Journal Search Engine

View PDF Download PDF Export Citation Korean Bibliography PMC Previewer
ISSN : 1738-1894(Print)
ISSN : 2288-5471(Online)
Journal of Nuclear Fuel Cycle and Waste Technology Vol.22 No.1 pp.67-80

Feasibility Study on Aluminum Under Laser Ablation for Corrosion Resistance in Molten Salt

Peggy T. Milota*, Supathorn Phongikaroon
Virginia Commonwealth University, 401 W Main St, Richmond, VA 23284, United States
* Corresponding Author. Peggy T. Milota, Virginia Commonwealth University, E-mail:, Tel: +1-804-827-2278

August 25, 2023 ; September 14, 2023 ; October 13, 2023


Fundamental aspects of creating passivation layers for corrosion resistance in nuclear engineering applications, specifically the ability to form complete layers versus porous ones, are being explored in this study. Utilizing a laser ablation technique, 1,064 nm fire at 10 Hz with 60 pulses per shot and 0.5 mm between impact points, aluminum samples are treated in an attempt to create a fully formed passivation layer that will be tested in a LiCl-KCl eutectic salt. By placing these samples into an electrochemical environment mimicking a pyroprocessing system, corrosion rates, resistances and material characteristics are tested for one week and then compared between treated and untreated samples. In initial testing, linear sweep voltammetry indicates corrosion current density for the untreated sample at −0.038 mA·cm−2 and treated samples at −0.024 mA·cm−2 and −0.016 mA·cm−2, respectively. This correlates to a control sample corrosion rate of −0.205 mm·yr−1 and treated rates of −0.130 mm·yr−1 and −0.086 mm·yr−1 for samples 1 and 2. In addition, electrochemical impedance spectroscopy circuits show application of a longer-lasting porous passivation layer on the treated metal, compared to the naturally forming layer. However, the current technique fails to create a uniform protection layer across the sample.


    1. Introduction

    Both pyroprocessing technology and molten salt reactors require compatible materials that can withstand the inherently aggressive environment of molten chloride salts. Raiman et al. suggested that the purity of the molten salt environment most affected corrosion rates [1]. Fission products, moisture, and oxide corrosion materials all play into the salt’s impurity and how it reacts within the system [2]. Because of this contamination, naturally occurring corrosion creates some protection via incomplete passivation layers, but these same impurities create instability in the film through localized attacks [3].

    There are many different explorable avenues to mitigate this corrosion. While high quality, high-cost alloy material may be implemented, controlling redox potential through chemical [2] and electrochemical [4] means has also proven to be viable. Alternative solutions have been posed by other works in terms of various metal treatments. For example, high temperature environments have been shown to facilitate alumina formation on alloys for use in molten salts [5] and laser ablation treatments in other research fields have proven useful in the improvement of corrosion resistance in various materials [6]. The latter approach has not been tested in molten salt, however.

    Thus, this becomes the motivation of our study to determine the feasibility of using the laser ablation technique to form a passivation layer on metals/alloys intended for use in a molten chloride environment. In addition, it is our intent to compare the treated corrosion rates/resistances to samples that have had no laser ablation pretreatment. We hope to provide a practical path of achieving a lower cost option and possibly another passivation layer method. High purity aluminum was selected as the initial test material (despite its unlikely use in molten salt systems as a structural material) due to its availability as part of an initial proof of concept; specifically, it can easily form an alumina layer between the abundance of aluminum present in the sample and atmospheric oxygen. Additionally, the first ionization energy of aluminum is comparable to that of iron, which ideally provides a similitude for stainless steel sample experiments in the future. The molten chloride system used is a eutectic LiCl-KCl at 500°C to simulate a pyroprocessing environment and due to an inherent basicity of the salt under these conditions.

    2. Experimental Programs

    2.1 Sample Preparation

    Aluminum pre-samples (approximately 5 mm × 10 mm × 6 mm) and test samples (approximately 50.5 mm × 10 mm × 6 mm) were sanded in succession from 100 grit to 1,500 grit with SiC paper, rinsed with deionized water and cleaned by alcohol wipe before sonication in ultra-pure water at room temperature. Control pre-samples were then examined using Scanning Electron Microscopy-Energy- Dispersive X-ray Spectroscopy (SEM-EDS; Tabletop SEM-Phenom ProX with EDX detector) for baseline conditions (Table 1).

    Table 1

    Initial elemental analysis

    Element symbol Atomic% conc. Weight% conc.

    Al 96.9 91.4
    Se 2.17 5.98
    As 0.990 2.59

    2.2 Laser Ablation Techniques

    A Quantel Q-Smart 450 mJ Nd:YAG pulsed laser at 1,064 nm wavelength was used to ablate the surface of the aluminum coupons. The laser was aligned such that the beam interacted with the metal samples mounted to an XYZ-axis manual linear translation stage that was bolted in place on a raised platform (Fig. 1). The translation stage allowed controlled movement across the length, width and depth of the sample for uniform coverage (Fig. 2).

    Fig. 1

    Q-Smart 450 laser and sample stage setup. Green arrows represent the path of the ablation laser beam.

    Fig. 2

    Sample verification of laser alignment and coverage achieved (5 mm beam).


    Initial treatment experiments began with 45° single shots of 1,064 nm wavelength, 5 mm diameter beam laser fire in the open atmosphere (approximately 21°C). This was repeated over the width and height of the pre-sample piece, with overlap sufficient to ensure total coverage. Additional pre-sample trials were attempted with 20, 60 and 100 pulses per shot, at 1 Hz and 10 Hz, with decreasing spacing between shots and after being focused and collimated through LA4148 UV Fused Silica Plano convex and Thorlabs LD2297-A Bi-Concave lenses, respectively (Table 2). Each sample was re-evaluated after treatment, via SEM, to analyze topography and material composition to assess the surface for oxygen content.

    Table 2

    Laser configuration trials for oxide creation treatment

    Trials Beam diam. (mm) Shots Freq (hz) Spacing (mm) Oxygen wt%

    1 5.0 Single 1.0 1.5 0
    2 5.0 20 1.0 1.5 0
    3 5.0 60 10 1.5 0
    4 5.0 100 10 1.5 0
    5 1.0 (ang)1 Single 1.0 1.5 4.9
    6 1.0 (ang)1 Single 1.0 0.50 5.3
    7 1.0 (ang)1 60 1.0 0.50 30
    8 <1.0 (col)2 Single 1.0 0.50 5.5
    9 <1.0 (col)2 Single (w/ DI)3 1.0 0.50 7.9
    10 <1.0 (col)2 10 1.0 0.50 3.1
    11 <1.0 (col)2 30 1.0 0.50 3.6
    12 <1.0 (col)2 10 10 0.50 11
    13 <1.0 (col)2 30 10 0.50 18
    14 <1.0 (col)2 60 10 0.50 23.7

    1: Laser impact at an angle other than 45° due to focusing of incoming beam

    2: Beam collimated after focusing to reestablish 45° impact angle

    3: Sample coated with drops of DI water to attempt to introduce additional oxygen

    Trial 5- first configuration where oxygen was discovered

    Trial 14- treatment configuration for initial experiment

    Upon verification of oxygen bonding on the surface, further laser parameter adjustments were made in order to find a balance between oxygen content and time to completion. An arrangement of a collimated beam of less than 1 mm (approximately 400 μm), with 60 pulses per shot, at 10 Hz and with 0.5 mm between impact points was used for treatment of the full-sized test samples (Fig. 3). This arrangement netted ~24wt% oxygen on the pre-samples that were ran through SEM for final analysis, where initial oxygen arrangements were only ~5% (see Table 3).

    Fig. 3

    Fully ablated test sample.

    Table 3

    Final elemental analysis

    Element symbol Atomic% conc. Weight% conc.

    Al 49.2 56.0
    O 35.0 23.7
    Th 0.81 7.97
    N 13.1 7.77
    Se 1.20 3.94
    Mg 0.630 0.65

    2.3 Electrochemical Testing in Molten LiCl-KCl Salt

    To calculate and compare corrosion rates, an untreated coupon of aluminum was subjected to electrochemical analysis in a LiCl-KCl salt bath inside a FB1400 Model Thermolyne muffle furnace within an argon filled glovebox. This electrochemical cell setup included a 30 g mixture of 55.70LiCl-44.30KCl eutectic mixture (12.54 g LiCl and 17.47 g KCl) prepared in an alumina crucible under an argon atmosphere that limited contaminants to <5 ppm for O2 and H2O. The salt mixture was then dried at 250°C for 2 hours, before being raised to just above eutectic temperature (506±5°C, to reach 500°C at the center of the salt) at a rate of approximately 100°C every 30 minutes. The furnace remained at 506°C for over 4 hours, which ensured sufficient temperature throughout the melt and allowed proper time for the salts to homogenize. Instrumentation was inserted into the molten salt as part of the electrochemical cell as follows: a tungsten rod counter electrode was used for current measurement, a 10mol% Ag/AgCl reference electrode inside a Pyrex tube allowed for potential reference measurements and the sample was connected to a molybdenum wire as the working electrode (Fig. 4). This setup was repeated for the test sample.

    Fig. 4

    Electrochemical cell setup used for OCV, LSV and EIS corrosion analysis.


    For both the control and subsequent test sample, each instrumentation group was connected to EC-Lab software (BioLogic - Science Instruments, software version V11.36) for measurement and analysis and remained within the salt bath for a period of one week to allow sufficient time for corrosion induction and electrochemical data collection. These electrochemical techniques included 24-hour rounds of open-circuit voltage (OCV), then electrochemical impedance spectroscopy (EIS) and linear sweep voltammetry (LSV). Final layer topography and elemental analysis for the treated sample will be included in future work.

    3. Results and Discussion

    Initial topography and elemental analysis of the presamples verified no oxygen was present in the control coupon (see Table 1 for initial analysis data), verifying that any future oxygen was likely the result of the ablation process. The trial images showed only a removal of imperfections and stray material present on the surface of the control sample (Figs. 5(a) and 5(b)). Elemental analysis of initial and subsequent samples reveals a lack of oxide formation at any point on the sample for any beam diameter over 1 mm (Fig. 5(c)). Results listed in Tables 1 and 3 indicate that there were selenium and arsenic contents across all samples, but this is likely due to errors in raw SEM data use.

    Fig. 5

    Aluminum sample SEM - EDS topographies and elemental analysis. (a) Untreated control sample (pure aluminum), (b/c) Single shot 1,064 nm wavelength treatment (no oxygen present), (d/e) Treated aluminum sample 1,064 nm wavelength, 60 shot, 10 Hz treatment (with surface oxygen discovered).


    Focusing of the laser beam to under 1 mm imparted sufficient energy to create the initial 5wt% oxide layer (see trial 5 of Table 2) and then further focus to ~400 μm, collimated to return the impact to 45°, with 60 shots at a 10 Hz frequency (trial 14 of Table 2) allowed a 24wt% oxygen content along the surface (Figs. 5(d) and 5(e), Table 3). The OCV, EIS and analyzed LSV data also appear to support this passivation creation assumption, as described below.

    Open circuit voltage readings were taken for the control sample and both treated samples at the beginning of the experiment and at the end of one week. The initial OCV data (see Figs. 6(a), 6(b), and 6(c) below) shows a decrease in corresponding equilibrium potential between the control sample (−1.42 V) and treated samples (−1.39 V and −1.27 V, respectively). The shape of the final OCV graphs trend to similar final voltages, but see a similarly decreasing initial voltage upon initiation of the final test (decreasing by approximately 11% for sample 1 and 21% for sample 2, Figs. 6(b) and 6(c), respectively). The difference in the initial and end OCV graph shapes is due to known noise in the furnace used, causing drift within the potential readings, a similar observation to those reported by Killinger [7]. Additional deviations from the general expected shape can be seen in sample 2, with the jump in potential at 2,400 and 2,700 seconds, likely due to electrical interference. Overall, these tests convey a similar but higher starting potential for the treated samples and a similar final ending potential, while a smaller differential over the treated sample testing. This general range was expected due to the natural creation of the oxidation layer on the untreated sample in the salt over time and suggests a lower susceptibility to corrosion [7], comparatively, at each starting point while approaching a similar susceptibility at the end.

    Fig. 6

    EC Lab results for day 1 initial OCV and day 7 final OCV, (a) untreated, (b) first treated and (c) second treated samples. Tests varied between −10 V and 10 V with dtR at 0.5 s.


    Additionally, EIS graphs (or Nyquist plots) show the change in impedance versus real resistance on treated and untreated samples, at the initial submersion into salt and one week after introduction (Fig. 7(a)). Similar initial resistance and plot shapes are seen for initial EIS tests on both the control and treated samples, although sample 2 shows a noticeably higher initial resistance, comparatively. Both treated samples’ data at the end of the week reveals a significantly more rounded comparison of resistance and impedance, where the control sample has a flatter, more linear comparison. Sample two resistance at the end of the weeks begins much higher, partially due to the higher initial resistance measurement. Ultimately, an increase in resistance to impedance is seen for both treated and untreated samples over the week. There is a similar differential between the first treated sample and the control, however, there is a much larger difference for sample 2. Utilizing equivalent electrical circuits as described in the Zeng, et al. 2001 work to analyze further, achieving the goal of a fully formed protective scale would have resembled the two semi-circle impedance circuits (Fig. 7(c)) [8]. However, both samples appear to show only porous scale formation, specifically like that of partial circles and the long ~45-degree diffusion line (Figs. 7(a) and 7(b)).

    Fig. 7

    (a) EC Lab EIS data from zero to 144 hours for untreated and treated samples. Depiction of (b) porous scale impedance and (c) fully formed protective layer impedance for comparison [8].

    Fig. 8

    (a) Encinas et. al’s depiction set of porous scale and fully formed protective layer impedance evolution on steel over 164 hours [9] versus EIS data, to include untreated and treated electrolyte resistance and passivation layer resistance for two samples (b/c, respectively).


    Combining this knowledge with the presumption that alumina degradation will mirror the steel corrosion evolution of Encinas’ group (from 2019) allows both a hypothesis of the changes occurring within the corrosion cell over the week and a subsequent assumption in terms of the material’s corrosion resistance [9]. Specifically, both samples appear to show an increase in electrolyte resistance, presumably due to corrosion causing material concentration increase in the salt (Figs. 8(b) and 8(c)). However, at the 144-hour mark, there is a distinct difference in shape between the untreated and treated samples’ impedance. Based on the work of Encinas, the untreated sample appears to be further along in the passivation layer deterioration, while both treated samples appear to remain just within the protective layer resistance semi-circle before transitioning to mass diffusion [9].

    The LSV analysis initially created graphs of current and potential proportionality; in addition the data was then used for Tafel plot creation and ultimately to approximate corrosion rate. This was done by utilizing the Stern-Geary equation through Faraday’s Law, and therefore corrosion rate was calculated by using the following expression:

    C o r r o s i o n R a t e [ m m y r ] = i c o r r K ρ a l l o y ( f i n i M W i )

    where K is the correlation constant defining units of corrosion rate, ρalloy is the alloy density (g·cm−3), fi is the mole fraction of the element i in the alloy, ni is the electrons transferred in element i, and MWi is the atomic weight of element i (amu). In this study, we let K = 3.27 (g·mm)/ (u·yr·mA·cm) [10] for mm·yr−1, ρalloy = 2.70 g·cm−3, f = 1 for pure aluminum, n = 6.00 for the six electrons transferred in the Al2O3 reaction and the MW = 27.0 u for 27Al.

    Derivation of the corrosion current density (icorr) was then done by finding the anodic and cathodic branch equations from each sample’s Tafel graph [10]. Tafel plots are the graphs of corrosion density (the log of dividing each current quantity by current average, from the LSV data files, over working electrode area) versus potential (also from the LSV data file) for each sample. These graphs create branch-like curves whose best-fit equations intercept at a point equal to icorr (Fig. 9). Utilizing Excel and Python to generate best fit equations and intercepts, this gave an untreated corrosion current density of −0.038 mA·cm−2 and treated current density of −0.024 mA·cm−2 for sample 1 and −0.016 mA·cm−2 for sample 2.

    Fig. 9

    Tafel plots created from LSV data including icorr derivation for pretreatment (control) sample and (a) treated sample 1 and (b) sample 2.


    Inputting this into the aforementioned Stern-Geary equation corresponded to an untreated corrosion rate of approximately −0.205 mm·yr−1 and treated rates of −0.130 mm·yr−1 and −0.086 mm·yr−1, a decrease of 36.5% and 58.2%, respectively. This relies, however, upon the interpretation of the best fit line for the Tafel data and how data is used within the best fit scenario (as shown in Fig. 10). By altering the data region to be used as part of the best fit, a series of line equations, intercepts and corresponding corrosion rates can be calculated (Figs. 10 and 11).

    Fig. 10

    (a) Tafel plot analysis of pretreated sample, with (b) examples of branch best fit equations due to data set manipulation.

    Fig. 11

    Tafel plot analysis of (a) sample 1 and (b) sample 2, with examples of branch best fit equations, prepared as shown in Fig. 10.


    Comparing the changes in anodic and cathodic branch data and how those changes affected branch equations ultimately gave a range of corrosion density values to input into the corrosion rate equation (Table 5). Specifically, depending upon the potential change of the sample, distinct changes were seen in the cathodic branch of the untreated sample (with only small variations in the anodic branch) while both branches of the treated samples saw similar corrosion rate changes with a change in data analysis (Table 4). Overall, even at the most extreme edges of data analysis in which the current density shows the largest deviations, the treated samples showed smaller corrosion rates than the untreated samples.

    Table 4

    Tafel branch best fit line equations (with differing sections of LSV data utilized) and the corresponding corrosion rates

    Pretreated corrosion rates (mm·yr−1)
    1 2 3 4 5
    Anodic branch equations: y=m‧x+bJNFCWT-22-1-67_T4-F1.gif
    m b m b m b m b m b
    Cathodic branches:
    y = m‧x+b
    6.535 −1.075 6.754 −1.072 6.340 −1.080 6.363 −1.079 6.317 −1.081
    JNFCWT-22-1-67_T4-F2.gif m b
    A −14.17 −1.853 −0.2048 −0.2034 −0.2054 −0.2054 −0.2054
    B −14.01 −1.848 −0.2050 −0.2035 −0.2056 −0.2056 −0.2057
    C −13.04 −1.818 −0.2068 −0.2053 −0.2075 −0.2075 −0.2076
    D −15.28 −1.880 −0.2009 −0.1996 −0.2014 −0.2014 −0.2015
    E −15.93 −1.898 −0.1995 −0.1983 −0.2001 −0.2001 −0.2002
    Sample 1 Treated corrosion rates (mm·yr−1)
    6 7 8 9 10
    Anodic branch equations: y=m‧x+bJNFCWT-22-1-67_T4-F3.gif
    m b m b m b m b m b
    Cathodic branches:
    y = m‧x+b
    17.33 −0.9786 12.48 −1.064 10.79 −1.095 8.769 −1.142 6.528 −1.197
    JNFCWT-22-1-67_T4-F4.gif m b
    F −15.73 −1.768 −0.1300 −0.1360 −0.1383 −0.1389 −0.1394
    G −15.72 −1.751 −0.1296 −0.1301 −0.1378 −0.1387 −0.1394
    H −13.12 −1.710 −0.1308 −0.1375 −0.1400 −0.1411 −0.1421
    I −8.049 −1.608 −0.1352 −0.1445 −0.1481 −0.1509 −0.1536
    J −5.770 −1.552 −0.1352 −0.1457 −0.1503 −0.1536 −0.1571
    Sample 2 Treated corrosion rates (mm·yr−1)
    11 12 13 14 15
    Anodic branch equations: y=m‧x+bJNFCWT-22-1-67_T4-F5.gif
    m b m b m b m b m b
    Cathodic branches:
    y = m‧x+b
    26.07 −1.124 15.47 −1.206 26.61 −1.119 24.62 −1.135 19.88 −1.171
    JNFCWT-22-1-67_T4-F6.gif m b
    K −6.322 −1.590 −0.0857 −0.0850 −0.0877 −0.0953 −0.1051
    L −5.659 −1.590 −0.0850 −0.0844 −0.0866 −0.0942 −0.1033
    M −4.973 −1.584 −0.0877 −0.0871 −0.0893 −0.0978 −0.1084
    N −4.156 −1.594 −0.0882 −0.0877 −0.0904 −0.0991 −0.1100
    O −6.868 −1.593 −0.0926 −0.0920 −0.0948 −0.1051 −0.1176
    Table 5

    Tafel branch best fit line intercepts, as icorr and Ecorr, and the corresponding corrosion rates

    Pretreated corrosion rates Treated 1 corrosion rates Treated 2 corrosion rates

    i(corr) E(corr) CR = (k-mass-icorr)
    i(corr) E(corr) CR % Diff i(corr) E(corr) CR % Diff

    mA-cm−2 V mm-y−1 mA-cm−2 V mm-y−1 (vs 1A) mA-cm−2 V mm-y−1 (vs 1A)

    1A −0.03761 −1.321 −0.2048 6F −0.02387 −1.392 −0.1300 36.52 11K −0.01573 −1.499 −0.08565 58.18
    1B −0.03764 −1.321 −0.2050 6G −0.02379 −1.391 −0.1296 36.74 11L −0.01560 −1.495 −0.08496 58.52
    1C −0.03798 −1.323 −0.2068 6H −0.02402 −1.395 −0.1308 36.13 11M −0.01610 −1.507 −0.08769 57.19
    1D −0.03688 −1.316 −0.2009 6I −0.02482 −1.409 −0.1352 34.00 11N −0.01620 −1.510 −0.08823 56.92
    1E −0.03664 −1.314 −0.1995 6J −0.02482 −1.409 −0.1352 34.00 11O −0.01700 −1.529 −0.09259 54.79
    2A −0.03734 −1.324 −0.2034 7F −0.02497 −1.375 −0.1360 33.60 12K −0.01560 −1.499 −0.08496 58.52
    2B −0.03737 −1.325 −0.2035 7G −0.02389 −1.374 −0.1301 36.47 12L −0.01549 −1.496 −0.08436 58.81
    2C −0.03770 −1.327 −0.2053 7H −0.02525 −1.379 −0.1375 32.85 12M −0.01600 −1.507 −0.08714 57.45
    2D −0.03664 −1.320 −0.1996 7I −0.02653 −1.395 −0.1445 29.45 12N −0.01610 −1.511 −0.08769 57.19
    2E −0.03641 −1.318 −0.1983 7J −0.02675 −1.398 −0.1457 28.87 12O −0.01690 −1.530 −0.09204 55.06
    3A −0.03771 −1.319 −0.2054 8F −0.02540 −1.369 −0.1383 32.46 13K −0.01610 −1.497 −0.08769 57.19
    3B −0.03775 −1.319 −0.2056 8G −0.02530 −1.368 −0.1378 32.72 13L −0.01590 −1.493 −0.08660 57.72
    3C −0.03810 −1.321 −0.2075 8H −0.02570 −1.373 −0.1400 31.66 13M −0.01640 −1.505 −0.08932 56.39
    3D −0.03698 −1.314 −0.2014 8I −0.02720 −1.389 −0.1481 27.67 13N −0.01660 −1.509 −0.09041 55.86
    3E −0.03674 −1.313 −0.2001 8J −0.02760 −1.393 −0.1503 26.61 13O −0.01740 −1.528 −0.09477 53.73
    4A −0.03772 −1.319 −0.2054 9F −0.02550 −1.366 −0.1389 32.19 14K −0.01750 −1.489 −0.09531 53.46
    4B −0.03775 −1.320 −0.2056 9G −0.02546 −1.366 −0.1387 32.30 14L −0.01730 −1.485 −0.09422 54.00
    4C −0.03810 −1.322 −0.2075 9H −0.02590 −1.370 −0.1411 31.13 14M −0.01796 −1.497 −0.09782 52.24
    4D −0.03698 −1.315 −0.2014 9I −0.02770 −1.385 −0.1509 26.34 14N −0.01820 −1.501 −0.09912 51.60
    4E −0.03674 −1.313 −0.2001 9J −0.02820 −1.389 −0.1536 25.01 14O −0.01930 −1.521 −0.1051 48.68
    5A −0.00372 −1.319 −0.2054 10F −0.02560 −1.365 −0.1394 31.92 15K −0.01930 −1.478 −0.1051 48.68
    5B −0.03776 −1.319 −0.2057 10G −0.02560 −1.364 −0.1394 31.92 15L −0.01896 −1.474 −0.1033 49.58
    5C −0.03811 −1.321 −0.2076 10H −0.02610 −1.368 −0.1421 30.59 15M −0.01990 −1.487 −0.1084 47.08
    5D −0.03699 −1.314 −0.2015 10I −0.02820 −1.381 −0.1536 25.01 15N −0.02020 −1.492 −0.1100 46.28
    5E −0.03675 −1.313 −0.2002 10J −0.02885 −1.386 −0.1571 23.28 15O −0.02160 −1.512 −0.1176 42.56

    *Bold column headers indicate anodic and cathodic branch pairs from Table 4 above.

    Additionally, corrosion area of the treated sample was measured and calculated via calipers and density/mass manipulation (Table 6). Initial size measurements and mass data were taken prior to the salt bath, corroded area and final measurements were taken after removal from the salt bath and density calculations were done for data unable to be directly measured (i.e. the corroded area density was calculated by finding the mass of the corrosion section divided by the measured corroded volume). Ultimately, these measurements showed a decrease in untreated sample density from 0.02630 g·mm−3 to 0.02580 g·mm−3, with an increase in comparable electrode surface area from 748.0 mm2 to 759.0 mm2 after testing. For the treated samples, the average density of treated sample 1 decreased from 0.02630 g·mm−3 to 0.02606 g·mm−3 and area increased from 982.7 mm2 to 1,004 mm2, where sample 2 average density decreased from 0.02630 g·mm−3 to 0.02609 g·mm−3 and area increased from 683.8 mm2 to 697.2 mm2 after removal from the molten salt. Initial area values were calculated by utilizing control width and depth measurements with the electrode height of each treated sample. This increase in area and overall mass but slight decrease in average density is consistent with growth of the passivation layer (that is larger and denser than the plain metal) but suggests a loss of material from within that volume due to anticipated, localized pitting.

    Table 6

    Mass and dimensional measurements for initial and corroded samples 1 and 2

    Corrosion data

    Initial Final (control) Final (treated 1) Final (treated 2)

    Mass (g) 8.318 Mass (g) Trials 8.266 Mass (g) Trials 8.410 Mass (g) Trials 8.370

    8.266 8.410 8.370
    8.266 8.410 8.370
    Avg: 8.266 Avg: 8.410 Avg: 8.370

    Height (mm) Trials 50.50 Height (mm) Trials 20.61 Height (mm) Trials 28.32 Height Trials 19.40

    50.50 21.33 28.48 19.25
    50.50 21.33 28.13 18.69
    50.50 Avg: 21.09 Avg: 28.31 Avg: 19.11

    Length (mm) 9.980 Length (mm) Trials 10.05 Length (mm) Trials 10.16 Length (mm) Trials 10.18

    10.23 10.19 10.16
    10.03 10.19 10.16
    Avg: 10.10 Avg: 10.18 Avg: 10.17

    Depth (mm) 6.270 Depth (mm) Trials 6.310 Depth (mm) Trials 6.450 Depth (mm) Trials 6.380

    6.410 6.300 6.370
    6.380 6.440 6.380
    Avg: 6.367 Avg: 6.397 Avg: 6.377

    Vm (mm3) 3,160 Vm (mm3) 1,840 Vm (mm3) 1,389 Vm (mm3) 1,964

    Vc (mm3) 0 Vc (mm3) 1,356 Vc (mm3) 1,843 Vc (mm3) 1,239

    Vt (mm3) 3,160 Vt (mm3) 3,197 Vt (mm3) 3,232 Vt (mm3) 3,203

    ρm (g‧mm−3) 2.632×10–3 ρm (g‧mm−3) 2.632×10–3 ρm (g‧mm−3) 2.632×10–3 ρm (g‧mm−3) 2.632×10–3

    ρc (g‧mm−3) 0.00 ρc (g‧mm−3) 2.522×10–3 ρc (g‧mm−3) 2.579×10–3 ρc (g‧mm−3) 2.579×10–3

    <ρ> (g‧mm−3) 2.632×10–3 <ρ> (g‧mm−3) 2.577×10–3 <ρ> (g‧mm−3) 2.606×10–3 <ρ> (g‧mm−3) 2.606×10–3

    Ac (mm2) 748.0 Ac (mm2) 759.0 Ac (mm2) 1,004 Ac (mm2) 697.2

    Vm: volume of the uncorroded metal    Vc: volume of the corroded metal

    Vt: total volume of the metal       ρm: density of the uncorroded metal

    ρc: density of the corroded metal     <ρ>: average density of the metal

    Ac: corroded area relative to height of corroded section of electrode

    4. Conclusion and Recommendations

    While the data seems to show a lack of fully formed passivation layer in the EIS circuit analysis, ablation as a method for formation of longer lasting porous passivation layers appears viable in terms of corrosion rate decrease. The corresponding increase in area, decrease in density, change in potential, and change in resistance all present similar possible interpretations of stronger, but still susceptible layers. These assumptions require experiment repetition, however.

    In addition, there were multiple anomalies noted through the data manipulation of this work: noticeably small corrosion rates derived by the equations, as well as a large jump in initial LSV resistance on both samples. The former will require verification based on future work that includes layer analysis by SEM, where the latter is likely due to either the naturally occurring passivation of each sample as it is introduced to salt or the small usable area of the counter electrode. Passivation of the samples is to be expected based on understanding of previous work and geometric changes made to adjust the counter electrode area by other groups are less likely to be viable when sized up into large scale pyroprocessing equipment and therefore seen as insufficient. While the passivation anomalies are expected to continue throughout the repetition and evolution of this work, contrast between the working and counter electrodes will be minimized via smaller working samples.

    To finalize the initial experiment, corroded sample SEM topography and Ionically Coupled Plasma – Mass Spectroscopy (ICP-MS; Agilent Technologies 7900 ICPMS detector) material identification will also be completed for layer analysis, specifically to better understand the density changes based on layer creation versus material loss. Future modifications to the experiment procedure include automation of the sample movement, as well as passivation layer analysis using Laser Induced Breakdown Spectroscopy (LIBS) rather than SEM-EDS. Upon solidification of the experimental procedure and setup with repeat results, additional laser-setting trials will be done in order to attempt to optimize layer uniformity and oxygen percentage. Finally, additional metals, alloys and coatings (SS, Hastelloy, etc.) will be tested in order to analyze the potential for increased corrosion resistance across multiple salt types/ environments.


    The Molten Salt Group would like to acknowledge and thank the contributions of Virginia Commonwealth University’s Graduate Teaching Assistantship funding for making this work possible.

    Conflict of Interest

    No potential conflict of interest relevant to this article was reported.




    1. S.S. Raiman and S. Lee, “Aggregation and Data Analysis of Corrosion Studies in Molten Chloride and Fluoride Salts”, J. Nucl. Mater., 511, 523-535 (2018).
    2. J. Zhang, C.W. Forsberg, M.F. Simpson, S. Guo, S.T. Lam, R.O. Scarlat, F. Carotti, K.J. Chan, P.M. Singh, W. Doniger, K. Sridharan, and J.R. Keiser, “Redox Potential Control in Molten Salt Systems for Corrosion Mitigation”, Corros. Sci., 144, 44-53 (2018).
    3. A.G. Fernández and L.F. Cabeza, “Corrosion Evaluation of Eutectic Chloride Molten Salt for New Generation of CSP Plants. Part 1: Thermal Treatment Assessment”, J. Energy Storage, 27, 101125 (2020).
    4. W. Ding, H. Shi, A. Jianu, Y. Xiu, A. Bonk, A. Weisenburger, and 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, 298-313 (2019).
    5. J.C. Gomez-Vidal, A.G. Fernandez, R. Tirawat, C. Turchi, and 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, 234-245 (2017).
    6. H. Wan, J. Lin, and J. Min, “Effect of Laser Ablation Treatment on Corrosion Resistance of Adhesive-Bonded Al Alloy Joints”, Surf. Coat. Technol., 345, 13-21 (2018).
    7. D. Killinger, “Molten Salt Technologies for Advanced Nuclear Fuel Cycles and Molten Salt Reactors”, Virginia Commonwealth University thesis, VCU Scholars Compass (2022).
    8. C.L. Zeng, W. Wang, and W.T. Wu, “Electrochemical Impedance Models for Molten Salt Corrosion”, Corros. Sci., 43(4), 787-801 (2001).
    9. V. Encinas-Sánchez, M.T. de Miguel, M.I. Lasanta, G. García-Martín, and 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, 157-163 (2019).
    10. D.A. Jones, Principles and Prevention of Corrosion, 2nd ed., 75-113, Prentice-Hall, New Jersey (1996).

    Editorial Office
    Contact Information

    - Tel: +82-42-861-5851, 866-4157
    - Fax: +82-42-861-5852
    - E-mail:

    SCImago Journal & Country Rank