Note: Descriptions are shown in the official language in which they were submitted.
I
TITLE OF THE INVENTION
Method and system for fabrication of suspension plasma sprayed TiB2 coatings
FIELD OF THE INVENTION
[0001] The present invention relates to TiB2-based coatings. More
specifically, the present invention is concerned
with a method and a system for fabrication of suspension plasma sprayed TiB2
coatings.
BACKGROUND
[0002] Titanium diboride (TiB2) is a high melting point (3225 C) material
characterized by high hardness and wear
resistance, good chemical stability in corrosive and high-temperature
environments, and high thermal and electrical
conductivities. Used in applications such as cutting tools, ballistic armors,
turbine blades, crucibles for molten metals,
thermocouple sheaths, nuclear rods. TiB2 is also a promising candidate as a
cathode material for primary aluminum
production due its good electrical conductivity, very low solubility in molten
Al, good resistance to corrosion by cryolitic
electrolyte and excellent aluminum wettability, to replace commonly used
graphite cathodes by wettable TiB2 cathodes
in aluminum smelters fo example. Moreover, when coupled with inert anodes that
emit 02 instead of CO2 during the
aluminum electrolysis process, a substantial decrease of GHG emissions is
expected in addition to potentially reduce
the operating and capital costs.
[0003] The manufacturing of massive TiB2 cathodes is however challenging
because of the poor sintering ability of
TiB2, requiring high temperature and pressure. The use of sintering aid
additives such as Ti and Fe allows the
pressureless sintering of TiB2 at a relatively low temperature, of about 1600
C, preventing exaggerated grain growth
detrimental for its mechanical properties. However, in contact with molten Al,
additives and remaining oxides at the
TiB2 grain boundaries may react and form phases with increased molar volume,
inducing crack formation and, in turn,
the failure of the TiB2 cathode during aluminum electrolysis. TiB2 composite
cathode material, such as TiB2-AIN and
TiB2-TiC for example have also been investigated to remedy the brittleness of
pure TiB2 ceramic material and improve
its resistance to thermal shock. However, their properties, especially
resistance to corrosion and electrical conductivity,
still need to be further improved.
[0004] Alternatively, wettable TiB2-based cathodes can be produced as TiB2-
coated materials. Carbon is usually
selected as a substrate mainly because of its low cost and thermal expansion
coefficient close to the thermal
expansion coefficient of TiB2. To date, different TiB2 coating methods have
been evaluated such as electrodeposition,
chemical vapor deposition, plasma spraying, colloidal and carbon-bonded
coatings, each coating method having its
Date Recue/Date Received 2022-01-26
2
own advantages and limitations.
[0005] Plasma spraying is a well-established method for producing with a
relatively high deposition rate, typically of a
few kg h-1, different types of protective coatings for various industrial
applications such as transport, energy, materials
extraction and processing, biomedical, and electronic applications for
example. Plasma spraying has been used to
deposit TiB2 in various conditions. For instance, TiB2 have been deposited on
alumina substrates by air plasma
spraying (APS). However, TiB2 undergoes partial oxidation during APS,
resulting in coatings with decreased electrical
conductivity. The use of argon as a shield gas, the addition of carbon in the
powder or the use of Ar-H2 plasma may
enhance the electrical conductivity of the coatings but oxidation cannot be
totally avoided. Air plasma spraying (APS)
was also used to deposit TiB2 on carbon cathodes, resulting in about 800 [tm-
thick coatings made of a mixture of fully
and semi melted TiB2 particles, partially oxidized to TiO2 and B203. The
microstructure and tribological properties of air
plasma spraying (APS) TiB2 coating were also investigated, showing a rough
coating with a lamellar microstructure, a
porosity of 12%, and an oxygen content of 3 wt %. TiB2-MoSi2 composites were
deposited on carbon cathodes by air
plasma spraying (APS) in which the presence of MoSi2 favours the densification
of the coating. Vacuum plasma
spraying (VPS) has also been used to fabricate TiB2 coatings on different
carbon materials, and the influences of
different process parameters on the microstructure of the deposits were
studied. It was shown that the coating density
and pore size distributions are sensitive to the chamber pressure. The vacuum
plasma spraying (VPS) TiB2 deposits
have a high hardness, good adherence to the carbon substrate, and a good
wettability by aluminum. Thus, although
plasma spraying has been successfully used to produce wettable TiB2-coated
cathodes, resulting TiB2 coatings are
usually porous because of the difficulty to fully melt the TiB2 powder, and
this porosity may cause aluminum and the
electrolyte to reach the substrate, leading to the deterioration of the
substrate/coating interface and degradation of the
substrate.
[0006] There is still a need for a method and a system for fabrication of TiB2
coatings.
SUMMARY OF THE INVENTION
[0007] More specifically, in accordance with the present invention, there is
provided a method comprising
selecting a powder suspension of TiB2 particules of a size of at most 10 pm;
positioning a plasma torch at a distance
from the substrate; and spraying the TiB2 particules on the substrate by
injecting the suspension of the powder of TiB2
particules into the plasma torch; wherein the method comprises selecting a
porosity of each layer deposited on the
substrate by selecting spraying parameters.
[0008] There is further provided a system comprising a plasma torch; with a
plasma current selected in a range
Date Recue/Date Received 2022-01-26
3
between 160 A and 250 A; a selected plasma gas comprising between 25 and 80%
argon, between 5 and 65%
nitrogen and between 10 and 15% hydrogen; a total gas flow of the plasma
selected in a range between 160 and 275
slpm, a feed rate of a powder suspension of TiB2 particules selected in a
range between 10 and 80 mL/min, a
suspension load of the TiB2 particules selected in a range between 5 and 30
weight%, and a temperature of the
substrate being controlled in a range between 200 and 900 C.
[0009] Other objects, advantages and features of the present invention will
become more apparent upon reading
of the following non-restrictive description of specific embodiments thereof,
given by way of example only with
reference to the accompanying drawings.
BRIEF DESCRIPTION OF THE DRAWINGS
[0010] In the appended drawings:
[0011] FIG. 1A shows a scanning electron microscope (SEM) image of as-received
TiB2 powder;
[0012] FIG. 1B shows particle size distribution (PSD) analysis of the as-
received TiB2 powder;
[0013] FIG. 2A shows a surface scanning electron microscope (SEM) image of a
TiB2 coating on a graphite substrate
obtained by suspension plasma spray (SPS) without shroud according to an
embodiment of an aspect of the present
disclosure;
[0014] FIG. 2B shows a cross-section scanning electron microscope (SEM) image
of a TiB2 coating on a graphite
substrate obtained by suspension plasma spray (SPS) without shroud according
to an embodiment of an aspect of the
present disclosure
[0015] FIG. 2C shows a cross-section scanning electron microscope (SEM) image
of a TiB2 coating on a graphite
substrate obtained by suspension plasma spray (SPS) without shroud according
to an embodiment of an aspect of the
present disclosure;
[0016] FIG. 2D shows a surface scanning electron microscope (SEM) images of a
TiB2 coating on a graphite
substrate obtained by suspension plasma spray (SPS) with a shroud according to
an embodiment of an aspect of the
present disclosure;
Date Recue/Date Received 2022-01-26
4
[0017] FIG. 2E shows a cross-section scanning electron microscope (SEM) image
of a TiB2 coating on a graphite
substrate obtained by suspension plasma spray (SPS) with a shroud according to
an embodiment of an aspect of the
present disclosure;
[0018] FIG. 2F shows a cross-section scanning electron microscope (SEM) image
of a TiB2 coating on a graphite
substrate obtained by suspension plasma spray (SPS) with a shroud according to
an embodiment of an aspect of the
present disclosure;
[0019] FIG. 3A show a cross-section scanning electron microscope (SEM)
micrograph of scratch test grooves made
at 5 N in a TiB2 coating deposited without shroud according to an embodiment
of an aspect of the present disclosure;
[0020] FIG. 3B show a cross-section scanning electron microscope (SEM)
micrograph of scratch test grooves made
at 5 N in a TiB2 coating deposited with a shroud according to an embodiment of
an aspect of the present disclosure;
[0021] FIG. 3C shows evolution of scratch width with applied load;
[0022] FIG. 4 shows X-ray diffraction (XRD) patterns of raw TiB2 powder and
TiB2 coatings obtained by suspension
plasma spray (SPS), without and with a shroud, according to an embodiment of
an aspect of the present disclosure;
[0023] FIG. 5A shows X-ray photoelectron spectroscopy (XPS) Ti 2p spectra of
TiB2 coatings prepared with (solid
line) and without (dotted line) a shroud according to an embodiment of an
aspect of the present disclosure;
[0024] FIG. 5B shows X-ray photoelectron spectroscopy (XPS) Ti Is spectra of
TiB2 coatings prepared with (solid
line) and without (dotted line) shroud according to an embodiment of an aspect
of the present disclosure;
[0025] FIG. 6 shows the evolution with time of the contact angle of molten
aluminum on TiB2 coatings deposited by
suspension plasma spray (SPS) with and without a shroud according to an
embodiment of an aspect of the present
disclosure;
[0026] FIG. 7A shows a cross-section scanning electron microscope (SEM)
micrograph of a SPS TiB2 coating
deposited with a shroud after 8 h of contact with molten aluminum at 1000 C;
Date Recue/Date Received 2022-01-26
5
[0027] FIG. 7B shows a cross-section scanning electron microscope (SEM)
micrograph of a TiB2 coating deposited
by suspension plasma spray (SPS) with a shroud after 8 h of contact with
molten aluminum at 1000 C;
[0028] FIG. 7C shows an energy dispersive spectroscopy (EDS) mapping image of
Ti element of a TiB2 coating
deposited by suspension plasma spray (SPS) with a shroud after 8 h of contact
with molten aluminum at 1000 C;
[0029] FIG. 7D shows an energy dispersive spectroscopy (EDS) mapping image of
aluminum element of a TiB2
coating deposited by suspension plasma spray (SPS) with a shroud after 8 h of
contact with molten aluminum at
1000 C;
[0030] FIG. 7E shows an energy dispersive spectroscopy (EDS) mapping image of
Zr element of a SPS TiB2 coating
made with shroud after 8 h of contact with molten aluminum at 1000 C;
[0031] FIG. 8 shows a cross-section scanning electron microscope (SEM)
micrograph after 8 h of contact with molten
aluminum of a TiB2 coating by suspension plasma spray (SPS) with a shroud;
[0032] FIG. 9A shows X-ray diffraction (XRD) pattern of TiB2 raw powder;
[0033] FIG. 9B shows particle size distribution histogram of the TiB2 raw
powder;
[0034] FIG. 9C shows scanning electron microscope (SEM) micrograph of the TiB2
raw powder;
[0035] FIG. 9D shows elemental Ti energy dispersive spectroscopy (EDS) mapping
of the TiB2 raw powder;
[0036] FIG. 9E shows elemental Zr energy dispersive spectroscopy (EDS) mapping
of the TiB2 raw powder;
[0037] FIG. 10A shows X-ray diffraction (XRD)pattems of S1 TiB2 coating;
[0038] FIG. 10B shows X-ray diffraction (XRD)pattems of S2 TiB2 coating;
Date Recue/Date Received 2022-01-26
6
[0039] FIG. 10C shows X-ray diffraction (XRD)pattems of S3 TiB2 coating;
[0040] FIG. 10D shows X-ray diffraction (XRD)pattems of S4 TiB2 coating;
[0041] FIG. 10E shows X-ray diffraction (XRD)pattems of S5 TiB2 coating;
[0042] FIGs. 11A-11E are top-views and FIGs. 11F-11J are cross-sectional
scanning electron microscope (SEM)
micrographs as follows: FIG. 11A, FIG. 11F: S1 (TiB2-P25); FIG. 11B, FIG. 11G:
S2 (TiB2-P19), FIG.11C, FIG. : 11H
S3 (TiB2-P13), FIG. 11D, FIG. 111: S4 (TiB2-P4) and FIG.11E, FIG.11J :S5 (TiB2-
P5/19) coatings;
[0043] FIG. 12 shows evolution with time of the contact angle of molten
aluminum at 1000 C on graphite, TiB2-P19,
TiB2-P13 and TiB2-P5/19 coatings;
[0044] FIGs. 13 show cross-section scanning electron microscope (SEM) images
and corresponding elemental C, Al
and Ti energy dispersive spectroscopy (EDS) maps of: FIG. 13A: TiB2-P19, FIG.
13B: TiB2-P13 and FIG. 13C: TiB2-
P5/19 coatings after sessile drop tests at 1000 C for 8 h;
[0045] FIG. 14 is a schematical sectional view of a system according to an
embodiment of an aspect of the present
invention; and
[0046] FIG. 15 is a schematical side view of the system of FIG. 14.
DESCRIPTION OF ILLUSTRATIVE EMBODIMENTS
[0047] The present invention is illustrated in further detail by the
following non-limiting examples.
[0048] A method for suspension plasma spray of TiB2-coated graphite cathodes
according to an embodiment of an
aspect of the present disclosure generally comprises injecting a suspension of
a powder of TiB2 particules in a solvent
into a plasma torch, in open air.
[0049] TiB2 particules of a size of at most 10 pm, in a range between about
100 nm and about 10 pm for example,
are selected for suspension in the powder. The spray parameters are selected
to adjust a residence time of the TiB2
particles in the plasma, to achieve complete melting of the particles, in
order to obtain a dense coating, with few
Date Recue/Date Received 2022-01-26
7
porosity, on the substrate.
[0050] The substrate may be in graphite, anthracite, or a combination for
example. The temperature of the substrate
is monitored using a thermocouple is connected to the back of the substrate
for example, and the length of time
between successive passes of the torch is decreased or increased to maintain
the temperature of the substrate in the
range between about 100 and 1000 C, for example between 400 C and 800 C.
[0051] Since chemical reactivity increases with the surface to volume ratio,
fine TiB2 particles are more sensitive to
oxidation during spraying than larger ones. To protect the particles from
oxidation, in a first set of experiments, a gas
shroud was used to isolate the plasma effluent from the ambient gas and for
retaining TiB2 particles of a size in a range
between about 100 nm and about 10 pm, within the plasma effluent, as
schematically illustrated for example in FIGs. 14
and 15. The characteristics, including morphology, crystalline structure,
chemical surface state and aluminum wettability,
of suspension plasma spray (SPS) TiB2 coatings prepared with and without an
argon shroud, using a commercially
available micrometric TiB2 powder (Hunan HuaWei Aerospace Special Materials,
China) were compared. Other inert,
non-reactive gas, such as helium for instance, could be used. TiB2 was
deposited on grit-blasted graphite substrates
made from graphite rods (GROO8G grade from Graphitestore, USA) cut in disks of
25.4 mm diameter and 5.5 mm
thickness. An axial III plasma spray system from Northwest Mettech Corp.
(Canada) was used to perform the
suspension plasma spray (SPS) deposition. A suspension of TiB2 powder (20 wt%)
in absolute ethanol containing
polyvinylpyrrolidone as dispersive agent (1.2 wt% was fed into a DC plasma
torch, with a feeding rate of the suspension
set to 45 mL min-1. The plasma gas was a mixture of 45% Ar, 45% N2 and 10% H2
with a flow rate of 180 L min-1. The
plasma power was 100 kW. Suspension plasma spray (SPS) depositions were
performed with and without the argon
shroud, setting a distance W between the output of the torch and the substrate
of 5.0 cm in the experiment without the
shroud and of 6.3 cm in the experiment with the shroud. The in-flight particle
speed was measured with an Accuraspray
4.0 device from Tecnar Automation Ltd. (Canada). In all cases, the particle
speed was between 550 and 570 m s-1. No
reliable temperature measurements were possible due to the high luminosity of
the plasma jet. For both deposition
conditions, namely with and without the shroud, ten substrates were positioned
on a vertically spinning sample holder,
and 15 passes were performed for the coatings, each pass consisting of one
back and forth displacement of the torch
over the substrates, each pass being separated by a pause of 10 s to keep the
substrate temperature below 400 C.
[0052] The particle size distribution (PSD) of the TiB2 powder was determined
by laser diffraction using a Malvern
Panalytical Spraytec instrument after dispersing the powder in ethanol using a
Malvern wet dispersion accessory.
[0053] The crystalline structure of the resulting samples was determined by X-
ray diffraction (XRD) using a Bruker D8
Advance diffractometer (Cu Ka radiation A = 0.15418 nm). Scanning electron
microscopy (SEM) micrographs were
Date Recue/Date Received 2022-01-26
8
obtained with a VEGA3 SEM from TESCAN coupled with XFlash 6110 energy
dispersive X-ray spectroscopy (EDS)
detector from Bruker for elemental analysis and mapping. Cross-section SEM
observations were done on resin-
embedded and polished samples. The apparent density of the coatings was
calculated using the relation: papp =
-Am 2e where Am is the weight difference before and after deposition, r is
the radius of the graphite disk substrate and t
nr
is the mean coating thickness measured from cross-section scanning electron
microscopy (SEM) images. X-ray
photoelectron spectroscopy (XPS) analyses of the TiB2 coatings were performed
with an Escalab 220i XL from
ThermoFisher Scientific. X-rays were issued from an aluminium monochromatic
source (hv = 1486.6 eV). For all
spectra, the energy scale was adjusted so that the Cis peak was at 284.8 eV.
XPS peak fitting was done with the
version 2.3.22 of CasaXPS software. Scratch test measurements were done with
the Micro Scratch Tester from CEM
instrument on a cross-section of resin-embedded samples. Constant normal loads
at 1, 2, 3, 5 and 10 N were applied
by a Rockwell C diamond type of 200 pm radius in ambient air. The diamond tip
moved along a straight line (1 to 1.5
mm) at a speed of 0.8 mm min-1 from the graphite substrate toward the TiB2
coating.
[0054] Aluminum wettability measurements were performed using a sessile drop
technique. Pieces of aluminium wire
(50 mg) were cleaned for 10 minutes in a 1M NaOH solution and then
ultrasonically cleaned in acetone for 10 min,
before being deposited on the substrate. Then, the substrates and the pieces
of aluminium wire were inserted in a
sealed furnace alumina tube that was vacuum-pumped and heated at 500 C
overnight. Then, the temperature was
brought to 1000 C at a heating rate of 4 C min-land was maintained at 1000 C
during the sessile drop measurement.
Both temperature and pressure were monitored during the experiment. Due to
sample outgassing, the pressure could
initially reach 10 to 50 mPa before slowly decreasing and stabilizing at less
than 4 mPa. A light was fixed at a first end
of the furnace tube and images of the aluminum drop were recorded by a camera
placed at a second end of the
furnace tube. The reported contact angle is the average of two measurements
made on the left and right hand-sides of
the aluminum drop. Angles were manually determined with ImageJ 1.52n software.
Scanning electron microscopy-
energy dispersive spectroscopy (SEM-EDS) cross-section observations of the
samples were performed after the
sessile drop experiment in order to evaluate aluminum infiltration into the
samples.
[0055] FIGs. 1 show a SEM micrograph (FIG. 1A) and PSD curves (FIG. 1B) of the
as-received TiB2 powder. The
TiB2 particles have an irregular shape with Dv10, Dv50 and Dv90 values of 0.9,
1A and 2.2 pm, respectively. Whiter
particles are observed on the SEM micrograph (FIG. 1A); localized energy
dispersive spectroscopy (EDS) analysis
suggests that these particles are zirconia-based particles, most probably
originating from the erosion of the milling
tools, typically made of zirconia-based materials, during the powder
fabrication process. From global energy dispersive
spectroscopy (EDS) analysis, the Zr/Ti atomic ratio of the as-received powder
is estimated in the range between about
2 and about 3%. B and 0 elements are too light to be quantified accurately
with the energy dispersive spectroscopy
Date Recue/Date Received 2022-01-26
9
(EDS) analyser.
[0056] Cross-section and top-view SEM micrographs of two SPS TiB2 coatings
obtained without and with the shroud
are shown in FIGs. 2. The thickness of the coatings is about 100 20 lam. Both
coatings exhibit a "cauliflower-type"
morphology, with cauliflower tops ranging in size from about 20 to about 100
pm. As seen in FIGs. 2C and 2F, the
morphology of a number of TiB2 particles in the coatings remains similar to
the morphology of the as-received powder
(FIG. 1A), which is indicative of limited melting during the spraying process,
and leads to a high porosity of the
coatings. The apparent density of both coatings is about 3 g.cm-3 or 66 % of
the TiB2 bulk density, confirming their high
porosity (34%). From this, it can be inferred that the small size of the
powder led to a shorter residence time of the TiB2
particles in the plasma plume. This fact, combined with the high melting point
and relatively low density of TiB2 has
greatly increased the porosity to levels higher than typical air plasma
spraying (APS) experiments. In both cases, with
and without the shroud, the interface between the coating and the graphite
substrate shows no defect and a good
anchorage of TiB2 particles (FIGs, 2B, 2E).
[0057] Scratch test measurements were performed to qualitatively investigate
the mechanical properties and the
adhesion strength of the coatings deposited on the graphite substrate.
Micrographs of the grooves formed by applying
a force of 5N are shown in FIGs. 3A and 3B for coatings made with and without
the shroud respectively. In both cases,
the stress is accommodated by a homogeneous and irreversible deformation of
the coating typical of a plastic
behavior. No delamination of the coating from the graphite substrate was
observed even at 10 N load, thus confirming
the good anchorage of the deposited layer on the graphite substrate. FIG. 3C
shows that the scratch width increases
with the load, but no difference is observed associated with the presence or
absence of the argon shroud.
[0058] X-ray diffraction (XRD) patterns of the as-received TiB2 powder and
TiB2 coatings prepared by suspension
plasma spray (SPS) with and without shroud are shown in FIG. 4. The XRD
pattern of the as-received powder exhibits
a series of peaks that can be indexed to TiB2, and no other peaks are
observed. The XRD patterns of TiB2 coatings
also display the same series of peaks, indicating that TiB2 has been
successfully deposited. No evidence of peaks that
could be associated with titanium oxide or boron oxide phases are observed.
The XRD traces of the TiB2 coatings
exhibit a new series of peaks at 20 = 30A5 and 50.3 that belongs to yttria
stabilized zirconia (YSZ) phase. A Rietveld
refinement analysis indicates that TiB2 coatings made with and without the
shroud have the same amount of YSZ,
typically between about 1.2 and about 1A mol%. The YSZ diffraction peaks are
not observed in the as-received
powder, most probably because it is amorphous or poorly crystallized. It may
be inferred that ZrO2 present as
contaminant in the as-received TiB2 powder crystallizes during the deposition,
giving rise to sharp diffraction peaks in
the XRD patterns of the resulting coatings. Alternatively, the higher
intensity of the YSZ peaks may relate to a higher
Date Recue/Date Received 2022-01-26
10
concentration of YSZ in the coatings compared to the as-received powder due to
a higher deposition efficiency (DE) of
YSZ as compared to TiB2. A higher deposition efficiency (DE) for YSZ can
result from its lower melting point and
higher density than TiB2. The TiB2 crystallite size increases from 36 nm for
the as-received powder to 52 nm and 99 nm
in the coatings with and without the shroud, respectively. This increase could
be attributed to in-flight grain growth due
to the great enthalpy of the plasma.
[0059] FIGs. 5A and 5B show X-ray photoelectron spectra (XPS) of Ti 2p and B
Is spectra of TiB2 coatings deposited
with and without the shroud. Both samples show the characteristic Ti 2p3/2 of
TiB2 and B Is of B203 at 458.7 and 192.4,
respectively. In the case of TiO2, the corresponding Ti 2p1/2 spin-orbit
coupled component is also observed at 464.9 eV.
For the coating deposited in absence of the shroud, only these peaks are
observed. Keeping in mind that the depth
probed at the surface of the coating is only in a range between about 3 and
about 5 nm, this indicates that the very top
surface of the TiB2 coating is oxidized. In contrast, the coating made with
the shroud exhibits another set of peaks
located at 454. 5 eV (Ti 2p3/2) and 187.5 eV (B 1s). The corresponding Ti
2p112 spin orbit coupled component is
expected at 460.7 eV, but it is hardly discernable since it falls on the high
binding energy side of the main Ti 2p3/2 peak
of TiO2. The Ti 2p3/2 peak at 454.5 eV and the B Is peak at 187.5 eV are
associated with TiB2. The presence of these
peaks is an indication that the thickness of the oxide layer at the surface of
the TiB2 coating made with the shroud is
reduced compared to coating made without the shroud.
[0060] Results of molten aluminum sessile drop test on SPS TiB2 coatings made
with and without shroud can be
seen in FIG. 6. A drastic difference in spreading kinetic of molten aluminum
over these two samples can be seen. On
sample deposited with the shroud, the aluminum drop contact angle decreases
rapidly from 150 to nearly 0 in about
30 min, confirming its excellent wettability for molten aluminum I. A good
aluminum wettability is also observed on TiB2
coating deposited without shroud, although at a much slower wetting kinetics.
After initiation period lasting about 30
min where the contact angle is large, the contact angle decreases from about
150 to about 2 in more than 250 min,
which is a factor of 8 longer than that observed for coatings sprayed with the
shroud. This difference in spreading
kinetics can be mainly explained by the difference in oxide concentration,
namely thickness, at the surface of SPS TiB2
particles, as supported by the XPS results. The initiation period observed for
coating made without shroud may
correspond to the time required for molten aluminum to dissolve/reduce the
TiO2 and B203 oxide layer beneath the
initial aluminum drop. Then the following slow spreading phase may originate
from the infiltration of the molten
aluminum through the porous TiB2 coating, inducing a progressive
reduction/dissolution of the oxides present onto the
inner SPS TiB2 particles and allowing the true aluminum contact angle on "in-
situ cleaned" TiB2 to be attained. Thus,
the aluminum spreading behavior is likely to be affected by both the TiB2
coating morphology, impacting on the
aluminum infiltration process, and its chemical surface state, impacting on
the oxide dissolution process. Since both
Date Recue/Date Received 2022-01-26
11
TiB2 coatings have a similar porous morphology, the surface oxide dissolution
seems to be the main process governing
the aluminum spreading kinetics. However, this difference in wetting kinetics
does not affect the good aluminum
wettability of the present SPS TiB2 coatings, regardless of the use or not of
the argon shroud during their preparation.
[0061] FIGs. 7 show cross-section SEM micrographs and EDS mappings of Ti, Al
and Zr elements for the TiB2
coating made with the shroud after 8 h of contact with molten aluminum at 1000
C. Similar SEM-EDS analyses were
performed on TiB2 coating made without shroud (not shown) and the same
phenomena were observed. The porous
TiB2 coating has been infiltrated by molten aluminum and the molten aluminum
was able to reach the coating/graphite
substrate interface and tends to accumulate in this area, as highlighted by
the aluminum mapping image (FIG. 7D).
The direct contact of molten aluminum with the graphite substrate may be
detrimental for the coating mechanical
integrity due to the possible formation of brittle A14C3 phase. However, no
delamination is observed in FIGs. 7,
suggesting a good adhesion of the SPS TiB2 coating on the graphite substrate
despite its significant infiltration with
molten aluminum. A few longitudinal and transversal macrocracks are
nonetheless observed in some regions of the
coatings (see FIG. 8). Prolonged contact with molten aluminum is required to
evaluate the long-term stability of the
SPS TiB2 coating. There is also an obvious concentration of Zr at the
interface between the coating and the substrate
(FIG. 7E), suggesting that YSZ initially present in the coating has been
reacted with molten aluminum and migrated to
the interface. This is confirmed by the fact that YSZ peaks are no longer
observed in the XRD pattern of the coating
impregnated with molten aluminum (data not shown).
[0062] Deposition of micrometric TiB2 powder on a graphite substrate by
suspension plasma spray (SPS) for use as a
wettable cathode for aluminum production is thus demonstrated. A shroud may be
used as discussed herein, instead of
deposition of the coatings in controlled atmosphere. The obtained coatings are
porous with loosely bounded TiB2
particles. This unexpected high level of porosity can be attributed to the
fact that even though suspension plasma spray
(SPS) theoretically gives a higher heat transfer to the particles than APS,
the shorter transit time spent by the particles
in the plasma results in a reduced amount of melting. Nonetheless, the SPS
TiB2 coatings show a good adhesion to the
graphite substrate. It was also shown that the use of a shroud limits the
amount of TiO2 and B203 formed at the surface
of the coating. This lower amount of oxides accelerates the initial spreading
kinetic of molten aluminum but no effect
whatsoever on the wettability reached at steady state after a few hours in
contact with the molten aluminum. The
molten aluminum was able to fully impregnate the TiB2 layer in all cases and
is then likely to react with the graphite
substrate to form aluminum carbide which may reduce the long-term stability of
the coating.
[0063] Suspension plasma spray (SPS) is thus successfully used for application
of TiB2 coatings, with high deposition
rates in the order of kg / hour, opening the way to industrial applications.
The use of a shroud with an axial injection
Date Recue/Date Received 2022-01-26
12
SPS torch, typically a relatively small plasma torch of a diameter of about 10
cm and length of about 15 cm for
example, mounted on a 6-axis robot which makes it possible to spray coatings
on parts of a size in the range between
a few mm and tens of meters, and of various shapes, was thus shown effective.
Moreover, suspension plasma
spraying at atmospheric pressure allows coating deposition on large components
at reduced costs compared to
vacuum plasma spraying (VPS) for example.
[0064] Other experiments were performed with focus on the densification of the
TiB2 coating by optimizing the
suspension plasma spray (SPS) parameters, namely by adjusting the nitrogen
(N2) concentration in the plasma gas,
suspension feed rate and suspension TiB2 load. It is shown that an increase of
the TiB2 coating density is possible by
increasing the N2/Ar ratio and by decreasing the suspension feed rate and
load.
[0065] The densest TiB2 coating was obtained for the B28 sample, using the
following SPS parameters:
Total gas flow
Power WD Feed Suspension load
Sample # (slpm)
Ar/N2/H2 (%) (kW) (cm) (ml/min) (wt% TiB2)
o
180
B28 109 6.3 30 10
30/60/10
Table I
[0066] The behavior of the dense TiB2 coatings after contact with molten
aluminum was studied. It is shown that
below the aluminum drop, TiB2 coating is not delaminated and molten aluminum
did not penetrate in the coating and
thus the SPS TiB2 coating effectively protects the graphite substrate from
direct contact with molten aluminum.
[0067] Other experiments were performed, with focus on TiB2 coating cracking.
[0068] As shown in experiments described hereinabove, dense TiB2 coatings may
be obtained by increasing the heat
transfer to in-flight TiB2 particles using optimized suspension plasma spray
(SPS) parameters.
[0069] A multilayer method is described hereinbelow to prevent stress building
up with densification, which may lead
to the formation of cracks in the coating. By depositing layers with varying
levels of porosity, the stresses associated
with the build-up of the dense layers, as well as the stresses thermally
induced during and after deposition, can be
Date Recue/Date Received 2022-01-26
13
dissipated through the more porous layers, allowing for a mechanically stable
coating, as described in relation to
experiments, based on experiments described hereinabove, and results described
in relation to FIGs. 9 and following.
[0070] Experiments were performed to assess the effects of spray parameters on
the density of TiB2 coatings
prepared by suspension plasma spray (SPS), as well as the impact of the
density of the coating on the behavior of the
coating in contact with molten aluminum at 1000 C, using TiB2 powder (Hunan
HuaWei Aerospace Special Materials,
China). The particle size distribution was characterized by laser diffraction
using a Malvern Panalytical Spraytec
instrument after dispersion of the powder in ethanol using a Malvern Wet
Dispersion Accessory. Coatings were
deposited on graphite substrates. Disc-shaped graphite substrates (254 mm
diameter and 5.5 mm thickness) were
prepared by cutting a graphite rod (GR008G grade from Graphitestore, USA).
Substrates were then grit-blasted prior to
their use. Only one side of the disc-shaped samples was coated, using plasma
spraying parameters are listed in Table
II below.
TiB2
Coating
Total gas distance Suspension Suspension Coating
Power powder
thickness
Sample # flow (SLPM) W feed rate load porosity
(kW) feed rate (1-1In)
Ar/N2/H2 (c/o) (cm) (mL min-1) (wt%) (c/o)
(g min-1)
51 180
100 6.3 45 20 7A0 25 110
(TiB2-P25) 45/45/10
S2 180
109 6,3 45 20 7A0 19 95
(TiB2-P19) 30/60/10
S3 180
109 6.3 30 20 4/0 13 80
(TiB2-P13) 30/60/10
S4 180
109 6.3 30 10 2.35 4 60
(TiB2-P4) 30/60/10
19
180
109 7,3 45 10 3.55 (bottom 65
S5 30/60/10
layer)
(TiB2-P5/19)
180 5
109 6,3 30 10 2.35 40
30/60/10 (top layer)
Table II: Suspension plasma spray process parameters
Date Recue/Date Received 2022-01-26
14
[0071] Samples are referred to as TiB2-P)0(, where PXX is the porosity of the
coatings thus prepared. The thickness
of the coatings ranges from 60 to 100 pm. For each spray conditions, 24
substrates were installed in a vertically
spinning sample holder rotating at 62 rpm. A Mettech's Axial III Plasma Spray
System equipped with an argon shroud
was used. The argon shroud (Northwest Mettech Corp., Canada) is a mechanical
piece attached to the exit of the SPS
torch where the argon gas is injected to minimize contact of the sprayed
particles with the atmosphere. Suspensions of
or 20 %wt of TiB2 powder in ethanol were prepared and t2 %wt of
polyvinylpyrrolidone was added as dispersive
agent. After the deposition process, the samples were left to cool down in
air.
[0072] The crystalline structure of TiB2 powder and as-sprayed TiB2 coatings
was determined by X-ray diffraction
(XRD) using a Bruker D8 diffractometer equipped with Cu Ka radiation. Cross-
section of the coatings were prepared by
using a precision cut-off machine (Secotom- 15, Struers A/S, Denmark). The
coatings were then embedded in an
epoxy resin and polished by standard metallographic procedures. A Tescan Vega3
scanning electron microscope
(SEM) equipped with a Bruker XFlash 6 110 Detector energy dispersive X-ray
spectroscopy (EDS) was used to
analyze the top-view and the cross-section of as-sprayed coatings and samples
after sessile drop test.
[0073] The porosity of TiB2 coatings was measured by ImageJ 1.52n software
using the Ti and Al SEM-EDS
mappings recorded on TiB2 coatings after sessile drop tests. The porosity
values of each coating are presented in
Table II above.
[0074] Aluminum contact angle measurements were carried out using sessile drop
technique. This test was chosen to
give relative molten aluminum wetting characteristics to compare to other
similar tests reported in the literature. Before
the sessile drop test, a piece of aluminum wire (about 100 mg) was cleaned in
a 1M NaOH solution for 15 min. It was
then rinsed with deionised water and finally cleaned using an ultrasound bath
for 10 min in acetone. The aluminum wire
was then twisted and compressed into a spring shape. Then, the sample and the
aluminum spring-shaped piece were
inserted in a tubular furnace equipped with a vacuum pump. The tube was vacuum-
pumped overnight prior to being
heated at 500 C for 2 hours. Then, the temperature was raised to 1000 C at a
heating rate of 4 C min-1 and was
maintained at 1000 C for 8 hours. The evolution of aluminum contact angle on
TiB2 coated samples was recorded
using a camera placed on one end of the tube. The recorded images were used to
measure the aluminum contact
angles using ImageJ 1.52n software.
[0075] The characterization results of the TiB2 starting powder are shown in
FIGs. 9. The XRD pattern of the TiB2
powder presents only one series of diffraction peaks that are indexed to TiB2,
as shown in FIG. 9A. The particle size
distribution of TiB2 powder is shown in FIG. 9B. The particle size is centered
along a single mode, with maximal volume
Date Recue/Date Received 2022-01-26
15
frequency at 1.5 pm, while the D10, Doo and Dgo values are 0.9, 1.4 and 2.1
pm, respectively. SEM micrographs and
EDS mapping results of the starting powder are also shown in FIGs. 9A-9E.
Spherical or near-spherical TiB2 particles
are observed. A few zirconia particles are also observed that arises most
probably from the erosion of milling media
used to prepare the TiB2 powder. Based on the EDS results, the atomic ratio
7z_Erzr is 2 %. The amount of zirconia
contamination is too low to be detected by XRD. Note that B element was too
light to be accurately quantified with the
EDS detector used.
[0076] The XRD patterns of samples S1-S5 are displayed in FIGs. 10. For all
coatings, only one set of peaks is
detected that are assigned to TiB2. SEM images of the surface of TiB2
coatings, shown in FIGs. 11A - 11E, reveal that
all coatings have the same "cauliflower-type" microstructure. Sample S4 shows
the presence of macrocracks (FIG.
11D) that are not observed in samples S1, S2, S3 and S5. FIGs. 11F ¨ 11J show
the cross-sectional SEM micrographs
of as-sprayed coatings. The coating thickness varies between 60 and 100 pm.
FIG. 111 shows that the cracks
previously observed at the surface of sample S4 extend through the thickness
of the layer to reach the interface
between the deposit and the substrate. There are obvious differences in the
porosity of the TiB2 coatings shown in the
SEM cross-section micrographs of FIGs. 11F ¨ 11J that can be rationalized by a
comparison of the spray deposition
parameters (Table II).
[0077] For sample S1, a plasma power of 100 kW, a total gas flow of 180 SLPM
and a gas composition with equal
amount of N2 and Ar (45% each) was used. A feed rate of 45 mL min-1 and a
suspension load (%wt. of TiB2 powder in
ethanol) of 20 wt.%) were used, which led to the injection into the plasma of
7.2 grams of TiB2 powder per minute. The
porosity of S1 is 25%. In these conditions, it is surmised that TiB2 particles
were not fully melted upon impact on the
substrate and were not able to form splats but were aggregated together, which
led to the formation of a porous
coating. To circumvent this limitation, sample S2 was prepared with a higher
ratio of N2 to Ar (2:1 instead of 1:1), while
keeping constant the total gas flow at the TiB2 powder feed rate. As is often
the case in plasma spray, parameters are
intertwined, and the higher amount of N2 in the plasma led to a 10% increase
of the power, from 100 to 109 kW. The
porosity of coating S2 is 19% %), significantly lower than S1. A higher plasma
power increases the heat available in the
plasma to melt the TiB2 particles. In addition, N2 has a higher thermal
conductivity and specific enthalpy than Ar, which
lead to an increase heat transfer to the TiB2 partides. These factors
contribute to increasing the energy available to
melt the TiB2 particles before they reached the substrate, therefore
decreasing the porosity of the coating.
[0078] For sample S3 and S4, all deposition conditions were kept the same as
sample S2 except for the TiB2 powder
feed rate that was progressively reduced from 7.10 g min-1 for S2 to 4.70 g
min-1 for S3, and then to 2.35 g min-1 for S4.
As a result, the amount of energy available that can be transferred to any
single TiB2 particle is increased because
Date Recue/Date Received 2022-01-26
16
gradually less TiB2 particles are injected in the plasma at the same time. As
a result, the porosity of TiB2 coatings
decreases from 19% for S2 to 13%) for S3, and then to 4% for S4. Again, it is
inferred that more energy is transferred
to each individual TiB2 particles as their concentration is decreased in the
plasma, which leads to a more complete
melting of these TiB2 particles and to denser coatings.
[0079] Sample S4 is the less porous coating (4%). However, as noted
previously, sample S4 exhibits several cracks
that extend from the top of the layer to the coating/substrate interface. In
dense coating, higher mechanical stress
during the build-up of the coating can lead to the formation of cracks upon
cooling. These cracks are not observed in
51, S2 and S3 coatings that are more porous. In these coatings, the presence
of numerous pores (porosity in the
range between 13 and 25%) helps to mitigate the// tensile stress building up
upon cooling, preventing the formation of
cracks. However, although porous TiB2 coatings (about 30% porous) are crack-
free, molten aluminum is able to
infiltrate the porous structure of TiB2 coating and reaches the
coating/substrate interface, causing the delamination in
some regions of the coating.
[0080] Accordingly, a method was developed to intercalate a porous TiB2 layer
between a denser TiB2 layer and the
substrate, in order to limit the infiltration of aluminum while keeping a
porous structure that allows accommodating the
tensile stress building up upon cooling. For example, sample S5 was prepared,
where a porous bottom layer was first
deposited, followed by a denser top layer. The porosity of the bottom layer
(which is about 65 lam thick) is 19%), while
the porosity of the top layer (which is about 40 lam thick) is 5% %). As seen
in FIGs. 11E-11J there is no crack in
sample S5. One can also observe that there is no clear line of demarcation
between these two layers. This is attributed
to the fact that the first TiB2 particles of the second layer can infiltrate
the porous structure of the first layer. This density
gradient at the junction of the two layers allows maintaining a good cohesion
in such a multilayer structure.
[0081] The sessile drop tests were carried out to investigate the behavior of
TiB2 coatings in contact with molten
aluminum at 1000 C, which is a temperature typical for aluminum electrolysis.
One of the main concerns of SPS TiB2
coatings in contact with molten aluminum is the infiltration of aluminum into
the TiB2 coating promoted by the good
wetting properties of TiB2. Such infiltration can result in longitudinal and
transversal cracks in the coating. Sample S4
(TiB2-P4) was not examined because it contains long and continuous crack
through the whole coating layer. Since
molten aluminum can easily penetrate through these cracks, there was no point
for performing sessile drop test on this
sample.
[0082] FIG. 12 shows the evolution with time of apparent contact angle of
molten aluminum on graphite as well as on
Date Recue/Date Received 2022-01-26
17
TiB2-P13, TiB2-P19 and TiB2-P5/19 coatings recorded at 1000 C in vacuum (< 10
mPa). All TiB2 coatings exhibit not
only a lower apparent aluminum contact angle, but also a higher spreading rate
than graphite. The TiB2-P19 and TiB2-
P13 coatings show an excellent apparent aluminum wettability as aluminum
droplet spreads very rapidly over the
sample to reach a contact angle of 00 in 195 and 300 minutes, respectively.
This behavior is favored by the relatively
high porosity of these coatings which allows molten aluminum to penetrate the
porous coatings. The TiB2-P5/19
coating, despite its slower wetting kinetics, still has a much better aluminum
wettability than graphite as the aluminum
contact angle reaches 0 in 460 minutes while for graphite, even after 480
minutes, the aluminum contact angle does
not decrease to less than 75 .
[0083] FIGs. 13 show the EDS mapping results of TiB2-P19, TiB2-P13 and TiB2-
P5/19 coatings after sessile
aluminum drop tests. In all cases, three layers are observed corresponding to
aluminum, TiB2 coating and graphite
substrate. In TiB2-P19 coating, only on some regions, a very thin layer of
aluminum remained on the top of TiB2 coating
and aluminum has completely penetrated into the pores of TiB2 layer. Also,
large Al-rich regions are formed at the
interface of the coating and the substrate and TiB2 coating has been
delaminated in some parts. In TiB2-P13 coating, a
thick layer of aluminum can still be seen above TiB2 coating but aluminum has
penetrated into the pores of TiB2 layer
too. Small Al-rich regions are also formed at the interface of the coating and
the substrate. However, TiB2-P13 coating
seems to be still well-attached to the graphite substrate and no obvious crack
or delamination is observed. With the
double-layer TiB2-P5/19 coating, molten aluminum has not infiltrated into the
coating and remained only on the surface
of TiB2. Hence, no Al-rich region is formed at the coating/substrate
interface, the coating layer is not delaminated after
sessile drop test and there is no evidence of cracks.
[0084] FIG. 12 shows differences in the spreading kinetics of the aluminum
droplet with the nature of the samples.
These differences are attributed to the difference in the porosity of TiB2
coatings. As mentioned above, in TiB2-P19 and
TiB2-P13 coatings, molten aluminum could infiltrate into the pores of TiB2
layer which accelerates the spreading
kinetics of aluminum droplet. Since the porosity is higher in TiB2-P19
coating, aluminum spreads with even faster rate
than in TiB2-P13 coating. On the other hand, with the double-layer coating,
since its top layer is dense, the infiltration of
the molten aluminum does not happen and aluminum wettability occurs with a
slower rate. However, the good
wettability of TiB2 coatings is not affected by this difference.
[0085] TiB2 coatings are thus deposited by suspension plasma spray on graphite
substrates using TiB2 powder as
starting material. Coatings with different levels of porosity are produced by
selecting spray parameters which impact on
the melting rate of the TiB2 particles. Crack-free TiB2 coatings can be
obtained when the coating layer is 13-25%
porous %. Deposition of a denser TiB2 layer (5% porous) with more than 60 pm
thickness results in macrocracks that
Date Recue/Date Received 2022-01-26
18
propagate through the whole coating layer. A mixed-layer approach may be
adopted based on the fact that molten
aluminum can penetrate into the pores in porous TiB2 coating and via the
cracks in the dense TiB2 coating. In a double-
layer approach for example, first a porous layer was deposited on graphite
substrate which was then coated with a
dense layer; thus, cracks were not formed in the sprayed TiB2 coating and
molten aluminum did not infiltrate into the
coating., which displayed an excellent aluminum wettability without any sign
of degradation after 8 h of contact with
molten aluminum.
[0086] The method thus allows production of suspension plasma spray (SPS)
coatings of porous and dense TiB2
layers, in which the porous layers favor mechanical or thermal stress release
in order to limit the coating cracking,
whereas the dense layers prevent molten aluminum penetration in the coating
and its direct contact with the substrate.
The substrate may be a graphite or carbon-based substrate as exemplified
herein or in other electrically conductive
materials, or metallic, ceramic or cermet substrates, depending on the target
application of the coating, for example as
anode or protective coatings.
[0087] As described hereinabove, suspension plasma spray (SPS) is used to
deposit TiB2 coatings, at high
deposition rates, allowing coating large components compared to vacuum based
processes. To counter oxidation
arising when depositing TiB2 by SPS, an argon gas shield was used to protect
the in-flight TiB2 particles from oxidation
with air during spraying. Thus, the titanium diboride particles are deposited
in air without oxidizing them; without
requiring monitoring the atmosphere, which reduces significantly the cost of
operation in comparison to vacuum plasma
spray for instance. Suspension plasma spray (SPS) using finer particles is
controlled to provide a high heat transfer per
particle. Coatings deposited on graphite substrate were analyzed by scanning
electron microscopy, X-ray diffraction
and X-ray photoelectron spectroscopy to characterize their morphology,
porosity, crystalline structure and surface
oxidation state. Aluminum wetting tests were also carried out to assess the
operational quality of the coatings.
[0088] TiB2 coatings (60-110 p.m thick) were deposited on graphite substrate
using suspension plasma spray (SPS).
The porosity of the coating are controlled by selecting the suspension plasma
spray (SPS) deposition parameters,
resulting in coating porosities ranging from 4 to 25%. A double-layered TiB2
coating consisting in a denser top layer
(4% porous, 40 p.m thick) and porous bottom layer (19% porous), 65 p.m thick)
was also prepared. No cracks were
observed in the as-sprayed porous and double-layered coatings, in contrast to
the denser (5 % porosity) coating.
Sessile aluminum drop tests were performed on TiB2 coatings to investigate
their behavior in contact with molten
aluminum. It was shown that the coating porosity impacts on the spreading
kinetics of aluminum drop, and that all TiB2
coatings have a much better molten aluminum wettability than the graphite
substrate. Moreover, after 8h of contact with
molten aluminum, no aluminum infiltration and no mechanical degradation were
observed in the double-layered TiB2
Date Recue/Date Received 2022-01-26
19
coating, in contrast to the single layer coatings.
[0089] The method and the system for fabrication of suspension plasma sprayed
TiB2 coatings may find a range of
applications, overcoming a number of issues in developing wettable TiB2-based
cathodes for aluminum production for
instance, such as the difficulty to sinter or melt this refractory material,
especially in large commercial units.
[0090] As people in the field will now be in a position to appreciate, the
present method may also be used for
depositing TiB2 protective coatings on a range of tools and components used in
the aluminum industry. A range of tools
and components essential for the production of primary aluminum are prone to
deteriorate prematurely upon exposure
to liquid aluminum and / or cryolite (Na3AIF6) due to the highly corrosive
nature of these media at high temperature,
about 1000 C. For example, steel stubs used to suspend and power the carbon
anodes are subject to the corrosive
effect of fluorinated effluents emitted by the electrolytic cell. In addition,
during an anode descent or during "waves" in
the electrolyte, molten cryolite bath can come into contact with the base of
the steel stubs, inducing their corrosion.
This corrosion has the effect of contaminating the aluminum produced with iron
and of reducing the life of the entire
anode suspension system including stubs and connecting bars between the stubs,
which are typically recovered and
reused after extraction of worn anodes.
[0091] Thermocouples measuring the temperature of the cryolite bath also
corrode rapidly and therefore cannot be
used continuously and must be changed after only a few hours of use. Other
components and tools, such as for
example sampling spoons, skimming blades, pouring basin filters, crucible
siphons, etc. are in prolonged contact with
liquid aluminum. A number of these pieces of equipment is also used for the
production, by smelting, of secondary
aluminum from recycling. The premature degradation of these various tools and
components, generally made of steel,
results in significant costs in consumables, in addition to causing
operational problems.
[0092] In order to respond to this problem, a protective coating is applied to
the surface of this equipment. High
resistance to corrosion in a cryolite medium or in the presence of liquid
aluminum, and high resistance to thermal and
mechanical shocks, are required.
[0093] There is thus provided a method and a system for spraying TiB2 coatings
in open air, using a gas shroud to
protect the particles from oxidation, or in a vacuum or controlled-atmosphere
chamber. To overcome the fact that TiB2,
characterized by a very high melting point, over 3200 C, and low density, may
be difficult to melt in the plasma jet, fine
TiB2 particles are selected, a suspension medium of these TiB2 particles are
injected in the plasma jet; the spray
Date Recue/Date Received 2022-01-26
20
parameters are selected so that the TiB2 particles have a residence time long
enough to fully melt in the plasma: the
plasma current, selected in the range between about 160 A and about 250 A,
determines a plasma power ranging
between about 70 kW and 150 about kW, for example between about 90 kW and
about 120 kW and the total gas flow
of the plasma is in the range between about 160 and about 275 slpm, the feed
rate of the suspension is selected
between about 10 and about 80 mL/min, and the suspension load, referring to
the solid content of titanium boride in the
suspension, is selected between about 1 about 40 weight%, for example between
5 and 20wt%; the substrate
temperature is controlled in the range between about 100 and 1000 C, for
example between 400 C and 800 C. The
plasma gases comprising between about 20 and about 80%, for example between
about 10 and about 40%, argon;
between about 5 and about 80%, for example between about 40 and about 70%,
nitrogen; and between about between
about 0 and about 25%, for example between about 10 and about 15% hydrogen,
depending on the percentage of
nitrogen in the plasma: since nitrogen has a higher heat conductivity than
argon, a higher percentage of nitrogen in the
plasma increases the heat conductivity of the plasma and is therefore selected
to optimize melting of the particles. The
spray deposition parameters may be controlled to provide selectively i)
conditions of a plasma with high heat
conductivity and high power, for an optimized heating and melting of the TiB2
particles so that a main part of the TiB2
particles impact on the substrate in a fully melted state, thereby building a
layer of a first density; and ii) conditions with
an overall less intense plasma heat conductivity, so that a higher percentage
of the TiB2 particles impact the substrate
while in a partly molten state or in a solid state, which translates to a
higher porosity thereby building a layer with a
second density lower than the first density, allowing the selective and
controlled deposition of series of dense and
porous layers. The dense layers reduce the penetration of the cryolite inside
the coating, while the porous layers, i.e. of
a lower density, mitigate the build-up of stresses through their higher
porosity.
[0094] A shroud may be used positioned at a shroud-to-substrate distance S,
selected in the range between about 5
mm and 100 mm, for example between about 5 mm and 30 mm. The distance S is
defined as the distance W minus
the length of the shroud itself, which in the example presented herein was 53
mm. A system as illustrated in FIGs. 14
and 15 for example comprises a plasma torch for injection of the powder
suspension of TiB2particules, and a substrate
at the distance W from the output of the plasma torch distance.
[0095]
As the selected TiB2 particles are deposited with an axial injection plasma
torch, the suspensions is
injected axially directly into the plasma; the suspension can be injected at
the onset of the plasma, thereby enhancing
the heat transfer to the particles. In contrast, in a conventional radial
injection system, the suspension has to be
injected further down the plasma stream; the suspension, being injected
radially, has to penetrate through the outer
gas layer of the plasma while not completely crossing the width of the plasma,
in order to be properly injected. As a
result, in the radial injection system, the residence time of the particles is
mathematically smaller than in the axial
Date Recue/Date Received 2022-01-26
21
injection system. The higher residence time achieved in the case of axial
injection, especially a higher residence time
at the plasma core, yields optimized heating conditions.
[0096] By controlling the deposition parameters, the porosity of the coating
may be controlled across various dense
and porous layers as described hereinabove, by controlling the aluminum
penetration up to the substrate using dense
layers, and reducing mechanical stress and cracking using porous layers.
Moreover, since the build-up of the coatings
in plasma spray is heavily influenced by the shadow effect, and thus a high
roughness and an inhomogeneous
roughness of the substrate can cause strong discrepancies in the porosity of
the coatings, the structure of the deposit
is also controlled by controlling the roughness of the substrate before
deposition in such a way to favorize deposition
and avoid the particles to follow the gas lines instead of depositing, by
selecting an arithmetic average roughness Ra in
the range between about 0.5 pm and about 12 pm, for example between about 0.5
and about 4 pm.
[0097] By such engineering of the microstructure of the deposited coating, the
present method allows to achieve
mechanical requirements for a wettable cathode for aluminium electrolysis, in
terms of mechanical strength and
resistance to cracking due to thermal or chemical forces, as well as to
potentially increase the lifetime of the wettable
cathode by increasing the amount of layers in the coating, thus lengthening
the time before there is a failure through
the entire coating.
[0098] In addition, the fine control of the microstructure permits to reach a
hierarchical roughness of the coating
surface, which further improves the wettability of titanium diboride by
aluminium. Indeed, the general microstructure of
the titanium diboride deposited, tends towards the so-called "cauliflower
microstructure, especially when not fully
molten. The cauliflower microstructure exhibits a hierarchical roughness, with
a micro roughness on each "cauliflower
structure , and a more macro roughness resulting from the plurality of
"cauliflowers" structures, which results in an
increase in wettability.
[0099] The present disclosure presents a method and a system for TiB2
coatings. Application to wettable cathodes for
aluminum production, and to protective coatings of various components likely
to be in contact with cryolitic or molten
aluminum medium, is described. The service life of various equipment used for
primary and secondary aluminum
production is thus increased.
[00100] The scope of the claims should not be limited by the embodiments set
forth in the examples, but should be
given the broadest interpretation consistent with the description as a whole.
Date Recue/Date Received 2022-01-26
