- Email: [email protected]
Tungsten borides layers deposited by a nanosecond laser pulse
Accepted Manuscript Tungsten borides layers deposited by a nanosecond laser pulse
Justyna Chrzanowska-Giżyńska, Piotr Denis, Jacek Hoffman, Maciej Giżyński, Tomasz Mościcki, Dariusz Garbiec, Zygmunt Szymański PII: DOI: Reference:
S0257-8972(17)31263-X https://doi.org/10.1016/j.surfcoat.2017.12.040 SCT 22958
To appear in:
Surface & Coatings Technology
Received date: Revised date: Accepted date:
21 August 2017 12 December 2017 16 December 2017
Please cite this article as: Justyna Chrzanowska-Giżyńska, Piotr Denis, Jacek Hoffman, Maciej Giżyński, Tomasz Mościcki, Dariusz Garbiec, Zygmunt Szymański , Tungsten borides layers deposited by a nanosecond laser pulse. The address for the corresponding author was captured as affiliation for all authors. Please check if appropriate. Sct(2017), https://doi.org/10.1016/j.surfcoat.2017.12.040
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.
ACCEPTED MANUSCRIPT Tungsten borides layers deposited by a nanosecond laser pulse Justyna Chrzanowska-Giżyńska*1, Piotr Denis1, Jacek Hoffman1, Maciej Giżyński2,Tomasz Mościcki1, Dariusz Garbiec3, Zygmunt Szymański1 Institute of Fundamental Technological Research, Polish Academy of Sciences, Pawińskiego 5B, 02106 Warsaw, Poland 2 Faculty of Material Science and Engineering of Warsaw University of Technology, Wołoska 141, 02-507 Warsaw, Poland 3 Metal Forming Institute, Jana Pawla II 14, 61-139 Poznan, Poland * corresponding author e-mail: [email protected], [email protected]
PT
1
AC
CE
PT E
D
MA
NU
SC
RI
Abstract Tungsten borides belong to the group of potentially superhard materials which hardness could be compared to cubic boron nitride and diamond. However, difficulty in fabrication of single phase material using conventional methods is the main drawback of this group of ceramics. In order to overcome this problem material can be deposited as a thin layer e. g. in the pulsed laser deposition process. In this paper, the effect of laser wavelength and energy density of nanosecond Nd:YAG laser on the WBx-type layers were analyzed using wavelengths 355 and 1064 nm with the energy density of laser beam from 1.7 to 5 J/cm2 and from 1.7 to 9.3 J/cm2, respectively. The WB2.5 and WB4.5 targets synthesized in Spark Plasma Sintering process were used and the layers were deposited onto Si (100) substrate heated to a temperature of 570°C. Layers’ microstructure were analyzed using X-ray diffraction and scanning electron microscope equipped with energy dispersive X-ray spectrometer. Change of laser wavelength and energy density resulted in variations of the chemical composition and morphology of deposited layers. Finally, W2B-βWB, αWB-WB-WB3 and WB3, and boron layers were deposited wherein WB3 structure is formed in a wide range of laser fluences and at both investigated wavelength. Next, WB3 layers were investigated in the indentation test at a load of 5-30 mN and its hardness was up to 50 ± 10 GPa.
ACCEPTED MANUSCRIPT 1. Introduction
PT E
D
MA
NU
SC
RI
PT
The modern cutting tools often work in harsh conditions, thus in order to increase their performance new group of materials with high hardness, fracture toughness, and chemically inert is desired. Materials which meet those requirements have a high density of valence electrons, covalent bonds and fill up its valence shell. Nowadays, transition metal borides are widely discussed as a group of materials possessing very high hardness, among them tungsten borides are one of the most interesting. However, theoretical calculations show mechanical properties vary significantly depending on its chemical and phase composition [1-5]. One of the first theoretical calculations concerning WBx type compounds [1] indicated high bulk modulus (BM) and hardness (H) of WB2 (BM = 372 GPa, H = 19.9 GPa) and WB4 (BM = 304 GPa, H = 38.4 GPa). Good mechanical properties of WB4 and WB2 were confirmed by Wang et al. [2] (BM = 292.7 – 324.3 GPa, H = 41.1 – 42.2 GPa) and by Zhao et al. [3] (BM = 338 GPa, H = 39.4 GPa). Also, a WB3 structure with a hardness of 38.3 GPa was presented as a superhard material in work [4]. In a comprehensive study of tungsten borides [5] authors reported that the hardest WBx-type structures belong to WB2 (H = 39.7 GPa) and WB3 (H = 36.9 GPa). Among mentioned compounds W2B [6, 7], WB [6, 8, 9, 11], WB2 [6, 14, 15], W2B5 [6, 7], WB3 [10] and WB4 [1, 13, 16] were already synthesized, however only few papers [10-11, 1416] have concerned on fabrication of layers. Moreover, none of the aforementioned papers concerns on the effect of deposition parameters on chemical composition, microstructure, and mechanical properties of layers. The pulsed laser deposition (PLD) has been successfully used for deposition of pure boron [17] and tungsten [18, 19]. In this work, we accomplished deposition of the WBx -type layers by nanosecond laser pulse. Due to the strict control of deposition parameters, pulsed laser deposition (PLD) allowed depositing materials having a complex chemical composition and high melting point. So far, merely few papers concerning tungsten borides layers deposited by laser pulse [10, 12, 16] were published. Moreover, in Ref. [10, 12] only single WB3 layer deposited with 355 nm wavelength and nanosecond irradiation, and in Ref. the [16] WB4 layers deposited with 527 nm wavelength and femtosecond irradiation were discussed. This work will focus on analyzing the effect of laser wavelength and energy density (fluence) on the chemical and phase composition, morphology and hardness of WB-type layers. 2. Materials, Experiment and Characterization
AC
CE
2.1. Targets During pulsed laser deposited two kinds of targets named WB2.5 and WB4.5 were used. In these targets boron to tungsten molar ratio was 2.5 and 4.5, respectively. Prompted by the fact that hardness of WBx -type compounds increases with boron content we aimed to deposit layers with boron to tungsten molar ratio at least two. Due to lack of commercially available targets with B/W ratio over 2, the targets were synthesized from pure powders of boron (625 mesh, 99.7% purity, Sigma Aldrich) and tungsten (625 mesh, 99.9% purity, Sigma Aldrich) by Spark Plasma Sintering. The WB2.5 target was composed of 57% WB phase and 43% WB2 phase, and the WB4.5 target was composed of 83% WB2 phase and 17% WB3 phase. Isolated grains of unsynthesized boron were embedded in the targets. Details of the sintering process as well as properties of the targets are presented elsewhere [10, 11]. 2.2. Pulsed laser deposition
ACCEPTED MANUSCRIPT
D
Laser Fluence spot size [J/cm2] [mm2] 3.5 9.3 5 1.7 5 1.7 9.3 5 1.7 5 1.7
AC
CE
PT E
Tab. 1. Deposition parameters Target Wavelength Pulse [nm] energy [J] WB2.5 1064 0.39 0.21 0.08 355 0.21 0.07 WB4.5 1064 0.39 0.21 0.08 355 0.21 0.07
MA
NU
SC
RI
PT
The deposition process occurred in a vacuum chamber evacuated to a working pressure of 2 ÷ 5·10-6 mbar. The material was evaporated using Q-switched Nd:YAG laser (Quantel YG 981 E10, 10 Hz pulses) emitting at the fundamental (1064 nm) and third (355 nm) harmonic producing 10 and 9 ns pulses, respectively. The laser beam was at an angle of 45° to the normal to the target surface and the target was continuously rotated to avoid deep drilling on its surface. In this paper we use the term “fluence” which means the energy of laser pulse reaching the target divided by the spot size. The energy of impulses was controlled by Q-switch attenuator and the spot size was 3.5 mm2 for each of laser wavelength. Worth to mention the total transmission coefficient of lens and quartz window was 0.84 and 0.83 at 355 and 1064 nm wavelength, respectively. In preliminary experiments, the deposition rate was measured using tuneable quartz oscillator (Maxtek Thickness Monitor TM-350/400, 6 MHz). For WB2.5 and WB4.5 targets, the deposition rate did not vary and therefore for both targets deposition process was carried under the same conditions. In the case of 1064 nm radiation the laser fluence was 1.7, 5 and 9.3 J/cm2 and in the case of 355 nm radiation, the fluence was 1.7 and 5 J/cm2. Fluence 1.7 J/cm2 was just above the ablation threshold and fluences 5 and 9.3 J/cm2 were maximum available fluence at 355 and 1064 nm radiation, respectively. It is worth noting that selected fluences are relatively low and as such should prevent phase explosion to take place (typical threshold values are above 10 J/cm2, e. g. for niobium and graphite are 15 and 22 J/cm2, respectively [20]). All layers were deposited on Si (100) substrate placed 42 mm away from the target. At the beginning of the PLD experiment, an attempt to deposit a layer at room temperature had been made, but as-deposited layer was amorphous and had very low adhesion. For this reason, layers were deposited on the substrate heated to 570 °C, the maximum available temperature of a substrate holder. The deposition time was 90 min in the case of layers deposited with fluence 9.3 and 5 J/cm2 at 1064 and 355 nm, respectively. Other layers were deposited 180 min. Deposition parameters are summarised in tab. 1.
Substrate temperature [°C] 570
Time [min]
Coating thickness [µm]
90 180 180 90 180 90 180 180 90 180
0.9 ± 0.1 0.9 ± 0.1 ≈ 0.03 0.9 ± 0.1 0.8 ± 0.1 0.9 ± 0.1 0.9 ± 0.1 ≈ 0.04 0.9 ± 0.1 0.8 ± 0.1
2.3. Characterization The thickness of deposited layers was measured by scanning profilometer T8000 (CSM Instruments) as a distance between surface of the substrate and the layer neglecting debris. The characterization of morphology of the deposited layers was made by scanning electron microscope (SEM) Hitachi SU8000 equipped with X-ray energy-dispersive spectrometer (EDS). The crystalline structure was determined on the basis of X-Ray Diffraction (XRD) with the use of diffractometer (Bruker D8 discover with Cu radiation source, λ = 1.5418 Å). To avoid the contribution of the substrate the 2θ scans with source fixed at 8° position were done. For this configuration the maximum angular range was 20–72°. The crystalline structure was determined on the basis of X-Ray Diffraction (XRD) with the use of diffractometer. The hardness of the selected layers was measured using Ultra-low Load Indentation system (CSM Instruments). For all measurements, Vickers-shaped diamond indenter and the indentation load from 5
ACCEPTED MANUSCRIPT to 30 mN was used. At each loading at least 8 indentations were done in areas with low density of droplets. Due to the high roughness of PLD coatings before each nanoindentation, preloading was applied in order to determine the initial level of the examined area. Young Modulus and hardness were specified using the Oliver-Pharr method.
3. Results and discussion
PT E
D
MA
NU
SC
RI
PT
3.1. Surface morphology and chemical composition In the case of deposition from the WB2.5 target with a wavelength of 1064 nm, the morphology of layer changes significantly depending on laser fluence. The layer deposited with high fluence 9.3 J/cm2 had wrinkled surface and droplets on its surface (fig. 1 a, b). The appearance of the layer may indicate low adhesion to the underlying substrate. The lower fluence of 5 J/cm2 caused the formation of the smoother layer, though increased the number and size of the droplets (fig. 1 c, d). Further decrease of fluence to 1.7 J/cm2 ultimately limited the number of deposited droplets (fig. 1 e, f), but it also significantly decreased the deposition rate (Tab. 1). Changing the wavelength to 355 nm resulted more droplets on the layer’s surface (fig. 2). Even though fluence has been reduced to 1.7 J/cm2, a lot of droplets were deposited on the surface of the layer.
AC
CE
Fig. 1. SEM images of layers deposited from the WB2.5 target at 1064 nm radiation with fluence 9.3 J/cm2 (a , b), 5 J/cm2 (c, d), and 1.7 J/cm2 (e, f). In image f) particle marked with A is composed of pure boron and B is composed of both boron and tungsten.
Fig. 2. SEM images of layers deposited from the WB2.5 target at 355 nm radiation with fluence 5 J/cm2 (a, b) and 1.7 J/cm2 (c, d) Layers originated from the WB4.5 target had better adhesion, even layer deposited at 1064 nm with fluence 9.3 J/cm2 did not spall (fig. 3). As one could expect the number of droplets decreased
ACCEPTED MANUSCRIPT
SC
RI
PT
with lower fluence. Moreover, at 1064 nm radiation and fluence 1.7 J/cm2, droplets were hardly observed (fig. 3 e). Again, as in the case of layers deposited from the WB2.5 target, the number and the size of droplets was higher for 355 nm wavelength (fig. 4).
PT E
D
MA
NU
Fig. 3. SEM images of layers deposited from the WB4.5 target at 1064 nm radiation with fluence 9.3 J/cm2 (a , b), 5 J/cm2 (c, d) ), and 1.7 J/cm2 (e, f)
CE
Fig. 4. SEM images of layers deposited from WB4.5 target at 355 nm radiation with fluence 5 J/cm2 (a, b) and 1.7 J/cm2 (c, d)
AC
With the exception of layers deposited with fluence just above ablation threshold at 1064 nm radiation, the chemical composition was the same for smooth part of the layers and droplets on its surface (fig. 5). Moreover, boron content in deposited layers was higher at lower fluences and at 355 nm radiation (fig. 6). It should be kept in mind that EDS method poorly deals with quantitative measurements of light elements including boron, thus presented results could be used just for comparison.
NU
SC
RI
PT
ACCEPTED MANUSCRIPT
AC
CE
PT E
D
MA
Fig. 5. SEM-EDS analysis of layers deposited from the WB2.5 (left) and WB4.5 (right) targets at 1064 nm radiation with fluence 5 J/cm2; SEM image of the investigated area (a, c), EDS spectrum of selected points at acceleration voltage 5 kV (b, d)
Fig. 6. SEM-EDS analysis of layers deposited from the WB2.5 target at 1064 nm radiation with fluence 9.3J/cm2 (a) and 5 J/cm2 (b); at 355 nm radiation with fluence 5 J/cm2 (c) The layer deposited from the WB2.5 target at 1064 nm radiation with fluence 1.7 J/cm2 is composed of two types of particles: composed of both boron and tungsten, and composed of pure boron (see fig. 1 f). The layer deposited from the WB4.5 target at 1064 nm radiation with fluence 1.7 J/cm2 is a special case because SEM-EDS analysis shows that it is composed of pure boron (fig. 7). Moreover, the image of the inclined sample indicates that the deposited grains grow perpendicularly to the substrate surface (fig. 8).
RI
PT
ACCEPTED MANUSCRIPT
PT E
D
MA
NU
SC
Fig. 7. SEM-EDS spectrum of the layer deposited from the WB4.5 target at 1064 nm with fluence 1.7 J/cm2
CE
Fig. 8. SEM image of the layer deposited from the WB4.5 target at 1064 nm with fluence 1.7 J/cm2, the surface inclined by 35°
AC
3.2. XRD results The layer deposited from WB2.5 target at 1064 nm radiation with fluence 9.3 J/cm2 was composed of tetragonal W2B and rhombohedral β-WB phase (fig. 9 a). Calculated cell parameters of W2B phases a = 5.592 Å and c = 4.750 Å were similar to theoretical one (a = 5.571 Å; c = 4.771 Å [5]). Calculated cell parameters of β-WB phases a = 3.177 Å; b = 8.478 Å; c = 3.096 Å were close similar to theoretical ones (a = 3.177 Å; b = 8.485 Å; c = 3.101 Å [5]). Fluence reduction to 5 J/cm2 leads to the formation of α-WB, β-WB and WB3 phases (fig. 9 b). Because of the small number of diffraction maxima the cell parameters were not calculated. It was unable to register XRD spectrum of the layer deposited with fluence 1.7 J/cm2.
SC
RI
PT
ACCEPTED MANUSCRIPT
NU
Fig. 9. XRD spectra of layers deposited from the WB2.5 target at 1064 nm with fluence 9.3 J/cm2 (a) and 5 J/cm2 (b). Solid circles – W2B phase, open circles – βWB phase, star – αWB phase, hearts – WB3 phase, asterisk – unidentified peak. The corresponding Miller Indices are given above each peak.
AC
CE
PT E
D
MA
Changing the radiation to 355 nm caused the formation of WB3 phase for both 5 and 1.7 J/cm2 fluence (fig. 12). The calculated cell parameters for layer deposited with fluence 5 J/cm2 are: a = 5.228 Å and c = 6.120 Å, for layer deposited with fluence 1.7 J/cm2 are a = 5.183 Å and c = 6.224 Å, while theoretical cell parameters are a = 5.2 Å and c = 6.313 Å. The difference between calculated and theoretical cell parameters is higher in case of fluence 5 J/cm2. Also, diffraction lines are broader at higher fluence. Both features can be a result of a higher residual strain and a larger number of defects in the crystalline structure.
Fig. 10. XRD spectra of layers deposited from the WB2.5 target at 355 nm radiation with fluence 5 J/cm2 (a) and 1.7 J/cm2 (b). Hearts represent the diffraction peaks of the WB3 phase, the corresponding Miller Index is given above each peak. All layers deposited from the WB4.5 target consist mainly of WB3 phase (fig. 11). However, on the XRD spectra some unidentified peaks, indicated with asterisk, can be observed. In the case of the
ACCEPTED MANUSCRIPT
CE
PT E
D
MA
NU
SC
RI
PT
layer deposited at 1064 nm radiation with fluence 9.3 J/cm2 the high broadening of diffraction lines was observed (fig. 11 a). Moreover, its calculated cell parameters vary from theoretical ones more than in case of the layer deposited with fluence 5 J/cm2. indicates small size of crystallites and high content of defects in the material. It was also noticed that layers deposited with 355 nm radiation have elevated background signal for the 2θ range from 20° to 50°.
AC
Fig. 11. XRD spectra of layers deposited from the WB4.5 target and calculated cell parameters of WB3 phase. The ablation at 1064 nm radiation with fluence 9.3 J/cm2 (a) and 5 J/cm2 (b). The ablation at 355 nm radiation with fluence 5 J/cm2 (c) and 1.7 J/cm2 (d). Hearts represent the diffraction peaks of the WB3 phase, the corresponding Miller Index is given above each peak. 3.3. Nanoindentation test The Young modulus did not vary significantly for all investigated layers and was 479.8 ± 28.5 GPa. The results of nanohardness are presented in fig. 12. Nanoindentation of the layer deposited from the WB2.5 target at 1064 nm radiation with fluence 9.3 J/cm2 was impossible due to low adhesion.
PT
ACCEPTED MANUSCRIPT
SC
RI
Fig. 12. Results of nanoindentation test. The layer deposited from the WB2.5 target at 355 nm with fluence 5 J/cm2 – blue line. The layer deposited from the WB4.5 target at 355 nm with fluence 5 J/cm2 – green line, and the layer deposited at 1064 nm with fluence 9.3 J/cm2 – red line. Silicon substrate – dashed line
MA
NU
The uncertainty of the nanoindentation hardness is due to the high roughness of the material. Despite the presence of droplets deposited layers were very hard. The hardness of layers deposited from WB4.5 target was up to 50 ± 10 GPa under the load 30 mN. Investigated layer deposited from the WB2.5 target at 355 nm with fluence 5 J/cm2 had slightly lower hardness 36 ± 10 GPa. Presented results may be even underestimated because the nanoindentation depth was of about 200 nm. Considering the layer thickness of 0.9 ± 0.1 µm and the substrate hardness of 10.7 ± 0.7 GPa, presented results are affected by the substrate hardness. The hardness of deposited layers is similar to the hardness of WB4 layers deposited by femtosecond laser [16]. It is worth noting that despite the presence of droplets on the surface of PLD layers, they are still very hard.
D
4. Discussion
AC
CE
PT E
Morphological defects, including macro-particles, ‘doughnut rings’ and ‘scars’ are commonly observed for films deposited by various Physical Vapour Deposition techniques, that is well described in Dearnley [20]. Also the layers reported in this manuscript have droplets on their surface. More droplets were observed onto the surface of layers deposited with higher fluence and 355 nm radiation (fig. 13). The higher fluence increases the temperature of the irradiated targets’ area and could result in the phase explosion [21-23]. The absorption of the laser radiation in plasma plume is lower at shorter wavelength, so more energy reaches the target’s surface [25]. Moreover, for WBx-type compounds the absorption is higher at 355 nm than at 1064 nm radiation [26], thus smaller volume of the target absorbs the energy and is heated to higher temperature.
PT
ACCEPTED MANUSCRIPT
RI
Fig. 13. Schematic view showing the trend between the number of droplets and fluence in the case of layers deposited from WB2.5 target (dashed line) and WB4.5 target (solid line) at radiation of 355 nm (blue line) and 1064 nm (red line).
AC
CE
PT E
D
MA
NU
SC
The difference of chemical composition between the target and the deposited layers was observed. Moreover, chemical and phase composition of the deposited layers varies depending on the laser wavelenght and fluence (fig. 14). This phenomena can be explained by the selective evaporation of the target’s components [24]. During the ablation of multicomponent target at lower fluences the energy absorbed by the target can be too low to evaporate the component having higher melting point and molar mass [12, 24]. In the case of WB2.5 and WB4.5 targets there is a significant difference in physical properties between the target’s components. Thus, boron as the element more volatile and having lower melting point (melting point of boron and tungsten is 2349 and 3695 K, respectively) will evaporate first. The effect of selective evaporation is more pronounced at 1064 nm radiation as the energy of laser pulse is absorbed in greater volume. The differences between layers deposited with 1064 and 355 nm can be explainned by diferent characteristics of plasma plume. In the case of the ablation with 1064 nm the plasma plume absorbtion is higher what cause wider radial expanssion of boron atoms and ions [12] resulting from its smaller than tungsten atomic mass. Moreover, boron atoms (and ions) are scattered in collisions witch much heavier tungsten. Therefore, ablation at 1064 nm radiation and with higher fluences will result in the decrease of boron atoms deposited on the layer.
Fig. 14. Schematic view showing the trend between the chemical composition and fluence of layers deposited from WB2.5 target (dashed line) and WB4.5 target (solid line) at radiation of 355 nm (blue line) and 1064 nm (red line).
ACCEPTED MANUSCRIPT 5. Conclusions
PT
In this work WBx-type layers deposited by nanosecond laser pulse at 355 and 1064 nm radiation were investigated. The W2B-βWB, αWB-WB-WB3 and WB3, and boron layers were deposited wherein WB3 structure is formed in a wide range of laser fluences and at both investigated wavelength. Thus, among tungsten borides formed in PLD process the WB3 phase is probably the most common one. The content of boron in deposited layer is higher in the case of ablation at 355 nm radiation and increases at lower fluences. The WB3 layers were investigated in the indentation test at a load of 5-30 mN and its hardness varied from 36 ± 10 GPa to 50 ± 10 GPa. Presented results confirm the possibility of using WB3 as a superhard material.
AC
CE
PT E
D
MA
NU
SC
RI
Acknowledgement This work was supported by the NCN (National Science Centre) Research Project: 2015/19/N/ST8/03928. Part of investigations was carried out with the use of CePT infrastructure financed by the European Union – the European Regional Development Fund within the Operational Programme "Innovative economy" for 2007-2013. Authors thank prof. Kucharski for nanoindentation test.
ACCEPTED MANUSCRIPT
AC
CE
PT E
D
MA
NU
SC
RI
PT
Bibliography 1. Q. Gu, G. Krauss, W. Steurer, Transition metal borides: superhard versus ultraincompressible, Adv. Mater. 20 (2008) 3620-3626 2. M. Wang, Y. Li, T. Cui, Y. Ma, G. Zou, Origin of hardness in WB4 and its implications for ReB2, TaB4, MoB4, TcB4, and OsB4, Appl. Phys. Lett. 93 (2008) 101905 3. E. Zhao, J. Meng, Y. Ma, Z. Wu, Phase stability and mechanical properties of tungsten borides from first principles calculations, Phys. Chem. Chem. Phys. 12 (2010) 13158-1316 4. Y. Liang, Y. Gou, X. Yuan, Z. Zhong, W. Zhang, Unexpectedly hard and highly stable WB3 with a noncompact structure, Chemical Physics Letters 580 (2013) 48-52 5. X. -Y. Cheng, X. –Q. Chen, D. –Z. Li, Y. –Y. Li, Computational materials Discovery: the case of the W-B system, Acta Cryst. C C70 (2014) 85-103 6. R. Kiessling, The Borides of Some Transition Elements, Acta Chem. Scand. 4 (1950) 209-227 7. L. Yang, Z. Kan, Y. Zeng, X. Wang, S. Du, C. Tao, P. Ren, X. Cui, M. Wen, Boron doped bcc -W films: Achieving excellent mechanical properties and tribological performance by regulating substrate bias voltage, Appl. Surf, Sci. 423 (2017) 275-282 8. Q. Li, D. Zhou, W. Zheng, Y. Ma, C. Chen, Global structural optimization of tungsten borides, Phys. Rev. Lett. 110 (2013) 136403 9. S. Okada, K. Kudou, T. Lundström, Preparations and Some Properties of W 2B, δ-WB and WB2 Crystals from High-Temperature Metal Solutions, Jpn. J. Appl. Phys. 34 (1995) 226-231 10. T. Mościcki, J. Radziejewska, J. Hoffman, J. Chrzanowska, N. Levintant- Zayonts, D. Garbiec, Z. Szymański, WB2 to WB3 phase change during reactive spark plasma sintering and pulsed laser ablation/deposition processes, Ceram. Int. 41 (7) (2015) 8273-8281 11. J. Chrzanowska, Ł. Kurpaska, M. Giżyński, J. Hoffman, Z. Szymański, T. Mościcki, Fabrication and characterization of superhard tungsten boride layers deposited by radio frequency magnetron sputtering, Ceram. Int. 10 (2016) 12221-12230 12. T. Mościcki, Differences in nanosecond laser ablation and deposition o tungsten, boron and WB2/B composite due to optical properties, Int. J. Opt. 5438721 (2016) 13. R. Mohammadi, A. T. Lech, M. Vie, B. E. Weaver, M. T. Yeung, S. H. Tolbert, R. Kaner, Tungsten tetraboride, an inexpensive superhard material, Proc. Natl. Acad. Sci. USA 108 (2011) 10958-10962 14. C. Jiang, Z. Pei, Y. Liu, J. Xiao, J. Gong, C. Sun, Preparation and characterization of superhard AlB2-type WB2 nanocomposite coatings, Phys. Status Solidi A 210 (6) (2013) 12211227 15. Y. M. Liu, C. L. Jiang, Z. L. Pei, H. Lei, J. Gong, C. Sun, Microstructure and properties of AlB2- type WB2 thin films deposited by direct- current magnetron sputtering, Surf. Coat. Technolog. 245 (2014) 108-116 16. J. V. Rau, A. Latini, R. Teghil, A. De Bonis, M. Fosca, R. Caminiti, V. R. Albertini, Superhard tungsten tetraboride films prepared by pulsed laser deposition method, Appl. Mater. Interfaces 3 (2011) 3738-3743 17. D. Dellasega, V. Russo, A. Pezzoli, C. Conti, N. Lecis, E. Besozzi, M. Beghi, C. E. Bottani, M. Passoni, Boron films produced by high energy Pulsed Laser Deposition, Mater. Des. 134 (2017) 35-43 18. D. Dellasega, G. Merlo, C. Conti, C. E. Bottani, M. Passoni, Nanostructured and amorphouslike tungsten films grown by pulsed laser deposition, J. Appl. Phys. 112 (2012) 084328 19. A. T. T. Mostako, C. V. S. Rao, A. Khare, Mirrorlike pulsed laser deposited tungsten thin film, Rev. Sci. Instrum. 82 (2011) 013101 20. P. A. Dearnley, Introduction to Surface Engineering, Cambridge University Press, New York, 2017, pp. 179-182, and 187-189 21. N. M. Bulgakova, A. V. Bulgakov, Pulsed laser ablation of solids: transition from normal vaporization to phase explosion, Appl. Phys. A 73 (2001) 199-208 22. W. O. Siew, W. K. Lee, H. Y. Wong, T. K. Yong, S. S. Yap, T. Y. Tou, Investigation of droplet formation in pulsed Nd: YAG laser deposition of metals and silicon, Appl. Phys. A 101 (2010) 627-632
ACCEPTED MANUSCRIPT
AC
CE
PT E
D
MA
NU
SC
RI
PT
23. J. Chrzanowska, J. Hoffman, P. Denis, M. Giżyński, T. Mościcki, The effect of process parameters on rhenium diboride films deposited by PLD, Surf. Coat. Technol. 277 (2015) 1522 24. C. Klamt, A. Dittrich, B. Jaquet, C. Eberl, F. Dӧring, H. U. Krebs, Largest possible deviations from stoichiometry transfer during pulsed laser deposition, Appl. Phys. A 122 (2016) 701 25. T. Mościcki, J. Hoffman, J. Chrzanowska, The absorption and radiation of a tungsten plasma plume during nanosecond laser ablation, Phys. Plasmas 22 (2015) 103303 26. M. Maździarz, T. Mościcki, Structural, mechanical and optical properties of potentially superhard WBx polymorphs from first principle calculations, Mater. Chem. Phys. 179 (2016) 92-102
ACCEPTED MANUSCRIPT Highlights:
PLD as a suitable technique for the deposition of tungsten triboride layers
PLD layers composed of tungsten triboride have a hardness from 36 to 46.8 GPa
Selective evaporation of boron from the composite WB4.5 target irradiated with 1064
AC
CE
PT E
D
MA
NU
SC
RI
PT
nm at low fluences
Justyna Chrzanowska-Giżyńska, Piotr Denis, Jacek Hoffman, Maciej Giżyński, Tomasz Mościcki, Dariusz Garbiec, Zygmunt Szymański PII: DOI: Reference:
S0257-8972(17)31263-X https://doi.org/10.1016/j.surfcoat.2017.12.040 SCT 22958
To appear in:
Surface & Coatings Technology
Received date: Revised date: Accepted date:
21 August 2017 12 December 2017 16 December 2017
Please cite this article as: Justyna Chrzanowska-Giżyńska, Piotr Denis, Jacek Hoffman, Maciej Giżyński, Tomasz Mościcki, Dariusz Garbiec, Zygmunt Szymański , Tungsten borides layers deposited by a nanosecond laser pulse. The address for the corresponding author was captured as affiliation for all authors. Please check if appropriate. Sct(2017), https://doi.org/10.1016/j.surfcoat.2017.12.040
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.
ACCEPTED MANUSCRIPT Tungsten borides layers deposited by a nanosecond laser pulse Justyna Chrzanowska-Giżyńska*1, Piotr Denis1, Jacek Hoffman1, Maciej Giżyński2,Tomasz Mościcki1, Dariusz Garbiec3, Zygmunt Szymański1 Institute of Fundamental Technological Research, Polish Academy of Sciences, Pawińskiego 5B, 02106 Warsaw, Poland 2 Faculty of Material Science and Engineering of Warsaw University of Technology, Wołoska 141, 02-507 Warsaw, Poland 3 Metal Forming Institute, Jana Pawla II 14, 61-139 Poznan, Poland * corresponding author e-mail: [email protected], [email protected]
PT
1
AC
CE
PT E
D
MA
NU
SC
RI
Abstract Tungsten borides belong to the group of potentially superhard materials which hardness could be compared to cubic boron nitride and diamond. However, difficulty in fabrication of single phase material using conventional methods is the main drawback of this group of ceramics. In order to overcome this problem material can be deposited as a thin layer e. g. in the pulsed laser deposition process. In this paper, the effect of laser wavelength and energy density of nanosecond Nd:YAG laser on the WBx-type layers were analyzed using wavelengths 355 and 1064 nm with the energy density of laser beam from 1.7 to 5 J/cm2 and from 1.7 to 9.3 J/cm2, respectively. The WB2.5 and WB4.5 targets synthesized in Spark Plasma Sintering process were used and the layers were deposited onto Si (100) substrate heated to a temperature of 570°C. Layers’ microstructure were analyzed using X-ray diffraction and scanning electron microscope equipped with energy dispersive X-ray spectrometer. Change of laser wavelength and energy density resulted in variations of the chemical composition and morphology of deposited layers. Finally, W2B-βWB, αWB-WB-WB3 and WB3, and boron layers were deposited wherein WB3 structure is formed in a wide range of laser fluences and at both investigated wavelength. Next, WB3 layers were investigated in the indentation test at a load of 5-30 mN and its hardness was up to 50 ± 10 GPa.
ACCEPTED MANUSCRIPT 1. Introduction
PT E
D
MA
NU
SC
RI
PT
The modern cutting tools often work in harsh conditions, thus in order to increase their performance new group of materials with high hardness, fracture toughness, and chemically inert is desired. Materials which meet those requirements have a high density of valence electrons, covalent bonds and fill up its valence shell. Nowadays, transition metal borides are widely discussed as a group of materials possessing very high hardness, among them tungsten borides are one of the most interesting. However, theoretical calculations show mechanical properties vary significantly depending on its chemical and phase composition [1-5]. One of the first theoretical calculations concerning WBx type compounds [1] indicated high bulk modulus (BM) and hardness (H) of WB2 (BM = 372 GPa, H = 19.9 GPa) and WB4 (BM = 304 GPa, H = 38.4 GPa). Good mechanical properties of WB4 and WB2 were confirmed by Wang et al. [2] (BM = 292.7 – 324.3 GPa, H = 41.1 – 42.2 GPa) and by Zhao et al. [3] (BM = 338 GPa, H = 39.4 GPa). Also, a WB3 structure with a hardness of 38.3 GPa was presented as a superhard material in work [4]. In a comprehensive study of tungsten borides [5] authors reported that the hardest WBx-type structures belong to WB2 (H = 39.7 GPa) and WB3 (H = 36.9 GPa). Among mentioned compounds W2B [6, 7], WB [6, 8, 9, 11], WB2 [6, 14, 15], W2B5 [6, 7], WB3 [10] and WB4 [1, 13, 16] were already synthesized, however only few papers [10-11, 1416] have concerned on fabrication of layers. Moreover, none of the aforementioned papers concerns on the effect of deposition parameters on chemical composition, microstructure, and mechanical properties of layers. The pulsed laser deposition (PLD) has been successfully used for deposition of pure boron [17] and tungsten [18, 19]. In this work, we accomplished deposition of the WBx -type layers by nanosecond laser pulse. Due to the strict control of deposition parameters, pulsed laser deposition (PLD) allowed depositing materials having a complex chemical composition and high melting point. So far, merely few papers concerning tungsten borides layers deposited by laser pulse [10, 12, 16] were published. Moreover, in Ref. [10, 12] only single WB3 layer deposited with 355 nm wavelength and nanosecond irradiation, and in Ref. the [16] WB4 layers deposited with 527 nm wavelength and femtosecond irradiation were discussed. This work will focus on analyzing the effect of laser wavelength and energy density (fluence) on the chemical and phase composition, morphology and hardness of WB-type layers. 2. Materials, Experiment and Characterization
AC
CE
2.1. Targets During pulsed laser deposited two kinds of targets named WB2.5 and WB4.5 were used. In these targets boron to tungsten molar ratio was 2.5 and 4.5, respectively. Prompted by the fact that hardness of WBx -type compounds increases with boron content we aimed to deposit layers with boron to tungsten molar ratio at least two. Due to lack of commercially available targets with B/W ratio over 2, the targets were synthesized from pure powders of boron (625 mesh, 99.7% purity, Sigma Aldrich) and tungsten (625 mesh, 99.9% purity, Sigma Aldrich) by Spark Plasma Sintering. The WB2.5 target was composed of 57% WB phase and 43% WB2 phase, and the WB4.5 target was composed of 83% WB2 phase and 17% WB3 phase. Isolated grains of unsynthesized boron were embedded in the targets. Details of the sintering process as well as properties of the targets are presented elsewhere [10, 11]. 2.2. Pulsed laser deposition
ACCEPTED MANUSCRIPT
D
Laser Fluence spot size [J/cm2] [mm2] 3.5 9.3 5 1.7 5 1.7 9.3 5 1.7 5 1.7
AC
CE
PT E
Tab. 1. Deposition parameters Target Wavelength Pulse [nm] energy [J] WB2.5 1064 0.39 0.21 0.08 355 0.21 0.07 WB4.5 1064 0.39 0.21 0.08 355 0.21 0.07
MA
NU
SC
RI
PT
The deposition process occurred in a vacuum chamber evacuated to a working pressure of 2 ÷ 5·10-6 mbar. The material was evaporated using Q-switched Nd:YAG laser (Quantel YG 981 E10, 10 Hz pulses) emitting at the fundamental (1064 nm) and third (355 nm) harmonic producing 10 and 9 ns pulses, respectively. The laser beam was at an angle of 45° to the normal to the target surface and the target was continuously rotated to avoid deep drilling on its surface. In this paper we use the term “fluence” which means the energy of laser pulse reaching the target divided by the spot size. The energy of impulses was controlled by Q-switch attenuator and the spot size was 3.5 mm2 for each of laser wavelength. Worth to mention the total transmission coefficient of lens and quartz window was 0.84 and 0.83 at 355 and 1064 nm wavelength, respectively. In preliminary experiments, the deposition rate was measured using tuneable quartz oscillator (Maxtek Thickness Monitor TM-350/400, 6 MHz). For WB2.5 and WB4.5 targets, the deposition rate did not vary and therefore for both targets deposition process was carried under the same conditions. In the case of 1064 nm radiation the laser fluence was 1.7, 5 and 9.3 J/cm2 and in the case of 355 nm radiation, the fluence was 1.7 and 5 J/cm2. Fluence 1.7 J/cm2 was just above the ablation threshold and fluences 5 and 9.3 J/cm2 were maximum available fluence at 355 and 1064 nm radiation, respectively. It is worth noting that selected fluences are relatively low and as such should prevent phase explosion to take place (typical threshold values are above 10 J/cm2, e. g. for niobium and graphite are 15 and 22 J/cm2, respectively [20]). All layers were deposited on Si (100) substrate placed 42 mm away from the target. At the beginning of the PLD experiment, an attempt to deposit a layer at room temperature had been made, but as-deposited layer was amorphous and had very low adhesion. For this reason, layers were deposited on the substrate heated to 570 °C, the maximum available temperature of a substrate holder. The deposition time was 90 min in the case of layers deposited with fluence 9.3 and 5 J/cm2 at 1064 and 355 nm, respectively. Other layers were deposited 180 min. Deposition parameters are summarised in tab. 1.
Substrate temperature [°C] 570
Time [min]
Coating thickness [µm]
90 180 180 90 180 90 180 180 90 180
0.9 ± 0.1 0.9 ± 0.1 ≈ 0.03 0.9 ± 0.1 0.8 ± 0.1 0.9 ± 0.1 0.9 ± 0.1 ≈ 0.04 0.9 ± 0.1 0.8 ± 0.1
2.3. Characterization The thickness of deposited layers was measured by scanning profilometer T8000 (CSM Instruments) as a distance between surface of the substrate and the layer neglecting debris. The characterization of morphology of the deposited layers was made by scanning electron microscope (SEM) Hitachi SU8000 equipped with X-ray energy-dispersive spectrometer (EDS). The crystalline structure was determined on the basis of X-Ray Diffraction (XRD) with the use of diffractometer (Bruker D8 discover with Cu radiation source, λ = 1.5418 Å). To avoid the contribution of the substrate the 2θ scans with source fixed at 8° position were done. For this configuration the maximum angular range was 20–72°. The crystalline structure was determined on the basis of X-Ray Diffraction (XRD) with the use of diffractometer. The hardness of the selected layers was measured using Ultra-low Load Indentation system (CSM Instruments). For all measurements, Vickers-shaped diamond indenter and the indentation load from 5
ACCEPTED MANUSCRIPT to 30 mN was used. At each loading at least 8 indentations were done in areas with low density of droplets. Due to the high roughness of PLD coatings before each nanoindentation, preloading was applied in order to determine the initial level of the examined area. Young Modulus and hardness were specified using the Oliver-Pharr method.
3. Results and discussion
PT E
D
MA
NU
SC
RI
PT
3.1. Surface morphology and chemical composition In the case of deposition from the WB2.5 target with a wavelength of 1064 nm, the morphology of layer changes significantly depending on laser fluence. The layer deposited with high fluence 9.3 J/cm2 had wrinkled surface and droplets on its surface (fig. 1 a, b). The appearance of the layer may indicate low adhesion to the underlying substrate. The lower fluence of 5 J/cm2 caused the formation of the smoother layer, though increased the number and size of the droplets (fig. 1 c, d). Further decrease of fluence to 1.7 J/cm2 ultimately limited the number of deposited droplets (fig. 1 e, f), but it also significantly decreased the deposition rate (Tab. 1). Changing the wavelength to 355 nm resulted more droplets on the layer’s surface (fig. 2). Even though fluence has been reduced to 1.7 J/cm2, a lot of droplets were deposited on the surface of the layer.
AC
CE
Fig. 1. SEM images of layers deposited from the WB2.5 target at 1064 nm radiation with fluence 9.3 J/cm2 (a , b), 5 J/cm2 (c, d), and 1.7 J/cm2 (e, f). In image f) particle marked with A is composed of pure boron and B is composed of both boron and tungsten.
Fig. 2. SEM images of layers deposited from the WB2.5 target at 355 nm radiation with fluence 5 J/cm2 (a, b) and 1.7 J/cm2 (c, d) Layers originated from the WB4.5 target had better adhesion, even layer deposited at 1064 nm with fluence 9.3 J/cm2 did not spall (fig. 3). As one could expect the number of droplets decreased
ACCEPTED MANUSCRIPT
SC
RI
PT
with lower fluence. Moreover, at 1064 nm radiation and fluence 1.7 J/cm2, droplets were hardly observed (fig. 3 e). Again, as in the case of layers deposited from the WB2.5 target, the number and the size of droplets was higher for 355 nm wavelength (fig. 4).
PT E
D
MA
NU
Fig. 3. SEM images of layers deposited from the WB4.5 target at 1064 nm radiation with fluence 9.3 J/cm2 (a , b), 5 J/cm2 (c, d) ), and 1.7 J/cm2 (e, f)
CE
Fig. 4. SEM images of layers deposited from WB4.5 target at 355 nm radiation with fluence 5 J/cm2 (a, b) and 1.7 J/cm2 (c, d)
AC
With the exception of layers deposited with fluence just above ablation threshold at 1064 nm radiation, the chemical composition was the same for smooth part of the layers and droplets on its surface (fig. 5). Moreover, boron content in deposited layers was higher at lower fluences and at 355 nm radiation (fig. 6). It should be kept in mind that EDS method poorly deals with quantitative measurements of light elements including boron, thus presented results could be used just for comparison.
NU
SC
RI
PT
ACCEPTED MANUSCRIPT
AC
CE
PT E
D
MA
Fig. 5. SEM-EDS analysis of layers deposited from the WB2.5 (left) and WB4.5 (right) targets at 1064 nm radiation with fluence 5 J/cm2; SEM image of the investigated area (a, c), EDS spectrum of selected points at acceleration voltage 5 kV (b, d)
Fig. 6. SEM-EDS analysis of layers deposited from the WB2.5 target at 1064 nm radiation with fluence 9.3J/cm2 (a) and 5 J/cm2 (b); at 355 nm radiation with fluence 5 J/cm2 (c) The layer deposited from the WB2.5 target at 1064 nm radiation with fluence 1.7 J/cm2 is composed of two types of particles: composed of both boron and tungsten, and composed of pure boron (see fig. 1 f). The layer deposited from the WB4.5 target at 1064 nm radiation with fluence 1.7 J/cm2 is a special case because SEM-EDS analysis shows that it is composed of pure boron (fig. 7). Moreover, the image of the inclined sample indicates that the deposited grains grow perpendicularly to the substrate surface (fig. 8).
RI
PT
ACCEPTED MANUSCRIPT
PT E
D
MA
NU
SC
Fig. 7. SEM-EDS spectrum of the layer deposited from the WB4.5 target at 1064 nm with fluence 1.7 J/cm2
CE
Fig. 8. SEM image of the layer deposited from the WB4.5 target at 1064 nm with fluence 1.7 J/cm2, the surface inclined by 35°
AC
3.2. XRD results The layer deposited from WB2.5 target at 1064 nm radiation with fluence 9.3 J/cm2 was composed of tetragonal W2B and rhombohedral β-WB phase (fig. 9 a). Calculated cell parameters of W2B phases a = 5.592 Å and c = 4.750 Å were similar to theoretical one (a = 5.571 Å; c = 4.771 Å [5]). Calculated cell parameters of β-WB phases a = 3.177 Å; b = 8.478 Å; c = 3.096 Å were close similar to theoretical ones (a = 3.177 Å; b = 8.485 Å; c = 3.101 Å [5]). Fluence reduction to 5 J/cm2 leads to the formation of α-WB, β-WB and WB3 phases (fig. 9 b). Because of the small number of diffraction maxima the cell parameters were not calculated. It was unable to register XRD spectrum of the layer deposited with fluence 1.7 J/cm2.
SC
RI
PT
ACCEPTED MANUSCRIPT
NU
Fig. 9. XRD spectra of layers deposited from the WB2.5 target at 1064 nm with fluence 9.3 J/cm2 (a) and 5 J/cm2 (b). Solid circles – W2B phase, open circles – βWB phase, star – αWB phase, hearts – WB3 phase, asterisk – unidentified peak. The corresponding Miller Indices are given above each peak.
AC
CE
PT E
D
MA
Changing the radiation to 355 nm caused the formation of WB3 phase for both 5 and 1.7 J/cm2 fluence (fig. 12). The calculated cell parameters for layer deposited with fluence 5 J/cm2 are: a = 5.228 Å and c = 6.120 Å, for layer deposited with fluence 1.7 J/cm2 are a = 5.183 Å and c = 6.224 Å, while theoretical cell parameters are a = 5.2 Å and c = 6.313 Å. The difference between calculated and theoretical cell parameters is higher in case of fluence 5 J/cm2. Also, diffraction lines are broader at higher fluence. Both features can be a result of a higher residual strain and a larger number of defects in the crystalline structure.
Fig. 10. XRD spectra of layers deposited from the WB2.5 target at 355 nm radiation with fluence 5 J/cm2 (a) and 1.7 J/cm2 (b). Hearts represent the diffraction peaks of the WB3 phase, the corresponding Miller Index is given above each peak. All layers deposited from the WB4.5 target consist mainly of WB3 phase (fig. 11). However, on the XRD spectra some unidentified peaks, indicated with asterisk, can be observed. In the case of the
ACCEPTED MANUSCRIPT
CE
PT E
D
MA
NU
SC
RI
PT
layer deposited at 1064 nm radiation with fluence 9.3 J/cm2 the high broadening of diffraction lines was observed (fig. 11 a). Moreover, its calculated cell parameters vary from theoretical ones more than in case of the layer deposited with fluence 5 J/cm2. indicates small size of crystallites and high content of defects in the material. It was also noticed that layers deposited with 355 nm radiation have elevated background signal for the 2θ range from 20° to 50°.
AC
Fig. 11. XRD spectra of layers deposited from the WB4.5 target and calculated cell parameters of WB3 phase. The ablation at 1064 nm radiation with fluence 9.3 J/cm2 (a) and 5 J/cm2 (b). The ablation at 355 nm radiation with fluence 5 J/cm2 (c) and 1.7 J/cm2 (d). Hearts represent the diffraction peaks of the WB3 phase, the corresponding Miller Index is given above each peak. 3.3. Nanoindentation test The Young modulus did not vary significantly for all investigated layers and was 479.8 ± 28.5 GPa. The results of nanohardness are presented in fig. 12. Nanoindentation of the layer deposited from the WB2.5 target at 1064 nm radiation with fluence 9.3 J/cm2 was impossible due to low adhesion.
PT
ACCEPTED MANUSCRIPT
SC
RI
Fig. 12. Results of nanoindentation test. The layer deposited from the WB2.5 target at 355 nm with fluence 5 J/cm2 – blue line. The layer deposited from the WB4.5 target at 355 nm with fluence 5 J/cm2 – green line, and the layer deposited at 1064 nm with fluence 9.3 J/cm2 – red line. Silicon substrate – dashed line
MA
NU
The uncertainty of the nanoindentation hardness is due to the high roughness of the material. Despite the presence of droplets deposited layers were very hard. The hardness of layers deposited from WB4.5 target was up to 50 ± 10 GPa under the load 30 mN. Investigated layer deposited from the WB2.5 target at 355 nm with fluence 5 J/cm2 had slightly lower hardness 36 ± 10 GPa. Presented results may be even underestimated because the nanoindentation depth was of about 200 nm. Considering the layer thickness of 0.9 ± 0.1 µm and the substrate hardness of 10.7 ± 0.7 GPa, presented results are affected by the substrate hardness. The hardness of deposited layers is similar to the hardness of WB4 layers deposited by femtosecond laser [16]. It is worth noting that despite the presence of droplets on the surface of PLD layers, they are still very hard.
D
4. Discussion
AC
CE
PT E
Morphological defects, including macro-particles, ‘doughnut rings’ and ‘scars’ are commonly observed for films deposited by various Physical Vapour Deposition techniques, that is well described in Dearnley [20]. Also the layers reported in this manuscript have droplets on their surface. More droplets were observed onto the surface of layers deposited with higher fluence and 355 nm radiation (fig. 13). The higher fluence increases the temperature of the irradiated targets’ area and could result in the phase explosion [21-23]. The absorption of the laser radiation in plasma plume is lower at shorter wavelength, so more energy reaches the target’s surface [25]. Moreover, for WBx-type compounds the absorption is higher at 355 nm than at 1064 nm radiation [26], thus smaller volume of the target absorbs the energy and is heated to higher temperature.
PT
ACCEPTED MANUSCRIPT
RI
Fig. 13. Schematic view showing the trend between the number of droplets and fluence in the case of layers deposited from WB2.5 target (dashed line) and WB4.5 target (solid line) at radiation of 355 nm (blue line) and 1064 nm (red line).
AC
CE
PT E
D
MA
NU
SC
The difference of chemical composition between the target and the deposited layers was observed. Moreover, chemical and phase composition of the deposited layers varies depending on the laser wavelenght and fluence (fig. 14). This phenomena can be explained by the selective evaporation of the target’s components [24]. During the ablation of multicomponent target at lower fluences the energy absorbed by the target can be too low to evaporate the component having higher melting point and molar mass [12, 24]. In the case of WB2.5 and WB4.5 targets there is a significant difference in physical properties between the target’s components. Thus, boron as the element more volatile and having lower melting point (melting point of boron and tungsten is 2349 and 3695 K, respectively) will evaporate first. The effect of selective evaporation is more pronounced at 1064 nm radiation as the energy of laser pulse is absorbed in greater volume. The differences between layers deposited with 1064 and 355 nm can be explainned by diferent characteristics of plasma plume. In the case of the ablation with 1064 nm the plasma plume absorbtion is higher what cause wider radial expanssion of boron atoms and ions [12] resulting from its smaller than tungsten atomic mass. Moreover, boron atoms (and ions) are scattered in collisions witch much heavier tungsten. Therefore, ablation at 1064 nm radiation and with higher fluences will result in the decrease of boron atoms deposited on the layer.
Fig. 14. Schematic view showing the trend between the chemical composition and fluence of layers deposited from WB2.5 target (dashed line) and WB4.5 target (solid line) at radiation of 355 nm (blue line) and 1064 nm (red line).
ACCEPTED MANUSCRIPT 5. Conclusions
PT
In this work WBx-type layers deposited by nanosecond laser pulse at 355 and 1064 nm radiation were investigated. The W2B-βWB, αWB-WB-WB3 and WB3, and boron layers were deposited wherein WB3 structure is formed in a wide range of laser fluences and at both investigated wavelength. Thus, among tungsten borides formed in PLD process the WB3 phase is probably the most common one. The content of boron in deposited layer is higher in the case of ablation at 355 nm radiation and increases at lower fluences. The WB3 layers were investigated in the indentation test at a load of 5-30 mN and its hardness varied from 36 ± 10 GPa to 50 ± 10 GPa. Presented results confirm the possibility of using WB3 as a superhard material.
AC
CE
PT E
D
MA
NU
SC
RI
Acknowledgement This work was supported by the NCN (National Science Centre) Research Project: 2015/19/N/ST8/03928. Part of investigations was carried out with the use of CePT infrastructure financed by the European Union – the European Regional Development Fund within the Operational Programme "Innovative economy" for 2007-2013. Authors thank prof. Kucharski for nanoindentation test.
ACCEPTED MANUSCRIPT
AC
CE
PT E
D
MA
NU
SC
RI
PT
Bibliography 1. Q. Gu, G. Krauss, W. Steurer, Transition metal borides: superhard versus ultraincompressible, Adv. Mater. 20 (2008) 3620-3626 2. M. Wang, Y. Li, T. Cui, Y. Ma, G. Zou, Origin of hardness in WB4 and its implications for ReB2, TaB4, MoB4, TcB4, and OsB4, Appl. Phys. Lett. 93 (2008) 101905 3. E. Zhao, J. Meng, Y. Ma, Z. Wu, Phase stability and mechanical properties of tungsten borides from first principles calculations, Phys. Chem. Chem. Phys. 12 (2010) 13158-1316 4. Y. Liang, Y. Gou, X. Yuan, Z. Zhong, W. Zhang, Unexpectedly hard and highly stable WB3 with a noncompact structure, Chemical Physics Letters 580 (2013) 48-52 5. X. -Y. Cheng, X. –Q. Chen, D. –Z. Li, Y. –Y. Li, Computational materials Discovery: the case of the W-B system, Acta Cryst. C C70 (2014) 85-103 6. R. Kiessling, The Borides of Some Transition Elements, Acta Chem. Scand. 4 (1950) 209-227 7. L. Yang, Z. Kan, Y. Zeng, X. Wang, S. Du, C. Tao, P. Ren, X. Cui, M. Wen, Boron doped bcc -W films: Achieving excellent mechanical properties and tribological performance by regulating substrate bias voltage, Appl. Surf, Sci. 423 (2017) 275-282 8. Q. Li, D. Zhou, W. Zheng, Y. Ma, C. Chen, Global structural optimization of tungsten borides, Phys. Rev. Lett. 110 (2013) 136403 9. S. Okada, K. Kudou, T. Lundström, Preparations and Some Properties of W 2B, δ-WB and WB2 Crystals from High-Temperature Metal Solutions, Jpn. J. Appl. Phys. 34 (1995) 226-231 10. T. Mościcki, J. Radziejewska, J. Hoffman, J. Chrzanowska, N. Levintant- Zayonts, D. Garbiec, Z. Szymański, WB2 to WB3 phase change during reactive spark plasma sintering and pulsed laser ablation/deposition processes, Ceram. Int. 41 (7) (2015) 8273-8281 11. J. Chrzanowska, Ł. Kurpaska, M. Giżyński, J. Hoffman, Z. Szymański, T. Mościcki, Fabrication and characterization of superhard tungsten boride layers deposited by radio frequency magnetron sputtering, Ceram. Int. 10 (2016) 12221-12230 12. T. Mościcki, Differences in nanosecond laser ablation and deposition o tungsten, boron and WB2/B composite due to optical properties, Int. J. Opt. 5438721 (2016) 13. R. Mohammadi, A. T. Lech, M. Vie, B. E. Weaver, M. T. Yeung, S. H. Tolbert, R. Kaner, Tungsten tetraboride, an inexpensive superhard material, Proc. Natl. Acad. Sci. USA 108 (2011) 10958-10962 14. C. Jiang, Z. Pei, Y. Liu, J. Xiao, J. Gong, C. Sun, Preparation and characterization of superhard AlB2-type WB2 nanocomposite coatings, Phys. Status Solidi A 210 (6) (2013) 12211227 15. Y. M. Liu, C. L. Jiang, Z. L. Pei, H. Lei, J. Gong, C. Sun, Microstructure and properties of AlB2- type WB2 thin films deposited by direct- current magnetron sputtering, Surf. Coat. Technolog. 245 (2014) 108-116 16. J. V. Rau, A. Latini, R. Teghil, A. De Bonis, M. Fosca, R. Caminiti, V. R. Albertini, Superhard tungsten tetraboride films prepared by pulsed laser deposition method, Appl. Mater. Interfaces 3 (2011) 3738-3743 17. D. Dellasega, V. Russo, A. Pezzoli, C. Conti, N. Lecis, E. Besozzi, M. Beghi, C. E. Bottani, M. Passoni, Boron films produced by high energy Pulsed Laser Deposition, Mater. Des. 134 (2017) 35-43 18. D. Dellasega, G. Merlo, C. Conti, C. E. Bottani, M. Passoni, Nanostructured and amorphouslike tungsten films grown by pulsed laser deposition, J. Appl. Phys. 112 (2012) 084328 19. A. T. T. Mostako, C. V. S. Rao, A. Khare, Mirrorlike pulsed laser deposited tungsten thin film, Rev. Sci. Instrum. 82 (2011) 013101 20. P. A. Dearnley, Introduction to Surface Engineering, Cambridge University Press, New York, 2017, pp. 179-182, and 187-189 21. N. M. Bulgakova, A. V. Bulgakov, Pulsed laser ablation of solids: transition from normal vaporization to phase explosion, Appl. Phys. A 73 (2001) 199-208 22. W. O. Siew, W. K. Lee, H. Y. Wong, T. K. Yong, S. S. Yap, T. Y. Tou, Investigation of droplet formation in pulsed Nd: YAG laser deposition of metals and silicon, Appl. Phys. A 101 (2010) 627-632
ACCEPTED MANUSCRIPT
AC
CE
PT E
D
MA
NU
SC
RI
PT
23. J. Chrzanowska, J. Hoffman, P. Denis, M. Giżyński, T. Mościcki, The effect of process parameters on rhenium diboride films deposited by PLD, Surf. Coat. Technol. 277 (2015) 1522 24. C. Klamt, A. Dittrich, B. Jaquet, C. Eberl, F. Dӧring, H. U. Krebs, Largest possible deviations from stoichiometry transfer during pulsed laser deposition, Appl. Phys. A 122 (2016) 701 25. T. Mościcki, J. Hoffman, J. Chrzanowska, The absorption and radiation of a tungsten plasma plume during nanosecond laser ablation, Phys. Plasmas 22 (2015) 103303 26. M. Maździarz, T. Mościcki, Structural, mechanical and optical properties of potentially superhard WBx polymorphs from first principle calculations, Mater. Chem. Phys. 179 (2016) 92-102
ACCEPTED MANUSCRIPT Highlights:
PLD as a suitable technique for the deposition of tungsten triboride layers
PLD layers composed of tungsten triboride have a hardness from 36 to 46.8 GPa
Selective evaporation of boron from the composite WB4.5 target irradiated with 1064
AC
CE
PT E
D
MA
NU
SC
RI
PT
nm at low fluences