Surface & Coatings Technology 318 (2017) 82–89

Contents lists available at ScienceDirect

Surface & Coatings Technology

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

Possibility of spraying of copper coatings on polyamide 6 with low

pressure cold spray method
Aleksandra Małachowska a,b,⁎, Marcin Winnicki b, Łukasz Konat b, Tomasz Piwowarczyk b, Lech Pawłowski a,
Andrzej Ambroziak b, Mateusz Stachowicz b
a
Science des Procédés Céramiques et Traitements de Surface UMR 7315, University of Limoges, CNRS 12, rue Atlantis, 87068 Limoges, France
b
Wroclaw University of Technology, ul. Lukasiewicza 5, 50-371 Wroclaw, Poland

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

Article history: This paper discusses metallization of polymers using a low-pressure cold spray (Dymet 413). Three commercial
Received 31 July 2016 copper powders: i) spherical and ii) dendritic were deposited on thermoplastic polymer - polyamide 6 (PA6). It
Received in revised form 30 January 2017 was difficult to successfully apply a copper coating directly on the polymer substrate, therefore interlayers were
Accepted 1 February 2017
applied. Additionally, the copper powder was pre-treated in hydrogen atmosphere to remove the oxide layer and
Available online 3 February 2017
reduce its critical velocity. Finally, the adhesion strength, electrical conductivity, oxygen content and microstruc-
Keywords:
ture of resulting coatings were determined. Coatings were characterized by one order of magnitude of lower con-
Cold spray ductivity than the bulk material and bond strength of 3.6 MPa. The powder shape turned out to have a decisive
Polymer metallization effect on the possibility on coatings formation.
Electrical conductivity © 2017 Elsevier B.V. All rights reserved.

1. Introduction
Deposition of metal coatings onto polymers has been widely studied
in order to improve thermal and electrical properties as well as to re-
duce surface degradation. Some well-established processes were used
include physical vapor deposition (PVD) [1–4] or electroless deposition
[5], but the thickness of the obtained layer was limited and deposition
rates were low. Therefore, thermal spraying processes [6,7] are some-
times applied, which allow for the deposition of a thick coating on var-
ious substrate geometries, and the recoating of damaged elements. The
main challenge is temperature-sensitivity of polymers. Recently, cold
spray process was tested a potential solution. In this process, material
is deposited in a solid state and therefore the temperature impact is
lower in comparison to traditional thermal spraying methods. The pow-
der particles are accelerated in stream of heated and pressurized gas
and projected towards the substrate. Critical velocity is a key concept
in the cold spray method [7]. It is defined as velocity that an individual
particle of powder must attain in order to be deposited after impact
with the substrate [8]. This definition is valid for ductile materials as
brittle materials will cause erosion for any velocity at temperatures
below their melting temperature [9]. Critical velocity depends mostly
on the sprayed material's mechanical properties but varies also with
particle size, particle morphology, particle impact temperature or pow-
der oxidation [10–12]. The critical value may be calculated with the
⁎ Corresponding author at: Wroclaw University of Technology, ul. Lukasiewicza 5, 50-
371 Wroclaw, Poland.
E-mail address: aleksandra.malachowska@pwr.wroc.pl (A. Małachowska).
equation proposed by Assadi et al. [11] and then developed by Schmidt
et al. [10]. It takes the following form:
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ffi
u  
u T −T R
u F 1 ∙4∙σ TS ∙ 1− i
t T m −T R
vcr ¼ þ F 2 ∙C pp ∙ðT m −T i Þ ð1Þ
ρ
where: ρ – density, σTS – tensile strength, Tm – melting point, Ti – impact
temperature, TR – reference temperature (293 K), cp – specific heat of
particle, F1 – mechanical calibration (for cold spray 1.2), F2 – thermal
calibration (for cold spray 0.3).
The model takes into account the specific heat, tensile strength, me-
chanical and thermal calibration, but particles size is not included [10]:
with decrease in particle size critical velocity increase. The possible rea-
son for higher critical velocity of small particles may be the higher con-
tent of oxides or adsorbents hindering the bonding. Usually, powder
contains a mixture of particles of varying diameters. In such cases, crit-
ical velocity is calculated for larger particles due to fact that smaller par-
ticles achieve a higher velocity [10].
Lupoi and O'Neill [13] pointed out that polymers might be coated
when the particle impact energy calculated from the critical velocity
of a given material and particle mass is sufficiently low. According to
this formula, tin and lead will be easy to deposit, for aluminium and ti-
tanium deposition and erosion process will be coexistent, and for cop-
per, the erosion process will be prevalent [13]. This assumption was
confirmed for tin, which was deposited on various substrates PC/ABS,
polypropylene, polystyrene and polyamide-6 [13]. The aluminium was
deposited on PEEK [14] and Lexan [15], however in case of Lexan, the

http://dx.doi.org/10.1016/j.surfcoat.2017.02.001
0257-8972/© 2017 Elsevier B.V. All rights reserved.
 A. Małachowska et al. / Surface & Coatings Technology 318 (2017) 82–89
Table 1
Spray conditions for coatings deposition.
Powders Gas temperature Gas pressure Standoff distance
[°C] [MPa] [mm]
Cu-S, 200; 300 0.9 10
Cu-D
Cu-S – spherical copper powder; Cu-D – dendritic copper powder.
deposition process was possible only for high powder feed rates [15].
Additional spraying passes turned out to densify the aluminium coat-
ings [16]. Thin copper coating was achieved on PA66 [17]. After deposi-
tion of the first layer, erosion occurred and further deposition was
scanty. This was probably caused by a change of substrate surface to a
metallic one and hence a change of parameters necessary for depositing
a coating. Ganesan et al. [18] proved that the erosion process might be
limited by taking advantage of different shapes of copper powders and
using tuned interlayers. Thick copper coatings were achieved on PVC
and epoxy [18]. They observed also that metal deposition on thermoset
is more difficult because of their greater brittleness. The impact of pow-
der particles causes erosion of the substrate instead of its plastic
deformation.
The deposition efficiency on polyamide 6 depends probably also on
the shape and oxidation level of the powder. It was already proved
that some stresses remain in powder after the manufacturing process
[19] which, together with a thin oxide layer, increases the critical veloc-
ity. These stresses might be reduced through heat treatment which re-
sults in lower critical velocity and hence better deposition properties.
Therefore, in this paper the influence of heat treatment on copper depo-
sition on polyamide 6 is studied.
2. Experimental procedure
2.1. Spraying conditions
The coatings were deposited using a low-pressure cold spray device
Dymet 413 with a standard de Laval nozzle. Air was applied as working
gas. The spraying parameters for coating are listed in Table 1 and for in-
terlayers in Table 2. Two interlayers were selected based on literature: i)
an interlayer from spherical copper particle and ii) tin with alumina ad-
mixture. Spherical copper particles have high inertia and therefore are
stable in contact with subsequent sprayed particles [18]. Tin is an easy
deformable material, with low critical velocity which might be deposit-
ed easily on various substrates [10,13]. These two interlayers should fa-
cilitate copper deposition. The final coatings were sprayed with
dendritic copper powder on an Sn + Al2O3 interlayer in three spraying
passes with gas temperature equal to 200 °C and gas pressure equal to
0.9 MPa.
2.2. Spraying materials
Three commercially available copper powders were used for
spraying (Fig. 1): (i) spherical Cu (Libra, Poland), size of D10-5.94 μm,
D50-13.98 μm, D90-25.41 μm (Fig. 1a), and (ii) dendritic Cu (Libra, Po-
land), size of D10-10.21 μm, D50-30.67 μm, D90-80.77 μm (Fig. 1c).
Coatings were sprayed using Cu powders in delivery state and after
heat treatment. Additionally, powder mixture of Sn + 50 wt% Al2O3
(Obninsk Centre for Powder Spraying, Russia), with particle size of
Table 2
Spray conditions for interlayers' deposition.
Powders Gas temperature Gas pressure Standoff distance
[K] [MPa] [mm]
Sn + Al2O3 200 0.9 10
Cu-S 200 0.9 10
83
Gun traverse speed, [mm/min] No. of spray gun passes Powder feeding rate, [g/min]
2000 1; 3 30
D10-11.3 μm, D50-22.3 μm, D90-40.6 μm (Fig. 1e), was used for
interlayers.
Semicrystalline polyamide 6 (PA6) (Plastics Group, Poland) with
glass transition temperature ~ 50 °C and melting temperature ~220 °C
was used as a substrate material. The substrates were in shape of a
plate with dimensions 3 × 50 × 100 mm for metallographic examina-
tion, in the shape of a disc with Ø 40 mm in diameter and a thickness
of 10 mm for bond strength test, and in the shape of a rectangle with
5 mm width and 80 mm length for electrical resistivity measurements.
Before spraying, the substrate was cleaned with acetone.
2.3. Temperature-programmed reduction measurement
The TPR measurements were done with a 15 mg sample placed into
a quartz microreactor. It was made with a mixture of 5 vol% H2 in Ar, and
the temperature was increased linearly to 950 °C at a rate of 10 °C/min.
The hydrogen consumption was monitored by TCD detector. The
highest hydrogen consumption in case of a dendritic powder was
±300 °C and in the case of spherical powder ±450 °C (Fig. 2). The ob-
tained values indicate that dendritic powders have higher oxygen con-
tent, which is caused by a developed surface. At the same time this
oxide starts to reduce at a lower temperature.
2.4. Reduction of powder oxides
Data obtained from TPR measurements was used to reduce the ox-
ides present in powders. The reduction was made with hydrogen (com-
mercial purity 99.995%) using a self-made set-up consisting of a steel
tube and furnace FCF7SM (Czylok, Poland), working in the temperature
range 20–1150 °C with precision ±3 °C. The powder was placed in the
tube through which hydrogen was flowing. The temperature was mea-
sured using a thermocouple placed in the tube. The dendritic powder
was reduced for 1 h in temperature ±300 °C and the spherical one for
1 h at a temperature ± 450 °C. After heat treatment the powder was
milled and sieved to get the particle size below 40 μm. The powders
size after reduction of powder oxides amounted to: i) spherical Cu
(Libra, Poland), D10-8.24 μm, D50-18.97 μm, D90-34.86 μm and (ii)
dendritic Cu (Libra, Poland), D10-11.18 μm, D50-34.40 μm, D90-
98.51 μm. The coatings were sprayed one day after preparing the
powder.
2.5. Oxygen content measurements
The oxygen content in the used powder and resulted coatings was
analysed using an O2 and N2 analyser - Leco TC436 (Leco, USA) with
the inert gas fusion method.
Gun traverse speed, [mm/min] No. of spray gun passes Powder feeding rate, [g/min]
2000 1 30
2000 1 30
84 A. Małachowska et al. / Surface & Coatings Technology 318 (2017) 82–89
Fig. 1. Morphology of powders used for deposition process a) spherical copper particles, b) spherical copper particles after reduction of powder oxides c) dendritic copper particles d)
dendritic copper particles after reduction of powder oxides e) mixture of Sn + Al2O3.
2.6. Electrical resistivity measurement
The resistance of the samples was measured using a custom-built
four-terminal sensing probe with spring-loaded pogo-pins. The
Fig. 2. Thermal programmed reduction of copper powder with hydrogen (TDC - thermal
conductivity detector).
distance between the voltage-sensing pins was 20 mm. The measure-
ment system consisted of an Agilent E3632A power supply and Keithley
2000 and 2001 multimeters. The multimeters did not provide a test cur-
rent big enough (less than 10 mA) to measure the resistance of the sam-
ples with satisfactory precision, therefore the power supply and one
multimeter were used to provide and measure the 100 mA test current
while the second multimeter was used to measure the voltage. Low re-
sistance of the samples (less than 100 mΩ) resulted in the power dissi-
pated in the sample being less than 1 mW. The electrical resistivity was
calculated according to the formula:
Sr U
r¼ ∙ ð2Þ
l I
where: ρ - resistivity (Ω m), U – measured voltage (V), I – measured
current (A), l - distance between measuring probe (m) Sr - the cross-
section of the conductive layer = a ∗ g, where a - width of the conduc-
tive layer (m), g – thickness of the conductive layer (m).
The thickness of the coatings for electrical resistivity amounted to
221 μm ± 22 (the thickness is given including Sn + Al2O3 interlayer).
The measured coatings were deposited in one spraying pass.
2.7. Bong strength
Bong strength of the coating to the substrate was determined from
five samples in a tensile pull-off test in accordance with PN-EN 582
 A. Małachowska et al. / Surface & Coatings Technology 318 (2017) 82–89
Fig. 3. Coating sprayed with spherical powder in delivery state with temperature 200 °C in a) 1 and b) 3 passes.
standard “Thermal spraying - Determination of Tensile Adhesive
Strength”. The bond strength value was determined as the ratio of the
maximum load to the cross-sectional area of the sample. The coating
was bonded with the counter-sample using the epoxy resin adhesive
Distal.
2.8. Microhardness measurements
Microhardness was measured with the use of a digital micro hard-
ness tester MMT-X7 MATSUZAWA CO., LTD (Akita, Japan) with a
Vickers indenter according to the standard PN-EN ISO 6507-3 of 2007.
The load of 0.9807 N were chosen for each type of coating. The
Fig. 4. Coating sprayed with spherical powder in delivery state with temperature 300 °C in
three passes.
Fig. 5. Coating sprayed with dendritic powder in delivery state with temperature 200 °C in a) one and b) three passes.
85
measurement was done on cross-sections in the middle of the coatings.
Microhardness was given as an average of five measurements.
2.9. Metallographic examination
The coatings' microstructure was examined on prepared cross-sec-
tions with electron scanning microscopes: Hitachi TM 3000 (Hitachi,
Japan) and Philips SEM XL 30 (Philips, The Netherlands).
3. Results and discussion
3.1. Spraying with powder in delivery state
The preliminary test indicates that copper powder does not deform
sufficiently in contact with a polymer surface to form coating when
using low pressure cold spray equipment. In the case of spherical pow-
der, the particles just stuck to the substrate independent on the number
of passes (Fig. 3). Probably, the hardness and stiffness of the substrate
was too low to cause deformation or the particles velocity particles
was well below critical velocity. Moreover, in the literature it was
found that bigger particles have lower critical velocity [10], but in the
case of soft substrate it seems that only very small particles can deform.
Additionally, although there is a high difference in tensile strength and
Young's modulus between the polyamide 6 and the copper, the copper
particles are not driven deep into the substrate and rather remain on the
surface.
Similar phenomena take place for higher gas inlet temperatures. The
continuous layer has not been formed (Fig. 4). However, the particles
are driven deeper into the substrate are thicker in comparison to the
lower inlet temperature.
Continuous coating was also not formed in case of spraying with
dendritic powder, but the powder particles are deformed more inten-
sively in compare to spherical particles (Fig. 5).
86 A. Małachowska et al. / Surface & Coatings Technology 318 (2017) 82–89
3.2. Spraying with powder in delivery state on interlayers
The solution proposed by Ganesan et al. [18] to use spherical parti-
cles as interlayers was tested. The dendritic particles were sprayed on
the previous deposited spherical ones but without success. It can be
seen that dendritic particles fragmented and partially deformed, but
no bonding may be found (Fig. 6a). The interlayer was probably too
loose and the velocity of the dendritic particles too low. As a conse-
quence, the spherical particles were displaced by subsequent dendritic
particles. There are only a few places in which the spherical particles
remained.
Tin was tested as the interlayer because it poses the lowest critical
velocity among usual cold spray feedstock material. As a variation,
Sn + A2O3 was also applied. The addition of A2O3 is commonly in the
low pressure cold spray method: it hammers additionally already de-
posited metal particles, sandblasts the substrate, and prevents nozzle
clogging [20]. Dendritic and spherical copper powders were used as
feedstock material. A continuous layer has been not achieved; upcom-
ing copper particles removed the interlayer from the substrate, leaving
only small amount of tin (Fig. 6b,c).
3.3. Spraying with powder after heat treatment
When using powders after deoxidation higher deformation was
clearly observed, but there is still no bonding between adjoining parti-
cles (Fig. 7). The highest deformation grade was observed for dendritic
powder especially for higher number of spraying passes, which enables
further for deposition adherent coating using Sn + Al2O3 interlayer
(Fig. 8). Despite using of Sn + Al2O3 interlayer it was not possible to ob-
tain continuous coating with spherical powder (Fig. 7f).
Dendritic copper was successfully deposited on PA6 with the
Sn + A2O3 interlayer. The influence of the substrate nature was
highlighted by Zhang [21]. He observed scant deposition of aluminium
particles on the tin substrate. This was attributed to a low melting tem-
perature point; during impact, tin melted, losing its strength, and there-
fore aluminium particles rebounded easily. Only small amounts of small
Fig. 6. Coating sprayed with dendritic powder in delivery state on: a) spherical copper particle, b) Sn + Al2O3 interlayer (gas temperature 200 °C) and c) spherical powder in delivery state
on Sn + Al2O3 interlayer (gas temperature 200 °C).
aluminium particles were deposited on the substrate, but whether they
formed a bond or were just embedded in soft tin remains unclear.
Deposited copper coatings were well-bonded with the substrate.
Their morphology depends mostly on the number of spraying passes;
coatings deposited in several passes were densified by a hammering ef-
fect. It is especially noticeable in coating deposited in three passes, in
which the main porosity area is situated in the top region, which is typ-
ical for the cold sprayed method [22,23] (Fig. 8). In coating deposited in
one pass, porosity near the substrate might be observed which unusual
for a cold spray process [23]. Also, morphology of the powder probably
contributed to the porosity as impact energy was not accumulated in
one contact point.
3.4. Oxidation measurements
Among powders, oxygen content was lowest for the spherical one -
0.042 wt% and increased with irregularity of the powder morphology to
0.065 wt% (Fig. 9) for the dendritic powder. Powders after heat treat-
ment in hydrogen and storing for one day exhibit about 1/3 lower oxi-
dation than the initial one. The oxygen value for dendritic copper
amounted to 0.043 wt%. The deposition efficiency measured for dendrit-
ic copper after deoxidation amounted to 30.95 ± 4%. The deposition ef-
ficiency for rest of the powders oscillated around the level of 0%.
An initial oxide layer, even at room temperature, forms rapidly on
copper. It was reported that after half an hour an initial oxide layer of
1 nm is formed, then oxide growth speed slows down and after
11 days the layer has a thickness of about 4.5 nm [24]. The changing
of oxidation rates is prescribed to the coalescence of initial oxide islands,
which block surface diffusion routes; further oxide growth requires
much slower bulk diffusion [25]. Initial oxide layer thickness equal to
1 nm might be recalculated for 20 μm spherical copper particle to
0.010 wt% (taking copper oxide density ~6.14 g/cm3). This oxygen con-
tent appears after half an hour contact with air so it can be hardly
avoided. Taking into account necessary milling and sieving of powder
after heat treatment, obtained values of oxygen content seem to be cor-
rect. The oxygen content derived from initial oxidation decreases with
increasing the size of the particle. Due to developed surface area, copper
 A. Małachowska et al. / Surface & Coatings Technology 318 (2017) 82–89
Fig. 7. Coating sprayed with: spherical copper particles after deoxidation in a) one and b) three passes and f) on Sn + Al2O3 interlayer; dendritic copper particles after deoxidation in c) one
and d) three passes and e) on spherical particles interlayer.
amounts in dendritic or globular powder are higher, which is also noted
in the measurement. Powder properties will change slowly after initial
oxide formation, following logarithmic law [26]. Such behaviour was
observed for dendritic copper particles by Ko et al. [27]. Powder
annealed in a vacuum reached 60% spraying efficiency, which decreased
to 40% for two weeks' air exposure and after which it remained a six-
week period. Oxygen content after annealing was 0.066 wt% [27].
Particle oxidation has notably influence on its deposition behaviour.
Increase of critical velocity from 310 m/s to 550 m/s for oxygen content
0.02 to 0.14 wt% was reported by Li et al. [28]. This might provide an ex-
planation for coating formation when spraying with powder after first
deoxidizing the powder with hydrogen, but the difference of 0.02 wt%
in oxygen content for heat-treated and untreated powder is small.
Kang et al. [29] claimed that the relation between oxygen content and
the increase in critical velocity is linear for aluminium powder. The low-
est critical velocity values were estimated to be 721 m/s for 0.001 wt%,
and increased to 867 m/s for 0.045 wt% oxygen content and are much
higher than given by Schmidt [9].
Taking into account the linear relation for data given by Li [28], the
increase of 0.02 wt% in oxygen content, which is visible between pow-
der in delivery state and after heat treatment, will cause critical velocity
to increase at the level of 40 m/s. Additionally, it must be noted that in
the case of a low-pressure system, air is used as the propellant gas.
This results in additional oxidation of powder particles despite the low
temperature of the process. It was claimed that oxygen content in cold
87
spray coatings is at the same level or even lower than feedstock material
due to separation of the oxide layer during impact. This crushed oxide is
partially removed and partially remains in formed coating [30]. Howev-
er, in this study oxygen amount in the coating was roughly two times
higher than for powder material in the case of copper. This additional
oxidation occurs partially during the spraying process and partially dur-
ing cooling of the sample, so it will also affect the deposition process. As
it was shown for Ti particles only during their presence in the nozzle
will the oxide growth be ~ 0.3 nm for a 10 μm particle [31]. A high
amount of oxides in coatings might also be the result of low velocity
which was insufficient to break them. A similar percent of oxygen con-
tent in all coating was reported by Li [12], however a much higher
spraying temperature was used at 520 °C.
Heat treatment of powder affects not only the oxide layer but also
the hardness of the powder [19]. The ultimate tensile strength of bulk
copper material amounted to 300 MPa and increased to 360 MPa for
copper in the form of powder [19]. Therefore, even heat treatment in
air, which causing additional oxidizing of the feedstock material, in-
creases spraying efficiency to a certain degree [32]. The increase of de-
position efficiency was, however, higher for powder treated in a
vacuum [32] The benefits of using heat treatment were extremely im-
portant for spraying copper coatings of polymers as it requires a very
narrow window of deposition and it was not possible to achieve the
coating without the heat treatment of powders. The oxygen content re-
corded for coatings should have limited influence on electrical
88 A. Małachowska et al. / Surface & Coatings Technology 318 (2017) 82–89

Fig. 8. Coating sprayed with dendritic powder on an Sn + Al2O3 interlayer: a) one spraying pass, b) three spraying passes.

conductivity, i.e. for copper, an oxygen content of ~0.20% decreased the
electrical conductivity value to 99% [33].
3.5. Electrical resistivity measurements
The electrical resistivity of copper coating amounted to ~26 μΩ cm
(approximately one order of magnitude higher than bulk material).
The low electrical conductivity was probably caused by presence of po-
rosity, which indicates small areas of metallic boding due to insufficient
velocity. The resistivity values for all copper coatings are in the range
given by Sudharshan [34] (24.39 μΩ cm) for following spraying param-
eters gas pressure: 1.4 MPa and a temperature of 300 °C, but two times
higher than this given by Ganesan [35] for copper coating on PVC at
200 °C [36] at a pressure of 2 MPa [18,35]. The difference could only
be partially explained through pressure difference because almost no
effect of gas pressure on deposition efficiency was observed [18]. It
could possibly be caused by lower oxidation of powder particles while
spraying (due to use of a nitrogen instead of air), which facilitates con-
tact to clean metallic surfaces. As a consequence, a greater number of
metallic bonds a created.
In the literature it is claimed that cold-sprayed coatings may reach
even 90% of conductivity of bulk copper [37]. On this background, ob-
tained results are limited in number. One on the ways to improve the
coatings' conductivity is heat treatment of coatings after spraying.
Koivuluoto et al. [38] reported the conductivity values to be 79% IACS
for HPCS, and 46% IACS for LPCS-sprayed coatings; they increased to
90% IACS for HPCS and 69% IACS after heat treatment (400 °C, 2 h).
The coatings were sprayed with spherical powders. However, heat
treatment was not applied due to the nature of the substrate.

Fig. 9. Oxygen content in feedstock material and coatings. Fig. 10. Fracture after bond strength test.
 A. Małachowska et al. / Surface & Coatings Technology 318 (2017) 82–89 89

3.6. Mechanical properties of coatings More studies are needed to further assess the influence of initial
feedstock properties on the deposition process on the polyamide 6 sub-
The bond strength of copper coatings amounted to 3.6 ± 0.2 MPa. strate. Future plans, including measurement of microhardness of pow-
The fractures occurred partially in the interlayer and partially in the der before and after spraying as a function of the oxidation level. It is
coatings (Fig. 10). Koivuluoto et al. [20] reported bond strength of den- also planned to develop a numerical model for investigation of the con-
dritic copper sprayed with a low-pressure cold spray (pressure 0.6 MPa tact mechanism of different powder particles.
and temperature 540 °C) to be ~7.5 MPa. This increased significantly to
~17.5 MPa with the addition of 50% Al2O3. Obtained bond strength value
amounts ~50% of it. Acknowledgements
The bonding mechanism and its strength is usually attributed to two
main mechanisms: formation of shear instabilities in particle/substrate This research was supported by NCN (Narodowe Centrum Nauki),
and particle/particles interface caused by extensive plastic deformation (grant no. UMO-2014/15/N/ST8/02661).
during impact [11,39] and mechanical interlocking [40]. However, in
the case of polymer/metal bonding, only the second one may appear References
due to the different nature of the coating and substrate material. Addi-
[1] C. Maurer, U. Schulz, Wear 317 (2014) 246–253.
tionally “interlocking might not be as strong as the one observed on [2] U. Schulz, Appl. Opt. 45 (2006) 1608–1618.
the metals substrate due to the soft nature of the polymer” [35]. The [3] J.-H. Yang, Y.-S. Han, J.-H. Choy, Thin Solid Films 495 (2006) 266–271.
other problem [35,41] when creating bonding was attributed to the [4] R. Suchentrunk, Kunststoff-Metallisierung: Mit 35 Tabellen, 3. Aufl, Schriftenreihe
Galvanotechnik und Oberflächenbehandlung, 37, Leuze, Bad Saulgau, 2007.
squeezed out polymer during particle impact which might partially en- [5] C.K. Lee, Mater. Chem. Phys. 114 (2009) 125–133.
close the metal particles and acts as a barrier against bonding formation. [6] R. Knight, M. Ivosevic, S. Kalinidi, G. Palmese, Proceedings of the ITSC 2003, 2003.
This was claimed to happen when using spherical copper as an interlay- [7] D.L. Gilmore, R.C. Dykhuizen, R.A. Neiser, M.F. Smith, T.J. Roemer, J. Therm. Spray
Technol. 8 (1999) 576–582.
er [35]. Tin turned out to have a more beneficial influence on adhesion [8] V.K. Champagne, The Cold Spray Materials Deposition Process: Fundamentals and
strength [18,35]. Ganesan et al. attributed this to the low mechanical Applications, Woodhead; CRC Press, Cambridge, Boca Raton, 2007 (http%3A//
properties of tin particles which were able to undergo the deformation www.worldcat.org/oclc/191557639).
[9] T. Schmidt, H. Assadi, F. Gärtner, H. Richter, T. Stoltenhoff, H. Kreye, T. Klassen, J.
instead of damaging the surface. This initial interlayer then enabled
Therm. Spray Technol. 18 (2009) 794–808.
bonding or interlocking for subsequent layers with higher mechanical [10] T. Schmidt, F. Gärtner, H. Assadi, H. Kreye, Acta Mater. 54 (2006) 729–742.
properties i.e. copper [18,35]. In the case of spherical copper particles, [11] H. Assadi, F. Gärtner, T. Stoltenhoff, H. Kreye, Acta Mater. 51 (2003) 4379–4394.
the shear strength of the copper coating was 1.93 ± 0.7 MPa in compar- [12] W.-Y. Li, C. Zhang, H.-T. Wang, X.P. Guo, H.L. Liao, C.-J. Li, C. Coddet, Appl. Surf. Sci.
253 (2007) 3557–3562.
ison to 5.4 ± 0.9 MPa for the tin interlayer [35]. Obtained bond strength [13] R. Lupoi, W. O'Neill, Surf. Coat. Technol. 205 (2010) 2167–2173.
lies in the middle of these two values. [14] X.L. Zhou, A.F. Chen, J.C. Liu, X.K. Wu, J.S. Zhang, Surf. Coat. Technol. 206 (2011)
Deposited copper coatings reached a microhardness of 125HV0.1 ± 132–136.
[15] H. Ye, J. Wang, Mater. Lett. 137 (2014) 21–24.
12 on PA6. This value is much lower than that given by Sudharshan - [16] D. Giraud, F. Borit, V. Guipont, M. Jeandin, J.M. Malhaire, Proceedings of the ITSC,
300 or even 450HV0.1 for copper coatings on aluminium [34] but simi- 2012 265–270 (Houston, Texas, USA).
lar to values given by [20,35]. However, due to different loads used for [17] M. Barletta, A. Gisario, V. Tagliaferri, Surf. Coat. Technol. 201 (2006) 296–308.
[18] A. Ganesan, M. Yamada, M. Fukumoto, J. Therm. Spray Technol. 22 (2013)
measurement, the values can be not be directly compared. The low 1275–1282.
values of copper microhardness might be prescribed to low a deforma- [19] H. Assadi, I. Irkhin, H. Gutzmann, F. Gärtner, M. Schulze, M. Villa Vidaller, T. Klassen,
tion grade of feedstock material. Additionally, lower microhardness of Adv. Powder Technol. 26 (2015) 1544–1554.
[20] H. Koivuluoto, P. Vuoristo, J. Therm. Spray Technol. 19 (2010) 1081–1092.
feedstock after heat treatment results in lower values of coatings micro- [21] D. Zhang, P.H. Shipway, D.G. McCartney, J. Therm. Spray Technol. 14 (2005)
hardness [32]. 109–116.
[22] L. Ajdelsztajn, J. Schoenung, B. Jodoin, G. Kim, Metall. Mater. Trans. A 36 (2005)
657–666.
3.7. Conclusions
[23] T.H. van Steenkiste, J.R. Smith, R.E. Teets, Surf. Coat. Technol. 154 (2002) 237–252.
[24] S. Suzuki, Y. Ishikawa, M. Isshiki, Y. Waseda, Mater. Trans. JIM 38 (1997) 1004–1009.
Copper coatings were successfully deposited on polyamide 6 with [25] J.C. Yang, B. Kolasa, J.M. Gibson, M. Yeadon, Appl. Phys. Lett. 73 (1998) 2841.
low-pressure cold spray using an Sn + Al2O3 interlayer and dendritic [26] J. Iijima, J.-W. Lim, S.-H. Hong, S. Suzuki, K. Mimura, M. Isshiki, Appl. Surf. Sci. 253
(2006) 2825–2829.
copper powder. The following specific conclusions have been drawn [27] K.H. Ko, J.O. Choi, H. Lee, B.J. Lee, Powder Technol. 218 (2012) 119–123.
based on the research: [28] C.-J. Li, H.-T. Wang, Q. Zhang, G.-J. Yang, W.-Y. Li, H.L. Liao, J. Therm. Spray Technol.
19 (2010) 95–101.
- The possibility of copper coating formation on the polymer is related [29] K. Kang, S. Yoon, Y. Ji, C. Lee, Mater. Sci. Eng. A 486 (2008) 300–307.
to powder morphology and oxygen content in the powder. Dendritic [30] W.-Y. Li, H. Liao, C.-J. Li, H.-S. Bang, C. Coddet, Appl. Surf. Sci. 253 (2007) 5084–5091.
[31] A. Małachowska, M. Winnicki, L. Pawłowski, B. Pateyron, A. Ambroziak, P.
powders, owing to having several contact places with the substrate Sokołowski, Proceedings of COMSOL Conference, Rotterdam, October 23–25, 2013.
surface, limit erosion processes during impact of the particles and [32] K.H. Ko, J.O. Choi, H. Lee, J. Mater. Process. Technol. 214 (2014) 1530–1535.
enable deposition of copper coatings on the polymer. On the other [33] J.R. Davis, A. Committee, Copper and Copper Alloys, ASM International, 2001
(https://books.google.pl/books?id=sxkPJzmkhnUC).
hand, this phenomenon causes porosity in the obtained coatings. [34] P. Sudharshan Phani, D. Srinivasa Rao, S.V. Joshi, G. Sundararajan, J. Therm. Spray
- Velocities obtained using a low-pressure cold spray are too small to Technol. 16 (2007) 425–434.
obtain porous-free copper coatings on polymers. This unfavourable [35] A. Ganesan, J. Affi, M. Yamada, M. Fukumoto, Surf. Coat. Technol. 207 (2012)
262–269.
effect is additionally strengthened by low substrate and interlayer
[36] W.B. Choi, L. Li, V. Luzin, R. Neiser, T. Gnäupel-Herold, H.J. Prask, S. Sampath, A.
hardness. Gouldstone, Acta Mater. 55 (2007) 857–866.
- Electrical conductivity amounted to 26 μΩ cm, which is one order of [37] T. Stoltenhoff, H. Kreye, H.J. Richter, J. Therm. Spray Technol. 11 (2002) 542–550.
[38] H. Koivuluoto, A. Coleman, K. Murray, M. Kearns, P. Vuoristo, J. Therm. Spray
magnitude higher than the bulk material. This low electrical conduc-
Technol. 21 (2012) 1065–1075.
tivity is caused by insufficient metallic bonding. It is probable that [39] M. Grujicic, C. Zhao, W. DeRosset, D. Helfritch, Mater. Des. 25 (2004) 681–688.
using nitrogen as the propellant gas could improve conductivity [40] T. Hussain, D.G. McCartney, P.H. Shipway, D. Zhang, J. Therm. Spray Technol. 18
owing to the reduction of oxygen content, and hence easier forma- (2009) 364–379.
[41] I. Burlacov, J. Jirkovský, L. Kavan, R. Ballhorn, R.B. Heimann, J. Photochem. Photobiol.
tion of metallic bonding. A Chem. 187 (2007) 285–292.