This may be the author’s version of a work that was submitted/acceptedfor publication in the following source:

Bell, Stuart, Will, Geoffrey, & Steinberg, Ted(2019)Corrosion testing under inert atmosphere with stainless steel crucibles.In Richter, Christoph (Ed.) Proceedings of the 24th SolarPACES Inter-national Conference (SolarPACES 2018) (AIP Conference Proceedings,Volume 2126).American Institute of Physics (AIP), United States of America, pp. 1-7.

This file was downloaded from: https://eprints.qut.edu.au/132333/

c© Consult author(s) regarding copyright matters

This work is covered by copyright. Unless the document is being made available under aCreative Commons Licence, you must assume that re-use is limited to personal use andthat permission from the copyright owner must be obtained for all other uses. If the docu-ment is available under a Creative Commons License (or other specified license) then referto the Licence for details of permitted re-use. It is a condition of access that users recog-nise and abide by the legal requirements associated with these rights. If you believe thatthis work infringes copyright please provide details by email to qut.copyright@qut.edu.au

Notice: Please note that this document may not be the Version of Record(i.e. published version) of the work. Author manuscript versions (as Sub-mitted for peer review or as Accepted for publication after peer review) canbe identified by an absence of publisher branding and/or typeset appear-ance. If there is any doubt, please refer to the published source.

https://doi.org/10.1063/1.5117719

Corrosion Testing Under Inert Atmosphere with Stainless

Steel Crucibles

Stuart Bella), Geoff Willb) and Ted Steinbergc)

School of Chemistry, Physics, Mechanical Engineering, Science and Engineering Faculty, Queensland University of

Technology

2 George St, Brisbane, QLD 4000, Australia

a)Corresponding author: s2.bell@qut.edu.au

b)g.will@qut.edu.au c)t.steinberg@qut.edu.au

Abstract. Molten salt eutectics have potential applications in Thermal Energy Storage (TES) systems as Phase Change

Materials (PCMs), however eutectic salts with melting temperatures above 400°C are highly corrosive under air

atmospheres. A method for testing under inert atmosphere is presented and results from testing under isothermal and

thermal cycling conditions for two salts (NaCl/Na2CO3 and NaCl/Na2SO4) with melting temperatures of around 630°C are

discussed. Chromium oxides were present at the salt/metal interface whilst iron and nickel remained in metal phase.

INTRODUCTION

Corrosion of structural metals in thermal energy storage and heat transfer fluids for solar thermal power plants is

an important consideration for the design and lifetime of these systems. Conducting corrosion studies involving

molten salts and other high temperature media for solar thermal power plants often requires low oxygen

environments to control the salt acidity, and solubility of alloying elements [1 - 3]. Sealing furnaces and chambers is

difficult at the temperatures called for (400 – 900 °C) as traditional high temperature sealing materials are

unsuitable. This paper presents the results of a method for testing under these conditions which doesn’t require a

sealed furnace or chamber.

This process uses a crucible with an interference fitted lid machined from the alloy of interest. The thermal

energy storage media is introduced to the crucible under controlled atmosphere in a glove box, where the lid is press

fitted into the crucible. The press fit lid could then be welded to ensure that there is no ingress of oxygen, or egress

of test media. The sealed crucible can then be tested in a standard furnace, without the need for atmosphere control.

Obviously, in any system which is sealed at low temperature and then heated to high temperature, the pressure

change must be considered. The testing presented here, was undertaken with high purity salt eutectics, which had

been dried before being added to the crucible. This means that there should be little or no pressure increase due to

vaporization of the salt or impurities. Also, the volume of salt in the crucible was kept to less than half of the

headspace, to ensure that the influence of any vaporization or decomposition was minimized. Minimising the danger

of pressure in these vessels is a topic of future investigations.

METHODOLOGY

Four crucibles and lids were machined from 316L stainless steel bar. The components of two eutectic salts -

NaCl/Na2CO3 (melting temperature of 632 °C) and NaCl/Na2SO4 (melting temperature of 628 °C) - were dried at

200°C overnight to ensure the moisture content was minimized. 20 grams of salt was added to the crucibles in a

glove box under an argon atmosphere, and the lids press fitted onto the crucibles to produce four sealed crucibles,

two loaded with each salt. One crucible of each salt was added to a furnace set to run isothermally 650°C, whilst the

other two crucibles were added to a furnace set to cycle between 600 and 650°C, over a two hour period. The

thermocouple data for the first two cycles are shown in Figure 1. This thermal cycle over a range of 50°C was

chosen to reflect the likely thermal cycle experienced by a phase change based thermal energy storage system,

where a majority of the energy is stored as latent heat.

These crucibles were left to run for 32 days (768 hours, 384 cycles), after which they were cooled and removed

from the furnaces and sectioned with a dry cut on a precision saw. Dry sanding was performed - to prevent the salt

from dissolving - with 360, 600, 1000, 1200 and 2000 grit sandpaper to obtain a flat surface. Final polishing was

undertaken with 6 micron, 3 micron and 0.25 micron diamond paste using kerosene as a polishing fluid. A JEOL

7001F scanning electron microscopy (SEM) and an Oxford XMax 80 X-ray energy dispersive spectroscopy (EDS)

detector were used to characterize the interface between the salt and the metal. Samples of the salt were analysed

with a Perkin Elmer Optima 8300 DV inductively coupled plasma optical emission spectroscopy (ICP-OES) to

determine the elemental concentrations of metal alloys dissolved in the salt.

FIGURE 1. Thermocouple data for the 1st and 2nd thermal cycle. Furnace control was based on the average temperature.

FIGURE 2. Sectioned Crucibles prepared for Imaging a) NaCl/Na2CO3 Isothermal, b) NaCl/Na2CO3 Cycled, c) NaCl/Na2SO4

Isothermal, d) NaCl/Na2SO4 Cycled

RESULTS

NaCl/Na2CO3 crucibles

The interfaces between the crucibles and NaCl/Na2CO3 salt are depicted in Fig. 3 and Fig. 4. Both crucibles have

primarily chromium oxide layers, with the other main alloying elements remaining as metal. This is expected in a

stainless steel alloy, as chromium is the most easily oxidized element and will scavenge any available oxygen

(oxidation potential of Cr is +0.73 V, compared to +0.44 for Fe and +0.23 for Ni).

The isothermally tested crucible/salt interface (Fig. 3 and Fig. 4, a) consists of a very thin (approximately 1 μm)

chromium oxide layer in contact with the salt, which has flaked away from the wall in certain places (Fig 3). Under

this layer is an iron/nickel metal layer 4 – 10 μm thick, which is the result of chromium depletion into the salt.

Underneath is another chromium oxide layer 3 – 8 μm thick, then a region where chromium oxide and iron/nickel

metal are finely intermingled. This corrosion layer structure is different to that developed in this salt with high levels

of oxygen present. In this situation, the outer layer in contact with the salt is iron oxide, as chromium oxide in

contact with the salt is dissolved into the salt [4].

A similar result was observed in a recent study conducted under a reduced-oxygen study with Mg2Cl/KCl/NaCl

salt where chromium depleted iron/nickel metal was spalled off the 310 stainless steel substrate [5]. However, this

study found oxygen attack penetrating the steel to over 50 µm, suggesting that oxygen or other impurities were

present in their experiment at higher levels than in these sealed crucibles.

FIGURE 3. SEM Electron Backscatter images of Crucible/Salt interface with NaCl/ Na2CO3 salt tested isothermally at 650 °C

The thermally cycled crucible (Fig 4, b) shows similar chromium oxide and metal phases to the isothermally

tested crucible, but arranged in multiple layers. The total damaged depth is also similar between the two samples.

The thin outer chromium oxide layer is again in contact with the salt, but there is no clear delineation between the

subsurface chromium oxide and iron/nickel metal phases. The chromium oxides appear to be connected with a

network of pathways – possibly cracks or grain boundaries – which have allowed oxygen to penetrate into the metal

and form subsurface chromium oxide layers.

Both crucibles have sodium present in all areas of chromium oxide, but no chloride, indicating that the sodium

may form a compound with the chromium and oxygen. The presence of sodium in 316L corrosion layers has been

noted before, in high oxygen environments where it was present in the outer iron oxide layer [4]. Also of note, is the

small particles present on the thin chromium oxide outer layer in both crucibles are primarily iron, with high levels

of manganese, which is a minor alloying element in 316L stainless steel.

(a) (b)

FIGURE 4. SEM Electron Backscatter images and EDS Maps (outlined in green) of Crucible/Salt interface of

NaCl/ Na2CO3 salt under a) 650 °C isothermal, b) 600 – 650 °C cycled, conditions

NaCl/Na2SO4 crucibles

The salt/metal interface of the crucibles filled with NaCl/Na2SO4 salt are depicted in Fig 5. Significantly less

oxide is present in both cases compared to the crucibles with the NaCl/Na2CO3 salt, although the thermally cycled

crucible does have some chromium oxide present at the interface. This could be caused by oxide solubility being

reduced at lower temperatures and when the salt is solid, which slows the dissolution of any oxides produced on the

metal surface.

Some chromium depletion is present at the metal surface, probably due to the oxide formation. What has

occurred in both of these samples is there is a high density of particles present in approximately the outer 8 μm of

the metal (seen as dark inclusions in the backscatter image). The exact nature of these particles is not clear from the

EDS as the emission voltage of sulphur Kα and molybdenum Lα shell overlap at 2.3 keV, and both are present in

the sample. As heavy elements backscatter electrons more strongly that light elements, and these particles show up

dark in these images, these particles could be a result of sulphur ingress and precipitation in the metal.

(a) (b)

FIGURE 5. SEM Electron Backscatter images and EDS Maps (outlined in green) of Crucible/Salt interface of NaCl/ Na2SO4

salt under a) 650 °C isothermal, b) 600 – 650 °C cycled, conditions

Alternatively, these particles could be a result of bulk precipitates moving toward the surface of the metal. This

would require the precipitates to be of lower molecular mass than the bulk metal solution, i.e. some form of metal

carbide or nitride. This is supported by the presence of manganese - which is a strong carbide forming element [6] -

in this area, particularly the particle in the bottom left of the EDS map in Fig 5. a). An EDS point spectrum is

included in Fig. 6 for this point, and indicates a high manganese and sulfur or molybdenum content.

FIGURE 6. EDS Point spectrum of Point 1 in Fig. 5, a)

Salt Impurities

Table 1 depicts the metal alloying element impurity content in the salt in each crucible after testing. As can be

seen, the NaCl/Na2CO3 salt tested isothermally has the highest impurity level, with the relative impurity level

compared to the thermally cycled sample 4 – 10 times greater. The significantly lower concentration of all alloy

elements in the salt for the cycled sample is probably due reduced time under molten conditions. Also the outer

oxide layer appears to remain attached to the wall in the thermally cycled samples, whereas it detaches and is

moving into the salt in the isothermally tested sample (Fig 3.).

The NaCl/Na2SO4 salt contained significantly less impurities, and the levels were fairly consistent between the

thermally cycled and isothermal samples. The lack of oxide scale in these crucibles may account for this difference.

TABLE 1. Concentration of elements present in salt of each crucible (ppm)

Salt Conditions Fe Cr Ni Mo Mn Si

Base alloy (SS316L) Bal 16.0 – 18.0 10.0 - 14.0 2.0 - 3.0 0 - 2.0 0.75

NaCl:Na2CO3 Isothermal 4887.477 1117.060 753.176 370.417 127.314 157.260

NaCl:Na2CO3 Cycled 788.744 173.394 79.525 64.784 29.756 155.439

NaCl:Na2SO4 Isothermal 369.120 78.905 37.810 2.657 11.652 24.399

NaCl:Na2SO4 Cycled 419.058 90.261 40.971 3.660 17.559 18.946

CONCLUSION

This work demonstrates that there is appreciable differences in the scale structure and impurities present in the

salt from thermally cycled testing compared to isothermal testing of stainless steel in NaCl/Na2CO3 salts. Primarily

chromium oxides were produced in these tests, with both iron and nickel remaining in the metal phase.

There was little difference observed for the samples tested in the NaCl/Na2SO4 salt, although both samples

produced higher numbers of particles in the surface layer of the metal. These may have been produced by sulfur

ingress into the sample, or movement of bulk precipitates toward the surface.

The findings presented here indicate that this sealed crucible method can be used to conduct corrosion tests

under controlled atmospheric conditions, for both isothermal and thermal cycling. These conditions reflect the likely

operational conditions of a thermal energy storage system, or heat transfer fluid. The results obtained differ

significantly from tests undertaken in air, suggesting that this method of testing may be of use for investigating

molten salt corrosion under low oxygen environments.

ACKNOWLEDGMENTS

The data reported in this paper were obtained at the Central Analytical Research Facility operated by the Institute

for Future Environments (QUT). This work is supported by the Australia Solar Thermal Research Institute (ASTRI)

and received funding from ARENA as part of ARENA's International Engagement Program. The views expressed

herein are not necessarily the views of the Australian Government, and the Australian Government does not accept

responsibility for any information or advice contained herein.

REFERENCES

1. H. Fathollahnejad, B. H. Tsao, R. Ponnappan, D. Jacobson, J. Mater. Eng. Perform, 2, 125-134 (1993)

2. K. Shridharan and T. R. Allen, "Corrosion in Molten Salts,” in Molten Salts Chemistry, edited by F. Lantelme

and H. Groult (Elsevier, Oxford, 2013), pp. 241-267.

3. A. M. Kruizenga, Corrosion Mechanisms in Chloride and Carbonate Salts (SAND2012-7594) (Sandia National

Laboratories, Albuquerque, 2012)

4. M. Liu, S. Bell, S. Segarra, N.H.S. Tay, G. Will, W. Saman, F. Bruno, Sol. Energy Mater. Sol. Cells, 170, 1-7

(2017).

5. W. Ding, H. Shi, Y. Xiu, A. Bonk, A. Weisenberger, A. Jianu, T. Bauer, Sol. Energy Mater. Sol. Cells, 184, 22-

30 (2018).

6. V. Lazic, D. Milosavljevic, S. Aleksandrovic, P. Marinkovic, G. Bogdanovic, and B. Nedeljkovic, Tribology in

Industry, 32, 11 – 20 (2010).