Corrosion Science 203 (2022) 110351

Contents lists available at ScienceDirect

Corrosion Science
journal homepage: www.elsevier.com/locate/corsci

Study of KCl-induced hot corrosion behavior of high velocity oxy-fuel

sprayed NiCrAlY and NiCrBSi coatings deposited on 12CrMoV boiler steel
at 700 ℃
Xia Liu a, 1, Kai Hu a, b, 1, Shihong Zhang a, b, *, Tong Xu c, Lin Chen c, Eungsun Byon d,
Dongguang Liu e
a
Key Laboratory of Green Fabrication and Surface Technology of Advanced Metal Materials, Ministry of Education, Anhui University of Technology, Ma’anshan
243002, PR China
b
School of Materials Science and Engineering, Anhui University of Technology, Ma’anshan 243002, PR China
c
State Key Laboratory for Mechanical Behavior of Materials, School of Materials Science and Engineering, Xi’an Jiaotong University, Xi’an 710049, PR China
d
Material Processing Research Department, Korea Institute of Materials Science (KIMS), 66 Sangnam-dong, Changwon, Kyungnam 641-010, Republic of Korea
e
Industry & Equipment Technology, Hefei University of Technology, Hefei 230009, PR China

A R T I C L E I N F O A B S T R A C T

Keywords: The hot corrosion behavior of two kind of HVOF-sprayed NiCr-based coatings with different oxide-forming el­
Hot corrosion ements (such as Al and Si) was investigated and compared in the presence of KCl salt at 700 ℃. The results
NiCrAlY showed that the hot corrosion resistance of NiCrBSi coating was superior to that of NiCrAlY coating. NiCrAlY
NiCrBSi
coating was seriously corroded and severe internal oxidation was produced along the splat boundaries, while the
KCl salt
corrosion of NiCrBSi coating was relatively mild and the inner coating remained intact, which was ascribed to the
denser structure and the generation of a SiO2 protective layer.

1. Introduction exhibit better resistance to sulfide corrosion in coal-fired boilers, but do

not perform well for alkali chlorides. Meiβner et al. [9] studied the hot
The large-scale CO2 emission from fossil fuel combustion is a direct corrosion of ferritic-martensitic X20CrMoV12–1 steel in KCl-rich
trigger for the global greenhouse effect [1,2]. The Paris Agreement biomass co-firing environments, showing that the addition of KCl to
agreed in 2016 has set targets that try to keep global average tempera­ sulfate salt accelerates degradation of the substrate. Pettersson et al.
ture rise to below 2 ◦ C above pre-industrial levels [3]. Biomass accounts [10] explored the hot corrosion behavior of 304 stainless steel in KCl,
for 70% of global renewable energy and its main form of energy con­ K2SO4 and K2CO3 salts. Their result indicated that K2SO4 is less corrosive
version is direct combustion, which can generate electricity in biomass to 304 stainless steel but the corrosion is aggravated by the reaction of
power plants [4]. Furthermore, biomass burning has become an KCl and K2CO3. Hence, there is a strong demand to improve the corro­
important way to achieve carbon-neutrality, which is of great signifi­ sion resistance of boiler materials in an alkali chloride corrosion envi­
cance in reducing CO2 emissions [5]. However, different from coal fuel, ronment, but the traditional metal materials with superior hot corrosion
most biomass fuels contain more alkali metals, chlorine and other ele­ are usually expensive.
ments. Alkali chlorides formed after combustion in the boiler remain in In recent years, with the development of high-temperature coating
the ash and slag, and readily adhere to the heating surface to form an technology, the application of high-temperature coatings to the boiler
accumulation of ash [6]. Under the high temperature environment in the heating surface to prevent high-temperature corrosion has been widely
boiler, the heating surface would be corroded by alkali chloride in the adopted in experimental research and engineering applications. Among
ash, causing thinning and spalling of the oxide layer on the heating various coatings, the HVOF-sprayed NiCr-based coatings has been
surface [7,8]. Traditional ferrite, martensitic, and austenitic steels proved to be a workable proposition to protect boiler tubes due to its low

* Corresponding author at: Key Laboratory of Green Fabrication and Surface Technology of Advanced Metal Materials, Ministry of Education, Anhui University of
Technology, Ma’anshan 243002, PR China.
E-mail address: [email protected] (S. Zhang).
1
Xia Liu and Kai Hu contributed equally to this manuscript.

https://doi.org/10.1016/j.corsci.2022.110351
Received 25 September 2021; Received in revised form 27 April 2022; Accepted 29 April 2022
Available online 4 May 2022
0010-938X/© 2022 Elsevier Ltd. All rights reserved.
X. Liu et al. Corrosion Science 203 (2022) 110351

porosity and good corrosion resistance [11–13]. To date, many 2.2. Preparation of coatings
NiCr-based coatings have been subjected to hot corrosion experiments,
and the results showed that the higher the Cr content, the stronger the ZB-2000 HVOF equipment (Beijing Zhenbang Aerospace Precision
hot corrosion performance of the coating [14,15]; when there is alkali Machinery Co., Ltd, Beijing, China), manipulated with a Yaskawa
chloride present in the corrosive medium (for example, KCl), the pro­ (Japan) robot could be used to deposit coatings on all six sides of the
tective oxide film Cr2O3 will be destroyed due to the formation of substrates (15 mm × 15 mm × 5 mm) with a thickness of about 300 µm.
alkaline chromate [16,17]. However, it is not only chromium oxide that Before spraying, all six sides of the samples were sand blasted at an angle
plays a role in the hot corrosion process, aluminum oxide, silicon oxide, of 90◦ relative to the sample position using corundum sand. The average
etc., can also play a good blocking role in the hot corrosion process roughness of the sample surface was about 5.4 µm. The samples were
[18–20]. Therefore, the presence of other oxide film-forming elements then ultrasonically cleaned in alcohol, followed by cool air-drying to
in NiCr-based coatings will also exert a significant influence on the hot prepare them for spraying. The spraying process parameters are dis­
corrosion resistance of the entire system, but there are few related played in Table 2.
studies thereof.
As a commonly used NiCr-based coating, the hot corrosion perfor­ 2.3. Hot corrosion tests
mance of NiCrAlY and NiCrBSi has often been studied [21–24]. The
protection capability of the coating is improved by the selective oxida­ The hot corrosion tests were performed in a horizontal tube furnace
tion of Cr, Al or Si to form an oxide layer at high temperatures. The for 100 h at 700 ℃ in ambient air. The amount of salt applied was
NiCrAlY coating is already recommended for use in oxidizing conditions controlled at about 5–6 mg/cm2. The samples were placed in individual
due to the formation of a continuous and integrated Al2O3 layer [25]. covered alumina crucibles throughout each hot-corrosion experiment to
NiCrBSi self-fluxing alloy coating exhibits excellent wear and corrosion prevent interference between different samples and the effect of peeling
resistance in virtue of borides and oxides [26,27]. Fu et al. [28] studied spatter on the change in mass. Before the test, each sample was weighed
the corrosion resistance of ERNiCrMo-13 and NiCrBSi coatings in together with the crucible. During the test, the samples were extracted
simulate coal-fired boiler conditions with fly-ash composition. They with their crucibles at 10 h intervals and cooled to room temperature in
concluded that the SiO2 and Cr2O3 formed at the initial stage showed air for weighing. The KCl salt on the coating surface was not removed at
great corrosion resistance, and then NaAlSiO4 formed through the re­ each weighing. The balance was calibrated before each weighing.
action of oxides and the deposited fly-ash, thereby losing their protec­
tion with prolonged corrosion. Previous studies have focused on the 2.4. Characterization of coatings
corrosion behavior of coatings in the environment of coal-fired boilers or
some conventional oxidizing environment [29–31]. Whereas, with the The surface and cross section of the coatings after hot corrosion were
promotion of biomass boilers, alkali chloride corrosion produced by observed by Phenom XL scanning electron microscope (SEM; Phenom,
biomass combustion has become a more severe threat and the behavior The Netherlands) equipped with an energy dispersive X-ray spectro­
and mechanism of alkali chloride corrosion of related materials are scope (EDS). The macroscopic morphologies of the hot-corrosion spec­
worth investigating [32,33]. In addition, whether Al in NiCrAlY and Si imens were observed by using stereomicroscope. The phase structures of
in NiCrBSi can form a stable and protective oxide film within a lower as-sprayed coatings and hot corrosion products were identified by Ul­
temperature range, and whether the oxide produced can prevent the tima IV X-ray diffractometer (XRD; Rigaku, Japan) equipped with Cu-Kα
corrosion of alkali metal chloride and what effect it has on the blocking radiation, which was operated with a step scanning rate of 10◦ /min and
effect of chromium oxide film remain unknown and are thus worth diffraction angle (2θ) between 20◦ ~80◦ . In order to further determine
exploring. the coating phase, the FEI Tecnai G2 F20 transmission electron micro­
Given the above, NiCrAlY and NiCrBSi coatings were selected as scope (TEM; USA) was used to observe and analyze the coating. The
target coatings and prepared by HVOF technology on 12CrMoV boiler focused ion beam (FIB; Germany) was employed to thin the sample to
steel. The hot corrosion behavior of the coatings was investigated in the meet requisite TEM test conditions. To reveal the microstructures of
presence of KCl salt at 700 ℃ in ambient air. Meanwhile, attention was cross sections, the coatings were cut through the middle with a low-
paid to the effects of alloy elements such as Al and Si on the corrosion speed precision cutting machine and then cold-mounted in a low-
properties of coatings. Finally, the hot corrosion mechanism of the shrinkage epoxy resin to prevent spallation of the resulting hot corro­
coating was elucidated. The research may provide some useful infor­ sion products. The porosity of coating was calculated by image analysis
mation for material designers and aid understanding of hot corrosion software (Image J) using SEM cross-sectional images with a scale of 80
mechanism related to corrosion environments with alkali chloride salts. µm. Five measurements at different positions were performed and the
measured values were averaged. The bond strength of coating was test
2. Experimental procedure by WDW-50 electronic universal tensile testing machine. Each coating
was tested thrice and the mean average value used.
2.1. Materials

Boiler steel 12CrMoV (nominal composition in wt%; 0.30 Cr, 0.25

Mo, 0.15 V, Fe bal.) was used as matrix material. Two Ni-based gas-
atomized feedstock powders used were NiCrAlY powder and NiCrBSi
powder, and both powders contained grains with an equivalent spher­
Table 2
ical diameter of 15–45 µm. The compositions of the two sprayed pow­ Parameters for HVOF spraying.
ders are summarized in Table 1.
Variable Value

Oxygen flow, L/min 250

Propane flow, L/min 70
Table 1 Air flow, L/min 350
Chemical compositions of powders (unit: wt%). Spray distance, mm 250
Feed rate, g/min 50
Powder Ni Cr Al Y Si B Fe C
Gun velocity, mm/min 800
NiCrAlY Bal. 22 10 1.0 / / / / Track pitch, mm 3
NiCrBSi Bal. 16 / / 4.0 3.2 < 5.0 0.9 Number of cycles 12

2
X. Liu et al. Corrosion Science 203 (2022) 110351

3. Results and discussion 3.2. Hot corrosion behavior of NiCrAlY and NiCrBSi coatings

3.1. Characterization of as-sprayed coatings 3.2.1. Surface morphology and component analysis
Fig. 5 shows the macroscopic morphologies of NiCrAlY and NiCrBSi
Fig. 1 shows the surface topographies of the as-sprayed NiCrAlY and coatings after corrosion for 10 h, 50 h, and 100 h: there was no peeling
NiCrBSi coatings. Some unmolten or partially molten powder particles, and cracking of the corroded layer and there were yellowish green and
as well as minor amounts of spreading molten droplets were observed on gray products seen on the surface of the NiCrAlY coating in the early
the NiCrAlY coating. A large area of spreading molten droplets was stage of hot corrosion. With the increase of hot corrosion time, the
observed on the surface of NiCrBSi coating, indicating that the state of yellowish-green products were reduced in proportion and the surface
melting of powder particles was better than that of the NiCrAlY coating, was covered with more gray products. For NiCrBSi coating, white par­
however, there were pores between powder particles and molten ticles and brown corrosion products were apparent on surface after 10 h
droplets on both coating surfaces. The cross-sections and fracture mor­ of hot corrosion. As the exposure time increased, the white particles
phologies of as-sprayed NiCrAlY and NiCrBSi coatings are shown in disappeared, and the surface was covered with a brownish corrosion
Fig. 2. All the coatings were well-bound to the substrate with a bond product. As a result, with increased corrosion time, the types and con­
strength exceeding 60 MPa. The porosities measured from the cross- tents of corrosion products on the surface also varied to some extent.
sectional micrographs of NiCrAlY and NiCrBSi coatings were 0.30 ± The XRD patterns of corroded specimens with different exposure
0.01 and 0.29 ± 0.01, respectively. As can be seen from cross-section times are demonstrated in Fig. 6: the mixed corrosion products were
and fracture morphologies, more pores, cracks and splat boundaries, mainly composed of oxides and spinel phases. For NiCrAlY coating, the
and some unmolten powder particles were observed in the NiCrAlY corrosion products after 10 h of corrosion mainly consisted of Cr2O3,
coating. On the contrary, the microstructure of NiCrBSi coating was Y2O3/YAlO3 and K2CrO4. Furthermore, a weak CrCl2 peak was detected
dense and no obvious splat boundaries were visible under SEM imaging. on the surface. After 50 h of hot corrosion, the overall peak strength of
It can also be seen from the fracture morphology that the powders were the corrosion products increased, indicating that the content of the
completely molten and the inner layer of the coating presented a good corrosion products such as Cr2O3 and Y2O3/YAlO3 increased. Mean­
bonding state between layers. while, a small amount of NiO, Al2O3 and NiCr2O4 were detected in the
The XRD patterns of feedstock powders and as-sprayed coatings are scales. The XRD pattern after 100 h of exposure was similar to that at
illustrated in Fig. 3. The three primary peaks detected in NiCrAlY 50 h. Comparatively, the XRD pattern of NiCrBSi coating for 10 h of
powder and as-sprayed coating represented a γ/γ′ phase. The NiCrBSi corrosion showed that SiO2 and Cr2O3 were the major corrosion prod­
powder was composed of γ-Ni, Ni3B, Cr2B and CrSi2. The as-sprayed ucts, followed by some residual KCl phase. As the hot corrosion time was
NiCrBSi coating retained the phases of feedstock powder and the in­ increased to 50 h, the KCl disappeared and the peak intensity of SiO2
tensity of Cr2B and CrSi2 peaks diminished. Furthermore, it can be increased. The hot corrosion products at 100 h were consistent with
observed that the XRD peaks of both two coatings were broader than those at 50 h. In conclusion, for NiCrAlY coating, the main corrosion
those of the powders, as attributed to grain refinement and micro-strain products were Cr2O3, with a lower amount of K2CrO4. Al2O3 was pro­
from the rapid solidification of molten powder during thermal spraying duced after a long period of corrosion albeit its content was low, indi­
[34,35]. To determine the phases present, TEM characterization was cating that the formation of alumina was affected to a certain extent at a
adopted for NiCrBSi coating. Fig. 4 displays TEM observations of the lower temperature in KCl salt. The main corrosion phase in the NiCrBSi
NiCrBSi coating and the selected area electron diffraction (SAED) coating was SiO2, with lower contents of Cr2O3 and K2CrO4. Since metal
pattern of the grains corresponding to marked regions. Region A with its disilicides (MSi2) had a high Si concentration, they could easily form
irregular structure is rich in Ni, Cr and B, which is identified as Ni3B and SiO2 films in principle [36]. According to Ellingham-Richardson dia­
Cr2B and the relationship between crystal orientations therein is Ni3B gram [37,38], the affinity of Si and O was greater than that of Cr and O,
[101] // Cr2B [404]. The EDS results showed the columnar area with the so Si in the CrSi2 phase would be preferentially oxidized in these con­
chemical compositions (in wt%) of 85.5% Ni, 6.9% Cr, 3.2% Si and 4.3% ditions. Furthermore, the PBR value of Cr to Cr2O3 was 2.07, and the
Fe corresponded to Region B. The SAED pattern of Region B indicated PBR value of Si to SiO2 was 1.72 [39], so SiO2 was denser, more stable,
that it consisted of γ-Ni and CrSi2. The TEM characterization results were and formed a better bond with the matrix, allowing it to block the
consistent with XRD patterns. infiltration of O and Cl to the matrix coating to a greater extent. The
preferentially formed SiO2 would slow the formation of Cr2O3 to a

Fig. 1. Surface topographies of the as-sprayed coatings: (a) NiCrAlY; (b) NiCrBSi.

3
X. Liu et al. Corrosion Science 203 (2022) 110351

Fig. 2. Section morphologies (left side) and fracture morphologies (right side) of the as-sprayed coatings: (a), (b) NiCrAlY; (c), (d) NiCrBSi.

each picture). As seen from Fig. 7, the NiCrAlY coating surface is loose
and porous after being corroded for 10 h. The EDS results indicated that
there are large amounts of Cr, Y, K, and Cl on the surface. Combined with
the results of EDS and XRD, the corrosion products at the initial stage
were found to be mainly composed of gray plate-shaped Cr2O3, white
prism-shaped Y2O3/YAlO3, spinel phase K2CrO4 and unconsumed KCl
salt [20,40]. As the exposure time increased to 50 h, the surface becomes
rough and uneven. Corresponding EDS results indicated that the content
of Cl was decreased and the content of O was increased, which dem­
onstrates that the residual KCl is gradually consumed and more corro­
sion products were generated. Meanwhile, white Y2O3/YAlO3 and gray
Cr2O3 particles were observed on the surface. After 100 h of hot corro­
sion, the coating surface was also rough and porous and it is covered
with more white Y2O3/YAlO3 particles and gray plate-shaped Cr2O3.
What’s more, Al2O3 also appeared on the coating surface from EDS
analysis of Area III (Fig. 7(f)).
For the NiCrBSi coating, the surface was smooth and dense after 10 h
of corrosion. The coating surface contained dark gray Si-rich areas,
Fig. 3. XRD patterns of the NiCrAlY and NiCrBSi powders and corresponding white Cr-rich areas, and, which indicated the formation of SiO2, Cr2O3,
as-sprayed coatings. and K2CrO4. The greater Cl content shown in EDS data pertaining to
Area II (Fig. 8(b)) may be either KCl or other chlorides. After 50 h of
certain extent, and part of the resulting Cr2O3 would also react with NiO corrosion, the coating surface became rough and porous. More dark gray
or KCl to form NiCr2O4 or K2CrO4. As a consequence, more SiO2 was SiO2 phases were formed on the surface and the white areas were rich in
detected in the NiCrBSi system. However, the specific corrosion prod­ Ni, Cr and K. As the corrosion time was increased to 100 h, the structure
ucts still need to be further determined by EDS analysis of the surface of corrosion products became more uniform and regular pores were
and cross-sectional of the corroded coatings. formed on the surface. Corresponding EDS results indicated that the
The evolution of surface morphologies during hot corrosion was surface contained high contents of Ni and Cr and a lower amount of Si,
implied in Figs. 7 and 8 (corresponding EDS results are listed next to which demonstrated that the amount of SiO2 phases decreased on the

4
X. Liu et al. Corrosion Science 203 (2022) 110351

Fig. 4. TEM morphologies (a), (c) and the SAED pattern of grains corresponding to (b) Region A, and (d) Region B for the as-sprayed NiCrBSi coating.

3.2.2. Cross-sectional morphology and component analysis

Fig. 9 illustrates the evolution of cross-sectional structure and EDS
mapping analysis of NiCrAlY coating after different exposure times with
KCl salt; the whole corrosion area includes two parts: a corrosion
product layer and internal oxidation zone. Based on the XRD and EDS
results, the outer corrosion product layer was found to be mainly
comprised of oxides rich in Cr, Y and K. K2CrO4 and Cr2O3 were
distributed through the corrosion layer and the main component of the
bottom dark gray layer was Al2O3. The coincidence of Al and Y on EDS
map analysis at the top of the corrosion product layer could be inferred
to arise from the presence of Y2O3 and YAlO3. Research shows that Y2O3
can improve adhesion of corrosion products and prevent peeling [41].
As seen from Fig. 9(a-c), a continuous Y2O3/YAlO3 layer was formed on
the surface when exposed for 100 h, which enhanced the adhesion of the
corrosion product layer. Corresponding inter layer mapping results at
10 h of corrosion indicated that O was mainly concentrated in corrosion
Fig. 5. Macroscopic morphologies of NiCrAlY and NiCrBSi coatings after product layer and Cl diffused along the splat boundaries voids. That is,
corrosion in KCl salt for 10 h, 50 h, and 100 h. the coating exhibited slight internal oxidation and chlorination, which
implied that the Al2O3 formed at the interface failed to block the
top of surface and Cr and Ni oxides were more concentrated thereon. diffusion of Cl2 and O2. When the exposure time was extended to 50 h
In summary, with increasing corrosion time, the types of corrosion and 100 h, O spreads into the coating resulting in the formation of in­
products on the NiCrAlY surface did not change to any significant extent; ternal oxide zones with a thickness of 75–100 µm and 130–165 µm,
it comprised Cr2O3, K2CrO4 and Y2O3/YAlO3, but the crystal grains grew respectively: the main internal oxide products were Al2O3 and Cr2O3.
gradually. However, both the types of corrosion products and surface The evolution of cross-sectional structure and EDS mapping analysis
morphology of NiCrBSi had changed, being firstly rich in SiO2, and later of NiCrBSi coating during hot corrosion process can be seen in Fig. 10.
evolving into porous oxides rich in Ni, Cr, and K. After 10 h of exposure to KCl salt, a corrosion product layer with a
thickness of 30–50 µm was formed on the surface. Strong signals of K
and Cl were detected on the surface, which could be attributed to the

5
X. Liu et al. Corrosion Science 203 (2022) 110351

Fig. 6. XRD patterns of NiCrAlY and NiCrBSi coatings after exposure to KCl salt for 10 h, 50 h, and 100 h.

Fig. 7. SEM topographic micrographs of the corroded NiCrAlY coatings at different hot corrosion times and the corresponding EDS analysis: (a), (b) 10 h; (c), (d)
50 h; (e), (f) 100 h.

presence of unreacted KCl. Meanwhile, signals corresponding to Cr and enriched in the corrosion product layer, suggesting that corrosion
O were detected, indicating that corrosion product layer also contained product layer mainly consisted of Cr2O3 and SiO2. SiO2 was distributed
Cr2O3 and K2CrO4. Furthermore, Fig. 10(a) indicates that a small through the whole corrosion product layer. With corrosion time
amount of SiO2 formed on the corrosion product layer surface. As the extending to 100 h, the thickness of the corrosion layer was increased to
exposure time increased to 50 h, the thickness of corrosion product layer 35–60 µm. As can be seen from Fig. 10(c), more Cr2O3 was formed in the
increased to 40–55 µm. The EDS results showed that Cr, Si and O were corrosion layer, and a continuous SiO2 layer formed at the corrosion

6
X. Liu et al. Corrosion Science 203 (2022) 110351

Fig. 8. SEM topographic micrographs of the corroded NiCrBSi coatings at different hot corrosion times and the corresponding EDS analysis: (a), (b) 10 h; (c), (d)
50 h; (e), (f) 100 h.

layer/metal interface. According to the EDS mapping result of inner (Δm)2

layer, after 10 h of corrosion, Cl and O were no longer detected within = Kp • t (1)
A
the coating, indicating that there was no internal oxidation or Cl diffu­
sion in the coating. When the corrosion time was extended to 50 h, small where Δm/A represents the mass gain per unit area (mg/cm2), Kp is
amounts of Cl and O were detected inside the coating, indicating that the average rate constant for hot corrosion mg2/(cm4⋅s), and t is the hot
coating had undergone slight internal oxidation and chlorination. With corrosion time (s). Fig. 11(b) demonstrates the linear fitting curves of
the further extension of the corrosion time, the internal oxidation of the the squared mass gain per unit area, which was obtained according to
coating and the diffusion of Cl increased to a certain extent. Eq. (1). The Kp values were calculated as 1.48 × 10− 3 mg2/(cm4⋅s) and
In summary, for NiCrAlY coating, the main corrosion products were 2.42 × 10− 4 mg2/(cm4⋅s) for NiCrAlY and NiCrBSi coatings, respec­
Cr2O3, K2CrO4 and the Al2O3 layer at the interface, which did not play a tively. Eventually, it should be noted that no reduction in mass change
good protective role, causing severe internal oxidation of the coating. during the 100 h exposure, indicating that no scale or coating peeling
Whereas, the main corrosion products Cr2O3 for NiCrBSi coating and the occurred during hot corrosion.
continuous SiO2 layer at the interface played a good protective role, Due to the high vapor pressure of chlorides generated during the
therefore, slight internal oxidation and Cl diffusion had occurred in the corrosion process, some of them may evaporate, causing certain errors in
coating. the mass change data. On the premise that the coating or oxide scale is
not peeled off, dimensional metrology is a more reliable method of
3.2.3. Corrosion kinetics of coatings in KCl salt corrosion damage to materials under high temperature. Hence, in order
The rate of gain of mass in the NiCrAlY and NiCrBSi coatings during to further accurately illustrate the corrosion kinetics of the coating, the
hot corrosion testing at 700 ℃ is shown in Fig. 11(a). The rate of gain of thickness of the oxide scale and the inner oxidation zone were measured
mass curves of the two coatings followed a parabolic trend, the mass in detail, as presented in Fig. 12. It can be seen that the thickness of the
rapidly increased during the early stages and the rate of gain thereof was oxide scale and the inner oxidation zone of the NiCrAlY coating is higher
reduced as the hot corrosion time increased. The gain of mass for than that of the NiCrBSi coating, implying that the NiCrAlY coating has
NiCrAlY and NiCrBSi coatings was 25.57 ( ± 1.89) and 10.04 ( ± 0.04) suffered more serious corrosion. Judging from the thickness of the oxide
mg/cm2, respectively. According to the Wagner theoretical model [30], scale, the two coatings are similar and both grow slowly with time
the growth rate under parabolic reaction kinetics can be simplified by progresses, and are stable at about 30 µm. But the depth of the inner
Eq. (1): oxidation zone varies greatly: the thickness of inner oxidation zone of

7
X. Liu et al. Corrosion Science 203 (2022) 110351

Fig. 9. Cross-sectional morphologies of NiCrAlY coatings after exposure to KCl salt as well as corresponding elemental mapping analysis with different exposure
times: (a) 10 h; (b) 50 h; (c) 100 h.

the NiCrAlY coating is larger and its oxidation degree is also greater
(Figs. 9 and 10) than that of NiCrBSi coating, which demonstrates that (2) Metal chloride has a higher vapor pressure at high temperatures.
the oxide scale of NiCrBSi coating can better block the infiltration of O or When the temperature was greater than 600 ℃, KCl would
Cl. The results confirmed that the NiCrBSi coating exhibited superior hot volatilize, and Cl2 appeared via the pathways listed in Eq. (4):
corrosion resistance than that of NiCrAlY in KCl salt, which is consistent
with the result of mass change data. 4KCl(g) + O2 (g) = 2K2 O(s) + 2Cl2 (g) (4)
The Cl2 produced through the two aforementioned pathways would
3.3. Hot corrosion mechanism for NiCrAlY and NiCrBSi coatings exposed diffuse into the coating through the oxide film and react with the metal
to KCl elements to produce the corresponding MCl2 (as shown in Eqs. (5)–(9)).
Then, according to the “activation oxidation mechanism” [20,30,45,46],
KCl deposition produced during biomass combustion causes severe the MCl2 would diffuse to where the partial pressure of oxygen is higher,
hot corrosion of boiler tubes. Based on studies concerning KCl and its regenerating unprotected oxides with a loose structure (as given by Eqs.
thermodynamics in biomass combustion, KCl can participate in all states (10)–(13)), thereby accelerating the oxidation and corrosion of the
(solid, gas, or liquid) [42,43]. Since the melting point of KCl is 770 ℃. coating. Therefore, whether the microstructure of the coating was dense
KCl is present in its solid state at the temperature of 700 ℃. The and whether the generated oxide could block the diffusion of Cl and O
corrosion process of the coating in the solid alkali metal chloride me­ directly determined the performance of the coating.
dium is mainly manifest in the following two ways [44]:
Ni + Cl2 (g) = NiCl2 (s)ΔG0f (NiCl2 ) = − 159.7kJ/mol at 700℃ (5)
(1) Exposed metallic elements (such as Ni, Al, and Si) in the sprayed
coating would react with O2 to produce oxides, then Cl2 was Cr + Cl2 (g) = CrCl2 (s)ΔG0f (CrCl2 ) = − 276.1kJ/mol at 700℃ (6)
produced during the reaction between the solid KCl/metal
interface involved in the following reactions: 3
Cr + Cl2 (g) = CrCl3 (s)ΔG0f (CrCl3 ) = − 350.5 kJ/mol at 700℃
y 2
xM(s) + O2 (g) = Mx Oy (2) (7)
2

1
KCl(s) + Mx Oy (s) = KMx Oy (s) + Cl2 (g) (3)
2

8
X. Liu et al. Corrosion Science 203 (2022) 110351

Fig. 10. Cross-sectional morphologies of NiCrBSi coatings after being corroded in KCl salt as well as corresponding elemental mapping analysis with different
exposure times: (a) 10 h; (b) 50 h; (c) 100 h.

Fig. 11. (a) Changes in mass of the NiCrAlY and NiCrBSi coatings exposed in KCl at 700 ℃, and (b) square of the changes in mass versus time.

3
Al + Cl2 (g) = AlCl3 (s) ΔG0f (AlCl3 ) = − 511.2 kJ/mol at 700℃ ΔG0f (NiO) = − 38.4kJ/mol at 700℃ (10)
2
(8)
3
2CrCl3 (g) + O2 (g) = Cr2 O3 (s) + 3Cl2 (g)
2
Si + 2Cl2 (g) = SiCl4 (s) ΔG0f (SiCl4 ) = − 534.5 kJ/mol at 700℃ (9)
ΔG0f (Cr2 O3 ) = − 232.6kJ/mol at 700℃ (11)
1
NiCl2 (g) + O2 (g) = NiO(s) + Cl2 (g)
2

9
X. Liu et al. Corrosion Science 203 (2022) 110351

chlorine (as predicted using Eq. (3)), and the Al2O3 generated in the
inner layer was also discontinuous, so a large amount of Cl2 and O2
diffused into the coating. With the extension of the corrosion time, the
chlorine diffused into the coating promoted the oxidation reaction, and
severe internal oxidation occurred inside the coating, consuming much
aluminum. However, the resulting aluminum oxide layer did not block
the diffusion of Cl2 and O2 especially well, so, with the further extension
of the corrosion time, the chlorine element further diffused into the
coating, but the amount of Al in the coating was limited, so an internal
Cr2O3 layer was produced underneath the Al2O3 layer. The corrosion
layer was relatively loose, and the oxygen partial pressure in the
corrosion layer was almost the same, so there was no obvious delami­
nation in the entire corrosion layer, and the main component was the
oxide of Cr. In summary, the defects and boundaries inside the sprayed
coating, the chromium oxide layer damaged by the formation of K2CrO4
and the discontinuous aluminum oxide layer caused the NiCrAlY coating
to produce severe internal chlorination and oxidation, so that the
coating exhibited poor corrosion resistance.
Fig. 12. A plot of scale thickness for NiCrAlY and NiCrBSi coatings after hot Different from NiCrAlY, as shown in Fig. 14, the structure of the
corrosion versus exposure time. The thickness of oxide scale and internal sprayed NiCrBSi coating was more compact, and the corroded surface
oxidation zone are plotted. was denser. In addition, since the Si concentration in CrSi2 was high and
the affinity of Si to O was greater than that of Cr and O, Si would be
3 preferentially oxidized. As shown in Fig. 9(a,b), dense and well-bonded
2AlCl3 (g) + O2 (g) = Al2 O3 (s) + 3Cl2 (g)
2 SiO2 was produced on the surface. Therefore, at the initial stage, Cl2 and
O2 did not diffuse into the coating, so there was a high partial pressure of
ΔG0f (Al2 O3 ) = − 299.49kJ/mol at 700℃ (12) oxygen and chlorine on the surface, which hindered the progress of
Reactions (2) and (3) to a certain extent. As a result, the coating had a
SiCl4 (g) + O2 (g) = SiO2 (s) + 2Cl2 (g) low oxidation rate, and a large amount of unreacted KCl remained on the
surface. With the extension of the corrosion time, since the dense oxide
ΔG0f (SiO2 ) = − 200.7kJ/mol at 700℃ (13) layer was neither complete nor continuous, Cl2 diffused into the interior
of the coating, but there was no obvious internal oxidation, indicating
A schematic representation of the corrosion mechanisms for NiCrAlY that the partial pressure of that internal oxygen contained in the NiCrBSi
coating is shown in Fig. 13. In terms of the microstructure before and coating was low, which hindered the progress of activation oxidation in
after corrosion, since there were more defects and splat boundaries in the matrix coating. Therefore, the resulting MCl2 would diffuse to the
the as-sprayed coating and the corrosion layer was relatively loose, Cl2 coating/corrosion layer interface for re-oxidation. According to the
diffused into the coating during the initial stage of corrosion. In accor­ Gibbs free energy value of Eqs. (5)–(9), SiCl4 formed more readily, fol­
dance with the value of Gibbs free energy in Eqs. (10)–(13), Al2O3 was lowed by CrCl3 and NiCl2, so a large amount of SiO2 was generated in the
formed in preference to Cr2O3, which required a lower partial pressure corrosion layer, and a small amount of Ni/Cr oxides were generated at
of oxygen, therefore, Cr2O3 was mainly distributed at the outer layer of the same time. Therefore, the coexistence of SiO2 and Ni/Cr oxides
the corrosion layer, and Al2O3 was formed at the corrosion layer/metal further interrupted the continuity of oxides. Since the diffusion coeffi­
interface. However, the cross-sectional micrograph implied that almost cient of Cr in Cr2O3 is much larger than that in SiO2 [47], Cr would
no Ni was detected in the corrosion layer, which was mainly because the preferentially diffuse outward from the metal matrix through Cr2O3.
Gibbs free energy of NiCl2 was lower than that of NiO, so the thermo­ While the diffusion coefficient of O in SiO2 is much larger than that in
dynamic stability of NiCl2 was higher, but NiCl2 was more readily Cr2O3 [47], so oxygen will diffuse inward preferentially through SiO2.
volatilized. Through combining XRD (Fig. 6(a)) and the cross-sectional Hence, during the corrosion process, Cr2O3 grows to the outside of the
view (Fig. 9(a)), the generated Cr2O3 was found to react with KCl to oxide scale, and SiO2 grows to the inside of the oxide scale. After a
form loosely structured K2CrO4, accompanied by a large amount of

Fig. 13. Schematic of hot corrosion mechanism for NiCrAlY coating exposed to KCl salt in air at 700 ℃.

10
X. Liu et al. Corrosion Science 203 (2022) 110351

Fig. 14. Schematic of hot corrosion mechanism for NiCrBSi coating exposed to KCl salt in air at 700 ℃.

period of time, as shown in Fig. 10(c), the corrosion layer was delami­ film, defects, and splat or grain boundaries, so as to accelerate the
nated: Cr2O3 accumulated on the outside of the oxide scale and the inner oxidation of the coating.
layer was mainly composed of SiO2. At the same time, it can be seen that (2) The NiCrAlY coating had been severely corroded and internally
Cr element is lacked at the oxide scale/metal matrix interface, but the oxidized. The corrosion products were Cr2O3, Y2O3/YAlO3 and a
concentration of Ni and Si elements increased. Since the affinity of Si small amount of K2CrO4; the internal oxidation products were
and O is greater than that of Cr and Ni, it is reasonable that SiO2 would mainly Al2O3. The splat boundaries were considered to provide a
replace Cr2O3 through the following reaction at the oxide scale/metal channel for the diffusion of chlorine and oxygen, and the main
interface: corrosion products Cr2O3 and the Al2O3 layer at the interface
failed to play a good protective role, leading to severe internal
2Cr2 O3 + 3Si = 3SiO2 + 4Cr (14)
oxidation.
Therefore, a continuous SiO2 layer is formed inside the oxide scale (3) The NiCrBSi coating showed mild corrosion, slight internal
(Fig. 10(c)). The continuous SiO2 layer hindered the outward migration oxidation, and chlorination. Cr2O3 and SiO2 were the main
of MCl2 to a certain extent and decreased the rate of diffusion of O2/Cl2 corrosion products, including a small quantity of K2CrO4. In
into the coating. Hence, the partial pressure of oxygen/chlorine inside addition to the compact structure, the continuous SiO2 layer also
the coating was relatively low, and the degree of internal oxygen/ played an essential role in blocking the diffusion of Cl2 and O2,
chlorination was relatively light. Under the same conditions, compared resulting in slight internal oxidation in the coating.
with the NiCrAlY coating, the partial pressure of that oxygen outside the
SiO2 layer was higher, so the corresponding metal chlorides, such as CRediT authorship contribution statement
NiCl2 and CrCl2, were oxidized outside the SiO2 layer, resulting in a
loose structure in the Ni and Cr oxides. This result explained why the Ni Xia Liu: Conceptualization, Methodology, Formal analysis, Writing –
content on the surface of the NiCrBSi corrosion layer was higher than review & editing. Kai Hu: Investigation, Data curation, Visualization,
that of the NiCrAlY coating. Writing – original draft. Shihong Zhang: Conceptualization, Resources,
It can be observed from the above research results that, in the KCl Supervision, Writing – review & editing. Tong Xu: Methodology, Re­
corrosion environment, Cr2O3 would be destroyed through the forma­ sources. Lin Chen: Methodology. Eungsun Byon: Conceptualization,
tion of K2CrO4, while Al2O3 and SiO2 would be generated at the interface Supervision. Dongguang Liu: Conceptualization, Methodology.
between the oxide scale and the matrix coating, which could block the
diffusion of O and Cl. Obviously, SiO2 layer shows a better effect in Declaration of Competing Interest
blocking the diffusion of chlorine and oxygen. Hence, NiCrBSi coating
has great potential for surface protection of metal substrates working in The authors declare that they have no known competing financial
alkali-chlorine environments. interests or personal relationships that could have appeared to influence
the work reported in this paper.
4. Conclusions
Data Availability
In this work, the NiCrAlY and NiCrBSi coatings were prepared on
12CrMoV boiler steel by HVOF spraying technology. The hot corrosion The raw data required to reproduce these findings are available from
behavior and mechanism of the two types of coatings were investigated the authors on request.
against KCl salt at 700 ℃ for different exposure times in relation to
structural characterizations. Based on the experimental data and anal­ Acknowledgments
ysis of such results, the main conclusions are as follows:
This work was supported by the University Synergy Innovation
(1) The as-sprayed NiCrAlY and NiCrBSi coatings had a dense Program of Anhui Province (Grant No. GXXT-2020-071), the
structure and were well bound to the substrate, while more Outstanding Youth Fund of Anhui Province (2108085J22) and Natural
unmolten particle and splat boundaries were present in the Science Research Project of Anhui Universities (Grant No.
NiCrAlY coating. During this hot corrosion process, the coating KJ2020A0243). The authors appreciate helpful comments for the
was corroded in accordance with the “activation oxidation” reviewers.
mechanism, and Cl2 would penetrate the coating through oxide

11
X. Liu et al. Corrosion Science 203 (2022) 110351

References [24] T.S. Sidhu, S. Prakash, R.D. Agrawal, Hot corrosion behaviour of HVOF-sprayed
NiCrBSi coatings on NiandFe-based superalloys in Na2SO4-60%V2O5 environment
at 900℃, Acta Mater. 54 (2006) 773–784.
[1] S.C. Okoro, M. Montgomery, F.J. Frandsen, P. Karen, High temperature corrosion
[25] K.A. Habib, M.S. Damra, J.J. Carpio, I. Cervera, J.J. Saura, Performance of NiCrAlY
under laboratory conditions simulating biomass-firing: a comprehensive
coatings deposited by oxyfuel thermal spraying in high temperature chlorine
characterization of corrosion products, Energy Fuels 28 (2014) 6447–6458.
environment, J. Mater. Eng. Perform. 23 (10) (2014) 3511–3522.
[2] A. Demirbas, Combustion characteristics of different biomass fuels, Prog. Energy
[26] J.M. Miguel, J.M. Guilemany, S. Vizcaino, Tribological study of NiCrBSi coating
Combust. Sci. 30 (2) (2004) 219–230.
obtained by different processes, Tribol. Int. 36 (2003) 181–187.
[3] A. Babin, C. Vaneeckhaute, M.C. Iliuta, Potential and challenges of bioenergy with
[27] X.Y. Dong, X.T. Luo, S.L. Zhang, C.J. Li, A novel strategy for depositing dense self-
carbon capture and storage as a carbon-negative energy source: a review, Biomass
fluxing alloy coatings with sufficiently bonded splats by one-step atmospheric
Bioenergy 146 (2021), 105968.
plasma spraying, J. Therm. Spray. Technol. 29 (2) (2019) 173–184.
[4] A.U. Syed, N.J. Simms, J.E. Oakey, Fireside corrosion of superheaters: effects of air
[28] C. Fu, Y. Li, Y.F. Wang, Microstructure and corrosion resistance of ERNiCrMo-13
and oxy-firing of coal and biomass, Fuel 101 (2012) 62–73.
and NiCrBSi coatings in simulated coal-fired boiler conditions: the effect of fly-ash
[5] X.F. Xing, R. Wang, N. Bauer, P. Ciais, J.J. Cao, J.M. Chen, et al., Spatially explicit
composition, Surf. Coat. Technol. 399 (2020), 126134.
analysis identifies significant potential for bioenergy with carbon capture and
[29] R.A. Antunes, M.C.L.D. Oliveira, Corrosion in biomass combustion: a material
storage in China, Nat. Commun. 12 (1) (2021) 3159.
selection analysis and its interaction with corrosion mechanisms and mitigation
[6] Y.T. Cai, K. Tay, Z.M. Zheng, W.M. Yang, H. Wang, G. Zeng, Z.W. Li, K.B. Siah,
strategies, Corros. Sci. 76 (2013) 6–26.
S. Prabakaran, Alkali metal chloride formed after combustion in the boiler exists in
[30] E. Sadeghimeresht, L. Reddy, T. Hussain, M. Huhtakangas, N. Markocsan, S. Joshi,
the ash and slag, which is easy to adhere to the surface of the heating surface to
Influence of KCl and HCl on high temperature corrosion of HVAF-sprayed NiCrAlY
form ash accumulation, Appl. Energy 230 (2018) 1447–1544.
and NiCrMo coatings, Mater. Des. 148 (2018) 17–29.
[7] J.Y. Zhang, Z.U. Rahman, X.B. Wang, Z. Wang, P. Li, Y.B. Wang, D. Bate, K. Zhao,
[31] S. Enestam, D. Bankiewicz, J. Tuiremo, K. Mäkelä, M. Hupa, Are NaCl and KCl
H.Z. Tan, Hot corrosion behaviors of TP347H and HR3C stainless steel with KCl
equally corrosive on superheater materials of steam boilers? Fuel 104 (2013)
deposit in oxy-biomass combustion, J. Environ. Manag. 263 (2020), 110411.
294–306.
[8] J.M. Johansen, J.G. Jakobsen, F.J. Frandsen, P. Glarborg, Release of K, Cl, and
[32] M. Bai, L. Reddy, T. Hussain, Experimental and thermodynamic investigations on
Sduring pyrolysis and combustion of high-chlorine biomass, Energy Fuels 25
the chlorine-induced corrosion of HVOF thermal sprayed NiAl coatings and
(2011) 4961–4971.
304stainless steels at 700 ◦ C, Corros. Sci. 135 (2018) 147–157.
[9] T.M. Meiβner, X. Montero, D. Fähsing, M.Z. Galetz, Cr diffusion coatings on a
[33] A. Agüero, I. Baráibar, M. Gutiérrez, M. Hernández, R. Muelas, S. Rodríguez,
ferritic-martensitic steel for corrosion protection in KCl-rich biomass co-firing
Biomass corrosion behavior of steels and coatings in contact with KCl/K2SO4 at 550
environments, Corros. Sci. 164 (2020), 108343. ◦
C under an oxy-fuel combustion atmosphere: a screening laboratory test, Surf.
[10] J. Pettersson, N. Folkeson, L.G. Johansson, J.E. Svensson, The effects of KCl, K2SO4
Coat. Technol. 350 (2018) 188–200.
and K2CO3 on the high temperature corrosion of a 304-type austenitic stainless
[34] N. Rana, M.M. Mahapatra, R. Jayaganthan, S. Prakash, High-temperature oxidation
steel, Oxid. Met. 76 (1) (2011) 93–109.
and hot corrosion studies on NiCrAlY coatings deposited by flame-spray technique,
[11] E.K. Hao, X.Q. Zhao, Y.L. An, W. Deng, H.D. Zhou, J.M. Chen, The effect of pre-
J. Therm. Spray. Technol. 24 (5) (2015) 769–777.
oxidation on microstructure, mechanical properties and high-temperature
[35] S.L. Liu, X.P. Zheng, G.Q. Geng, Dry sliding wear behavior and corrosion resistance
tribological behaviors of HVOF-sprayed NiCoCrAlYTa coating, Appl. Surf. Sci. 489
of NiCrBSi coating deposited by activated combustion-high velocity air fuel spray
(2019) 187–197.
process, Mater. Des. 31 (2010) 913–917.
[12] N. Abu-warda, A.J. López, F. Pedraza, M.V. Utrilla, Fireside corrosion on T24 steel
[36] K. Kurokawa, T. Sudiro, T. Sano, S. Kyo, O. Ishibashi, M. Nakamori, High-
pipes and HVOF NiCr coatings exposed to different salt mixtures, Corros. Sci. 173
temperature corrosion resistance of SiO2-forming materials, Corros. Rev. 36 (1)
(2020), 108747.
(2018) 65–74.
[13] A.J. López, M. Proy, V. Utrilla, E. Otero, J. Rams, High-temperature corrosion
[37] H.J.T. Ellingham, Reducibility of oxides and sulphides in metallurgical processes,
behavior of Ni-50Cr coating deposited by high velocity oxygen-fuel technique on
J. Soc. Chem. Ind. 63 (1944) 125–133.
low alloy ferritic steel, Mater. Des. 59 (2014) 94–102.
[38] F.D. Richardson, J.H.E. Jeffes, The thermodynamics of substances of interest in ion
[14] G.Y. Wang, H. Liu, T.Z. Chen, X.J. Zhang, H.Y. Li, Y. Yu, H. Yao, Comparative
and steel making from 0◦ C to 2400◦ C: I-oxides, J. Iron Steel Inst. 160 (1948)
investigation on thermal corrosion of alloy coatings ins imulated waste incinerator
261–273.
environment, Corros. Sci. 189 (2021), 109592.
[39] N.B. Pilling, R.E. Bedworth, The oxidation of metals at high temperatures, J. Inst.
[15] S. Mahajan, R. Chhibber, Hot corrosion studies of boiler steels exposed to different
Met. 29 (3) (1923) 529–591.
molten salt mixtures at 950 ℃, Eng. Fail. Anal. 99 (2019) 210–224.
[40] J. Lehmusto, D. Lindberg, P. Yrjas, B.J. Skrifvars, M. Hupa, Thermogravimetric
[16] J. Lehmusto, B.J. Skrifvars, P. Yrjas, M. Hupa, Comparison of potassium chloride
studies of high temperature reactions between potassium salts and chromium,
and potassium carbonate with respect to their tendency to cause high temperature
Corros. Sci. 59 (2012) 55–62.
corrosion of stainless 304L steel, Fuel Process. Technol. 105 (2013) 98–105.
[41] A. Gil, D. Naumenko, R. Vassen, J. Toscano, M. Subanovic, L. Singheiser, W.
[17] F. Davide, M. Ville, U. Mikko, K. Heli, V. Petri, Chlorine-induced high temperature
J. Quadakkers, Y-rich oxide distribution in plasma sprayed MCrAlY-coatings
corrosion of Inconel 625 sprayed coatings deposited with different thermal spray
studied by SEM with a cathodoluminescence detector and Raman spectroscopy,
techniques, Surf. Coat. Technol. 318 (2017) 233–243.
Surf. Coat. Technol. 204 (2009) 531–538.
[18] M. Schütze, M. Malessa, V. Rohr, T. Weber, Development of coatings for protection
[42] Y.B. Wang, H.Z. Tan, Condensation of KCl(g) under varied temperature gradient,
in specific high temperature environments, Surf. Coat. Technol. 201 (2006)
Fuel 237 (2019) 1141–1150.
3872–3879.
[43] N. Folkeson, T. Jonsson, M. Halvarsson, L.G. Johansson, J.E. Svensson, The
[19] L.K. Wu, J.J. Wu, W.Y. Wu, H.J. Yan, M.Y. Jiang, F.H. Cao, Hot corrosion behavior
influence of small amounts of KCl(s) on the high temperature corrosion of a Fe-
of electrodeposited SiO2 coating on TiAl alloy, Corros. Sci. 174 (2020), 108827.
2.25Cr-1Mo steel at 400 and 500 ◦ C, Mater. Corros. 62 (2011) 606–615.
[20] E. Sadeghimereshta, L. Reddyb, T. Hussainb, N. Markocsana, S. Joshia, Chlorine-
[44] H.P. Nielsen, F.J. Frandsen, K. Dam-Johansen, L.L. Baxter, The implications of
induced high temperature corrosion of HVAF-sprayed Ni-based alumina and
chlorine-associated corrosion on the operation of biomass-fired boilers, Prog.
chromia forming coatings, Corros. Sci. 132 (2018) 170–184.
Energy Combust. Sci. 26 (2000) 283–298.
[21] R. Jafaria, E. Sadeghi, High-temperature corrosion performance of HVAF-sprayed
[45] M.A. Olivas-Ogaz, J. Eklund, J.E. Svensson, J. Liske, T. Jonsson, Microstructural
NiCr, NiAl, and NiCrAlY coatings with alkali sulfate/chloride exposed to ambient
study of the influence of KCl and HCl on preformed corrosion product layers on
air, Corros. Sci. 160 (2019), 108066.
stainless steel, Oxid. Met. 87 (2017) 801–811.
[22] T.S. Sidhu, S. Prakash, R.D. Agrawal, Hot corrosion studies of HVOF NiCrBSi and
[46] M.A. Uusitalo, P.M.J. Vuoristo, T.A. Mäntylä, High temperature corrosion of
Stellite-6 coatings on a Ni-based superalloy in an actual industrial environment of a
coatings and boiler steels below chlorine-containing salt deposits, Corros. Sci. 46
coal fired boiler, Surf. Coat. Technol. 201 (2006) 1602–1612.
(6) (2004) 1311–1331.
[23] J. Eklund, J. Phother, E. Sadeghi, S. Joshi, J. Liske, High‑temperature corrosion of
[47] G. Welsch, P.D. Desai. Oxidation and Corrosion of Intermetallic Alloys, Purdue
HVAF‑sprayed Ni‑based coatings for boiler applications, Oxid. Met. 91 (2019)
Univ. Off. Pubns., Indiana, 1996.
729–747.

12