Fuel Processing Technology 104 (2012) 253–264

Contents lists available at SciVerse ScienceDirect

Fuel Processing Technology

journal homepage: www.elsevier.com/locate/fuproc

High temperature corrosion of superheater steels by KCl and K2CO3 under dry and
wet conditions
J. Lehmusto ⁎, P. Yrjas, B.-J. Skrifvars, M. Hupa
Åbo Akademi University, Piispankatu 8, 20500 Turku, Finland

a r t i c l e i n f o a b s t r a c t

Article history: The influence of both solid KCl and solid K2CO3 on the oxidation behavior of a ferritic 10CrMo type steel and
Received 22 February 2012 a nickel-based Alloy 625 type steel was investigated in the laboratory. Samples were exposed for a week at
Received in revised form 9 May 2012 temperatures of 500 °C, 550 °C, and 600 °C. The role of water vapor was also studied by carrying out exposures
Accepted 15 May 2012
under two different atmospheres: dry air, and air with a water content of 30%. Analyses were performed with
Available online 8 June 2012
SEM/EDXA to determine the extent of corrosion, the elemental distribution, and the identification of corrosion
Keywords:
products.
Ferritic steel Both salts were found to be corrosive towards both steels, but the influence of KCl was significantly greater
High temperature oxidation than that of K2CO3. In terms of durability, Alloy 625 exhibited higher corrosion resistance than 10CrMo. In
Nickel-based alloy the case of 10CrMo, the oxide formed always consisted of iron oxide with a chromium-rich region near the
Potassium chloride original steel surface. In the case of Alloy 625, the oxide consisted mainly of nickel oxide when it formed in
Potassium carbonate the presence of KCl, and chromium oxide when it formed in the presence of K2CO3.
Water vapor © 2012 Elsevier B.V. All rights reserved.

1. Introduction between potassium carbonate and the protective chromium oxide is

Chloride-induced corrosion at elevated temperatures is a problem
that often occurs in both biomass and waste combustion [1–3]. The
corrosive nature of biomass ashes is thought to originate from the
high content of alkali metal compounds such as potassium chloride
(KCl), and from the presence of chlorine combined with low sulfur
content [4–7]. Corrosion can be prevented or reduced by operating
boilers at lower steam temperatures or by using steels of higher
quality. Alloying with chromium and nickel is known to improve steel
by encouraging the formation of a protective oxide layer consisting of
a corundum-type solid solution (Cr, Fe)2O3, or even pure Cr2O3 if the
chromium content of the steel is high enough.
The corrosion tendencies of alkali chlorides, in particular potassium
chloride, towards various steels at high temperatures have been studied
extensively [1,8–19]. In addition to the corrosivity of alkali chlorides,
the reaction mechanism of chloride-induced high temperature corro-
sion has been investigated. Although the detailed reaction mechanism
remains elusive, it is generally accepted that the reaction proceeds via
intermediates rather than via direct conversion of metals into oxides
(Eqs. (1a), (1b)) [9,17,20]. The corrosivity of potassium carbonate
(K2CO3), another potassium salt occasionally found in fly ashes from
biomass fired boilers, has been less well explored than the corrosivity
of potassium chloride [21–23]. It has been suggested that the reaction
⁎ Corresponding author. Tel.: + 358 2 215 4038.
E-mail addresses: juho.lehmusto@abo.fi (J. Lehmusto), patrik.yrjas@abo.fi (P. Yrjas),
bengt-johan.skrifvars@topanalytica.com (B-J. Skrifvars), mhupa@abo.fi (M. Hupa).
analogous to the reaction between potassium chloride and chromium
oxide: the protective oxide layer is destroyed because of the formation
of chromate by potassium (Eq. (2)) [21]. The presence of water vapor
is required for reaction (1a) to take place, whereas reactions (1b) and
(2) take place under dry conditions. When reactions (1a) and (1b) are
considered, it should be noted, that at water vapor levels above
1 ppm, the formation of HCl is favored over the formation of molecular
chlorine [24]. This makes reaction (1a) more probable under the condi-
tions found in present-day biomass combustion.
2Cr2 O3 ðsÞþ 8KClðsÞ þ 4H2 Oðg Þ þ 3O2 ðg Þ↔4K2 CrO4 ðsÞ þ 8HClðg Þ ð1aÞ
2Cr2 O3 ðsÞ þ 8KClðsÞ þ 5O2 ðg Þ↔4K2 CrO4 ðsÞ þ 4Cl2 ðg Þ ð1bÞ
2Cr2 O3 ðsÞ þ 4K2 CO3 ðsÞ þ 3O2 ðg Þ↔4K2 CrO4 ðsÞ þ 4CO2 ðg Þ ð2Þ
At temperatures above 400 °C, water vapor is known to destroy
the protective chromium oxide layer, even when no other corrosive
compounds are present [25]. The loss of protective properties is due
to the formation of Cr(VI)oxide hydroxide vapor, which, when it
reacts further, converts the protective chromium oxide-rich surface
into a poorly protective iron-rich oxide [26,27].
2Cr2 O3 ðsÞ þ 3O2 ðg Þ þ 4H2 Oðg Þ↔4CrO2 ðOHÞ2 ðg Þ ð3Þ
The synergistic effect of alkali chlorides and water vapor on the
corrosion behavior of metals has also been examined. Various studies

0378-3820/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
doi:10.1016/j.fuproc.2012.05.020
254 J. Lehmusto et al. / Fuel Processing Technology 104 (2012) 253–264

Table 1 agree on the accelerating effect of the simultaneous presence of a

Chemical compositions of the steels in wt %. chloride and water vapor [28,29]. A similar effect has been reported
Material Cr Fe Mn Mo Ni Si nCr/nFe for potassium carbonate in the presence of water vapor [21]. Never-
theless, few studies address the synergistic effect of potassium car-
10CrMo9–10 2.2 96 0.4 1.0 – 0.2 0.02
Alloy 625 21.3 2.2 0.2 8.9 62.8 0.2 9.68 bonate and water vapor, since most of them focus on dry conditions
and partially molten mixtures [30–32].
This work has two main objectives. The first is to compare two po-
tassium salts to understand the corrosive tendency of the less studied
Table 2 potassium carbonate towards superheater materials. The second is
Measured oxide layer thicknesses of exposed 10CrMo and Alloy 625. to obtain more information about the role of water vapor in high
Salt Temperature Atmosphere Average oxidation Maximal oxidation temperature corrosion, especially in the presence of potassium carbon-
(°C) (μm) (μm) ate. Potassium chloride and potassium carbonate, salts commonly found
10CrMo Alloy 625 10CrMo Alloy 625 in fly ashes, were compared in terms of their ability to corrode a nickel-
based Alloy 625 type steel and a ferritic 10CrMo type steel under typical
No salt 500 Dry 5 b1 18 b1
No salt 500 Wet 20 b1 50 b1 biomass combustion conditions. The tests were performed under both
No salt 550 Dry 9 b1 16 b1 dry and humid atmospheres. The extent of corrosion, the elemental dis-
No salt 550 Wet 34 b1 40 b1 tribution, and the corrosion products were determined using a scanning
No salt 600 Dry 18 b1 40 b1 electron microscope equipped with an energy dispersive x-ray analyzer
No salt 600 Wet 92 b1 108 b1
KCl 500 Dry 17 12 49 70
(SEM/EDXA). An earlier paper [23] introduced a novel approach for
KCl 500 Wet 28 b1 67 b1 assessing corrosion at temperatures relevant to biomass combustion,
KCl 550 Dry 90 10 127 84 and will be explored further in this work. The new method examines
KCl 550 Wet 38 b1 90 b1 the nature and mechanism of corrosion based on a qualitative analysis
KCl 600 Dry 124 40 167 110
of the oxide layer thickness distributions.
KCl 600 Wet 89 2 112 15
K2CO3 500 Dry 4 b1 18 b1
K2CO3 500 Wet 15 b1 39 b1 2. Experimental
K2CO3 550 Dry 20 b1 33 b1
K2CO3 550 Wet 29 b1 51 b1
The present study investigated the corrosion of a nickel-based
K2CO3 600 Dry 38 7 59 12
K2CO3 600 Wet 24 10 58 19
Alloy 625 type steel and a ferritic 10CrMo type covered either with
KCl or K2CO3, in both dry and humid air. Reference exposures with

Fig. 1. Average oxide layer thicknesses of 10 CrMo (upper) and Alloy 625 steels (lower) as functions of temperature under the two atmospheric conditions: dry air (left) and air with
30% H2O (right).
 J. Lehmusto et al. / Fuel Processing Technology 104 (2012) 253–264
Fig. 2. Elemental maps for 10CrMo steel after exposure in dry air at 600 °C for 168 h.
Fig. 3. Oxide layer thickness distributions of 10CrMo steel after exposure in dry air for
168 h.
no salt were also performed. Table 1 shows the compositions of the
steels. All the experiments were run at 500°, 550°, and 600 °C with
an exposure time of 168 h. The surface area of each steel sample
was 4 cm 2 (2 × 2 cm), the thickness was 0.5 cm (10CrMo) or 1.0 cm
(Alloy 625), and the particle size of the salts was between 125 and
297 μm. The air for the exposures under dry conditions was dried by
passing it through silica gel. The remaining amount of water was
not measured, but the air was regarded as dry air. The water content
under wet conditions was fixed at approximately 30%.
Fig. 4. Elemental maps for 10CrMo steel after exposure in wet air with 30% H2O at 600 °C for 168 h.
255
Prior to exposure, the steel samples were polished in water with a
320 grit SiC grinding paper, followed by polishing in ethanol with a
1000 grit SiC grinding paper. The samples were further cleaned in
ethanol using an ultrasonic bath before the salts were deposited on
them. Using a mould, 0.250 g of salt was placed on the surface of each
sample by tapping it gently with a glass rod. No salt was added onto
the samples during the exposures.
The exposures were carried out in temperature-controlled hori-
zontal tube furnaces; temperature was recorded with thermocou-
ples attached to the sample holder, and sample temperatures were
maintained within ± 2 °C from the target temperature. Samples
were raised to the final exposure temperature at a rate of
5 °C min − 1. For the experiments conducted under dry conditions,
the air was passed through a drying tube filled with silica gel. To en-
sure that the atmosphere remained as dry as possible, the steel sam-
ples were placed inside a quartz glass tube positioned inside the
furnace. The humidity for the wet conditions was created by
pumping water into a preheater, from which the evaporated water
entered the tube furnace. Under both test conditions, the gas flow
was 4 dm 3 min − 1 at the inlet temperature. The samples were posi-
tioned horizontally, parallel to the flow direction.
After exposure, the samples were cast into chlorine-free epoxy
and cut for cross-sectional analyses. Prior to the EDX analyses,
they were polished with 600 and 1200 grit SiC grinding papers
using kerosene and then washed with petroleum ether in an ultra-
sonic bath.
Analytical scanning electron microscopy was carried out using a
Leo 1530 Gemini SEM under high vacuum. The SEM was equipped
with a Thermo Scientific UltraDry energy-dispersive X-ray (EDX)
system, which was used for elemental mapping and chemical quanti-
fication. The microscope was operated at 20 kV for general imaging,
256 J. Lehmusto et al. / Fuel Processing Technology 104 (2012) 253–264
Fig. 5. Oxide layer thickness distributions of 10CrMo steel after exposure in air with
30% H2O for 168 h.
15 kV for panoramic imaging, and either 15 or 20 kV for the EDX
analyses. The back scattered electron mode was used to obtain the
images required to analyze the thickness of the oxide layer, while
Fig. 6. Elemental maps for Alloy 625 (upper) and 10CrMo steels (lower) under KCl after exposure in dry air at 500 °C for 168 h.
the secondary electron mode was used for detailed structural studies
of this layer viewed from the side. A panoramic SEM picture was
obtained by combining 17–21 separate SEM micrographs. The oxide
layer thickness was determined using automatic image analysis: the
oxide layer was identified by contrast analysis and then measured.
The thickness was calculated as the mean thickness value based on
measurements at 10,000 points along the part of the cross section
covered with the salt deposit.
The test method and the experimental setup have been described
in detail in, e.g., [7].
3. Results and discussion
In this section, the results for both steels at all temperatures in-
vestigated are categorized according to the type of salt, and then
secondarily according to the atmosphere. Sample SEM images
used for the EDX elemental mapping are presented. The elemental
distribution is studied primarily in terms of atomic and weight per-
centages, but the images from the count measurements are used
because they are clearer. The count mode considers only those x-
ray photons which have arrived at the detector and have been
recorded by the analyzer.
Both the average and the maximal oxide layer thickness values
for each test are presented in Table 2 and are summarized in the
bar charts in Fig. 1.
 J. Lehmusto et al. / Fuel Processing Technology 104 (2012) 253–264 257

3.1. Samples with no salt

3.1.1. Dry conditions

In the tests performed under dry conditions, with no salt depos-
ited onto the steel specimens, no oxidation was detected on Alloy
625 at any of the temperatures investigated The oxide layer on
the steel surface was too thin to be detected from the cross-
sectional samples analyzed with SEM, nor could it be detected in
the cross-sectional EDX analyses.
In the case of the ferritic 10CrMo steel, oxidation was observed
at all temperatures. The oxide layer thickness grew as a function of
temperature, with average and maximum thicknesses of 5 and
18 μm at 500 °C; 9 and 16 μm at 550 °C; and 18 and 40 μm at
600 °C. The iron was distributed evenly throughout the oxide layer,
whereas chromium was found only in the region just above the
unreacted steel (Fig. 2). This is attributed to the fact that chromium
diffuses more slowly than iron in the magnetite and wüstite lattice
[33].
An examination of the distribution profiles of the oxide layer
thickness values showed that the oxide layer thickness increases
with increasing temperature (Fig. 3). Additionally, the distribution
curve for the highest temperature is clearly broader, indicating an
oxide layer with varying thickness values. This might be attributed
to the occurrence of a different oxidation mechanism, such as pitting,
at higher temperatures; another possibility is accelerated oxidation
due to temperature-dependent diffusion involving structural disparities
such as grain boundaries.

3.1.2. Wet conditions

The nickel-based Alloy 625 also exhibited high corrosion resis-
tance under wet conditions: no oxide layer could be detected at the
temperatures investigated. This is in agreement with earlier studies
showing that, under the same conditions, the corrosion resistance
of a nickel-based steel is higher than that of an austenitic steel [23].
Because the chromium content of these steels did not differ greatly,
the improved corrosion resistance was due to the presence of other
elements, such as nickel and/or molybdenum.
In the case of the ferritic 10CrMo steel, oxidation was observed
at all temperatures. The average corrosion layer thicknesses were
20 μm at 500 °C, 34 μm at 550 °C, and 92 μm at 600 °C, whereas the Fig. 7. Oxide layer thickness distributions for Alloy 625 (A) and 10CrMo steels (B) under
maximum corrosion layer thicknesses were 50 μm at 500 °C, 40 μm KCl after exposure in dry air for 168 h.
at 550 °C, and 108 μm at 600 °C. Despite the increased layer thickness,
the elemental distribution of the oxide formed under wet conditions
was similar to that formed under dry conditions (Fig. 4). The
outward-grown oxide above the original steel surface contained
iron alone, whereas the chromium was located in the inward-grown
layer in the form of chromium oxide hydroxide has been reported
mixed oxide close to the original steel surface. Although the composi-
earlier [26], but since the ferritic 10CrMo contained only 2.2 wt-%
tion of the iron oxide could not be determined, based on its appear-
chromium, the role of water vapor in the volatilization of chromium
ance, it is most likely magnetite (Fe3O4), as reported earlier [33].
species is minimal.
The voids are thought to have formed via the condensation of inward
moving vacancies.
When the thickness distribution profiles of the oxide layer were 3.2. Samples under KCl
examined, it could be seen that the oxide layers formed at the two
lower temperatures have thicknesses falling within a narrow region 3.2.1. Dry conditions
of the scale, indicating homogeneous oxidation (Fig. 5). Because of In the presence of KCl, Alloy 625 exhibited severe corrosion at all
the temperature dependence of diffusion, the maximum was shifted test temperatures. The measured mean oxide layer thicknesses were
up 15 μm as the temperature increased from 500° to 550 °C. Apart 12 μm (500 °C); 10 μm (550 °C); and 40 μm (600 °C). The oxide con-
from the location of the maxima, a comparison of the profiles indi- sisted mainly of nickel, with smaller amounts of chromium (Fig. 6).
cates that oxidation occurs in a similar manner at 500 °C and 550 °C. There were also regions containing potassium, chromium, and
The distribution profile for the sample maintained at 600 °C is shifted oxygen, but no chlorine. This indicates the formation of potassium
towards higher thickness values, and is also broader, indicating more chromate (K2CrO4), a solid intermediate, as described by Eq. (1b);
heterogeneous oxidation than at lower temperatures. An analysis of the same compound was also detected in earlier studies [23,29].
the effects of dry and wet air showed similarities in the two distribu- Interestingly, there were other regions in which chlorine was found
tion profiles at 600 °C (Figs. 3 and 5), and in the oxide structures without potassium, but with chromium. This indicates the formation
(Figs. 2 and 4), suggesting that oxidation occurs in a similar manner of another solid intermediate, chromium chloride (CrCl2), which has
under both conditions. The volatilization of the protective chromia been reported to form at low oxygen partial pressures [9,17] and has
258 J. Lehmusto et al. / Fuel Processing Technology 104 (2012) 253–264
Fig. 8. Elemental maps for Alloy 625 (upper) and 10CrMo steels (lower) under KCl after exposure in wet air at 600 °C for 168 h.
been found in experiments with an austenitic 304 L steel [23]. Small
amounts of K2CrO4 were also detected at the surface of the salt. Chro-
mium drifted from the steel surface most likely in the form of gaseous
chromium chloride and was converted into solid chromium oxide at
higher oxygen partial pressures. The chromium oxide then reacted
with potassium chloride to form K2CrO4.
In the case of Alloy 625 at the two lowest temperatures, the oxide
layer had roughly the same thickness as that on the austenitic 304 L
reported in [23]. At 600 °C, both the average and the maximal oxide
layer thicknesses of Alloy 625 were higher than the corresponding
values for 304 L. There are two possible explanations: the nickel oxide
on Alloy 625 is more porous than the chromium oxide on 304 L,
which makes the oxide layer thicker; or the oxide layer on Alloy 625
is thicker because there was greater material loss, in other words,
more severe corrosion.
The most severe corrosion was observed with ferritic 10CrMo
steel: 17 μm (500 °C); 90 μm (550 °C); and 124 μm (600 °C). At
all temperatures, the oxide consisted almost solely of iron with
smaller amounts of chromium (Fig. 6), and it had a multilayered
structure.
The distribution profiles for both steels are similar: apart from the
distribution curve of 10CrMo at 500 °C, none of the curves has a clear
maximum, and the oxide layer thicknesses are scattered over a broad
range (Fig. 7). The broad distribution of the thicknesses indicates
inhomogeneous oxidation, which is apparent from the panoramic
images, especially in the case of Alloy 625, for which pit-type, local-
ized material loss can be detected. The distribution profiles for the
two steels differ somewhat from those for 304 L, which have clear max-
ima, except for the highest test temperature [23]. The reason for the
broadening of the profiles for 10CrMo and Alloy 625 remains elusive:
the chromium content of 304 L and Alloy 625 is nearly the same, and al-
though the manganese content is higher for 304 L than for 10CrMo or
Alloy 625, it should not have an effect on the oxidation behavior of
the steels in the presence of a salt.
3.2.2. Wet conditions
Alloy 625 had high corrosion resistance in the presence of KCl in a
wet atmosphere. Detectable oxidation occurred only at the highest
temperature, 600 °C; under these conditions, an oxide layer with a
thickness of 2 μm was formed. The formation of a thinner oxide layer
under wet conditions than under dry conditions was also observed
in an earlier study with the austenitic 304 L steel [23]. There are two
possible explanations: in the presence of water vapor, a denser and
thus thinner oxide layer might form; or water vapor might inhibit
the reaction between the steel and the salt. The oxide layers formed
under wet and dry conditions have similar structures: the uppermost
 J. Lehmusto et al. / Fuel Processing Technology 104 (2012) 253–264
zone is almost pure nickel oxide, followed by a zone with a chromium-
rich oxide (Fig. 8). The chromium-rich region most likely originated
from the diffusion of nickel to the upper oxide, rather than from the
diffusion of chromium. In contrast to what occurred under dry condi-
tions, there were no clear indications of the formation of potassium
chromate or chromium chloride.
The ferritic 10CrMo experienced severe corrosion at all test tem-
peratures; the oxide layer thicknesses were 28 μm (500 °C); 38 μm
(550 °C); and 89 μm (600 °C). The oxide layers formed under wet
conditions were thinner than those formed under dry conditions,
which is the same as the pattern observed for Alloy 625. The elemen-
tal distribution of the oxide layer is analogous, as can best be seen in
the sample without any salt under wet conditions (Fig. 4): the oxide
layer consists almost completely of iron oxide, whereas chromium
can be found in the lower parts of the oxide (Fig. 8). The oxide does
not have the multilayered structure which forms in the presence
of KCl under dry conditions; rather, it resembles the structure of
the oxide formed under wet conditions with no salt. The oxide
layer thickness for the case with KCl under dry conditions is rel-
atively close to the oxide layer thickness with no salt under wet
Fig. 9. Oxide layer thickness distributions for Alloy 625 (A) and 10CrMo steels (B) under
KCl after exposure in wet air for 168 h.
259
conditions. This might indicate that, with a ferritic 10CrMo steel
under wet conditions, the presence of KCl does not change the
oxidation behavior of the steel significantly. A possible explana-
tion is that the influence of KCl blends into the impact of water
on the accelerated oxidation of iron-based steel, so they cannot
be distinguished.
The oxide layer thickness distribution profile for Alloy 625 has a
narrow peak at around 1 μm, which contains 45% of all the measured
values (Fig. 9). In addition to the peak indicating slight and homoge-
neous oxidation, some minor tailing can be observed, corresponding
to higher oxide layer thickness values. Because there were few higher
thickness values, it is not clear whether significant pitting occurred. In
the case of 10CrMo, the distribution curves are broad, so no unambigu-
ous peak maximum can be determined at any of the test temperatures.
The oxide layer thicknesses are shifted towards higher values with
increasing temperature, but the shapes of the profiles do not change
substantially as a function of temperature. The oxide layer thickness
values are not as widely distributed as those measured under dry
conditions. Although detailed information is still unavailable, the results
suggest that water vapor plays a role in the oxidation of steels in
the presence of potassium chloride: it might result in the formation of
denser, and thus thinner oxide layers, or it might interact with the
salt, which could have an effect on the interaction between the salt
and the steel.
3.3. Samples under K2CO3
3.3.1. Dry conditions
Accelerated oxidation was observed in the case of Alloy 625 only
at the highest test temperature, 600 °C, at which a 7 μm oxide layer
formed. The oxide contained mainly chromium, with a nickel-rich re-
gion at the interface of the oxide and the bulk steel (Fig. 10). Although
Alloy 625 contains only 2.2 wt.% iron, a continuous layer of iron
formed along the surface of the oxide.
When the reaction between iron and potassium carbonate was
studied thermogravimetrically, no reaction-based DTA signal could
be detected at 600 °C. The relative mass of the reaction mixture
increased by 3%, most likely due to the oxidation of iron (Fig. 11).
Because the reaction between iron and potassium carbonate occurred
on a very small scale, the cause of iron diffusion to the uppermost
part of the oxide layer of Alloy 625 is unclear. The chromium-rich
composition of the oxide suggests that, although the oxide thickened,
it did not lose its protective properties. The thickness and the struc-
ture of the oxide layer formed on Alloy 625 are similar to those
of the protective chromium oxide layer formed on 304 L at 600 °C
under wet conditions without salt [23]. It should be noted that oxides
formed in the presence of potassium chloride (Fig. 6) instead of
potassium carbonate consist of nickel oxide rather than chromium
oxide.
Accelerated oxidation was observed for 10CrMo at all test tem-
peratures: 4 μm (500 °C); 20 μm (550 °C); and 38 μm (600 °C). The
oxides are thinner than those formed in the presence of potassium
chloride. The structure of the oxides is closer to that of the oxides
formed in dry air with no salt than it is to the structure of the ox-
ides formed in the presence of potassium chloride. The oxides do
not have a multilayered structure and consist largely of iron
oxide, with chromium located near the original steel surface
(Fig. 10).
For both steels, the oxide layer is thicker than that formed in dry
air with no salt. This indicates that potassium carbonate, or more pre-
cisely potassium, plays an active role in accelerated oxidation. Traces
of a potassium-containing compound, most likely potassium chro-
mate, were found in both steels. This is in good agreement with
the results from the thermogravimetric studies, which showed that
potassium chromate is formed in the reaction between chromium
oxide and potassium carbonate [24].
260 J. Lehmusto et al. / Fuel Processing Technology 104 (2012) 253–264
Fig. 10. Elemental maps for Alloy 625 (upper) and 10CrMo steels (lower) under K2CO3 after exposure in dry air at 600 °C for 168 h.
The narrow distribution profile for Alloy 625 suggests homoge-
neous oxidation (Fig. 12). This is supported by the panoramic SEM
image, in which no localized corrosion or pits can be seen. Although
the distribution profiles for the austenitic 304 L steel [23] are broader
than those for Alloy 625, homogeneous oxidation occurs in both
cases. However, based on the distribution profiles alone, it is not
Fig. 11. The relative mass and DTA curves for the Fe + K2CO3 mixture (mole ratio 1:1)
as functions of temperature in synthetic air at a heating rate of 2 °C min− 1.
possible to determine whether the reaction mechanism is the same
for both steels. The oxide formed on Alloy 625 consists of cone-shaped
chromium oxide crystals, between which the grain boundaries can
be clearly seen (Fig. 13); these might function as diffusion paths for
chromium and oxygen, or as possible passageways for intermediate
compounds.
At all test temperatures, 10CrMo experienced accelerated oxida-
tion, and although the distribution profiles are broader than those
for Alloy 625, they are not substantially broader than the profiles for
304 L [23]. The small tails suggest that oxidation is relatively homoge-
neous. Once again, however, based on the distribution profiles alone,
it is not possible to determine whether the reaction mechanism is the
same for the various steels; it is unlikely since the compositions of the
steels are substantially different.
3.3.2. Wet conditions
Accelerated oxidation was observed in the case of Alloy 625 only
at the highest test temperature, 600 °C, at which a 10 μm oxide
layer was formed. The thickness of the layer is very close to that of
the layer formed under K2CO3 in dry air at 600 °C. In addition, the
elemental distribution of the oxides is nearly identical under the
two atmospheres: a continuous layer of iron at the surface, followed
by an oxide containing mainly chromium, underneath which lies a
nickel-rich region at the interface of the oxide and the bulk steel
(Fig. 14). The chromium oxide layer consists of cone-shaped crystals
similar to those in the layer formed under dry conditions. The similar-
ities in the oxide layer thicknesses, elemental distributions, and oxide
morphologies suggest that in the case of potassium carbonate, water
 J. Lehmusto et al. / Fuel Processing Technology 104 (2012) 253–264
Fig. 12. Oxide layer thickness distributions of Alloy 625 (A) and 10CrMo (B) steels
under K2CO3 after exposure in dry air for 168 h.
vapor does not have an effect on the reaction mechanism or on the
oxidation behavior of the steel.
For 10CrMo, accelerated oxidation was observed at all test
temperatures. The layer thicknesses were 15 μm (500 °C); 29 μm
(550 °C); and 24 μm (600 °C). The oxide layers are thicker than
those formed under dry conditions, except at the highest test temper-
ature. The oxide layers formed in the presence of potassium carbonate
are thinner than those formed under wet conditions either without
salt or in the presence of potassium chloride. This suggests that potas-
sium chloride promotes corrosion more aggressively than potassium
carbonate. The thinner oxide layers formed in the presence of potas-
sium carbonate could be attributed to the fact that water vapor has
reduced access to the steel surface in the presence of potassium car-
bonate. However, this does not make the presence of potassium car-
bonate desirable because these oxide layers adhere poorly to the
steel (Fig. 14). Apparently, the presence of potassium carbonate in-
terferes with oxide layer growth, thus preventing a protective layer
from forming. The oxide layer consists almost completely of iron
oxide, whereas chromium can be found in the lower parts of the
oxide near the original steel surface. Although the oxide adheres
261
Fig. 13. A backscattering image of the oxide layer formed on Alloy 625 under K2CO3
after exposure in dry air at 600 °C for 168 h.
poorly to the steel surface, it lacks the multilayered structure found
in the case of potassium chloride under dry conditions (Fig. 6).
The narrow distribution profile for Alloy 625 supports the idea
that the steel has similar oxidation behavior in both dry and wet
air (Fig. 15). Although some minor tailing can be seen, the SEM im-
ages do not reveal any significant localized corrosion, and the oxide
layer appears to have a constant thickness. The shape of the distribu-
tion profile is the same as that for 304 L at 500 °C under the same
conditions, whereas the other profiles for 304 L are clearly broader
[23].
The distribution profiles for 10CrMo are broader than those
under dry conditions, but the trend is the same: the maximum
is shifted towards higher layer thickness values as the tempera-
ture rises (Fig. 15). The shape of the profile for 550 °C is different:
it is clearly narrower than the profiles for the other two temper-
atures. No unambiguous reason for this could be found. Compared
to the profiles for 304 L, those for 10CrMo are broader at lower
temperatures, but the shift towards higher layer thickness values
with increasing temperature is greater in the case of 304 L than in
the case of 10CrMo. This could be attributed to the different com-
positions of the steels: the low chromium content of 10CrMo
provides less source material for the temperature-dependent re-
action between chromium and potassium carbonate, whereas
with 304 L, there is enough chromium in the steel for the reaction
to be affected by the higher temperature.
In the case of the ferritic 10CrMo steel, the oxide layers
formed under wet conditions in the presence of KCl are thicker
than those formed in the presence of K2CO3. Interestingly, the
opposite is true for the nickel-based Alloy 625 steel: the presence
of K2CO3 results in thicker oxide layers under wet conditions. In
terms of the mass gain [21] and oxide layer thickness [23], simi-
lar behavior was observed for an austenitic 304 L steel. The low
chromium content and the resulting limited formation of potassi-
um chromate on 10CrMo could explain the minor effect of K2CO3
on accelerated oxidation. In the case of KCl, accelerated oxidation
can proceed due to the presence of the chloride anion, regardless
of the limited formation of potassium chromate. The higher
chromium contents of the 304 L and Alloy 625 steels enable po-
tassium chromate formation, which could explain the more no-
ticeable effect of K2CO3 on oxide layer growth. The reason for
the inhibiting effect of water vapor on potassium chloride-
induced oxidation remains elusive, so it needs to be studied in
greater detail. One explanation could be the formation of denser
262 J. Lehmusto et al. / Fuel Processing Technology 104 (2012) 253–264
Fig. 14. Elemental maps for Alloy 625 (upper) and 10CrMo steels (lower) under K2CO3 after exposure in wet air at 600 °C for 168 h.
and thus thinner oxide layers in the presence of water. Another
possibility is the interaction between potassium chloride and
water vapor, which might compete with the interaction between
potassium chloride and chromium-rich steels.
4. Conclusions
A comparison of the corrosive nature of potassium chloride
(KCl) and potassium carbonate (K2CO3) on a 10CrMo-type ferritic
steel and on an Alloy 625-type nickel-based steel showed that
presence of potassium is sufficient to initiate accelerated oxida-
tion. However, the differences in the thicknesses and compositions
of the oxide layers formed depend not only on the temperature
and atmosphere, but also on the potassium compound: in the
case of Alloy 625, the oxide layers formed in the presence of KCl
are significantly thicker. In addition, oxide formed in the presence
of KCl consists mainly of nickel oxide, whereas in the presence of
K2CO3, a thinner and more homogeneous layer forms consisting
mostly of chromium oxide, suggesting that the oxide at least partly
retains its protective properties. Although 10CrMo experienced ac-
celerated oxidation under every test condition, the oxides formed
in the presence of K2CO3 are well-adhered, whereas the oxides
formed in the presence of KCl are multilayered, with poorly adher-
ing layers. This suggests that after accelerated oxidation is initiated
by the presence of potassium, the presence of chloride is essential
for it to continue.
The presence of water vapor has a clear effect on the oxide
layer thickness. When no salt is present, Alloy 625 strongly resists
accelerated oxidation, whereas in the case of 10CrMo, water vapor
accelerates the oxidation of the steel. When the two atmospheres
were compared, it was observed that thinner oxides form under
wet conditions in the presence of KCl. One possible explanation
is that water causes the oxide layer to be denser; another is that
KCl might form volatile species such as HCl in the presence of
water, which would decrease the amount of reactive chloride in
the system. However, in the case of K2CO3, no volatile compound
was observed. If the latter explanation holds true, in future
studies of high temperature corrosion, more attention should be
paid to the synergistic effect of water vapor and salt-induced
corrosion.
Acknowledgements
The authors would like to thank Mr. Linus Silvander for operat-
ing the SEM apparatus, Mr. Peter Backman for operating the DTA/
TG apparatus, and Mrs. Piia Leppäsalo for coloring the SEM pano-
rama images. This work was carried out within the ChemCom 2.0
(2008–2010) and FUSEC (2011–2014) projects, as part of the ac-
tivities of the Åbo Akademi Process Chemistry Centre. Other re-
search partners are VTT, Lappeenranta University of Technology,
Aalto University, and Tampere University of Technology. The au-
thors also gratefully acknowledge the support provided by the
 J. Lehmusto et al. / Fuel Processing Technology 104 (2012) 253–264
Fig. 15. Oxide layer thickness distributions for Alloy 625 (A), 10CrMo (B), and 304 L
steels (C) under K2CO3 after exposure in wet air for 168 h. Partly reproduced from [23].
National Technology Agency of Finland (TEKES), Andritz Oy, Metso
Power Oy, Oy Metsä-Botnia Ab, Foster Wheeler Energia Oy, UPM-
Kymmene OYj, Clyde Bergemann GmbH, International Paper Inc.,
263
and Top Analytica Oy Ab. Additional funding was received from
the Finnish Graduate School of Chemical Engineering.
References
[1] H.J. Grabke, E. Reese, M. Spiegel, The effects of chlorides, hydrogen chloride and
sulfur dioxide in the oxidation of steels below deposits, Corrosion Science 37
(1995) 1023–1043.
[2] K. Salmenoja, K. Makela, M. Hupa, R. Backman, Superheater corrosion in environments
containing potassium and chlorine, Journal of the Institute of Energy 69 (1996)
155–162.
[3] H.P. Nielsen, F.J. Frandsen, K. Dam-Johansen, L.L. Baxter, The implications of
chlorine-associated corrosion on the operation of biomass-fired boilers, Progress
in Energy and Combustion 26 (2000) 283–298.
[4] H.P. Nielsen, L.L. Baxter, G. Sclippa, C. Morey, F.J. Frandsen, K. Dam-Johansen,
Deposition of potassium salts on heat transfer surfaces in straw-fired boilers:
a pilot-scale study, Fuel 79 (2000) 131–139.
[5] F.J. Frandsen, Utilizing biomass and waste for power production—a decade of
contributing to the understanding, interpretation and analysis of deposits and
corrosion products, Fuel 84 (2005) 1277–1294.
[6] S.C. Cha, High temperature corrosion of superheater materials below deposited
biomass ashes in biomass combusting atmospheres, Corrosion Engineering
Science and Technology 42 (2007) 50–60.
[7] B.-J. Skrifvars, R. Backman, M. Hupa, K. Salmenoja, E. Vakkilainen, Corrosion
of superheater steel materials under alkali salt deposits, Part 1: the effect
of salt deposit composition and temperature, Corrosion Science 50 (2008)
1274–1282.
[8] E. Reese, H.J. Grabke, Einfluss von Natriumchlorid auf die Oxidation von
hochlegierten Chrom- und Chrom-Nickel-Stählen, Werkstoffe und Korrosion 44
(1993) 41–47.
[9] M. Spiegel, Salt melt induced corrosion of metallic materials in waste incineration
plants, Materials and Corrosion 50 (1999) 373–393.
[10] A. Zahs, M. Spiegel, H.J. Grabke, The influence of alloying elements on the
chlorine-induced high temperature corrosion of Fe– Cr alloys in oxidizing atmo-
spheres, Materials and Corrosion 50 (1999) 561–578.
[11] A. Zahs, M. Spiegel, H.J. Grabke, Chloridation and oxidation of iron, chromium,
nickel and their alloys in chloridizing and oxidizing atmospheres at 400–700°C,
Corrosion Science 42 (2000) 1093–1122.
[12] T. Ishitsuka, K. Nose, Stability of protective oxide films in waste incineration
environment—solubility measurements of oxides in molten chlorides, Corrosion
Science 44 (2002) 247–263.
[13] Y. Kawahara, High temperature corrosion mechanisms and effect of alloying elements
for materials used in waste incineration environment, Corrosion Science 44 (2002)
223–245.
[14] M.A. Uusitalo, P.M.J. Vuoristo, T.A. Mäntylä, High temperature corrosion of coatings and
boiler steels below chlorine-containing salt deposits, Corrosion Science 46 (2004)
1311–1331.
[15] S.C. Cha, M. Spiegel, Fundamental studies on alkali chloride induced corrosion during
combustion of biomass, Materials Science Forum 461–464 (2004) 1055–1062.
[16] Y.S. Li, M. Spiegel, S. Shimada, Corrosion behaviour of various model alloys with
NaCl–KCl coating, Materials Chemistry and Physics 93 (2005) 217–223.
[17] J. Pettersson, H. Asteman, J.-E. Svensson, L.-G. Johansson, KCl induced corrosion of
a 304-type austenitic stainless steel at 600°C; The role of potassium, Oxidation of
Metals 64 (2005) 23–41.
[18] A. Ruh, M. Spiegel, Thermodynamic and kinetic consideration on the corrosion of
Fe, Ni and Cr beneath a molten KCl— ZnCl2 mixture, Corrosion Science 48 (2006)
679–695.
[19] I.V. Oryshich, High-temperature chloride corrosion of industrial ЭИ827 and ВЖ-85
alloys, Protection of Metals 42 (2006) 596–600.
[20] Y. Shinata, Y. Nishi, NaCl-induced accelerated oxidation of chromium, Oxidation
of Metals 26 (1986) 201–212.
[21] J. Pettersson, N. Folkeson, L.-G. Johansson, J.-E. Svensson, The effects of KCl, K2SO4
and K2CO3 on the high temperature corrosion of a 304-type austenitic stainless
steel, Oxidation of Metals 76 (2011) 93–109.
[22] T. Blomberg, A thermodynamic study of the gaseous potassium chemistry in the con-
vection sections of biomass fired boilers, Materials and Corrosion 62 (2011) 635–641.
[23] J. Lehmusto, B.–J. Skrifvars, P. Yrjas, M. Hupa, Comparison of potassium chloride
and potassium carbonate with respect to their tendency to cause high tempera-
ture corrosion of stainless 304L steel, Original Research Article, Fuel Processing
Technology, (in press), (Available online 11 January 2012), http://dx.doi.org/10.
1016/j.fuproc.2011.12.016.
[24] J. Lehmusto, D. Lindberg, P. Yrjas, B.–J. Skrifvars, M. Hupa, Thermogravimetric
studies of high temperature reactions between potassium salts and chromium,
Corrosion Science 59 (2012) 55–62.
[25] E.J. Opila, Volatility of common protective oxides in high-temperature water
vapor: current understanding and unanswered questions, Materials Science
Forum 461–464 (2004) 765–774.
[26] H. Asteman, J.-E. Svensson, L.-G. Johansson, M. Norell, Indication of chromium
oxide hydroxide evaporation during oxidation of 304L at 873 K in the presence
of 10% water vapor, Oxidation of Metals 52 (1999) 95–111.
[27] H. Asteman, J.-E. Svensson, L.-G. Johansson, Evidence for chromium evaporation
influencing the oxidation of 304L: the effect of temperature and flow rate, Oxidation
of Metals 57 (2002) 193–216.
[28] F. Wang, Y. Shu, Influence of Cr content on the corrosion of Fe–Cr alloys: the synergistic
effect of NaCl and water vapor, Oxidation of Metals 59 (2003) 201–214.
264 J. Lehmusto et al. / Fuel Processing Technology 104 (2012) 253–264
[29] J. Pettersson, J.-E. Svensson, L.-G. Johansson, KCl-induced corrosion of a 304-type
austenitic stainless steel in O2 and in O2 + H2O environment: the influence of
temperature, Oxidation of Metals 72 (2009) 159–177.
[30] M. Spiegel, P. Biedenkopf, H.J. Grabke, Corrosion on iron base alloys and high
alloy steels in the Li2CO3–K2CO3 eutectic mixture, Corrosion Science 39 (1997)
1193–1210.
[31] S.A. Salih, A.N. El-Masri, A.M. Baraka, Corrosion behaviour of some stainless
steel alloys in molten alkali carbonates, Journal of Materials Science 36 (2001)
2547–2555.
[32] K. Takeuchi, A. Nishijima, K. Ui, N. Koura, C.-K. Loong, Corrosion behavior of Fe–Cr
alloys in Li2CO3–K2CO3 molten carbonate, Journal of the Electrochemical Society
152 (2005) B364–B368.
[33] J. Żurek, E. Wessel, L. Niewolak, F. Schmitz, T.-U. Kern, L. Singheiser, W.J. Quadakkers,
Anomalous temperature dependence of oxidation kinetics during steam oxidation of
ferritic steels in the temperature range 550–600°C, Corrosion Science 46 (2004)
2301–2317.