- Email: [email protected]
Laser-induced selective metallization of polypropylene doped with multiwall carbon nanotubes
Accepted Manuscript Title: Laser-induced selective metallization of polypropylene doped with multiwall carbon nanotubes Author: Karolis Ratautas Mindaugas Gedvilas Ina Stankeviˇciene Aldona Jagminien˙e Eugenijus Norkus Nello Li Pira Stefano Sinopoli Gediminas Raˇciukaitis PII: DOI: Reference:
S0169-4332(17)30934-0 http://dx.doi.org/doi:10.1016/j.apsusc.2017.03.238 APSUSC 35614
To appear in:
APSUSC
Received date: Revised date: Accepted date:
28-11-2016 17-2-2017 26-3-2017
Please cite this article as: K. Ratautas, M. Gedvilas, I. Stankeviˇciene, A. Jagminien˙e, E. Norkus, N.L. Pira, S. Sinopoli, G. Raˇciukaitis, Laser-induced selective metallization of polypropylene doped with multiwall carbon nanotubes, Applied Surface Science (2017), http://dx.doi.org/10.1016/j.apsusc.2017.03.238 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Laser-induced selective metallization of polypropylene doped with multiwall carbon
1,*
1
ip t
nanotubes
1
1
1
Karolis Ratautas , Mindaugas Gedvilas , Ina Stankevič iene , Aldona Jagminienė, Eugenijus Norkus , 2 3 1 Nello Li Pira , Stefano Sinopoli , Gediminas Rač iukaitis
cr
Center for Physical Sciences and Technology, Savanoriu Ave. 231, Vilnius, LT-02300, Lithuania 2 Centro Ricerche Fiat, Strada Torino 50, Orbassano (TO), 10043, Italy 3 BioAge Srl, Via Dei Glicini 25, Lamezia Terme (CZ), 88046, Italy * [email protected]
an
us
1
Highlights
M
1. PP doped with multiwall CNT can be activated with ns laser for electroless plating 2. Developed material is cheap decision for MID applications
d
3. Activation mechanism was preliminary proposed
Ac ce p
te
4. Demo for automotive application has been manufactured
Moulded interconnect devices (MID) offer the material, weight and cost saving by integration electronic circuits directly into polymeric components used in automotive and other consumer products. Lasers are used to write circuits directly by modifying the surface of polymers followed by an electroless metal plating. A new composite material – the polypropylene doped with multiwall carbon nanotubes was developed for the laser-induced selective metallization. Mechanism of surface activation by laser irradiation was investigated in details utilising picoand nanoseconds lasers. Deposition of copper was performed in the autocatalytic electroless plating bath. The laser-activated polymer surfaces have been studied using the Raman spectroscopy and scanning electron microscope (SEM). Microscopic images revealed that
1
Page 1 of 24
surface becomes active only after its melting by a laser. Alterations in the Raman spectra of the D and G bands indicated the clustering of carbon additives in the composite material. Optimal laser parameters for the surface activation were found by measuring a sheet resistance of the
ip t
finally metal-plated samples. A spatially selective copper plating was achieved with the smallest conductor line width of 22 μm at the laser scanning speed of 3 m/s and the pulse repetition rate
cr
of 100 kHz. Finally, the technique was validated by making functional electronic circuits by this
us
MID approach.
M
an
Keywords: Laser direct structuring; selective electroless plating; polypropylene; moulded interconnect devices; multiwall carbon nanotube
d
1. Introduction
te
The technology for selective metal plating of polymers has a high prospect for nowadays and future electronics. Particularly in the fields where usage of electronics increases rapidly, as the
Ac ce p
automotive industry, a huge amount of electronic components should fit into the certain volume of a device. In that case, printed circuit boards of 2D shape are inconvenient to use, and, therefore, the technology of moulded interconnect devices becomes more important [1]. The laser writing for selective plating of polymers can also be used on 3D surfaces since localised, and selective activation of a polymer is made with a laser beam. Scanning of the laser beam on 3D surfaces can be simply achieved technologically [2]. Another huge advantage of the laser writing for selective plating against the printed circuit board technology is the production costs of a prototype. The MID technology can be easily applied even for several new layouts of electronic devices. Moreover, it is a material-saving process as the metallization works selectively.
2
Page 2 of 24
There are several techniques of laser writing for selective plating of polymers: metal nano-ink printing [3, 4], laser-induced selective activation [5] and laser direct structuring (LDS) [6]. First one is a technique that uses silver or gold nanoparticle paste as an ink [7]. The method works in
ip t
two ways: using inkjet and laser heating [8] and laser-induced forward transfer method which uses a donor substrate near the specimen and applies a laser from the back side of the donor
cr
substrate for the lift-off process [4]. The nano-ink printing technology has some unsolved
us
problem when applied on 3D curved surfaces [3]. Laser-induced selective activation is a 3-step process. The first step is the laser surface structuring in a liquid environment. This step is used to
an
make porous, a sponge-like surface structure on a polymer. The structure is needed to keep the activation particles inside cavities [9]. The second step is chemical activation. The activation is
M
performed in a colloidal solution, usually using the palladium-tin chloride. This step includes not
d
only immersion a workpiece in the activation solution but also rinsing it after the activation.
te
Rinsing is the critical and not easy task. All the colloidal particles must be washed away from the laser-untreated areas but should remain inside the porous surface structures. The last, third step is
Ac ce p
the electroless metal plating, usually with copper [5]. The process is very unstable regarding its spatial selectivity, and laser processing in a liquid complicates application for 3D surfaces. LDS is the method using precursors mixed in a polymer matrix. These precursor additives are activated during the laser writing process in order to become a catalyst for electroless deposition of the metal, and thus the laser-scanned area can be selectively plated [6]. LDS is a 2-step process. There are a few commercial materials for LDS available on the market, but most of them are based on expensive metal-organic fillers, usually including palladium [6]. Compounds which are currently used in LDS are too expensive for a broad range of industrial applications. Therefore, the main objective of this research was to search for the technology
3
Page 3 of 24
which is capable of reducing the material costs. A new composite material – polypropylene (PP) with multiwall carbon nanotube (MWCNT) additives was investigated for the LDS approach using picosecond and nanosecond laser pulses. The concentration of MWCNT additives as low
ip t
as 2% by mass was tested. A usage of MWCNT in such low quantities enables to reduce a cost of the material (comparing with materials on the market) at least 2-5 times depending on a
cr
polymer. The sheet resistance measurement of the selectively plated samples, prepared using
us
different laser processing parameters and the same plating conditions, was used to find the optimal laser process window. Various laser processing parameters (pulse duration, laser energy,
an
wavelength, scanning speed, the number of scans) were applied. The surface activation process was also investigated using the Raman spectroscopy and scanning electron microscope imaging.
M
Mechanism of surface activation by laser irradiation was investigated in details. The specially
d
designed selectivity tests showed a high spatial resolution of the metal plating. The narrowest
te
width of the plated copper line was less than 22 μm, and other metals can be easy add on top of it. Finally, the new composite material was applied to make prototypes of the electronic circuit
Ac ce p
boards on 3D-shaped surfaces. 2. Experimental setup 2.1 Materials
A blend of polypropylene (PP) plastic (Lyondell Basell Hostacom CR1171G black) and industrial MWCNT (XNRI-7 from Mitsui and Co. LTD) was used for injection moulding. Grains of PP were previously mixed with MWCNT in an extruder at the concentration of 2.5%; 5.0%; 7.5% (of MWCNT) by mass. The PIOVAN mixer tool was used together with the Sandretto 330 tonnes injection moulding tool. The extruding temperature ranged 165-175 ºC. The injection
4
Page 4 of 24
pressure was 74 bars, and the temperature in a chamber ranged 175-195 ºC. Injected polymeric plates were used as a substrate for laser-induced selective metallization experiments.
ip t
2.2 Laser surface activation Two lasers: Baltic HP and Atlantic from Ekspla with a different pulse durations of 10 ns and
cr
10 ps at full width at half maximum (FWHM), respectively, were used for the polymer surface
us
activation. Translation of the laser beam was performed with a galvanometric scanner (Scanlab AG). The experimental setup is shown in Fig. 1(a). The F-theta lens of 80 mm focal length was
an
used to focus the laser beam on the surface of the polymer sample. The laser beam was scanned over the area to be metallised by hatching as shown in Fig. 1(b). The spatially selective
M
activation was tested depending on the used laser wavelength (1064 or 532 nm), scanning speed
d
(from 0.1 to 3 m/s), and the number of scans (from 1 to 50). 50% overlapping of the scanned
te
lines was used, and the average laser power was varied from 0.1 to 1 W. The pulse repetition rate was set to 50 or100 kHz. In all tests except the tests on spatial selectivity, a defocused laser beam
Ac ce p
with the spot size of 120 µm in diameter (Gaussian intensity level 1/e 2) was used.
(a)
(b)
5
Page 5 of 24
ip t
Fig. 1. (a) Experimental scheme of the laser activation of a polymer: LASER - picosecond or nanosecond laser, HWP - half wavelength wave plate, PBC - polarizing beam splitter cube, LBT - laser beam trap, L1 and L2 lens of the beam expander, M - highly reflecting mirror, GS galvanometer scanner, TFTL - telecentric-f-theta lens, S – sample, xy - denotes beam positioning directions. (b) Laser beam scanning path, z - beam propagation direction, D – laser spot size diameter on the sample, d – distance between overlapping neighbour scan lines (hatch distance)
cr
2.3 Sheet resistance measurement
The quality of the surface activation was investigated by testing samples after the copper plating.
us
The sheet resistance was measured using the four-probe method [10]. The results of sheet
an
resistance measurement were an indicator in estimating the best laser processing parameters.
M
2.4 Raman analysis
Raman spectroscopy was used to investigate a mechanism of the polymer activation. A
d
continuous-wave laser with a power of 10 mW and the wavelength of 632.8 nm was applied for
te
excitation of a sample. A microscope objective with the magnification factor of 50X was used to focus the excitation beam on the surface of the specimen. Each spectrum was averaged by
Ac ce p
collecting it ten times. The spectra were measured in the range of 250-3200 cm-1 using four different detectors. Analysis of the D and G Raman band before and after the laser treatment of a specimen was utilised to investigate the crystalline structure of the PP with MWCNT (PPMWCNT) additives. 2.5 Metal plating
After the laser activation, samples were copper-plated by immersing them in a chemical bath. Wayne Mayers [11] solution which composition is presented in Table 1 was used for the plating.
6
Page 6 of 24
The temperature of the bath was set to 40o C, and the laser-activated samples were immersed for 30 min.
ip t
Table 1. The composition of the copper electroless plating bath. Ingredient
Assignment
Formula
Concentration
Potassium sodium tartrate
Complexant
NaKC4H4O6·4H2O
Copper sulphate
Copper donor
CuSO4·5H2O
Sodium hydroxide
Buffer
NaOH
Sodium carbonate
Buffer
Na2 CO3
0.3 M
Formaldehyde
Reducing agent
HCOH
3.41 M
0.12 M 1.25 M
M
an
us
cr
0.35 M
te
d
3. Results
Ac ce p
3.1 Analysis of plated surfaces for different parameters of laser activation Initial results for laser treatment of the PP doped with MWCNT additives showed that electroless plating took place only after the surface was irradiated with nanosecond pulses. No plating was observed for samples structured with the picosecond laser in the investigated irradiation range. Therefore, the later experiments were continued only with the nanosecond laser. As known from the literature [12], picosecond pulses refer to “cold” photo-chemical ablation process, while the interaction of nanosecond pulses with the material leads to its thermal decomposition. The laseractivation process of PP with MWCNT additives seems to be based more on thermal effect than selective ablation as claimed in [13].
7
Page 7 of 24
The sheet resistance measurements were performed for the laser-treated areas after their electroless metallization to estimate the effect of laser processing parameters on the surface
1.000 100 10
1E-4 1E-5 1E-6 1E-7 1E-8 1E-9
0.2
0.4
0.6
0.8
1.0
P [W]
1E-3 1E-4 1E-5 1E-6 1E-7 1E-8 1E-9 1E-10 1E-11 0,0
us
0,1 m/s 0,2 m/s 0,3 m/s 0,4 m/s 0,5 m/s 0,6 m/s 0,7 m/s 0,8 m/s 0,9 m/s 1 m/s
cr
1 0,1 0,01
an
100 10 1 0.1 0.01 1E-3
1/ Rs [s sq ]
1 /Rs [S sq]
1000
1E-10 1E-11 0.0
ip t
activation.
0,2
0,4
0,6
0,1 m/s 0,2 m/s 0,3 m/s 0,4 m/s 0,5 m/s 0,6 m/s 0,7 m/s 0,8 m/s 0,9 m/s 1 m/s 0,8
1,0
P [W]
d
M
(a) (b) Fig. 2. Dependence of the sheet conductivity on the average laser power for various scanning speeds. The plating parameters were fixed. The measurement data were fit with spline curves, (a) represent the results of the laser activation using the 1064 nm and (b) of 532 nm wavelengths.
te
Initially, the influence of the average laser power and scanning speed for the surface activation was investigated at the pulse repetition rate of 50 kHz. The test results are presented by plots for
Ac ce p
both used wavelengths in Fig. 2, where the sheet resistance is replaced by the sheet conductance (1/Rs). The sheet conductance increased with the growth of laser fluence (power) until it reached the saturation. The reason why the sheet conductance stoped growing and saturated could be explained that the whole laser-activated surface was completely covered by a metal (Fig. 3d), while, in the growing stage of the curve, the metal layer contained not covered areas shown in Fig. 3 (a, b, c). It is clear from Fig. 2 that for each scanning speed, there was a certain laser power value when the sheet conductance saturated. The relation between the scanning speed, average laser power, and hatching distance were combined into a single parameter – the irradiation dose in J/cm2.
8
Page 8 of 24
ip t
(b)
(c)
an
us
cr
(a)
M
(d)
(e)
te
d
Fig. 3. Optical microscope images of metallised PP-MWCNT surfaces structured with the 532 nm wavelength, at the 0.8 m/s scanning speed and a different power of laser: (a) 0.1 W, (b) 0.2 W, (c) 0.3 W, (d) 0.4 W and (e) 4 W. All measurement results were combined and presented as the sheet conductance in dependence
Ac ce p
on the irradiation dose (Fig. 4). Although the same equivalent irradiation dose values were calculated from different parameter sets of laser processing (scanning speed and laser power), the samples processed at those regimes exhibited similar values of the sheet conductance (Fig. 4.). That results indicate that the laser-activation of the polymer surface has a threshold character. The drastic increase in the sheet conductance started at the dose of 0.55 J/cm2 and stabilised at the high (S‧sq) value when the irradiation dose was higher than 1.4 J/cm2 . The threshold value was considered to keep at 0.55 J/cm2, where the linear growth of the sheet conductance started.
9
Page 9 of 24
1000
(3)
100 10 1 0,1
ip t
0,1 W 0,2 W 0,3 W 0,4 W 0,5 W 0,6 W 0,7 W 0,8 W 0,9 W Average
1E-3
(2)
1E-4
cr
1/Rs [Ssq]
0,01
1E-5 1E-6
1E-8 1E-9
us
1E-7
(1)
Threshold: 0,55
an
1E-10 1
10
M
Irradiation dose [J/cm 2 ]
d
Fig. 4. Dependence of the sheet conductance on the irradiation dose. Laser radiation wavelength was 1064 nm.
te
For better understanding of the laser-activation step, the scanning electron microscope (SEM) imaging of the surfaces treated with various laser irradiation doses before the metallization was
Ac ce p
performed. Any noticeable changes in the polymer surface were observed for the doses below the threshold value. However, the SEM pictures revealed melting of the polymer surface when the irradiation dose was above the threshold. The steep increase in the sheet conductance started at approximately 0.55 J/cm2 and ended at 1.4 J/cm2. In this range of irradiation doses (from 0.55 to 1.4 J/cm2), the growing amount of re-melted PP-MWCNT material was evident on the surface by increasing the irradiation dose (Fig. 5.).
10 Page 10 of 24
ip t
us
cr
Fig. 5. Top-side SEM images of polymer surface processed with the indicated irradiation dose value around the activation threshold. Numbers 1-3 in the pictures correspond to the irradiation dose values numbered in Fig. 4. The sheet conductance was the highest when the surface of PP-MWCNT looks totally re-molten.
an
The SEM analysis proved that irradiation conditions leading to the surface activation also induce the melting of the surface. The activation threshold of the irradiation dose correlated with the
M
melting temperature of the PP-MWCNT composite. Both used laser wavelengths showed the similar behaviour of the activation. We suggest that the activation process is related with
d
polymer surface melting. However, the threshold irradiation dose for the 532 nm wavelength was
te
slightly lower as is shown in Fig. 6. This difference in the irradiation dose threshold can be
Ac ce p
affected by absorption coefficient of PP which is larger for the shorter 532 nm wavelength [14].
11 Page 11 of 24
1000 100 10 1
0,1 W 0,2 W 0,3 W 0,4 W 0,5 W 0,6 W 0,7 W 0,8 W 0,9 W Average
ip t
0.01 1E-3 1E-4 1E-5
cr
1/R s[S sq]
0.1
1E-6 1E-7
us
1E-8 1E-9
1
Threshold: 0,5
an
1E-10
10
2
M
Irradiation dose [J/cm ] Fig. 6. Sheet conductance dependence on the irradiation dose for the 532 nm wavelength.
d
The maximum sheet conductance of the copper-plated samples was found after activation with
te
the dose of 1.4 J/cm2 in the case of 1064 nm and 1.3 J/cm2 for the 532 nm wavelength (Fig. 4 and
Ac ce p
Fig. 6, respectively). The further slights decrease in the sheet conductance with the irradiation dose was probably related to the rougher surface of the polymer after its activation using a higher dose of irradiation. Using much high irradiation dose (~ 13 J/cm2), burning of the polymer surface took place, and, as a result, the surface was not fully covered by a metal (see Fig. 3 (e)). The sheet conductance measurement of the metal-plated specimens for various concentrations of MWCNT additives (from 2% to 7.5% by mass) did not show obvious changes in the final resistance value. Influence of the number of scans on the polymer surface activation was also investigated in our experiments. The same investigation method was used – measuring the sheet resistance of the finally plated samples. Multiple scanning over polymer surface did not provide any benefits.
12 Page 12 of 24
Several passes by a laser beam reduced the conductivity of a coated surface for the irradiation doses above the threshold. However, it increased also the conductivity of metallised copper for irradiation dose below the threshold. The reason is that later scans accumulated the temperature
ip t
of the surface and thus surface exceeds the melting point. Unfortunately, multiple scanning for the doses below the threshold did not allow to reach better metal plating quality comparing with
us
cr
a single scan when using the dose above the threshold.
For optimisation of the laser processing time, the maximum scanning speed of 3 m/s (limited by
an
the used galvoscanner) was tested for activation as well. The pulse repetition rate was increased to 100 kHz in order to have sufficient overlap of laser pulses. Irradiation dose value was kept
M
optimal for the 1064 nm wavelength – 1.4 J/cm2. No significant changes in sheet resistance were observed after the copper plating. That indicates further potentials for upscaling the laser
te
d
activation process speed.
Ac ce p
3.2 Raman spectroscopy for activated surface analysis In order to understand what happens with the polymer surface during the laser activation, Raman spectroscopy was applied. Differences in the Raman spectra of the polymer before and after laser activation were analysed. Raman spectra of the initial and treated with a laser polymer surfaces are shown in Fig. 7.
Raman spectrum of the PP with MWCNT sample before the laser treatment shows a wellrecognizable pattern of PP modes (Fig. 7 (a)). The middle-intensity band near 398 cm1 belongs to the deformational vibration of the C-C-C network [13]. The Raman bands sensitive to the crystalline phase of PP appear at 808 and 840 cm1 [14]. The presence of the band near 808 cm1 indicates that the studied sample contains crystalline phase, while the line near 840 cm1 is a
13 Page 13 of 24
marker of the defected phase which contains helical chains [13]. A stretching vibration of the CC bond is visible at 1166 cm1[13]. The doublet near 1439/1458 cm1 is associated with the deformation scissoring modes of methylene and methyl groups, respectively. In the high
ip t
frequency spectral region, the CH stretching vibrational modes are located at 2848 (CH2 symmetric), 2869 (CH3 symmetric), 2883 (CH3 symmetric), 2906 (CH2 symmetric), 2923 (CH2
cr
asymmetric), and 2958 cm1 (CH3 asymmetric) [14]. The presence of MWCNT in PP can be
us
recognised from three broad features located at 1329 (D band), 1597 (G band), and 2654 cm 1
Ac ce p
te
d
M
an
(2D band) [14]. The intensity of these bands is enhanced due to the resonant Raman scattering.
Fig. 7. Raman spectra of the PP-MWCNT composite: (a) before, and (b) after laser treatment (0.2 W, 0.2 m/s, 1064 nm). The Raman excitation wavelength was 632.8 nm. Spectra were corrected by using polynomial baseline function fit. Treatment of the PP-MWCNT sample with the laser radiation (Fig. 7(b)) resulted in several noticeable changes of the Raman bands. The relative intensity of the PP crystalline phasesensitive bands I808/I840 decreased from 2.30 to 1.26. Thus, the laser treatment increased the
14 Page 14 of 24
relative amount of the defected phase of PP comparing to the crystalline one. However, the laser treatment induced a strong decrease in the relative intensity of the PP bands comparing with the
ip t
carbon-related Raman features. Thus, the relative intensity I2883/IG decreased from 1.05 to 0.44. For analysis of the MWCNT precursor activation, the changes in the D and G Raman band
cr
intensities, responsible for carbon structure were investigated. Analysis of the D and G bands,
us
characteristic to carbon allotropes, showed several observable alterations in the spectra after the laser irradiation. Firstly, the relative intensity ID/IG increased after the activation, and, according
an
to Ferrari et al. [16], that indicates to the grow in the crystalline cluster size and formation of disordered nanocrystalline structure [17]. Secondly, the narrowing of the D and G bands was
M
observed. This result refers to the decrease in the number of defects – increased structural order. Narrowing of the bands was determinate by measuring FWHM of the fitted Gaussian curves as
te
d
shown in Fig 12. Moreover, a positive shift of the G band and negative shift of the D band were observed. The
Ac ce p
positive shift of the G band also indicates clustering of the carbon structure [18]. Characteristics of the Raman spectra before and after activation are represented in Table 2.
15 Page 15 of 24
D
1,0
laser treated untreated
G
ip t
0,6
cr
0,4
0,2
0,0 1000
1200
1400
1600
an
800
us
Raman intensity [a.u.]
0,8
1800
2000
M
Wavenumber [cm ]
d
Fig. 12. The Gaussian fit of the normalised spectral D and G bands, a surface treated with the laser (black colour), untreated (red colour). Arrows indicate the behaviour of the spectra after the laser activations.
te
The sheet resistance tests of the laser-treated but non-metalized and untreated surfaces were
Ac ce p
performed as well. The results indicated a decrease in the resistance more than 130 times after the laser activation (see values in Table 3). The equivalent reduction in the resistance was observed for all tested concentration of MWCNT additives in PP (2; 5; 7.5 %) Table 2. Characteristics of the carbon D and G bands before and after the activation with a laser (0.2 W, 0.2 m/s, 1064 nm).
I(D)/I(G)
Untreated
Treated
1.37
1.46
Dposition
1337.2 cm
-1
1333.6 cm-1
Gposition
1586.4 cm-1
1592.4 cm-1
DFWHM
108.4 cm-1
97.3 cm-1
GFWHM
108.5 cm-1
95.1 cm-1
16 Page 16 of 24
Table 3. Value of sheet resistance of non-metalized polymer surface Conc. of additivities
2,5 %
5%
yes
no
yes
Sheet resistance
1.66· 105 2.3· 107
no
yes
no
1.65· 105 2.2· 107 1.63· 105 2.2· 107
ip t
Laser treatment
7,5 %
cr
All these spectroscopy analysis results indicate structural changes of the material. The
us
destruction of the crystalline PP polymeric order was observed. However, the prime results for the selective plating are obtained from analysis of the D and G bands behaviour. It was shown
an
that average size of the crystalline MWCNT clusters was increased, and nano-crystalline disordered structure formation was initiated by the laser activation. The necessity of surface
M
melting for activation was proofed by the SEM images. Moreover, the change in conductance after the laser activation was shown as well. Therefore, all these results refer to the MWCNT
d
reorientation (reorganisation) in a molten pool of PP at the surface. When the laser beam melted
te
surface of the composite, MWCNT started to interact in a liquid PP and forms MWCNT
Ac ce p
connecting structure resulting in an increase of the surface conductivity by 130 times. The sheet resistance is still too high for application in electronics directly, but it can be a crucial parameter for selective metallization of the surface. The electroless plating procedure which was applied after the laser activation of the surface is a catalytic process. The main driven force in the electroless plating is the formaldehyde reduction reaction. This reaction is catalytic, and the final product of this reduction is a free electron on the catalyst surface. The following reaction is oxidation of metal ion. For this reaction, the free electron substitutes a ligand (sodium potassium tartrate) in the complexant compound. Therefore, the surface for the plating should be not only catalysing but also electro-conductive in order the
17 Page 17 of 24
electron could freely move on the surface of the catalyst. The increase in conductivity of the laser-activated surface more than 130 times is the key factor for a highly selective metal
ip t
deposition using the catalytic electroless plating method. 3.3 Test of the plating spatial selectivity
cr
Selectivity of metal plating was tested by varying diameter of the laser beam affecting the
us
polymer surface. The experiments were performed by scanning lines with a defocused laser beam. A sample position was vertically shifted relative to the focus position. Therefore, lines
an
with different width were modified on the polymer surface. The laser irradiation dose was kept constant during these experiments. After the metal plating procedure, the width of the metal lines
M
was measured. The range of defocusing was stepwise varied by 1 mm up to 18 mm. The
d
diameter of the activated area diameter was determined for the defocused laser spot size used in all previous tests. The diameter of the activated area was considered to keep the width of plated
te
metal line, and for the 134 µm beam (diameter at Gaussian intensity level 1/e2), and for the
Ac ce p
1.4 J/cm2 irradiation dose, it was 60 µm. A precise control of the line width was performed by adjusting the defocussing distance. The width of the line from 22 µm to more than 100 µm was achieved in a single laser scan without hatching as shown in Fig. 9.
18 Page 18 of 24
ip t cr us an M
d
Fig. 9. Results of the spatial selectivity test of plating by changing the focus position.
te
3.4 Fabrication of electronic circuit prototype
Ac ce p
Finally, the new PP-MWCNT material with the laser surface activation approach was applied for making the working prototypes of electronic circuits. Moulded interconnect device trials for FIAT 500 gloves box cover was prepared. The electronic circuit layout of the touch button for the cover opening mechanism was fabricated using the PP-MWCNT material and LDS process. Examples of conductive tracks for the touch sensitive button are shown in Fig. 10.
19 Page 19 of 24
ip t cr us an M
te
4. Conclusions
d
Fig. 10. Conductive tracks for the FIAT gloves box cover touch button demonstrator.
Ac ce p
Selective plating of the polypropylene (PP) doped with multiwall carbon nanotubes (MWCNT) was achieved after its surface activation with nanosecond laser pulses, but no metal plating was achieved after the picosecond laser treatment. Activation of the polymer with a nanosecond laser has a threshold for the irradiation dose, and this activation threshold is related to melting of the polymer surface. The optimal activation value of irradiation dose was higher using 1064 nm (1.4 J/cm2) than 532 nm (1.3 J/cm 2) wavelength because of the higher absorption due to a shorter wavelength, and, therefore, the lower melting threshold. The final sheet resistance values of the plated metal line did not show a significant difference between the specimens with a various concentration of the MWCNT additives in the tested range of 2.5 to 7.5% by weight. The surface
20 Page 20 of 24
treatment with the ns-laser also resulted in a decrease of the electrical sheet resistance of the nonmetalized PP-MWCNT.
ip t
Raman spectroscopy used in the activation chemistry analysis of PP-MWCNT revealed narrowing of the D and G bands, corresponding to vibration modes of carbon allotropes after the
cr
treatment with the ns-laser, indicating a decrease in the number of defects in crystalline phase of
us
carbon. The increase of the I(D)/I(G) ratio after the laser activation referred to the growth in the crystalline cluster size and formation of disordered nano-crystalline structure. The positive shift
an
of the G band confirmed the clustering as well. All Raman spectroscopy results, SEM analysis, and sheet resistance measurements test confirmed reorientation of the MWCNT additives into
M
the connecting structure when PP was melted by the laser. The increased electrical conductivity of the laser-activated areas enabled catalytic reaction of reducer in the electroless plating bath.
te
d
The conductive copper lines as narrow as 22 μm were achieved using PP-MWCNT. Polypropylene doped with MWCNT was shown to be suitable for applications in electronics, and
Ac ce p
that was proved by making several prototypes of functional electronic circuits. Acknowledgements
The authors acknowledge G. Niaura for the consultations and help in the Raman spectra analysis. Funding: the research leading to these results was partially funded by the European Union FP7 Program under grant agreement No 609355 (APPOLO). REFERENCES
21 Page 21 of 24
1.
P. Amend, C. Pscherer, T. Rechtenwald, T. Frick, and M. Schmidt, A fast and flexible
method for manufacturing 3D moulded interconnect devices by the use of a rapid prototyping
2.
G. Tsoukantas, K. Salonitis, P. Stavropoulos, and G. Chryssolouris, Overview of 3D laser
cr
materials processing concepts, Proc. SPIE 5131 (2003) 224-228.
E. Biver, L. Rapp, A.-P. Alloncle, P. Serra, and P. Delaporte, High-speed multi-jets
us
3.
ip t
technology, Phys. Procedia 5 (2010) 561-572.
printing using laser forward transfer: time-resolved study of the ejection dynamics, Opt. Express
L. Rapp, J. Ailuno, A. P. Alloncle, and P. Delaporte, Pulsed-laser printing of silver
M
4.
an
22 (2014) 17122-17134.
nanoparticles ink: control of morphological properties, Opt. Express 19 (2011) 21563-21574. Y. Zhang, H. Hansen, A. De Grave, P. Tang, and J. Nielsen, Selective metallization of
d
5.
te
polymers using laser induced surface activation (LISA)—characterization and optimization of
6.
Ac ce p
porous surface topography, Int. J. Adv. Manuf. Technol. 55 (2011) 573-580. J. K. M. Huske, J. Mller, and G. Esser, Laser Supported Activation and Additive
Metallization of Thermoplastics for 3D-MIDs, Proc. LANE (2001). 7.
R. Ramakrishnan, N. Saran, and R. J. Petcavich, Selective Inkjet Printing of Conductors
for Displays and Flexible Printed Electronics, J. Display Technol. 7 (2011) 344-347. 8.
S. H. Ko, J. Chung, Y. Choi, C. P. Grigoropoulos, N. R. Bieri, T.-y. Choi, C. Dockendorf,
and D. Poulikakos, Laser based hybrid inkjet printing of nano ink for flexible electronics, Proc. SPIE 5713 (2005) 97-104.
22 Page 22 of 24
9.
Y. Zhang, G. M. M. Kontogeorgis, and H. N. Hansen, An Explanation of the Selective
Plating of Laser Machined Surfaces Using Surface Tension Components, J. Adhes. Sci. Technol.
10.
ip t
25 (2011) 2101-2111. . M. Hansen, K. Stokbro, O. Hansen, T. Hassenkam, I. Shiraki, S. Hasegawa, and P.
11.
us
dimensional systems, Rev. Sci. Instrum. 74 (2003) 3701-3708.
T. Ogura, M. Malcomson, and Q. Fernando, Mechanism of copper deposition in
an
electroless plating, Langmuir 6 (1990) 1709-1710. 12.
cr
Bøggild, Resolution enhancement of scanning four-point-probe measurements on two-
Dieter W. Bäuerle, in Laser processing and chemistry, 4th ed., vol. 13, Springer: London,
M
United Kingdom 2011, Ch. 13.
F. S. Boydag, S. V. Mamedov, V. A. Alekperov, and G. Kandemir, A study of the optical
Ac ce p
14.
te
2448383.
d
13. A. Zecchina, F. Bardelli, S. Bertarione, G. Caputo, P. Castelli (Centro Researche Fiat), EP
properties of polypropylene based polymer composite films, Macromolecular Symposia 146 (1999) 187-192. 15.
F. R. Dollish, W. G. Fateley, and F. F. Bentley, Characteristic Raman frequencies of
organic compounds, Wiley, New York, 1974. 16.
E. E. Ibrahim, D. M. Chipara, R. Thapa, K. Lozano, and M. Chipara, Raman
Spectroscopy of Isotactic Polypropylene-Halloysite Nanocomposites, J. Nanomater. 2012 (2012), 8-16. 17.
A. C. Ferrari, and J. Robertson, Interpretation of Raman spectra of disordered and
amorphous carbon, Phys. Rev. B 61 (2000) 14095-14107. 23 Page 23 of 24
18.
R. Celiesiute, R. Trusovas, G. Niaura, V. Svedas, G. Raciukaitis, Z. Ruzele, and R.
Pauliukaite, Influence of the laser irradiation on the electrochemical and spectroscopic
Ac ce p
te
d
M
an
us
cr
ip t
peculiarities of graphene-chitosan composite film, Electrochim. Acta, 132 (2014) 265–276.
24 Page 24 of 24
S0169-4332(17)30934-0 http://dx.doi.org/doi:10.1016/j.apsusc.2017.03.238 APSUSC 35614
To appear in:
APSUSC
Received date: Revised date: Accepted date:
28-11-2016 17-2-2017 26-3-2017
Please cite this article as: K. Ratautas, M. Gedvilas, I. Stankeviˇciene, A. Jagminien˙e, E. Norkus, N.L. Pira, S. Sinopoli, G. Raˇciukaitis, Laser-induced selective metallization of polypropylene doped with multiwall carbon nanotubes, Applied Surface Science (2017), http://dx.doi.org/10.1016/j.apsusc.2017.03.238 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Laser-induced selective metallization of polypropylene doped with multiwall carbon
1,*
1
ip t
nanotubes
1
1
1
Karolis Ratautas , Mindaugas Gedvilas , Ina Stankevič iene , Aldona Jagminienė, Eugenijus Norkus , 2 3 1 Nello Li Pira , Stefano Sinopoli , Gediminas Rač iukaitis
cr
Center for Physical Sciences and Technology, Savanoriu Ave. 231, Vilnius, LT-02300, Lithuania 2 Centro Ricerche Fiat, Strada Torino 50, Orbassano (TO), 10043, Italy 3 BioAge Srl, Via Dei Glicini 25, Lamezia Terme (CZ), 88046, Italy * [email protected]
an
us
1
Highlights
M
1. PP doped with multiwall CNT can be activated with ns laser for electroless plating 2. Developed material is cheap decision for MID applications
d
3. Activation mechanism was preliminary proposed
Ac ce p
te
4. Demo for automotive application has been manufactured
Moulded interconnect devices (MID) offer the material, weight and cost saving by integration electronic circuits directly into polymeric components used in automotive and other consumer products. Lasers are used to write circuits directly by modifying the surface of polymers followed by an electroless metal plating. A new composite material – the polypropylene doped with multiwall carbon nanotubes was developed for the laser-induced selective metallization. Mechanism of surface activation by laser irradiation was investigated in details utilising picoand nanoseconds lasers. Deposition of copper was performed in the autocatalytic electroless plating bath. The laser-activated polymer surfaces have been studied using the Raman spectroscopy and scanning electron microscope (SEM). Microscopic images revealed that
1
Page 1 of 24
surface becomes active only after its melting by a laser. Alterations in the Raman spectra of the D and G bands indicated the clustering of carbon additives in the composite material. Optimal laser parameters for the surface activation were found by measuring a sheet resistance of the
ip t
finally metal-plated samples. A spatially selective copper plating was achieved with the smallest conductor line width of 22 μm at the laser scanning speed of 3 m/s and the pulse repetition rate
cr
of 100 kHz. Finally, the technique was validated by making functional electronic circuits by this
us
MID approach.
M
an
Keywords: Laser direct structuring; selective electroless plating; polypropylene; moulded interconnect devices; multiwall carbon nanotube
d
1. Introduction
te
The technology for selective metal plating of polymers has a high prospect for nowadays and future electronics. Particularly in the fields where usage of electronics increases rapidly, as the
Ac ce p
automotive industry, a huge amount of electronic components should fit into the certain volume of a device. In that case, printed circuit boards of 2D shape are inconvenient to use, and, therefore, the technology of moulded interconnect devices becomes more important [1]. The laser writing for selective plating of polymers can also be used on 3D surfaces since localised, and selective activation of a polymer is made with a laser beam. Scanning of the laser beam on 3D surfaces can be simply achieved technologically [2]. Another huge advantage of the laser writing for selective plating against the printed circuit board technology is the production costs of a prototype. The MID technology can be easily applied even for several new layouts of electronic devices. Moreover, it is a material-saving process as the metallization works selectively.
2
Page 2 of 24
There are several techniques of laser writing for selective plating of polymers: metal nano-ink printing [3, 4], laser-induced selective activation [5] and laser direct structuring (LDS) [6]. First one is a technique that uses silver or gold nanoparticle paste as an ink [7]. The method works in
ip t
two ways: using inkjet and laser heating [8] and laser-induced forward transfer method which uses a donor substrate near the specimen and applies a laser from the back side of the donor
cr
substrate for the lift-off process [4]. The nano-ink printing technology has some unsolved
us
problem when applied on 3D curved surfaces [3]. Laser-induced selective activation is a 3-step process. The first step is the laser surface structuring in a liquid environment. This step is used to
an
make porous, a sponge-like surface structure on a polymer. The structure is needed to keep the activation particles inside cavities [9]. The second step is chemical activation. The activation is
M
performed in a colloidal solution, usually using the palladium-tin chloride. This step includes not
d
only immersion a workpiece in the activation solution but also rinsing it after the activation.
te
Rinsing is the critical and not easy task. All the colloidal particles must be washed away from the laser-untreated areas but should remain inside the porous surface structures. The last, third step is
Ac ce p
the electroless metal plating, usually with copper [5]. The process is very unstable regarding its spatial selectivity, and laser processing in a liquid complicates application for 3D surfaces. LDS is the method using precursors mixed in a polymer matrix. These precursor additives are activated during the laser writing process in order to become a catalyst for electroless deposition of the metal, and thus the laser-scanned area can be selectively plated [6]. LDS is a 2-step process. There are a few commercial materials for LDS available on the market, but most of them are based on expensive metal-organic fillers, usually including palladium [6]. Compounds which are currently used in LDS are too expensive for a broad range of industrial applications. Therefore, the main objective of this research was to search for the technology
3
Page 3 of 24
which is capable of reducing the material costs. A new composite material – polypropylene (PP) with multiwall carbon nanotube (MWCNT) additives was investigated for the LDS approach using picosecond and nanosecond laser pulses. The concentration of MWCNT additives as low
ip t
as 2% by mass was tested. A usage of MWCNT in such low quantities enables to reduce a cost of the material (comparing with materials on the market) at least 2-5 times depending on a
cr
polymer. The sheet resistance measurement of the selectively plated samples, prepared using
us
different laser processing parameters and the same plating conditions, was used to find the optimal laser process window. Various laser processing parameters (pulse duration, laser energy,
an
wavelength, scanning speed, the number of scans) were applied. The surface activation process was also investigated using the Raman spectroscopy and scanning electron microscope imaging.
M
Mechanism of surface activation by laser irradiation was investigated in details. The specially
d
designed selectivity tests showed a high spatial resolution of the metal plating. The narrowest
te
width of the plated copper line was less than 22 μm, and other metals can be easy add on top of it. Finally, the new composite material was applied to make prototypes of the electronic circuit
Ac ce p
boards on 3D-shaped surfaces. 2. Experimental setup 2.1 Materials
A blend of polypropylene (PP) plastic (Lyondell Basell Hostacom CR1171G black) and industrial MWCNT (XNRI-7 from Mitsui and Co. LTD) was used for injection moulding. Grains of PP were previously mixed with MWCNT in an extruder at the concentration of 2.5%; 5.0%; 7.5% (of MWCNT) by mass. The PIOVAN mixer tool was used together with the Sandretto 330 tonnes injection moulding tool. The extruding temperature ranged 165-175 ºC. The injection
4
Page 4 of 24
pressure was 74 bars, and the temperature in a chamber ranged 175-195 ºC. Injected polymeric plates were used as a substrate for laser-induced selective metallization experiments.
ip t
2.2 Laser surface activation Two lasers: Baltic HP and Atlantic from Ekspla with a different pulse durations of 10 ns and
cr
10 ps at full width at half maximum (FWHM), respectively, were used for the polymer surface
us
activation. Translation of the laser beam was performed with a galvanometric scanner (Scanlab AG). The experimental setup is shown in Fig. 1(a). The F-theta lens of 80 mm focal length was
an
used to focus the laser beam on the surface of the polymer sample. The laser beam was scanned over the area to be metallised by hatching as shown in Fig. 1(b). The spatially selective
M
activation was tested depending on the used laser wavelength (1064 or 532 nm), scanning speed
d
(from 0.1 to 3 m/s), and the number of scans (from 1 to 50). 50% overlapping of the scanned
te
lines was used, and the average laser power was varied from 0.1 to 1 W. The pulse repetition rate was set to 50 or100 kHz. In all tests except the tests on spatial selectivity, a defocused laser beam
Ac ce p
with the spot size of 120 µm in diameter (Gaussian intensity level 1/e 2) was used.
(a)
(b)
5
Page 5 of 24
ip t
Fig. 1. (a) Experimental scheme of the laser activation of a polymer: LASER - picosecond or nanosecond laser, HWP - half wavelength wave plate, PBC - polarizing beam splitter cube, LBT - laser beam trap, L1 and L2 lens of the beam expander, M - highly reflecting mirror, GS galvanometer scanner, TFTL - telecentric-f-theta lens, S – sample, xy - denotes beam positioning directions. (b) Laser beam scanning path, z - beam propagation direction, D – laser spot size diameter on the sample, d – distance between overlapping neighbour scan lines (hatch distance)
cr
2.3 Sheet resistance measurement
The quality of the surface activation was investigated by testing samples after the copper plating.
us
The sheet resistance was measured using the four-probe method [10]. The results of sheet
an
resistance measurement were an indicator in estimating the best laser processing parameters.
M
2.4 Raman analysis
Raman spectroscopy was used to investigate a mechanism of the polymer activation. A
d
continuous-wave laser with a power of 10 mW and the wavelength of 632.8 nm was applied for
te
excitation of a sample. A microscope objective with the magnification factor of 50X was used to focus the excitation beam on the surface of the specimen. Each spectrum was averaged by
Ac ce p
collecting it ten times. The spectra were measured in the range of 250-3200 cm-1 using four different detectors. Analysis of the D and G Raman band before and after the laser treatment of a specimen was utilised to investigate the crystalline structure of the PP with MWCNT (PPMWCNT) additives. 2.5 Metal plating
After the laser activation, samples were copper-plated by immersing them in a chemical bath. Wayne Mayers [11] solution which composition is presented in Table 1 was used for the plating.
6
Page 6 of 24
The temperature of the bath was set to 40o C, and the laser-activated samples were immersed for 30 min.
ip t
Table 1. The composition of the copper electroless plating bath. Ingredient
Assignment
Formula
Concentration
Potassium sodium tartrate
Complexant
NaKC4H4O6·4H2O
Copper sulphate
Copper donor
CuSO4·5H2O
Sodium hydroxide
Buffer
NaOH
Sodium carbonate
Buffer
Na2 CO3
0.3 M
Formaldehyde
Reducing agent
HCOH
3.41 M
0.12 M 1.25 M
M
an
us
cr
0.35 M
te
d
3. Results
Ac ce p
3.1 Analysis of plated surfaces for different parameters of laser activation Initial results for laser treatment of the PP doped with MWCNT additives showed that electroless plating took place only after the surface was irradiated with nanosecond pulses. No plating was observed for samples structured with the picosecond laser in the investigated irradiation range. Therefore, the later experiments were continued only with the nanosecond laser. As known from the literature [12], picosecond pulses refer to “cold” photo-chemical ablation process, while the interaction of nanosecond pulses with the material leads to its thermal decomposition. The laseractivation process of PP with MWCNT additives seems to be based more on thermal effect than selective ablation as claimed in [13].
7
Page 7 of 24
The sheet resistance measurements were performed for the laser-treated areas after their electroless metallization to estimate the effect of laser processing parameters on the surface
1.000 100 10
1E-4 1E-5 1E-6 1E-7 1E-8 1E-9
0.2
0.4
0.6
0.8
1.0
P [W]
1E-3 1E-4 1E-5 1E-6 1E-7 1E-8 1E-9 1E-10 1E-11 0,0
us
0,1 m/s 0,2 m/s 0,3 m/s 0,4 m/s 0,5 m/s 0,6 m/s 0,7 m/s 0,8 m/s 0,9 m/s 1 m/s
cr
1 0,1 0,01
an
100 10 1 0.1 0.01 1E-3
1/ Rs [s sq ]
1 /Rs [S sq]
1000
1E-10 1E-11 0.0
ip t
activation.
0,2
0,4
0,6
0,1 m/s 0,2 m/s 0,3 m/s 0,4 m/s 0,5 m/s 0,6 m/s 0,7 m/s 0,8 m/s 0,9 m/s 1 m/s 0,8
1,0
P [W]
d
M
(a) (b) Fig. 2. Dependence of the sheet conductivity on the average laser power for various scanning speeds. The plating parameters were fixed. The measurement data were fit with spline curves, (a) represent the results of the laser activation using the 1064 nm and (b) of 532 nm wavelengths.
te
Initially, the influence of the average laser power and scanning speed for the surface activation was investigated at the pulse repetition rate of 50 kHz. The test results are presented by plots for
Ac ce p
both used wavelengths in Fig. 2, where the sheet resistance is replaced by the sheet conductance (1/Rs). The sheet conductance increased with the growth of laser fluence (power) until it reached the saturation. The reason why the sheet conductance stoped growing and saturated could be explained that the whole laser-activated surface was completely covered by a metal (Fig. 3d), while, in the growing stage of the curve, the metal layer contained not covered areas shown in Fig. 3 (a, b, c). It is clear from Fig. 2 that for each scanning speed, there was a certain laser power value when the sheet conductance saturated. The relation between the scanning speed, average laser power, and hatching distance were combined into a single parameter – the irradiation dose in J/cm2.
8
Page 8 of 24
ip t
(b)
(c)
an
us
cr
(a)
M
(d)
(e)
te
d
Fig. 3. Optical microscope images of metallised PP-MWCNT surfaces structured with the 532 nm wavelength, at the 0.8 m/s scanning speed and a different power of laser: (a) 0.1 W, (b) 0.2 W, (c) 0.3 W, (d) 0.4 W and (e) 4 W. All measurement results were combined and presented as the sheet conductance in dependence
Ac ce p
on the irradiation dose (Fig. 4). Although the same equivalent irradiation dose values were calculated from different parameter sets of laser processing (scanning speed and laser power), the samples processed at those regimes exhibited similar values of the sheet conductance (Fig. 4.). That results indicate that the laser-activation of the polymer surface has a threshold character. The drastic increase in the sheet conductance started at the dose of 0.55 J/cm2 and stabilised at the high (S‧sq) value when the irradiation dose was higher than 1.4 J/cm2 . The threshold value was considered to keep at 0.55 J/cm2, where the linear growth of the sheet conductance started.
9
Page 9 of 24
1000
(3)
100 10 1 0,1
ip t
0,1 W 0,2 W 0,3 W 0,4 W 0,5 W 0,6 W 0,7 W 0,8 W 0,9 W Average
1E-3
(2)
1E-4
cr
1/Rs [Ssq]
0,01
1E-5 1E-6
1E-8 1E-9
us
1E-7
(1)
Threshold: 0,55
an
1E-10 1
10
M
Irradiation dose [J/cm 2 ]
d
Fig. 4. Dependence of the sheet conductance on the irradiation dose. Laser radiation wavelength was 1064 nm.
te
For better understanding of the laser-activation step, the scanning electron microscope (SEM) imaging of the surfaces treated with various laser irradiation doses before the metallization was
Ac ce p
performed. Any noticeable changes in the polymer surface were observed for the doses below the threshold value. However, the SEM pictures revealed melting of the polymer surface when the irradiation dose was above the threshold. The steep increase in the sheet conductance started at approximately 0.55 J/cm2 and ended at 1.4 J/cm2. In this range of irradiation doses (from 0.55 to 1.4 J/cm2), the growing amount of re-melted PP-MWCNT material was evident on the surface by increasing the irradiation dose (Fig. 5.).
10 Page 10 of 24
ip t
us
cr
Fig. 5. Top-side SEM images of polymer surface processed with the indicated irradiation dose value around the activation threshold. Numbers 1-3 in the pictures correspond to the irradiation dose values numbered in Fig. 4. The sheet conductance was the highest when the surface of PP-MWCNT looks totally re-molten.
an
The SEM analysis proved that irradiation conditions leading to the surface activation also induce the melting of the surface. The activation threshold of the irradiation dose correlated with the
M
melting temperature of the PP-MWCNT composite. Both used laser wavelengths showed the similar behaviour of the activation. We suggest that the activation process is related with
d
polymer surface melting. However, the threshold irradiation dose for the 532 nm wavelength was
te
slightly lower as is shown in Fig. 6. This difference in the irradiation dose threshold can be
Ac ce p
affected by absorption coefficient of PP which is larger for the shorter 532 nm wavelength [14].
11 Page 11 of 24
1000 100 10 1
0,1 W 0,2 W 0,3 W 0,4 W 0,5 W 0,6 W 0,7 W 0,8 W 0,9 W Average
ip t
0.01 1E-3 1E-4 1E-5
cr
1/R s[S sq]
0.1
1E-6 1E-7
us
1E-8 1E-9
1
Threshold: 0,5
an
1E-10
10
2
M
Irradiation dose [J/cm ] Fig. 6. Sheet conductance dependence on the irradiation dose for the 532 nm wavelength.
d
The maximum sheet conductance of the copper-plated samples was found after activation with
te
the dose of 1.4 J/cm2 in the case of 1064 nm and 1.3 J/cm2 for the 532 nm wavelength (Fig. 4 and
Ac ce p
Fig. 6, respectively). The further slights decrease in the sheet conductance with the irradiation dose was probably related to the rougher surface of the polymer after its activation using a higher dose of irradiation. Using much high irradiation dose (~ 13 J/cm2), burning of the polymer surface took place, and, as a result, the surface was not fully covered by a metal (see Fig. 3 (e)). The sheet conductance measurement of the metal-plated specimens for various concentrations of MWCNT additives (from 2% to 7.5% by mass) did not show obvious changes in the final resistance value. Influence of the number of scans on the polymer surface activation was also investigated in our experiments. The same investigation method was used – measuring the sheet resistance of the finally plated samples. Multiple scanning over polymer surface did not provide any benefits.
12 Page 12 of 24
Several passes by a laser beam reduced the conductivity of a coated surface for the irradiation doses above the threshold. However, it increased also the conductivity of metallised copper for irradiation dose below the threshold. The reason is that later scans accumulated the temperature
ip t
of the surface and thus surface exceeds the melting point. Unfortunately, multiple scanning for the doses below the threshold did not allow to reach better metal plating quality comparing with
us
cr
a single scan when using the dose above the threshold.
For optimisation of the laser processing time, the maximum scanning speed of 3 m/s (limited by
an
the used galvoscanner) was tested for activation as well. The pulse repetition rate was increased to 100 kHz in order to have sufficient overlap of laser pulses. Irradiation dose value was kept
M
optimal for the 1064 nm wavelength – 1.4 J/cm2. No significant changes in sheet resistance were observed after the copper plating. That indicates further potentials for upscaling the laser
te
d
activation process speed.
Ac ce p
3.2 Raman spectroscopy for activated surface analysis In order to understand what happens with the polymer surface during the laser activation, Raman spectroscopy was applied. Differences in the Raman spectra of the polymer before and after laser activation were analysed. Raman spectra of the initial and treated with a laser polymer surfaces are shown in Fig. 7.
Raman spectrum of the PP with MWCNT sample before the laser treatment shows a wellrecognizable pattern of PP modes (Fig. 7 (a)). The middle-intensity band near 398 cm1 belongs to the deformational vibration of the C-C-C network [13]. The Raman bands sensitive to the crystalline phase of PP appear at 808 and 840 cm1 [14]. The presence of the band near 808 cm1 indicates that the studied sample contains crystalline phase, while the line near 840 cm1 is a
13 Page 13 of 24
marker of the defected phase which contains helical chains [13]. A stretching vibration of the CC bond is visible at 1166 cm1[13]. The doublet near 1439/1458 cm1 is associated with the deformation scissoring modes of methylene and methyl groups, respectively. In the high
ip t
frequency spectral region, the CH stretching vibrational modes are located at 2848 (CH2 symmetric), 2869 (CH3 symmetric), 2883 (CH3 symmetric), 2906 (CH2 symmetric), 2923 (CH2
cr
asymmetric), and 2958 cm1 (CH3 asymmetric) [14]. The presence of MWCNT in PP can be
us
recognised from three broad features located at 1329 (D band), 1597 (G band), and 2654 cm 1
Ac ce p
te
d
M
an
(2D band) [14]. The intensity of these bands is enhanced due to the resonant Raman scattering.
Fig. 7. Raman spectra of the PP-MWCNT composite: (a) before, and (b) after laser treatment (0.2 W, 0.2 m/s, 1064 nm). The Raman excitation wavelength was 632.8 nm. Spectra were corrected by using polynomial baseline function fit. Treatment of the PP-MWCNT sample with the laser radiation (Fig. 7(b)) resulted in several noticeable changes of the Raman bands. The relative intensity of the PP crystalline phasesensitive bands I808/I840 decreased from 2.30 to 1.26. Thus, the laser treatment increased the
14 Page 14 of 24
relative amount of the defected phase of PP comparing to the crystalline one. However, the laser treatment induced a strong decrease in the relative intensity of the PP bands comparing with the
ip t
carbon-related Raman features. Thus, the relative intensity I2883/IG decreased from 1.05 to 0.44. For analysis of the MWCNT precursor activation, the changes in the D and G Raman band
cr
intensities, responsible for carbon structure were investigated. Analysis of the D and G bands,
us
characteristic to carbon allotropes, showed several observable alterations in the spectra after the laser irradiation. Firstly, the relative intensity ID/IG increased after the activation, and, according
an
to Ferrari et al. [16], that indicates to the grow in the crystalline cluster size and formation of disordered nanocrystalline structure [17]. Secondly, the narrowing of the D and G bands was
M
observed. This result refers to the decrease in the number of defects – increased structural order. Narrowing of the bands was determinate by measuring FWHM of the fitted Gaussian curves as
te
d
shown in Fig 12. Moreover, a positive shift of the G band and negative shift of the D band were observed. The
Ac ce p
positive shift of the G band also indicates clustering of the carbon structure [18]. Characteristics of the Raman spectra before and after activation are represented in Table 2.
15 Page 15 of 24
D
1,0
laser treated untreated
G
ip t
0,6
cr
0,4
0,2
0,0 1000
1200
1400
1600
an
800
us
Raman intensity [a.u.]
0,8
1800
2000
M
Wavenumber [cm ]
d
Fig. 12. The Gaussian fit of the normalised spectral D and G bands, a surface treated with the laser (black colour), untreated (red colour). Arrows indicate the behaviour of the spectra after the laser activations.
te
The sheet resistance tests of the laser-treated but non-metalized and untreated surfaces were
Ac ce p
performed as well. The results indicated a decrease in the resistance more than 130 times after the laser activation (see values in Table 3). The equivalent reduction in the resistance was observed for all tested concentration of MWCNT additives in PP (2; 5; 7.5 %) Table 2. Characteristics of the carbon D and G bands before and after the activation with a laser (0.2 W, 0.2 m/s, 1064 nm).
I(D)/I(G)
Untreated
Treated
1.37
1.46
Dposition
1337.2 cm
-1
1333.6 cm-1
Gposition
1586.4 cm-1
1592.4 cm-1
DFWHM
108.4 cm-1
97.3 cm-1
GFWHM
108.5 cm-1
95.1 cm-1
16 Page 16 of 24
Table 3. Value of sheet resistance of non-metalized polymer surface Conc. of additivities
2,5 %
5%
yes
no
yes
Sheet resistance
1.66· 105 2.3· 107
no
yes
no
1.65· 105 2.2· 107 1.63· 105 2.2· 107
ip t
Laser treatment
7,5 %
cr
All these spectroscopy analysis results indicate structural changes of the material. The
us
destruction of the crystalline PP polymeric order was observed. However, the prime results for the selective plating are obtained from analysis of the D and G bands behaviour. It was shown
an
that average size of the crystalline MWCNT clusters was increased, and nano-crystalline disordered structure formation was initiated by the laser activation. The necessity of surface
M
melting for activation was proofed by the SEM images. Moreover, the change in conductance after the laser activation was shown as well. Therefore, all these results refer to the MWCNT
d
reorientation (reorganisation) in a molten pool of PP at the surface. When the laser beam melted
te
surface of the composite, MWCNT started to interact in a liquid PP and forms MWCNT
Ac ce p
connecting structure resulting in an increase of the surface conductivity by 130 times. The sheet resistance is still too high for application in electronics directly, but it can be a crucial parameter for selective metallization of the surface. The electroless plating procedure which was applied after the laser activation of the surface is a catalytic process. The main driven force in the electroless plating is the formaldehyde reduction reaction. This reaction is catalytic, and the final product of this reduction is a free electron on the catalyst surface. The following reaction is oxidation of metal ion. For this reaction, the free electron substitutes a ligand (sodium potassium tartrate) in the complexant compound. Therefore, the surface for the plating should be not only catalysing but also electro-conductive in order the
17 Page 17 of 24
electron could freely move on the surface of the catalyst. The increase in conductivity of the laser-activated surface more than 130 times is the key factor for a highly selective metal
ip t
deposition using the catalytic electroless plating method. 3.3 Test of the plating spatial selectivity
cr
Selectivity of metal plating was tested by varying diameter of the laser beam affecting the
us
polymer surface. The experiments were performed by scanning lines with a defocused laser beam. A sample position was vertically shifted relative to the focus position. Therefore, lines
an
with different width were modified on the polymer surface. The laser irradiation dose was kept constant during these experiments. After the metal plating procedure, the width of the metal lines
M
was measured. The range of defocusing was stepwise varied by 1 mm up to 18 mm. The
d
diameter of the activated area diameter was determined for the defocused laser spot size used in all previous tests. The diameter of the activated area was considered to keep the width of plated
te
metal line, and for the 134 µm beam (diameter at Gaussian intensity level 1/e2), and for the
Ac ce p
1.4 J/cm2 irradiation dose, it was 60 µm. A precise control of the line width was performed by adjusting the defocussing distance. The width of the line from 22 µm to more than 100 µm was achieved in a single laser scan without hatching as shown in Fig. 9.
18 Page 18 of 24
ip t cr us an M
d
Fig. 9. Results of the spatial selectivity test of plating by changing the focus position.
te
3.4 Fabrication of electronic circuit prototype
Ac ce p
Finally, the new PP-MWCNT material with the laser surface activation approach was applied for making the working prototypes of electronic circuits. Moulded interconnect device trials for FIAT 500 gloves box cover was prepared. The electronic circuit layout of the touch button for the cover opening mechanism was fabricated using the PP-MWCNT material and LDS process. Examples of conductive tracks for the touch sensitive button are shown in Fig. 10.
19 Page 19 of 24
ip t cr us an M
te
4. Conclusions
d
Fig. 10. Conductive tracks for the FIAT gloves box cover touch button demonstrator.
Ac ce p
Selective plating of the polypropylene (PP) doped with multiwall carbon nanotubes (MWCNT) was achieved after its surface activation with nanosecond laser pulses, but no metal plating was achieved after the picosecond laser treatment. Activation of the polymer with a nanosecond laser has a threshold for the irradiation dose, and this activation threshold is related to melting of the polymer surface. The optimal activation value of irradiation dose was higher using 1064 nm (1.4 J/cm2) than 532 nm (1.3 J/cm 2) wavelength because of the higher absorption due to a shorter wavelength, and, therefore, the lower melting threshold. The final sheet resistance values of the plated metal line did not show a significant difference between the specimens with a various concentration of the MWCNT additives in the tested range of 2.5 to 7.5% by weight. The surface
20 Page 20 of 24
treatment with the ns-laser also resulted in a decrease of the electrical sheet resistance of the nonmetalized PP-MWCNT.
ip t
Raman spectroscopy used in the activation chemistry analysis of PP-MWCNT revealed narrowing of the D and G bands, corresponding to vibration modes of carbon allotropes after the
cr
treatment with the ns-laser, indicating a decrease in the number of defects in crystalline phase of
us
carbon. The increase of the I(D)/I(G) ratio after the laser activation referred to the growth in the crystalline cluster size and formation of disordered nano-crystalline structure. The positive shift
an
of the G band confirmed the clustering as well. All Raman spectroscopy results, SEM analysis, and sheet resistance measurements test confirmed reorientation of the MWCNT additives into
M
the connecting structure when PP was melted by the laser. The increased electrical conductivity of the laser-activated areas enabled catalytic reaction of reducer in the electroless plating bath.
te
d
The conductive copper lines as narrow as 22 μm were achieved using PP-MWCNT. Polypropylene doped with MWCNT was shown to be suitable for applications in electronics, and
Ac ce p
that was proved by making several prototypes of functional electronic circuits. Acknowledgements
The authors acknowledge G. Niaura for the consultations and help in the Raman spectra analysis. Funding: the research leading to these results was partially funded by the European Union FP7 Program under grant agreement No 609355 (APPOLO). REFERENCES
21 Page 21 of 24
1.
P. Amend, C. Pscherer, T. Rechtenwald, T. Frick, and M. Schmidt, A fast and flexible
method for manufacturing 3D moulded interconnect devices by the use of a rapid prototyping
2.
G. Tsoukantas, K. Salonitis, P. Stavropoulos, and G. Chryssolouris, Overview of 3D laser
cr
materials processing concepts, Proc. SPIE 5131 (2003) 224-228.
E. Biver, L. Rapp, A.-P. Alloncle, P. Serra, and P. Delaporte, High-speed multi-jets
us
3.
ip t
technology, Phys. Procedia 5 (2010) 561-572.
printing using laser forward transfer: time-resolved study of the ejection dynamics, Opt. Express
L. Rapp, J. Ailuno, A. P. Alloncle, and P. Delaporte, Pulsed-laser printing of silver
M
4.
an
22 (2014) 17122-17134.
nanoparticles ink: control of morphological properties, Opt. Express 19 (2011) 21563-21574. Y. Zhang, H. Hansen, A. De Grave, P. Tang, and J. Nielsen, Selective metallization of
d
5.
te
polymers using laser induced surface activation (LISA)—characterization and optimization of
6.
Ac ce p
porous surface topography, Int. J. Adv. Manuf. Technol. 55 (2011) 573-580. J. K. M. Huske, J. Mller, and G. Esser, Laser Supported Activation and Additive
Metallization of Thermoplastics for 3D-MIDs, Proc. LANE (2001). 7.
R. Ramakrishnan, N. Saran, and R. J. Petcavich, Selective Inkjet Printing of Conductors
for Displays and Flexible Printed Electronics, J. Display Technol. 7 (2011) 344-347. 8.
S. H. Ko, J. Chung, Y. Choi, C. P. Grigoropoulos, N. R. Bieri, T.-y. Choi, C. Dockendorf,
and D. Poulikakos, Laser based hybrid inkjet printing of nano ink for flexible electronics, Proc. SPIE 5713 (2005) 97-104.
22 Page 22 of 24
9.
Y. Zhang, G. M. M. Kontogeorgis, and H. N. Hansen, An Explanation of the Selective
Plating of Laser Machined Surfaces Using Surface Tension Components, J. Adhes. Sci. Technol.
10.
ip t
25 (2011) 2101-2111. . M. Hansen, K. Stokbro, O. Hansen, T. Hassenkam, I. Shiraki, S. Hasegawa, and P.
11.
us
dimensional systems, Rev. Sci. Instrum. 74 (2003) 3701-3708.
T. Ogura, M. Malcomson, and Q. Fernando, Mechanism of copper deposition in
an
electroless plating, Langmuir 6 (1990) 1709-1710. 12.
cr
Bøggild, Resolution enhancement of scanning four-point-probe measurements on two-
Dieter W. Bäuerle, in Laser processing and chemistry, 4th ed., vol. 13, Springer: London,
M
United Kingdom 2011, Ch. 13.
F. S. Boydag, S. V. Mamedov, V. A. Alekperov, and G. Kandemir, A study of the optical
Ac ce p
14.
te
2448383.
d
13. A. Zecchina, F. Bardelli, S. Bertarione, G. Caputo, P. Castelli (Centro Researche Fiat), EP
properties of polypropylene based polymer composite films, Macromolecular Symposia 146 (1999) 187-192. 15.
F. R. Dollish, W. G. Fateley, and F. F. Bentley, Characteristic Raman frequencies of
organic compounds, Wiley, New York, 1974. 16.
E. E. Ibrahim, D. M. Chipara, R. Thapa, K. Lozano, and M. Chipara, Raman
Spectroscopy of Isotactic Polypropylene-Halloysite Nanocomposites, J. Nanomater. 2012 (2012), 8-16. 17.
A. C. Ferrari, and J. Robertson, Interpretation of Raman spectra of disordered and
amorphous carbon, Phys. Rev. B 61 (2000) 14095-14107. 23 Page 23 of 24
18.
R. Celiesiute, R. Trusovas, G. Niaura, V. Svedas, G. Raciukaitis, Z. Ruzele, and R.
Pauliukaite, Influence of the laser irradiation on the electrochemical and spectroscopic
Ac ce p
te
d
M
an
us
cr
ip t
peculiarities of graphene-chitosan composite film, Electrochim. Acta, 132 (2014) 265–276.
24 Page 24 of 24