TiO2 doped salt. The similar mechanism was found in our previous work with Al2O3 and SiO2 nanoparticles [32]. On the other hand, for the nitrate salt doped with TiO2 nanoparticles titanium was detected instead of magnesium, which indicates the predominance of the TiO2-iron reaction over the magnesium-iron reaction. TiO2 was found all over the surface of the carbon steel. In some places, titanium forms large agglomerates ∼5 μm in size, as it is evidenced from Fig. 4. XPS spectra shown in Fig. 5 confirm the above mentioned observations, namely, the presence of titanium and the absence of magnesium in the corrosion layer of carbon steel for the case of nitrate salt doped with TiO2 nanoparticles (green line in Fig. 5) while the presence of magnesium was detected in the case of nanoparticle-free nitrate salt (red line in Fig. 5). XRD diffractograms presented in Fig. 6 reveal the evolution of the chemical phase composition of carbon steel after the corrosion tests with the nanoparticle-free nitrate salt and with the TiO2 nanoparticles doped nitrate salt. In particular, the surface of carbon steel after 1500h corrosion tests with nanoparticles-free nitrate salt contains two phases, namely, iron (Fe) and hematite (Fe2O3). On the other hand, the corrosion products after the tests with nitrate salt doped with TiO2 are formed by magnetite (Fe3O4), which is a lower degree of oxidation compared to hematite, and iron-titanium mixed oxide (Fe1.7Ti0.3O3). The main mechanism for the reduced corrosion rates in the molten salt doped with TiO2 was found to be the formation of mixed iron-titanium oxide phase (Fe1.7Ti0.3O3) on the CS surfaces (Fig. 6). This mixed oxide layer was found to be more stable as compared to iron oxide layer, which prevented the peel-off of the corrosion scale (Fig. 3) and consequently reduced the rate of further oxidation. The formed phase suggests that upon oxidation of carbon steel in molten nitrate salt some Fe3+ iron atoms are substituted by titanium in hematite phase. Formation of such mixed oxide can be an explanation of the stabilization effect, which prevents corrosion layer from peeling off (Fig. 3). Quantitative XPS analysis of the carbon steel surface elemental composition after corrosion tests (Table 1) gives Ti:Fe ratio ∼0.20 which is in accord with the XRD derived result: Ti:Fe = 0.18. In order to quantify the effect of iron-titanium mixed oxide formation in the corrosion layer, the rate of oxidation was determined using three independent techniques. First, the mass gain evolution was examined, which was found to be more than twice lower for nitrate salt doped with nanoparticles than that for the nanoparticles-free salt after 1500h corrosion tests – Fig. 7a. Most probably, the lower mass gain and corrosion layer thickness at 500h were obtained for the pure molten salt due to the corrosion layer peel-off (Fig. 3). Apparently, such peel-off becomes predominant at 500h. This question is directly related with the kinetics of the corrosion layer formation, which requires more study and is planned in the future. Next, a similar trend was observed for the corrosion layer evolution (Fig. 7b), which was determined from the statistical examination of the cross-sections of carbon steel (Fig. 8). For example, it was demonstrated by Wang et al. that the corrosion rate of CS is decreasing after certain period of time while being continuously exposed to the molten salt [18]. This was related with the reduction of oxygen and water vapor compared to their initial state and also formation of a new mixed oxide phases [18]. Finally, the obtained results were confirmed by the XPS depth profiling technique, which revealed the enhanced thickness of the oxidation layer for nanoparticles-free nitrate salt compared to the salt doped with TiO2 nanoparticles – Fig. 9. In particular, the average corrosion layer after 1500h tests for pure molten salt is 11.6 μm and for TiO2 doped molten salt is 5.2 μm. These results reveal that the addition of 1 wt% of TiO2 nanoparticles to molten binary nitrate salt reduced the corrosion rate of carbon steel more than twice at 390 °C and had a stabilizing effect on the corrosion scale. Additional important observation can be extracted from XPS depth profiling experiments. In the case of nanoparticles-free nitrate salt magnesium incorporation into the corrosion layer of carbon steel can be
Fig. 9. XPS depth profiling for carbon steel after 1500h corrosion test at 390 °C with a) nitrate salt and b) nitrate salt doped with 1 wt% of TiO2 nanoparticles. Red lines present depth profile of the iron oxide component (FeOx) of the Fe2p XPS spectral line. FeOx component intensity represents contribution from several oxidation states of iron excluding metallic iron. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)
Fig. 10. Schematic representation of corrosion layer of carbon steel after corrosion test at 390 °C with molten a) nitrate salt and b) nitrate salt doped with TiO2 nanoparticles based on XPS depth profiling and XRD analysis.
the corrosion layer. This was demonstrated using XPS (Fig. 5) and EDX (Fig. 3) techniques. While when TiO2 is present in the salt, TiO2 reacts with the carbon steel instead of magnesium. This was demonstrated by EDX (Fig. 3), XPS (Fig. 5) and XRD (Fig. 6). The reactions of TiO2 with carbon steel is preferable to the reaction of magnesium with carbon 6
Solar Energy Materials and Solar Cells 203 (2019) 110171
U. Nithiyanantham, et al.
Fig. 11. SEM images of the cross-section view of carbon steel after 1500h corrosion test at 390 °C with a) nitrate salt and b) nitrate salt doped with 1 wt% of TiO2 nanoparticles. Obtained with focused ion beam technique.
Fig. 13. XRD pattern of binary nitrate salt doped with 1 wt% TiO2 nanoparticles after 1500h corrosion test with carbon steel at 390 °C.
Fig. 12. Contact angles of a) molten nitrate salt, b) molten binary nitrate salt + 1 wt% of TiO2 NPs and c) their temperature dependence.
tracked all along the oxide layer (Fig. 9a, green line). This suggests that Mg forms mixed oxide with iron upon oxidation in molten nitrate salt. We were not able to determine the Mg containing phase by means of XRD due to its very low content as can be seen from the EDX (Fig. 3a) and from the XPS (Figs. 5 and 9a) data. On the contrary, titanium containing phase (Fe1.7Ti0.3O3) was found to be located mostly in the topmost surface of the corrosion layer (Fig. 9b). In particular, the concentration of Ti drops exponentially after sputtering away ∼1 μm of material and continues to decrease with depth in the corrosion scale. Such observation allows suggesting that after the formation of the mixed iron-titanium oxide further oxidation of carbon steel takes place by the diffusion of oxygen through the Fe1.7Ti0.3O3 phase. Such diffusion is apparently slower as compared to the diffusion through the mixed iron-magnesium oxide, which results in the lower corrosion rates in the case of nitrate salt doped with nanoparticles compared to nanoparticles-free salt.
Fig. 14. Thermogravimetric curves of as prepared nitrate salt, nitrate salt + 1 wt% TiO2 nanoparticles and nitrate salt + 1 wt% TiO2 nanoparticles after 1500h corrosion test at 390 °C.
Based on the XPS depth profiling and XRD analysis presented above, while taking into account EDX mapping, one can schematically represent the corrosion scale of carbon steel after the corrosion tests with nanoparticles-free molten nitrate salt and nitrate salt doped with TiO2 nanoparticles (Fig. 10). 7
Solar Energy Materials and Solar Cells 203 (2019) 110171
U. Nithiyanantham, et al.
Finally, focused ion beam technique was used to obtain textural properties of the corrosion layers' cross-section. Due to the low-voltage ion polishing the cross-section of the corrosion scale reveals the textural features (like cracks, voids, etc), which are typically lost during the mechanical polishing applied in conventional cross-section preparation. Fig. 11 clearly shows that corrosion layer of the carbon steel after the corrosion tests with molten nitrate salt poses more cracks (Fig. 11a) as compared to the case of molten salt doped with TiO2 nanoparticles (Fig. 11b). All the described above observations provide strong evidence of anticorrosion effect of TiO2 nanoparticles doping into the molten nitrate salt. This effect we assign to formation of mixed iron-titanium oxides. It is known from literature that the doping by other nanoparticles like silica [21,28,32,51,52] or alumina [27,28,32] also results in their incorporation into the corrosion layer. In particular, the incorporation of Al2O3 [28] and SiO2 [51,52] nanoparticles was reported to decrease the corrosivity of a molten salt. While in other work SiO2 nanoparticles incorporation had a negligible effect on the corrosion rates [28]. In the case when corrosion tests were performed at lower temperature (310 °C) micro-bubbling of the nanofluids were found to be a dominating effect which led to increased corrosion rates, even though, the incorporation of Al2O3 and SiO2 nanoparticles was also reported [27,32]. In this work, we have found that the mechanism of TiO2 nanoparticles incorporation involves the formation of mixed oxide with iron. Such results may also shed some light on the possible mechanism of SiO2 and Al2O3 nanoparticles doping. In particular, in work [28] iron-silicon mixed oxide was reported as a corrosion product after corrosion test of Solar salt with SS 347. However, this Fe2SiO4 phase did not have a predominant role for the corrosion rates and the overall corrosivity of the salt did not change despite of nanoparticles doping [28]. It is generally known that the nanoparticle doping can change the wetting properties of a fluid [53], which pertains to corrosion phenomena. With these regards, high-temperature contact angle measurements were performed for molten salts and salt doped with nanoparticles. It was found that in the investigated temperature range, nanoparticles doping did not change the wetting properties of molten nitrate salt (Fig. 12). Therefore, obtained reduction of corrosivity can be attributed to the mixed oxide formation, rather than to the wetting properties of the nanofluid. It should be noted that both salts and nanoparticles were examined in terms of possible degradation after the corrosion tests by XRD (Fig. 13) and no evidence of such modifications were detected. It was also verified that nanoparticles doping does not affect the decomposition temperature (and hence the operational temperature range) of the salt (Fig. 14).
• •
Obtained results provide important insights into the corrosion mechanisms of molten salts based nanofluids and allow to reconsider their feasibility for CSP applications due to the improved compatibility. Further work should be focused on the optimization of the concentration of TiO2 nanoparticles, as well as on the feasibility analysis of such nanofluids. Acknowledgements UN gratefully acknowledges the Nanouptake COST Action CA15119 for the funding of a short-term scientific mission (STSM-Grant No 42643), between CIC Energigune and University of Birmingham Center for Energy storage. YG gratefully acknowledges the Nanouptake COST Action CA15119 for the funding of a short-term scientific mission (STSM-Grant No 42475), between CIC Energigune and International Iberian Nanotechnology Laboratory. The authors express their sincere thanks to Yagmur Polat, Leticia Martinez and Cristina Luengo for their technical support. Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.solmat.2019.110171. References [1] E. Zarza-Moya, A Comprehensive Guide to Solar Energy Systems, first ed., Academic Press, 2018. [2] H. Zhang, J. Baeyens, G. Cáceres, J. Degrève, Y. Lv, Thermal energy storage: recent developments and practical aspects, Prog. Energy Combust. Sci. 53 (2016) 1–40 https://doi.org/10.1016/j.pecs.2015.10.003. [3] J. Lilliestam, T. Barradi, N. Caldés, M. Gomez, S. Hanger, J. Kern, N. Komendantova, M. Mehos, W.M. Hong, Z. Wang, A. Patt, Policies to keep and expand the option of concentrating solar power for dispatchable renewable electricity, Energy Policy 116 (2018) 193–197 https://doi.org/10.1016/j.enpol.2018. 02.014. [4] S. Kuravi, J. Trahan, D.Y. Goswami, M.M. Rahman, E.K. Stefanakos, Thermal energy storage technologies and systems for concentrating solar power plants, Prog. Energy Combust. 39 (2013) 285–319 https://doi.org/10.1016/j.pecs.2013.02.001. [5] B. chen Zhao, M. song Cheng, C. Liu, Z. min Dai, System-level performance optimization of molten-salt packed-bed thermal energy storage for concentrating solar power, Appl. Energy 226 (2018) 225–239 https://doi.org/10.1016/j.apenergy. 2018.05.081. [6] T. Esence, A. Bruch, J. Fourmigué, B. Stutz, A versatile one-dimensional numerical model for packed-bed heat storage systems, Renew. Energy 133 (2019) 190–204 https://doi.org/10.1016/j.renene.2018.10.012. [7] G. Peiró, C. Prieto, J. Gasia, A. Jové, L. Miró, L.F. Cabeza, Two-tank molten salts thermal energy storage system for solar power plants at pilot plant scale: lessons learnt and recommendations for its design, start-up and operation, Renew. Energy 121 (2018) 236–248 https://doi.org/10.1016/j.renene.2018.01.026. [8] X. Li, E. Xu, S. Song, X. Wang, G. Yuan, Dynamic simulation of two-tank indirect thermal energy storage system with molten salt, Renew. Energy 113 (2017) 1311–1319 https://doi.org/10.1016/j.renene.2017.06.024. [9] A. Gottschalk, U. Ramamoorthi, Parametric simulation and economic estimation of thermal energy storage in solar power tower, Mater. Today: Proc. 5 (2018) 1571–1577 https://doi.org/10.1016/j.matpr.2017.11.248. [10] M. Walczak, F. Pineda, Á.G. Fernández, C. Mata-Torres, R.A. Escobar, Materials corrosion for thermal energy storage systems in concentrated solar power plants, Renew. Sustain. Energy Rev. 86 (2018) 22–44 https://doi.org/10.1016/j.rser.2018. 01.010. [11] A.G. Fernández, H. Galleguillos, E. Fuentealba, F.J. Pérez, Corrosion of stainless
4. Conclusions In this work corrosivity of molten salts based nanofluids has been explored experimentally by means of XRD, XPS, EDX, FIB/SEM techniques. In particular, the effect of TiO2 nanoparticles on the corrosion rates of eutectic NaNO3–KNO3 molten salt was investigated for the first time. The following conclusions can be formulated:
• Addition of 1 wt% of TiO • •
corrosion rates reduction and corrosion scale stabilization. By means of XRD, EDX and XPS it was shown that growth of surface irontitanium oxide is competitive to formation of iron-magnesium oxide - a chemical phase formed due to the presence of Mg impurity in the nitrate salt. By means of XPS depth profiling, it was demonstrated that such mixed oxide is located on the surface of corrosion scale on top of magnetite phase, hindering the diffusion of molten salt and further oxidation of carbon steel. The investigated molten salt based nanofluid was found to have similar decomposition temperature and surface properties (contact angle on the carbon steel) as nanoparticles-free nitrate salt.
2 nanoparticles to molten binary nitrate salt reduced corrosion rate of carbon steel tested at 390 °C more than twice, which was demonstrated by mass gain, cross-section analysis and XPS depth profiling. The presence of TiO2 nanoparticles in the salt resulted in a homogeneous oxidation of the surface of carbon steel at 390 °C, which is in drastic contrast to the case of nanoparticles-free salt, where pronounced corrosion scale separation and peel-off were evident from SEM experiments. Formation of mixed iron-titanium oxide (Fe1.7Ti0.3O3) was demonstrated by XRD analysis and suggested to be responsible for the
8
Solar Energy Materials and Solar Cells 203 (2019) 110171
U. Nithiyanantham, et al.
[12]
[13] [14]
[15]
[16] [17] [18]
[19]
[20]
[21] [22]
[23]
[24]
[25] [26]
[27]
[28]
[29] [30]
[31]
[32]
steels and low-Cr steel in molten Ca(NO3)2–NaNO3–KNO3 eutectic salt for direct energy storage in CSP plants, Sol. Energy Mater. Sol. Cells 141 (2015) 7–13 https:// doi.org/10.1016/j.solmat.2015.05.004. K. Federsel, J. Wortmann, M. Ladenberger, High-temperature and corrosion behavior of nitrate nitrite molten salt mixtures regarding their application in concentrating solar power plants, Energy Procedia 69 (2015) 618–625 https://doi.org/ 10.1016/j.egypro.2015.03.071. A.S. Dorcheh, R.N. Durham, M.C. Galetz, High temperature corrosion in molten solar salt: the role of chloride impurities, Mater. Corros. (2017) 1–9 https://doi.org/ 10.1002/maco.201609300. W. Wang, B. Guan, X. Li, J. Lu, J. Ding, Corrosion behavior and mechanism of austenitic stainless steels in a new quaternary molten salt for concentrating solar power, Sol. Energy Mater. Sol. Cells 194 (2019) 36–46 https://doi.org/10.1016/j. solmat.2019.01.024. F.J. Ruiz-cabañas, C. Prieto, R. Osuna, V. Madina, A.I. Fernández, L.F. Cabeza, Corrosion testing device for in-situ corrosion characterization in operational molten salts storage tanks: a516 Gr70 carbon steel performance under molten salts exposure, Sol. Energy Mater. Sol. Cells 157 (2016) 383–392 https://doi.org/10.1016/ j.solmat.2016.06.005. W. Cheng, D. Chen, C. Wang, High-temperature corrosion of Cr-Mo steel in molten LiNO3-NaNO3-KNO3 eutectic salt for thermal energy storage, Sol. Energy Mater. Sol. Cells 132 (2015) 563–569 https://doi.org/10.1016/j.solmat.2014.10.007. S. Goods, R. Bradshaw, Corrosion of stainless steels and carbon steel by molten mixtures of commercial nitrate salts, J. Mater. Eng. Perform. 13 (2004) 78–87 https://doi.org/10.1361/10599490417542. J. Wang, Y. Jiang, Y. Ni, A. Wu, J. Li, Investigation on static and dynamic corrosion behaviors of thermal energy transfer and storage system materials by molten salts in concentrating solar power plants, Mater. Corros. 70 (2019) 102–109 https://doi. org/10.1002/maco.201810362. V. Encinas-Sánchez, M.T. de Miguel, M.I. Lasanta, G. García-Martín, F.J. Pérez, Electrochemical impedance spectroscopy (EIS): an efficient technique for monitoring corrosion processes in molten salt environments in CSP applications, Sol. Energy Mater. Sol. Cells 191 (2019) 157–163 https://doi.org/10.1016/j.solmat. 2018.11.007. A. Gomes, M. Navas, N. Uranga, T. Paiva, I. Figueira, T.C. Diamantino, High-temperature corrosion performance of austenitic stainless steels type AISI 316L and AISI 321H, in molten solar salt, Sol. Energy 177 (2019) 408–419 https://doi.org/ 10.1016/j.solener.2018.11.019. J.W. Slusser, J.B. Titcomb, M.T. Heffelfinger, B.R. Dunbobbin, Corrosion in molten nitrate-nitrite salts, J. Met. 37 (1985) 24–27 https://doi.org/10.1007/BF03259692. J. Piquot, U. Nithiyanantham, Y. Grosu, A. Faik, Spray-graphitization as a protection method against corrosion by molten nitrate salts and molten salts based nanofluids for thermal energy storage applications, Sol. Energy Mater. Sol. Cells 200 (2019) 110024 https://doi.org/10.1016/j.solmat.2019.110024. M.T. de Miguel, V. Encinas-Sánchez, M.I. Lasanta, G. García-Martín, F.J. Pérez, Corrosion resistance of HR3C to a carbonate molten salt for energy storage applications in CSP plants, Sol. Energy Mater. Sol. Cells 157 (2016) 966–972 https://doi. org/10.1016/j.solmat.2016.08.014. Y. Grosu, O. Bondarchuk, A. Faik, The effect of humidity, impurities and initial state on the corrosion of carbon and stainless steels in molten HitecXL salt for CSP application, Sol. Energy Mater. Sol. Cells 174 (2018) 34–41 https://doi.org/10.1016/ j.solmat.2017.08.026. A.M. Kruizenga, D.D. Gill, M.E. LaFord, G. McConohy, Corrosion of High Temperature Alloys in Solar Salt at 400, 500, and 680oC, SANDIA REPORT, SAND 2013-8256 California, (2013). M. Zhu, S. Zeng, H. Zhang, J. Li, B. Cao, Electrochemical study on the corrosion behaviors of 316 SS in HITEC molten salt at different temperatures, Sol. Energy Mater. Sol. Cells 186 (2018) 200–207 https://doi.org/10.1016/j.solmat.2018.06. 044. Y. Grosu, N. Udayashankar, O. Bondarchuk, L. González-Fernández, A. Faik, Unexpected effect of nanoparticles doping on the corrosivity of molten nitrate salt for thermal energy storage, Sol. Energy Mater. Sol. Cells 178 (2018) 91–97 https:// doi.org/10.1016/j.solmat.2018.01.002. A.G. Fernández, B. Muñoz-Sánchez, j. Nieto-Maestre, A. García-Romero, High temperature corrosion behavior on molten nitrate salt-based nanofluids for CSP plants, Renew. Energy 130 (2019) 902–909 https://doi.org/10.1016/j.renene. 2018.07.018. A. Agüero, P. Audigié, S. Rodríguez, V. Encinas-Sánched, Protective coatings for high temperature molten salt heat storage systems in solar concentration power plants, AIP Conf. Proc. 2033 (2017) 090001, , https://doi.org/10.1063/1.5067095. Y. Grosu, U. Nithiyanantham, A. Zaki, A. Faik, A simple method for the inhibition of the corrosion of carbon steel by molten nitrate salt for thermal storage in concentrating solar power applications, npj Mater. Degrad. 2 (2018) 34 https://doi. org/10.1038/s41529-018-0055-0. V. Encinas-Sánchez, E. Batuecas, A. Macías-García, C. Mayo, R. Díaz, F.J. Pérez, Corrosion resistance of protective coatings against molten nitrate salts for thermal energy storage and their environmental impact in CSP technology, Sol. Energy 176 (2018) 688–697 https://doi.org/10.1016/j.solener.2018.10.083. U. Nithiyanantham, Y. Grosu, L. González-Fernández, A. Zaki, J.M. Igartua, A. Faik,
[33]
[34]
[35]
[36] [37] [38] [39]
[40]
[41] [42] [43]
[44]
[45]
[46] [47]
[48] [49] [50] [51]
[52] [53]
9
Corrosion aspects of molten nitrate salt-based nanofluids for thermal energy storage applications, Sol. Energy 189 (2019) 219–227 https://doi.org/10.1016/j.solener. 2019.07.050. Á.G. Fernández, L.F. Cabeza, Molten salt corrosion mechanisms of nitrate based thermal energy storage materials for concentrated solar power plants: a review, Sol. Energy Mater. Sol. Cells 194 (2019) 160–165 https://doi.org/10.1016/j.solmat. 2019.02.012. G. Mohan, M. Venkataraman, J. Gomez-Vidal, J. Coventry, Thermo-economic analysis of high-temperature sensible thermal storage with different ternary eutectic alkali and alkaline earth metal chlorides, Sol. Energy 176 (2018) 350–357 https://doi.org/10.1016/j.solener.2018.10.008. J.C. Gomez-Vidal, A.G. Fernandez, R. Tirawat, C. Turchi, W. Huddleston, Corrosion resistance of alumina forming alloys against molten chlorides for energy production. II: electrochemical impedance spectroscopy under thermal cycling conditions, Sol. Energy Mater. Sol. Cells 166 (2017) 234–245 https://doi.org/10.1016/j. solmat.2017.03.025. J.C. Gomez-Vidal, R. Tirawat, Corrosion of alloys in a chloride molten salt (NaClLiCl) for solar thermal technologies, Sol. Energy Mater. Sol. Cells 157 (2016) 234–244 https://doi.org/10.1016/j.solmat.2016.05.052. B. Liu, X. Wei, W. Wang, J. Lu, J. Ding, Corrosion behavior of Ni-based alloys in molten NaCl-CaCl2-MgCl2 eutectic salt for concentrating solar power, Sol. Energy Mater. Sol. Cells 170 (2017) 77–86 https://doi.org/10.1016/j.solmat.2017.05.050. W. Ding, A. Bonk, J. Gussone, T. Bauer, Electrochemical measurement of corrosive impurities in molten chlorides for thermal energy storage, J. Energy Storage 15 (2018) 408–414 https://doi.org/10.1016/j.est.2017.12.007. K. Vignarooban, P. Pugazhendhi, C. Tucker, D. Gervasio, A.M. Kannan, Corrosion resistance of Hastelloys in molten metal-chloride heat-transfer fluids for concentrating solar power applications, Sol. Energy 103 (2014) 62–69 https://doi.org/ 10.1016/j.solener.2014.02.002. K. Vignarooban, X. Xu, K. Wang, E.E. Molina, P. Li, D. Gervasio, A.M. Kannan, Vapor pressure and corrosivity of ternary metal-chloride molten-salt based heat transfer fluids for use in concentrating solar power systems, Appl. Energy 159 (2015) 206–213 https://doi.org/10.1016/j.apenergy.2015.08.131. M. Sarvghad, T.A. Steinberg, G. Will, Corrosion of stainless steel 316 in eutectic molten salts for thermal energy storage, Sol. Energy 172 (2018) 198–203 https:// doi.org/10.1016/j.solener.2018.03.053. H. Sun, J. Wang, Z. Li, P. Zhang, X. Su, Corrosion behavior of 316SS and Ni-based alloys in a ternary NaCl-KCl-MgCl2 molten salt, Sol. Energy 171 (2018) 320–329 https://doi.org/10.1016/j.solener.2018.06.094. J. Porcayo-Calderon, O. Sotelo-Mazon, V.M. Salinas-Bravo, C.D. Arrieta-Gonzalez, J.J. Ramos-Hernandez, C. Cuevas-Arteaga, Electrochemical performance of Ni20Cr coatings applied by combustion powder spry in ZnCl2-KCl molten salts, Int. J. Electrochem. Sci. 7 (2012) 1134–1148. H. Cho, J.W. Van Zee, S. Shimpalee, B.A. Tavakoli, J.W. Weidner, B.L. Garcia-Diaz, M.J. Martinez-Rodriguez, L. Olson, J. Gray, Dimensionless analysis for predicting Fe-Ni-Cr alloy corrosion in molten salt systems for concentrated solar power systems, Corrosion 72 (2016) 742–760 https://doi.org/10.5006/1865. J.C. Gomez-Vidal, A.G. Fernandez, R. Tirawat, C. Turchi, W. Huddleston, Corrosion resistance of alumina-forming alloys against molten chlorides for energy production. I: pre-oxidation treatment and isothermal corrosion tests, Sol. Energy Mater. Sol. Cells 166 (2017) 222–233 https://doi.org/10.1016/j.solmat.2017.02.019. J.C. Gomez-Vidal, Corrosion resistance of MCrAlX coatings in a molten chloride for thermal storage in concentrating solar power applications, npj Mater. Degrad. 7 (2017) 1 https://doi.org/10.1038/s41529-017-0012-3. W. Ding, H. Shi, A. Jianu, Y. Xiu, A. Bonk, A. Weisenburger, T. Bauer, Molten chloride salts for next generation concentrated solar power plants: mitigation strategies against corrosion of structural materials, Sol. Energy Mater. Sol. Cells 193 (2019) 298–313 https://doi.org/10.1016/j.solmat.2018.12.020. G. Mohan, M.B. Venkataraman, J. Coventry, Sensible energy storage options for concentrating solar power plants operating above 600°C, Renew. Sustain. Energy Rev. 107 (2019) 319–337 https://doi.org/10.1016/j.rser.2019.01.062. M. Arthur, A. Karim, An investigation into the thermophysical and rheological properties of nanofluids for solar thermal applications, Renew. Sustain. Energy Rev. 55 (2016) 739–755 https://doi.org/10.1016/j.rser.2015.10.065. M. Lasfargues, A. Bell, Y. Ding, In situ production of titanium dioxide nanoparticles in molten salt phase for thermal energy storage and heat transfer fluid applications, J. Nanoparticle Res. 18 (2016) 150 https://doi.org/10.1007/s11051-016-3460-8. M. Schuller, F. Little, D. Malik, M. Betts, Q. Shao, J. Luo, W. Zhong, S. Shankar, A. Padmanaban, Molten Salt-Carbon Nanotube Thermal Energy Storage for Concentrating Solar Power Systems, United States: N. P. (2012), https://doi.org/10. 2172/1036948 Web https://www.osti.gov/servlets/purl/1036948. A. Padmanaban, The Effect of Silica Nanoparticles on Corrosion of Steel by Molten Carbonate Eutectics, Master thesis (2011). M. Hernaiz, V. Alonso, P. Estellé, Z. Wu, B. Sundén, L. Doretti, S. Mancin, N. Çobanoğlu, Z.H. Karadeniz, N. Garmendia, M. Lasheras-Zubiate, L. Hernández López, R. Mondragón, R. Martínez-Cuenca, S. Barison, A. Kujawska, A. Turgut, A. Amigo, G. Huminic, A. Huminic, M.R. Kalus, K.G. Schroth, M.H. Buschmann, The contact angle of nanofluids as thermophysical property, J. Colloid Interface Sci. 547 (2019) 393–406 https://doi.org/10.1016/j.jcis.2019.04.007.
Contents lists available at ScienceDirect
Solar Energy Materials and Solar Cells journal homepage: www.elsevier.com/locate/solmat
Nanoparticles as a high-temperature anticorrosion additive to molten nitrate salts for concentrated solar power
T
Udayashankar Nithiyananthama,b, Yaroslav Grosua,∗, Argyrios Anagnostopoulosc, Enrique Carbó-Argibayd, Oleksandr Bondarchukd, Luis González-Fernándeza, Abdelali Zakia, Josu Mirena Igartuab, María Elena Navarroc, Yulong Dingc, Abdessamad Faika,∗∗ a
CIC Energigune, Albert Einstein 48, Miñano, Álava, 01510, Spain Applied Physics II Department, Faculty of Science and Technology, University of the Basque Country (UPV/EHU), PO Box 644, Bilbao, 48080, Spain c BCES Birmingham Centre of Energy Storage, University of Birmingham, United Kingdom d International Iberian Nanotechnology Laboratory, Av. Mestre José Veiga, s/n, 4715-330, Braga, Portugal b
ARTICLE INFO
ABSTRACT
Keywords: Corrosion Molten salt Thermal energy storage Concentrated solar power
Hot corrosion is profoundly detrimental to construction elements in Concentrated Solar Power (CSP) plants affecting their lifetime, running costs and safety. In this work we have studied the anticorrosion effect of TiO2 nanoparticles additives on carbon steel using XRD, XPS with depth profiling, EDX and FIB/SEM techniques. The results revealed that the addition of 1 wt% of TiO2 nanoparticles to molten binary nitrate salt reduces the corrosion rate of carbon steel more than twice and stabilizes the corrosion scale at 390 °C. The anticorrosion effect of TiO2 nanoadditive was attributed to the formation of iron-titanium mixed oxide on the carbon steel surface. It was confirmed by XRD and TGA techniques that addition of TiO2 nanoparticles does not alter the stability of the salt. In view of presented results, the feasibility of molten salts based nanofluids in the CSPs can be reconsidered in terms of improved compatibility with construction materials.
1. Introduction Reaching necessary dispatchability is a major challenge for renewable technologies, which is typically addressed through efficient energy storage. Concentrated solar power (CSP) technology [1] possesses several important advantages in this sense, as it allows for reliable, lowcost and relatively simple thermal energy storage (TES) system [2,3] instead of more expensive and complex chemical or mechanical storage systems [4]. There is a great deal of effort from scientific and engineering community for further improvement of TES technologies at CSP plants by using two-tank or single-tank thermocline configurations [5], where different storage materials and heat transfer fluids (HTF), like molten salts, thermal oils, air, phase change materials, etc., can be used [6]. Corrosion issues [7], particularly with most common storage material NaNO3–KNO3 molten salt [8], are considered important not only in terms of safety and power plant reliability, but also in terms of the electricity cost reduction [9,10]. With these regards, a rapidly growing number of works dedicated to the corrosion of molten salts are evident in the literature [11–47].
∗
Nitrate salts are often the object of investigation due to the high level of maturity of this storage material. HitecXL salt (48 wt% Ca(NO3)2–7wt% NaNO3–45 wt% KNO3) was studied in terms of corrosion with AISI 304, 430 stainless steels, low-Cr alloy steel T22 [11] and with carbon steel A516Gr70, stainless steels AISI 304, AISI 316 [24]. The increase in corrosion rates due to the chlorides in molten nitrate salts was demonstrated for Hitec salt (53 wt% KNO3–40 wt% NaNO2–7wt% NaNO3) [12], Solar salt (60 wt% NaNO3 – 40 wt% KNO3) [13] and KNO3–NaNO2–NaNO3–KCl salt [14]. On the contrary, the positive effect of chromium in steels on the compatibility between different molten nitrate salts and construction materials has been reported [16,21]. The corrosion rates generally increase with temperature as was demonstrated for stainless steels SS-304 and SS-316 at 570 °C and for carbon steel A36 at 316 °C with several nitrate salts mixtures [17]. The same negative effect of temperature stands for Solar salt, which was tested at 550 °C [20], 650 °C [21], 680 °C [25] and for Hitec salt tested at 450, 600 and 680 °C [26]. It should be noted that corrosion rate is often nonlinearly depends on temperature, for example, SS-316 steel was demonstrated to be resistant at 450 °C, while being severely corroded at
Corresponding author. Corresponding author. E-mail addresses: [email protected] (Y. Grosu), [email protected] (A. Faik).
∗∗
https://doi.org/10.1016/j.solmat.2019.110171 Received 17 June 2019; Received in revised form 26 August 2019; Accepted 8 September 2019 0927-0248/ © 2019 Elsevier B.V. All rights reserved.
Solar Energy Materials and Solar Cells 203 (2019) 110171
U. Nithiyanantham, et al.
600 and 680 °C by Hitec salt [26]. It is highly important to carry out the corrosion tests under the conditions close to those of real CSP plants. Ruiz-Cabañas et al. investigated the corrosion of carbon steel (CS) A516.Gr70 at the pilot plant scale by continuously or partially exposing it to solar salt at 390 °C [15]. From the analysis, the continuously exposed CS shows 5.46 μm/ year and partially exposed CS shows 3.57 μm/year corrosion rates [15]. Wang et al. studied compatibility of X80 carbon steel, 304, 316L stainless steels, and 600, 825 nickel alloys with Solar salt and HITEC salt under static and dynamic conditions [18]. The authors demonstrated the corrosion resistance behaviour of the stainless steel and nickel based alloy at 450 °C in the following order 304 < 316L < 600 < 825. They also found that the stainless steel has lower corrosion resistance compared to nickel based alloy [18]. Encinas-Sanchez et al. presented the electrochemical impedance spectroscopy method for monitoring corrosion process at the CSP plants [19]. The corrosion test was performed by using P91 steel at 580 °C for the time period of 1000h and the observed corrosion rate of ∼300 μm/ year. They also found the maximum mass gain of 0.59 mg/cm2 after 72 h corrosion test [19]. Interestingly the mass gain reduced with further exposure time due to the detachments of the corrosion layer from the P91 steel [19]. Walczak et al. described the corrosion problems relevant to TES systems using different combinations of construction materials and molten salts [10]. A recent review on the corrosion aspects of molten nitrate salts is available in Ref. [33]. Corrosion issues for molten salts other than nitrates are also actively discussed. In particular, carbonate [23,41] and chloride [34–42] salts in view of their potential to be used at the third generation CSP plants at temperatures above 600 °C. A review on sensible storage materials at temperatures above 600 °C, including corrosion aspects, can be found in Ref. [48]. Several anticorrosion methods were proposed to cope with compatibility issue in the CSP plants. In particular, Al slurry coating reduces corrosion rates in nitrate salts at 580 °C and in carbonate salts at 650 °C [29]. Recently we proposed a spry-graphitization coating for carbon steel against corrosion attack of molten nitrate salts under cycling conditions up to 500 °C [22,30]. Encinas-Sanchez et al. tested ZrO2–Y2O3 coating for protecting P91 against molten nitrate salt at 500 °C, demonstrating environmental benefits of such combination compared to SS-304 alloy [31]. Porcayo-Calderon et al. compared Ni20Cr coating to SS-304 in molten ZnCl2–KCl salt at 350, 400 and 450 °C [43]. Gomez-Vidal et al. used pre-oxidation for alumina forming alloys to improve the corrosion resistance under isothermal [45] and cycling [35] conditions. MCrAlX coatings were proposed against corrosion of molten chlorides salts [46]. Several strategies, namely, the addition of Mg and pre-oxidation, were suggested by Ding et al. to reduce the corrosion attack of chloride salts at 700 °C [47]. On the other hand, enhancement of thermophysical properties of molten salts due to nanoparticles doping (Al2O3, Graphite, SiO2, TiO2, etc.) are actively discussed in the literature. In particular, the enhancement of thermophysical properties for Al2O3 and SiO2 based nanofluids are well-known and studied by several researchers [48,49]. Furthermore, the in-situ preparation of the TiO2 based nanofluids shows 7.5% enhancement of its specific heat capacity [50]. However, there is no information available regarding the corrosion effect of TiO2 nanoparticles doped molten nitrate salts. With these regards, corrosion aspects of such molten salts based nanofluids become the point of interest [27,28]. At the moment the understanding of the influence of the nanoparticles on the corrosivity of molten salt is very limited as well as the number of corresponding articles [21,28,32,51,52]. Adding 1 wt% of silica nanoparticles into 42.7%Li2CO3–K2CO3 salt was reported to reduce the corrosion rates at 520 °C based on mass variation [51,52]. Fernandez et al. detected the incorporation of alumina and silica nanoparticles into the corrosion scale, which reduced the corrosion rates of SS-347 with Solar salt at 565 °C [28]. Earlier incorporation of Al2O3 and SiO2 nanoparticles into
the corrosion scale was also reported for HitecXL salt explored with carbon steel at 310 °C [27]. Most recently, our research group investigated the corrosion effect of Al2O3 and SiO2 based molten salts nanofluids at 310 and 390 °C with CS [32]. Both Al2O3 and SiO2 doped molten salts demonstrated around 3 times lower corrosion layer thickness compared to pure molten salt at 390 °C [32]. However, the corrosivity was found to be higher when nanoparticles are added to the salt at 310°C due to numerous microbubbles of air entrapped in the interparticle porosity, which increased the rates of oxidation [27,32]. From this limited number of works it is clear that nanoparticles have pronounced effect on the corrosivity of molten salts, however, the mechanism behind this effect is not yet understood. Understanding this mechanism has tremendous potential to increase the feasibility of molten salts based nanofluids at CSP plants, bringing the benefits not only in terms of thermophysical properties enhancement, but also in terms of cost reduction of the construction materials. The aim of this work is to define the dominant phenomena affecting the corrosivity of molten salts based nanofluids. For this purpose, a set of techniques was applied to carbon steel tested with molten binary nitrate salt doped with TiO2 nanoparticles. The corrosion test conditions as well as materials were chosen to be representative for the real CSP application. In particular, 390 °C is the maximum temperature for the cold tank in commercial parabolic through system. While carbon steel is a typical material for such tank. The corrosion rates were found to decrease considerably when nanoparticles were added, which was confirmed by three independent techniques (mass gain, cross-section analysis and XPS depth profiling). It was identified that the formation of mixed iron-titanium oxide is responsible for the stabilization of the corrosion layer and reduction of the corrosion rates. 2. Materials and techniques 2.1. Material The eutectic salt was prepared by mixing pre-dried NaNO3 and KNO3 in 51:49 wt% ratio. Next, the mixture was melted at 360 °C for 6 h and grinded using agate mortar to obtain a homogeneous fine powder of the eutectic salt. Both the NaNO3 and KNO3 were received from SQM with a purity of 99% and the impurities present in the salts are given in our previous work [27]. The TiO2 nanoparticles with a particle size of 21 nm were purchased from Sigma Aldrich. The nanofluids were prepared by using 99 wt% of the eutectic salt with 1 wt% of nanoparticle by physical shaking method. The physical shaking process was carried out through SPEX Sample Prep 9000-series High-Energy shaker mill for the time period of 15 min. This method has been demonstrated to be suitable for molten salts based nanofluids preparation [27]. A516 Gr70 carbon steel (CS) used for the corrosion tests has the following chemical composition: 98.68% Fe, 0.97% Mn, 0.31% C and 0.04% P (weight percentage). 2.2. Corrosion test protocol For the corrosion tests, pieces of CS A516 Gr70 having a size of around 3 × 12 × 14 mm were used. These pieces were cleaned with ultra-sonication for 15 min using acetone followed by ethanol and finally with DI water. It is worth to mention here that, similar to previous work [24,27,30], the samples were used under the conditions close to practical applications without applying any polishing or other external treatment of the CS surface, which may enormously affect the evolution of the corrosion rate. The mass of the CS was controlled before and after the corrosion tests. The corrosion tests were conducted in alumina crucible by immersing CS pieces in 5 g of eutectic salt or TiO2 based nanofluid. The corrosion tests were conducted under air at 390 °C for the periods of 250, 500, 1000 and 1500 h under static immersion condition. After the corrosion tests the samples were carefully washed with hot water. 2
Solar Energy Materials and Solar Cells 203 (2019) 110171
U. Nithiyanantham, et al.
Fig. 1. SEM images of the main steps of cross-section view preparation using FIB technique: a) carbon steel; b) Pt protective layer of ∼500 nm deposited on carbon steel; c) staircase-shaped cavity excavated using focused Ga ion beam; d) 52° tilted view of the cross-section.
2.3. Methods 2.3.1. Scanning electron microscopy - SEM FEI Helios NanoLab 450S DualBeam – Focused ion beam (FIB) with FEG SEM was used to record micrographs of the surfaces of carbon steel samples after the corrosion tests. Also, the FIB technique was deployed to observe cross-sectional view of the corrosion layer. Fig. 1 shows the steps of cross-section preparation procedure. First, Pt protective layer (∼1 μm) was deposited above the spot of interest on the carbon steel surface (Fig. 1a) via electron beam (3 keV) stimulated decomposition of a Pt-containing gaseous precursor (Fig. 1b). Next, a staircase-shaped cavity was excavated using a focused Ga ion beam (30 keV, 9.6 nA) (Fig. 1c). The dimensions of this cavity were around 15 μm wide, 30 μm long and 20 μm deep at the cross-sectional plane. Finally, polishing of the cross-section surface was performed by low-current (30 keV, 100 pA) focused Ga ion beam followed by 52° tilt of the sample for the cross-section analysis (Fig. 1d). Using a low-voltage (3 keV) and lowcurrent polishing combined with Pt protective layer, allowed to preserve the corrosion layer in its initial form and to gain the details of its topology usually inaccessible when mechanical polishing is applied. Additionally, Quanta 200 FEG SEM was used in high vacuum mode
Fig. 2. Raman spectrum of TiO2–nanoparticles. Insert: SEM micrograph.
Fig. 3. SEM micrographs and EDX-analysis of the surface of carbon steel after 1500h corrosion test at 390 °C with nanoparticles-free nitrate salt (a, c, d) and nitrate salt doped with 1 wt% of TiO2 nanoparticles (b, e, f). 3
Solar Energy Materials and Solar Cells 203 (2019) 110171
U. Nithiyanantham, et al.
Fig. 4. SEM micrographs and EDX-mapping of the surface of carbon steel after 1500h corrosion test at 390 °C with nitrate salt doped with 1 wt% of TiO2 nanoparticles.
Fig. 6. XRD diffractograms of carbon steel before and after the 1500h corrosion tests at 390 °C with nitrate salt and nitrate salt doped with 1 wt% of TiO2 nanoparticles.
Fig. 5. XPS spectra of the surfaces of carbon steel (CS) after the 1500h corrosion tests with nitrate salt and nitrate salt doped with 1 wt% of TiO2 nanoparticles (np).
Table 1 Quantitative XPS analysis (at%) of carbon steel after 1500h corrosion tests at 390 °C with nitrate salt (CS) and nitrate salt doped with nanoparticles (CS + TiO2).
with electron beam energies 10 kV, 20 kV and 30 kV with BSED and ETD detectors. Energy-Dispersive X-Ray Spectroscopy (EDX) was performed for elemental mapping. For observing cross-section in SEM, the samples were put in a resin holder made of Aka-Resin and Aka-Cure catalyst (1 mL + 0,135 mL proportion respectively) to maintain the corrosion layer intact during the cutting and polishing processes.
CS + TiO2 CS
Ti2p
Fe2p
O1s
C1s
Mg
5 –
26 32
47 66
22 1
– 1
2.3.3. Thermogravimetric analysis – TGA For the thermogravimetric analysis NETZSCH STA 449 F3 Jupiter was used with a constant airflow of 60 mL/min in the temperature range from 40 °C to 700 °C with the heating rate of 10°C / min.
2.3.2. X-ray diffraction - XRD Bruker D8 Discover X-ray diffractometer was used with a LYNXEYEXE detector using CuKα1 radiation (λ = 1.5418 Å) and Bragg-Brenato θ:2θ geometry. The data collection was carried out at room temperature, between 10° and 80° with a step of 0,02° and dwell time of 1,03s per step. The EVA program was used to determine the phase composition of the material.
2.3.4. X-ray photoelectron spectroscopy characterization – XPS The XPS measurements were performed in an ultra-high vacuum (UHV) system ESCALAB250Xi (Thermo Fisher Scientific). The base pressure in the system was below 5x10−10 mbar. XPS spectra were 4
Solar Energy Materials and Solar Cells 203 (2019) 110171
U. Nithiyanantham, et al.
Figure 7. a) mass gain and b) corrosion layer thickness for carbon steel tested with molten nitrate salt and nitrate salt doped with 1 wt% of TiO2 nanoparticles at 390 °C.
2.3.6. Contact angle measurements A KRUSS HT-2 Contact Angle was used to perform the contact angle (CA) measurements. All the CA measurements were conducted under ambient atmosphere, with a 2 °C/min heating rate. The presented CA values are the average of 3 measurements. CA values dispersion is subject to several factors, among which surface roughness and drop dimension are the most pertinent. To minimize their effect, a sample preparation protocol was followed. The salt eutectic powder was weighed, then milled and afterwards compressed at 40 MPa for 3 min in a 4 mm diameter round compression dye, to form same size pellets of 0.1g. The measurements were conducted on a carbon steel substrate (20 mm × 20 mm x 3 mm), which roughness was measured before and after each measurement using an EPS Interferometer. The surface roughness is found to have a constant value of S = 0.43 ± 0.20 μm. 3. Results
Fig. 8. Cross-section of carbon steel after 1500h corrosion test at 390 °C with a) nitrate salt and b) nitrate salt doped with 1 wt% of TiO2 nanoparticles.
SEM imaging of TiO2 nanoparticles (Fig. 2, insert) confirmed the average size of 21 nm specified by the supplier. Raman spectroscopy is the technique of choice for TiO2 polymorphs characterization. Fig. 2 shows a Raman spectrum from the TiO2 nanoparticles. The main peaks centred at 148 cm −1, ∼400 cm −1, ∼520 cm −1 and ∼640 cm −1 are close to the peaks’ positions in anatase, thus concluding that the TiO2 nanoparticles used in this study are mainly anatase polymorph. Fig. 3 presents SEM images of the carbon steel surface after corrosion test with pure nitrate salt (Fig. 3a) and with nanoparticles doped nitrate salt (Fig. 3b). A pronounced separation of the corrosion layer (peel-off) in the case of pure nitrate salt is clearly seen (Fig. 3a), while homogeneous corrosion appears when TiO2 nanoparticles were added (Fig. 3b). The high-magnification micrographs reveal the difference in the corrosion layer morphology (Fig. 3c–d versus Fig. 3e–f). Specifically, one can see that the nanoparticles are merging with the typical flakes of iron oxide (Fig. 3f) - the nanoparticles no longer possess the well-defined crystal shape as compared to the pristine nanoparticles (Fig. 2, insert). Such observation is in favour of the chemical reaction between the nanoparticles and the corrosion layer. Another argument to support the idea about possible chemical reactions between TiO2 and carbon steel comes from the EDX-analysis. Tables in Fig. 3 list elemental concentrations for the corresponded surfaces. In case of nanoparticles-free nitrate salt apart from iron and oxygen, magnesium was detected. Magnesium is a common impurity in nitrate salts and is often detected in the corrosion layers [11,15,24,27,30]. When salt without nanoparticles is used, Mg reacts with carbon steel during the corrosion test and incorporates into
acquired with a hemispherical analyzer and a monochromated X-ray source (Al Kα radiation, hν=1486.6 eV) operated at 15 keV and power 200 W. The XPS spectra were recorded with pass energies 20 eV and 200 eV for high resolution and survey spectra, respectively. The spectrometer was calibrated by setting the Au 4f7/2 level to 84.0 eV measured on a gold foil. XPS spectra were peak-fitted using Avantage data processing software. For peak fitting, the Shirley-type background subtraction was used. Quantification was done using elemental sensitivity factors provided by the Avantage library. The XPS system features a dual mode Ar ion source MAGCIS (Thermo Fisher Scientific) which can operate in monoatomic and cluster modes. The samples’ surfaces before XPS characterization were cleaned by using Ar cluster ions of 1000 atoms. For Ar cluster cleaning an accelerating voltage of 6 keV was used. For depth profiling the MAGCIS was operated in the monoatomic mode with beam energy 4 keV and beam current 3 μA. 2.3.5. Raman spectroscopy Raman spectra were recorded with Alpha300R spectrometer from WITec using 532 nm laser operated at 9x10−4 W. No radiation induced effects were observed during the measurements. The accumulations of 100 cycles were used for the measurements to achieve a sufficient signal-to-noise ratio. 5
Solar Energy Materials and Solar Cells 203 (2019) 110171
U. Nithiyanantham, et al.
steel. That is why Mg is not found in the corrosion layer after the corrosion tests with TiO2 doped salt. The similar mechanism was found in our previous work with Al2O3 and SiO2 nanoparticles [32]. On the other hand, for the nitrate salt doped with TiO2 nanoparticles titanium was detected instead of magnesium, which indicates the predominance of the TiO2-iron reaction over the magnesium-iron reaction. TiO2 was found all over the surface of the carbon steel. In some places, titanium forms large agglomerates ∼5 μm in size, as it is evidenced from Fig. 4. XPS spectra shown in Fig. 5 confirm the above mentioned observations, namely, the presence of titanium and the absence of magnesium in the corrosion layer of carbon steel for the case of nitrate salt doped with TiO2 nanoparticles (green line in Fig. 5) while the presence of magnesium was detected in the case of nanoparticle-free nitrate salt (red line in Fig. 5). XRD diffractograms presented in Fig. 6 reveal the evolution of the chemical phase composition of carbon steel after the corrosion tests with the nanoparticle-free nitrate salt and with the TiO2 nanoparticles doped nitrate salt. In particular, the surface of carbon steel after 1500h corrosion tests with nanoparticles-free nitrate salt contains two phases, namely, iron (Fe) and hematite (Fe2O3). On the other hand, the corrosion products after the tests with nitrate salt doped with TiO2 are formed by magnetite (Fe3O4), which is a lower degree of oxidation compared to hematite, and iron-titanium mixed oxide (Fe1.7Ti0.3O3). The main mechanism for the reduced corrosion rates in the molten salt doped with TiO2 was found to be the formation of mixed iron-titanium oxide phase (Fe1.7Ti0.3O3) on the CS surfaces (Fig. 6). This mixed oxide layer was found to be more stable as compared to iron oxide layer, which prevented the peel-off of the corrosion scale (Fig. 3) and consequently reduced the rate of further oxidation. The formed phase suggests that upon oxidation of carbon steel in molten nitrate salt some Fe3+ iron atoms are substituted by titanium in hematite phase. Formation of such mixed oxide can be an explanation of the stabilization effect, which prevents corrosion layer from peeling off (Fig. 3). Quantitative XPS analysis of the carbon steel surface elemental composition after corrosion tests (Table 1) gives Ti:Fe ratio ∼0.20 which is in accord with the XRD derived result: Ti:Fe = 0.18. In order to quantify the effect of iron-titanium mixed oxide formation in the corrosion layer, the rate of oxidation was determined using three independent techniques. First, the mass gain evolution was examined, which was found to be more than twice lower for nitrate salt doped with nanoparticles than that for the nanoparticles-free salt after 1500h corrosion tests – Fig. 7a. Most probably, the lower mass gain and corrosion layer thickness at 500h were obtained for the pure molten salt due to the corrosion layer peel-off (Fig. 3). Apparently, such peel-off becomes predominant at 500h. This question is directly related with the kinetics of the corrosion layer formation, which requires more study and is planned in the future. Next, a similar trend was observed for the corrosion layer evolution (Fig. 7b), which was determined from the statistical examination of the cross-sections of carbon steel (Fig. 8). For example, it was demonstrated by Wang et al. that the corrosion rate of CS is decreasing after certain period of time while being continuously exposed to the molten salt [18]. This was related with the reduction of oxygen and water vapor compared to their initial state and also formation of a new mixed oxide phases [18]. Finally, the obtained results were confirmed by the XPS depth profiling technique, which revealed the enhanced thickness of the oxidation layer for nanoparticles-free nitrate salt compared to the salt doped with TiO2 nanoparticles – Fig. 9. In particular, the average corrosion layer after 1500h tests for pure molten salt is 11.6 μm and for TiO2 doped molten salt is 5.2 μm. These results reveal that the addition of 1 wt% of TiO2 nanoparticles to molten binary nitrate salt reduced the corrosion rate of carbon steel more than twice at 390 °C and had a stabilizing effect on the corrosion scale. Additional important observation can be extracted from XPS depth profiling experiments. In the case of nanoparticles-free nitrate salt magnesium incorporation into the corrosion layer of carbon steel can be
Fig. 9. XPS depth profiling for carbon steel after 1500h corrosion test at 390 °C with a) nitrate salt and b) nitrate salt doped with 1 wt% of TiO2 nanoparticles. Red lines present depth profile of the iron oxide component (FeOx) of the Fe2p XPS spectral line. FeOx component intensity represents contribution from several oxidation states of iron excluding metallic iron. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)
Fig. 10. Schematic representation of corrosion layer of carbon steel after corrosion test at 390 °C with molten a) nitrate salt and b) nitrate salt doped with TiO2 nanoparticles based on XPS depth profiling and XRD analysis.
the corrosion layer. This was demonstrated using XPS (Fig. 5) and EDX (Fig. 3) techniques. While when TiO2 is present in the salt, TiO2 reacts with the carbon steel instead of magnesium. This was demonstrated by EDX (Fig. 3), XPS (Fig. 5) and XRD (Fig. 6). The reactions of TiO2 with carbon steel is preferable to the reaction of magnesium with carbon 6
Solar Energy Materials and Solar Cells 203 (2019) 110171
U. Nithiyanantham, et al.
Fig. 11. SEM images of the cross-section view of carbon steel after 1500h corrosion test at 390 °C with a) nitrate salt and b) nitrate salt doped with 1 wt% of TiO2 nanoparticles. Obtained with focused ion beam technique.
Fig. 13. XRD pattern of binary nitrate salt doped with 1 wt% TiO2 nanoparticles after 1500h corrosion test with carbon steel at 390 °C.
Fig. 12. Contact angles of a) molten nitrate salt, b) molten binary nitrate salt + 1 wt% of TiO2 NPs and c) their temperature dependence.
tracked all along the oxide layer (Fig. 9a, green line). This suggests that Mg forms mixed oxide with iron upon oxidation in molten nitrate salt. We were not able to determine the Mg containing phase by means of XRD due to its very low content as can be seen from the EDX (Fig. 3a) and from the XPS (Figs. 5 and 9a) data. On the contrary, titanium containing phase (Fe1.7Ti0.3O3) was found to be located mostly in the topmost surface of the corrosion layer (Fig. 9b). In particular, the concentration of Ti drops exponentially after sputtering away ∼1 μm of material and continues to decrease with depth in the corrosion scale. Such observation allows suggesting that after the formation of the mixed iron-titanium oxide further oxidation of carbon steel takes place by the diffusion of oxygen through the Fe1.7Ti0.3O3 phase. Such diffusion is apparently slower as compared to the diffusion through the mixed iron-magnesium oxide, which results in the lower corrosion rates in the case of nitrate salt doped with nanoparticles compared to nanoparticles-free salt.
Fig. 14. Thermogravimetric curves of as prepared nitrate salt, nitrate salt + 1 wt% TiO2 nanoparticles and nitrate salt + 1 wt% TiO2 nanoparticles after 1500h corrosion test at 390 °C.
Based on the XPS depth profiling and XRD analysis presented above, while taking into account EDX mapping, one can schematically represent the corrosion scale of carbon steel after the corrosion tests with nanoparticles-free molten nitrate salt and nitrate salt doped with TiO2 nanoparticles (Fig. 10). 7
Solar Energy Materials and Solar Cells 203 (2019) 110171
U. Nithiyanantham, et al.
Finally, focused ion beam technique was used to obtain textural properties of the corrosion layers' cross-section. Due to the low-voltage ion polishing the cross-section of the corrosion scale reveals the textural features (like cracks, voids, etc), which are typically lost during the mechanical polishing applied in conventional cross-section preparation. Fig. 11 clearly shows that corrosion layer of the carbon steel after the corrosion tests with molten nitrate salt poses more cracks (Fig. 11a) as compared to the case of molten salt doped with TiO2 nanoparticles (Fig. 11b). All the described above observations provide strong evidence of anticorrosion effect of TiO2 nanoparticles doping into the molten nitrate salt. This effect we assign to formation of mixed iron-titanium oxides. It is known from literature that the doping by other nanoparticles like silica [21,28,32,51,52] or alumina [27,28,32] also results in their incorporation into the corrosion layer. In particular, the incorporation of Al2O3 [28] and SiO2 [51,52] nanoparticles was reported to decrease the corrosivity of a molten salt. While in other work SiO2 nanoparticles incorporation had a negligible effect on the corrosion rates [28]. In the case when corrosion tests were performed at lower temperature (310 °C) micro-bubbling of the nanofluids were found to be a dominating effect which led to increased corrosion rates, even though, the incorporation of Al2O3 and SiO2 nanoparticles was also reported [27,32]. In this work, we have found that the mechanism of TiO2 nanoparticles incorporation involves the formation of mixed oxide with iron. Such results may also shed some light on the possible mechanism of SiO2 and Al2O3 nanoparticles doping. In particular, in work [28] iron-silicon mixed oxide was reported as a corrosion product after corrosion test of Solar salt with SS 347. However, this Fe2SiO4 phase did not have a predominant role for the corrosion rates and the overall corrosivity of the salt did not change despite of nanoparticles doping [28]. It is generally known that the nanoparticle doping can change the wetting properties of a fluid [53], which pertains to corrosion phenomena. With these regards, high-temperature contact angle measurements were performed for molten salts and salt doped with nanoparticles. It was found that in the investigated temperature range, nanoparticles doping did not change the wetting properties of molten nitrate salt (Fig. 12). Therefore, obtained reduction of corrosivity can be attributed to the mixed oxide formation, rather than to the wetting properties of the nanofluid. It should be noted that both salts and nanoparticles were examined in terms of possible degradation after the corrosion tests by XRD (Fig. 13) and no evidence of such modifications were detected. It was also verified that nanoparticles doping does not affect the decomposition temperature (and hence the operational temperature range) of the salt (Fig. 14).
• •
Obtained results provide important insights into the corrosion mechanisms of molten salts based nanofluids and allow to reconsider their feasibility for CSP applications due to the improved compatibility. Further work should be focused on the optimization of the concentration of TiO2 nanoparticles, as well as on the feasibility analysis of such nanofluids. Acknowledgements UN gratefully acknowledges the Nanouptake COST Action CA15119 for the funding of a short-term scientific mission (STSM-Grant No 42643), between CIC Energigune and University of Birmingham Center for Energy storage. YG gratefully acknowledges the Nanouptake COST Action CA15119 for the funding of a short-term scientific mission (STSM-Grant No 42475), between CIC Energigune and International Iberian Nanotechnology Laboratory. The authors express their sincere thanks to Yagmur Polat, Leticia Martinez and Cristina Luengo for their technical support. Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.solmat.2019.110171. References [1] E. Zarza-Moya, A Comprehensive Guide to Solar Energy Systems, first ed., Academic Press, 2018. [2] H. Zhang, J. Baeyens, G. Cáceres, J. Degrève, Y. Lv, Thermal energy storage: recent developments and practical aspects, Prog. Energy Combust. Sci. 53 (2016) 1–40 https://doi.org/10.1016/j.pecs.2015.10.003. [3] J. Lilliestam, T. Barradi, N. Caldés, M. Gomez, S. Hanger, J. Kern, N. Komendantova, M. Mehos, W.M. Hong, Z. Wang, A. Patt, Policies to keep and expand the option of concentrating solar power for dispatchable renewable electricity, Energy Policy 116 (2018) 193–197 https://doi.org/10.1016/j.enpol.2018. 02.014. [4] S. Kuravi, J. Trahan, D.Y. Goswami, M.M. Rahman, E.K. Stefanakos, Thermal energy storage technologies and systems for concentrating solar power plants, Prog. Energy Combust. 39 (2013) 285–319 https://doi.org/10.1016/j.pecs.2013.02.001. [5] B. chen Zhao, M. song Cheng, C. Liu, Z. min Dai, System-level performance optimization of molten-salt packed-bed thermal energy storage for concentrating solar power, Appl. Energy 226 (2018) 225–239 https://doi.org/10.1016/j.apenergy. 2018.05.081. [6] T. Esence, A. Bruch, J. Fourmigué, B. Stutz, A versatile one-dimensional numerical model for packed-bed heat storage systems, Renew. Energy 133 (2019) 190–204 https://doi.org/10.1016/j.renene.2018.10.012. [7] G. Peiró, C. Prieto, J. Gasia, A. Jové, L. Miró, L.F. Cabeza, Two-tank molten salts thermal energy storage system for solar power plants at pilot plant scale: lessons learnt and recommendations for its design, start-up and operation, Renew. Energy 121 (2018) 236–248 https://doi.org/10.1016/j.renene.2018.01.026. [8] X. Li, E. Xu, S. Song, X. Wang, G. Yuan, Dynamic simulation of two-tank indirect thermal energy storage system with molten salt, Renew. Energy 113 (2017) 1311–1319 https://doi.org/10.1016/j.renene.2017.06.024. [9] A. Gottschalk, U. Ramamoorthi, Parametric simulation and economic estimation of thermal energy storage in solar power tower, Mater. Today: Proc. 5 (2018) 1571–1577 https://doi.org/10.1016/j.matpr.2017.11.248. [10] M. Walczak, F. Pineda, Á.G. Fernández, C. Mata-Torres, R.A. Escobar, Materials corrosion for thermal energy storage systems in concentrated solar power plants, Renew. Sustain. Energy Rev. 86 (2018) 22–44 https://doi.org/10.1016/j.rser.2018. 01.010. [11] A.G. Fernández, H. Galleguillos, E. Fuentealba, F.J. Pérez, Corrosion of stainless
4. Conclusions In this work corrosivity of molten salts based nanofluids has been explored experimentally by means of XRD, XPS, EDX, FIB/SEM techniques. In particular, the effect of TiO2 nanoparticles on the corrosion rates of eutectic NaNO3–KNO3 molten salt was investigated for the first time. The following conclusions can be formulated:
• Addition of 1 wt% of TiO • •
corrosion rates reduction and corrosion scale stabilization. By means of XRD, EDX and XPS it was shown that growth of surface irontitanium oxide is competitive to formation of iron-magnesium oxide - a chemical phase formed due to the presence of Mg impurity in the nitrate salt. By means of XPS depth profiling, it was demonstrated that such mixed oxide is located on the surface of corrosion scale on top of magnetite phase, hindering the diffusion of molten salt and further oxidation of carbon steel. The investigated molten salt based nanofluid was found to have similar decomposition temperature and surface properties (contact angle on the carbon steel) as nanoparticles-free nitrate salt.
2 nanoparticles to molten binary nitrate salt reduced corrosion rate of carbon steel tested at 390 °C more than twice, which was demonstrated by mass gain, cross-section analysis and XPS depth profiling. The presence of TiO2 nanoparticles in the salt resulted in a homogeneous oxidation of the surface of carbon steel at 390 °C, which is in drastic contrast to the case of nanoparticles-free salt, where pronounced corrosion scale separation and peel-off were evident from SEM experiments. Formation of mixed iron-titanium oxide (Fe1.7Ti0.3O3) was demonstrated by XRD analysis and suggested to be responsible for the
8
Solar Energy Materials and Solar Cells 203 (2019) 110171
U. Nithiyanantham, et al.
[12]
[13] [14]
[15]
[16] [17] [18]
[19]
[20]
[21] [22]
[23]
[24]
[25] [26]
[27]
[28]
[29] [30]
[31]
[32]
steels and low-Cr steel in molten Ca(NO3)2–NaNO3–KNO3 eutectic salt for direct energy storage in CSP plants, Sol. Energy Mater. Sol. Cells 141 (2015) 7–13 https:// doi.org/10.1016/j.solmat.2015.05.004. K. Federsel, J. Wortmann, M. Ladenberger, High-temperature and corrosion behavior of nitrate nitrite molten salt mixtures regarding their application in concentrating solar power plants, Energy Procedia 69 (2015) 618–625 https://doi.org/ 10.1016/j.egypro.2015.03.071. A.S. Dorcheh, R.N. Durham, M.C. Galetz, High temperature corrosion in molten solar salt: the role of chloride impurities, Mater. Corros. (2017) 1–9 https://doi.org/ 10.1002/maco.201609300. W. Wang, B. Guan, X. Li, J. Lu, J. Ding, Corrosion behavior and mechanism of austenitic stainless steels in a new quaternary molten salt for concentrating solar power, Sol. Energy Mater. Sol. Cells 194 (2019) 36–46 https://doi.org/10.1016/j. solmat.2019.01.024. F.J. Ruiz-cabañas, C. Prieto, R. Osuna, V. Madina, A.I. Fernández, L.F. Cabeza, Corrosion testing device for in-situ corrosion characterization in operational molten salts storage tanks: a516 Gr70 carbon steel performance under molten salts exposure, Sol. Energy Mater. Sol. Cells 157 (2016) 383–392 https://doi.org/10.1016/ j.solmat.2016.06.005. W. Cheng, D. Chen, C. Wang, High-temperature corrosion of Cr-Mo steel in molten LiNO3-NaNO3-KNO3 eutectic salt for thermal energy storage, Sol. Energy Mater. Sol. Cells 132 (2015) 563–569 https://doi.org/10.1016/j.solmat.2014.10.007. S. Goods, R. Bradshaw, Corrosion of stainless steels and carbon steel by molten mixtures of commercial nitrate salts, J. Mater. Eng. Perform. 13 (2004) 78–87 https://doi.org/10.1361/10599490417542. J. Wang, Y. Jiang, Y. Ni, A. Wu, J. Li, Investigation on static and dynamic corrosion behaviors of thermal energy transfer and storage system materials by molten salts in concentrating solar power plants, Mater. Corros. 70 (2019) 102–109 https://doi. org/10.1002/maco.201810362. V. Encinas-Sánchez, M.T. de Miguel, M.I. Lasanta, G. García-Martín, F.J. Pérez, Electrochemical impedance spectroscopy (EIS): an efficient technique for monitoring corrosion processes in molten salt environments in CSP applications, Sol. Energy Mater. Sol. Cells 191 (2019) 157–163 https://doi.org/10.1016/j.solmat. 2018.11.007. A. Gomes, M. Navas, N. Uranga, T. Paiva, I. Figueira, T.C. Diamantino, High-temperature corrosion performance of austenitic stainless steels type AISI 316L and AISI 321H, in molten solar salt, Sol. Energy 177 (2019) 408–419 https://doi.org/ 10.1016/j.solener.2018.11.019. J.W. Slusser, J.B. Titcomb, M.T. Heffelfinger, B.R. Dunbobbin, Corrosion in molten nitrate-nitrite salts, J. Met. 37 (1985) 24–27 https://doi.org/10.1007/BF03259692. J. Piquot, U. Nithiyanantham, Y. Grosu, A. Faik, Spray-graphitization as a protection method against corrosion by molten nitrate salts and molten salts based nanofluids for thermal energy storage applications, Sol. Energy Mater. Sol. Cells 200 (2019) 110024 https://doi.org/10.1016/j.solmat.2019.110024. M.T. de Miguel, V. Encinas-Sánchez, M.I. Lasanta, G. García-Martín, F.J. Pérez, Corrosion resistance of HR3C to a carbonate molten salt for energy storage applications in CSP plants, Sol. Energy Mater. Sol. Cells 157 (2016) 966–972 https://doi. org/10.1016/j.solmat.2016.08.014. Y. Grosu, O. Bondarchuk, A. Faik, The effect of humidity, impurities and initial state on the corrosion of carbon and stainless steels in molten HitecXL salt for CSP application, Sol. Energy Mater. Sol. Cells 174 (2018) 34–41 https://doi.org/10.1016/ j.solmat.2017.08.026. A.M. Kruizenga, D.D. Gill, M.E. LaFord, G. McConohy, Corrosion of High Temperature Alloys in Solar Salt at 400, 500, and 680oC, SANDIA REPORT, SAND 2013-8256 California, (2013). M. Zhu, S. Zeng, H. Zhang, J. Li, B. Cao, Electrochemical study on the corrosion behaviors of 316 SS in HITEC molten salt at different temperatures, Sol. Energy Mater. Sol. Cells 186 (2018) 200–207 https://doi.org/10.1016/j.solmat.2018.06. 044. Y. Grosu, N. Udayashankar, O. Bondarchuk, L. González-Fernández, A. Faik, Unexpected effect of nanoparticles doping on the corrosivity of molten nitrate salt for thermal energy storage, Sol. Energy Mater. Sol. Cells 178 (2018) 91–97 https:// doi.org/10.1016/j.solmat.2018.01.002. A.G. Fernández, B. Muñoz-Sánchez, j. Nieto-Maestre, A. García-Romero, High temperature corrosion behavior on molten nitrate salt-based nanofluids for CSP plants, Renew. Energy 130 (2019) 902–909 https://doi.org/10.1016/j.renene. 2018.07.018. A. Agüero, P. Audigié, S. Rodríguez, V. Encinas-Sánched, Protective coatings for high temperature molten salt heat storage systems in solar concentration power plants, AIP Conf. Proc. 2033 (2017) 090001, , https://doi.org/10.1063/1.5067095. Y. Grosu, U. Nithiyanantham, A. Zaki, A. Faik, A simple method for the inhibition of the corrosion of carbon steel by molten nitrate salt for thermal storage in concentrating solar power applications, npj Mater. Degrad. 2 (2018) 34 https://doi. org/10.1038/s41529-018-0055-0. V. Encinas-Sánchez, E. Batuecas, A. Macías-García, C. Mayo, R. Díaz, F.J. Pérez, Corrosion resistance of protective coatings against molten nitrate salts for thermal energy storage and their environmental impact in CSP technology, Sol. Energy 176 (2018) 688–697 https://doi.org/10.1016/j.solener.2018.10.083. U. Nithiyanantham, Y. Grosu, L. González-Fernández, A. Zaki, J.M. Igartua, A. Faik,
[33]
[34]
[35]
[36] [37] [38] [39]
[40]
[41] [42] [43]
[44]
[45]
[46] [47]
[48] [49] [50] [51]
[52] [53]
9
Corrosion aspects of molten nitrate salt-based nanofluids for thermal energy storage applications, Sol. Energy 189 (2019) 219–227 https://doi.org/10.1016/j.solener. 2019.07.050. Á.G. Fernández, L.F. Cabeza, Molten salt corrosion mechanisms of nitrate based thermal energy storage materials for concentrated solar power plants: a review, Sol. Energy Mater. Sol. Cells 194 (2019) 160–165 https://doi.org/10.1016/j.solmat. 2019.02.012. G. Mohan, M. Venkataraman, J. Gomez-Vidal, J. Coventry, Thermo-economic analysis of high-temperature sensible thermal storage with different ternary eutectic alkali and alkaline earth metal chlorides, Sol. Energy 176 (2018) 350–357 https://doi.org/10.1016/j.solener.2018.10.008. J.C. Gomez-Vidal, A.G. Fernandez, R. Tirawat, C. Turchi, W. Huddleston, Corrosion resistance of alumina forming alloys against molten chlorides for energy production. II: electrochemical impedance spectroscopy under thermal cycling conditions, Sol. Energy Mater. Sol. Cells 166 (2017) 234–245 https://doi.org/10.1016/j. solmat.2017.03.025. J.C. Gomez-Vidal, R. Tirawat, Corrosion of alloys in a chloride molten salt (NaClLiCl) for solar thermal technologies, Sol. Energy Mater. Sol. Cells 157 (2016) 234–244 https://doi.org/10.1016/j.solmat.2016.05.052. B. Liu, X. Wei, W. Wang, J. Lu, J. Ding, Corrosion behavior of Ni-based alloys in molten NaCl-CaCl2-MgCl2 eutectic salt for concentrating solar power, Sol. Energy Mater. Sol. Cells 170 (2017) 77–86 https://doi.org/10.1016/j.solmat.2017.05.050. W. Ding, A. Bonk, J. Gussone, T. Bauer, Electrochemical measurement of corrosive impurities in molten chlorides for thermal energy storage, J. Energy Storage 15 (2018) 408–414 https://doi.org/10.1016/j.est.2017.12.007. K. Vignarooban, P. Pugazhendhi, C. Tucker, D. Gervasio, A.M. Kannan, Corrosion resistance of Hastelloys in molten metal-chloride heat-transfer fluids for concentrating solar power applications, Sol. Energy 103 (2014) 62–69 https://doi.org/ 10.1016/j.solener.2014.02.002. K. Vignarooban, X. Xu, K. Wang, E.E. Molina, P. Li, D. Gervasio, A.M. Kannan, Vapor pressure and corrosivity of ternary metal-chloride molten-salt based heat transfer fluids for use in concentrating solar power systems, Appl. Energy 159 (2015) 206–213 https://doi.org/10.1016/j.apenergy.2015.08.131. M. Sarvghad, T.A. Steinberg, G. Will, Corrosion of stainless steel 316 in eutectic molten salts for thermal energy storage, Sol. Energy 172 (2018) 198–203 https:// doi.org/10.1016/j.solener.2018.03.053. H. Sun, J. Wang, Z. Li, P. Zhang, X. Su, Corrosion behavior of 316SS and Ni-based alloys in a ternary NaCl-KCl-MgCl2 molten salt, Sol. Energy 171 (2018) 320–329 https://doi.org/10.1016/j.solener.2018.06.094. J. Porcayo-Calderon, O. Sotelo-Mazon, V.M. Salinas-Bravo, C.D. Arrieta-Gonzalez, J.J. Ramos-Hernandez, C. Cuevas-Arteaga, Electrochemical performance of Ni20Cr coatings applied by combustion powder spry in ZnCl2-KCl molten salts, Int. J. Electrochem. Sci. 7 (2012) 1134–1148. H. Cho, J.W. Van Zee, S. Shimpalee, B.A. Tavakoli, J.W. Weidner, B.L. Garcia-Diaz, M.J. Martinez-Rodriguez, L. Olson, J. Gray, Dimensionless analysis for predicting Fe-Ni-Cr alloy corrosion in molten salt systems for concentrated solar power systems, Corrosion 72 (2016) 742–760 https://doi.org/10.5006/1865. J.C. Gomez-Vidal, A.G. Fernandez, R. Tirawat, C. Turchi, W. Huddleston, Corrosion resistance of alumina-forming alloys against molten chlorides for energy production. I: pre-oxidation treatment and isothermal corrosion tests, Sol. Energy Mater. Sol. Cells 166 (2017) 222–233 https://doi.org/10.1016/j.solmat.2017.02.019. J.C. Gomez-Vidal, Corrosion resistance of MCrAlX coatings in a molten chloride for thermal storage in concentrating solar power applications, npj Mater. Degrad. 7 (2017) 1 https://doi.org/10.1038/s41529-017-0012-3. W. Ding, H. Shi, A. Jianu, Y. Xiu, A. Bonk, A. Weisenburger, T. Bauer, Molten chloride salts for next generation concentrated solar power plants: mitigation strategies against corrosion of structural materials, Sol. Energy Mater. Sol. Cells 193 (2019) 298–313 https://doi.org/10.1016/j.solmat.2018.12.020. G. Mohan, M.B. Venkataraman, J. Coventry, Sensible energy storage options for concentrating solar power plants operating above 600°C, Renew. Sustain. Energy Rev. 107 (2019) 319–337 https://doi.org/10.1016/j.rser.2019.01.062. M. Arthur, A. Karim, An investigation into the thermophysical and rheological properties of nanofluids for solar thermal applications, Renew. Sustain. Energy Rev. 55 (2016) 739–755 https://doi.org/10.1016/j.rser.2015.10.065. M. Lasfargues, A. Bell, Y. Ding, In situ production of titanium dioxide nanoparticles in molten salt phase for thermal energy storage and heat transfer fluid applications, J. Nanoparticle Res. 18 (2016) 150 https://doi.org/10.1007/s11051-016-3460-8. M. Schuller, F. Little, D. Malik, M. Betts, Q. Shao, J. Luo, W. Zhong, S. Shankar, A. Padmanaban, Molten Salt-Carbon Nanotube Thermal Energy Storage for Concentrating Solar Power Systems, United States: N. P. (2012), https://doi.org/10. 2172/1036948 Web https://www.osti.gov/servlets/purl/1036948. A. Padmanaban, The Effect of Silica Nanoparticles on Corrosion of Steel by Molten Carbonate Eutectics, Master thesis (2011). M. Hernaiz, V. Alonso, P. Estellé, Z. Wu, B. Sundén, L. Doretti, S. Mancin, N. Çobanoğlu, Z.H. Karadeniz, N. Garmendia, M. Lasheras-Zubiate, L. Hernández López, R. Mondragón, R. Martínez-Cuenca, S. Barison, A. Kujawska, A. Turgut, A. Amigo, G. Huminic, A. Huminic, M.R. Kalus, K.G. Schroth, M.H. Buschmann, The contact angle of nanofluids as thermophysical property, J. Colloid Interface Sci. 547 (2019) 393–406 https://doi.org/10.1016/j.jcis.2019.04.007.