9              DEPARTMENT OF MATERIAL STUDIES

 

 

Head of  Department:

Assoc. Prof. Zbigniew Werner

phone:

(22) 718-05-45

e-mail:

wernerz@ipj.gov.pl

 

 

 

Overview

The technology of modifying surfaces of practical-use materials by means of continuous and pulsed energy and particle beams has been intensely studied for more than 20 years. In some fields it is presently utilized on a wide scale in industry. Continuous or pulsed ion and plasma beams play a significant role among various approaches used in this area. The research carried at by Department P-IX is centered around the use of two ion implantation machines (ion implanters) of different kinds and several unique sources of high-intensity intense plasma pulses, utilized jointly with Department P-V. The Department cooperates with Forschungszentrum Rossendorf (FZR, Dresden, Germany) in the field of ion-beam-based analytical techniques and the use of ion implantation facilities. The main objectives of the Department are:

·        A search for new ways of modifying the surface properties of solid materials by means of continuous or pulsed ion and plasma beams and

·        implementation of ion implantation techniques in national industries as a method of improving the lifetime of machine parts and tools utilized in industry.

In 2004 these objectives were being accomplished in many ways, particularly by research on:

·         interaction of nitrogen atoms in expanded austenite formed in pure iron

·         formation of superconducting MgB2 phases

·         surface layer modification of ion-bombarded HDPE

·         nanolayers in AlN for direct bonding with copper

·         surface alloying of titanium with nickel and palladium for increased corrosion resistance 

·         nanolayers in alloys and ceramic coatings for  improved resistance to high-temperature corrosion 

·         improvement of cutting tools lifetime by means of nitrogen implantation.

The research was conducted in cooperation with Department P-V of IPJ, Institute of Nuclear Chemistry and Technology (Warsaw), Warsaw University of Technology, Institute of Technology of Materials for Electronics (Warsaw), Forschungszentrum Rossendorf FZR (Dresden, Germany), as well as with some industrial companies.

 

 

 

 

 

 

 

 

 

9.1         Interaction of Nitrogen Atoms in Expanded Austenite Formed in Iron by Intense Nitrogen Plasma Pulses

by J.Piekoszewski, B.Sartowska1), L.Waliś1), Z.Werner, M.Kopcewicz2), F.Prokert3), J.Stanisławski, J.Kalinowska2), W.Szymczyk

 

Several authors [e.g. 1, 2] have shown that it is possible to nitride stainless steel in such a way that a metastable phase is formed, in which nitrogen in solid solution increases the surface hardness and wear resistance without compromising corrosion behavior. This phase is referred to as nitrogen-expanded austenite and is denoted as γN. Some authors [e.g. 3, 4] claim that the γN phase can only be formed if Fe, Cr and Ni are available in the system, but the initial material needs not be of fcc austenite structure. However, it has been also demonstrated [5] that high intensity nitrogen plasma pulses can form γN phase even in pure α-iron (ARMCO). On the other hand, it has been shown [6] that γN in carbon steels also improves their tribological properties.

In the present work we tried to determine the character of interaction between nitrogen atoms in γN phase formed in pure iron treated by high-intensity nitrogen plasma pulses. ARMCO samples were irradiated with 20 intense nitrogen plasma pulses of about 1 μs duration each. At about 5 J/cm2 deposited energy the pulses were melting sample surface to about 1-2 μm and doping the molten layer with nitrogen. Retained dose after 20 pulses was 5.5x1017 N/cm2, and nitrogen concentration was 6.5-8.5 at% as found by nuclear reaction analysis (NRA). Samples were also examined by X-ray diffraction in grazing incident geometry (ω=2º) using CuKα rays (XRD analysis depth≈500 nm), and by conversion electron Mössbauer spectroscopy (CEMS depth≈300 nm).

 

Table 1

XRD parameters of ARMCO sample treated with 20 pulses of nitrogen plasma.

(hkl)

2Q (deg)

FWHM (deg)

a
(nm)

(a-a0)/a0(%)

a0 (1)

a0 (2)

(200)

50.180

0.622

0.36332

2.056

1.29

(220 )

73.652

0.838

0.36349

2.104

-

(311 )

89.276

1.189

0.36361

2.138

-

a0(1)=0.356 nm extrapolated for pure austenit [7]

a0(2)=0.3587 nm for untreated 310 stainless steel [8]

 

Expansion of the crystal lattice is confirmed [7, 8]. However, lattice transformed to austenite from α-Fe by us is less susceptible for expansion than that of an originally austenite steel (e.g. compare lattice parameter a relative change (a-a0)/a0 ≈ 2% with 2.9% observed in [8] for 310 steel containing 8at.% of nitrogen). The difference is due to different structures and compositions of both materials.

CEMS spectra were fitted with spectra of various phases. Fγ0/FγN fraction ratio derived from the fit was used to semi-qualitatively estimate character of interactions between nitrogen atoms present in the austenitic phases. Two models discussed in [9] were considered: (i) occupation of each of 6 octahedral sites occurs randomly, i.e. there is no interaction between nitrogen atoms (model A) and (ii) strong repulsive forces act between both first- and second-nearest nitrogen atoms, so they tend to separate from each other (model B).

Fig. 1   The box represents value of Fg0/FgN fraction ratio observed experimentally in ARMCO sample treated with 20 pulses of nitrogen plasma. The horizontal sides represent the range of experimen­tally measured concentrations, the vertical sides represent the estimated error of the Fg0 /FgN fraction ratio. The solid lines show theoretical predictions assuming no interaction and strong repulsion between nitrogen atoms in the nearest-neighbour positions.

 

The experimental result is much closer to model B than to A. Therefore, we conclude that strong repulsion forces act between nitrogen atoms in the fcc austenitic structure formed as a result of nitriding of pure iron by intense nitrogen plasma pulses.

 

[1]     C.Blawert, et al., Surf. Engin., 15(1999)469

[2]     O.Özturk, et al., J Appl. Phys., 77, 8(1995)3839

[3]     C.Blawert, et al., Surf. Coat. Technol. 1891 (1999)116

[4]     E.Menthe, et al., Surf. Coat. Tech. 412 (1995)74

[5]     J.Piekoszewski, J.Langner, Z.Werner, et al., NIM B 80/81, 344 (1993)

[6]     B.Sartowska, et al. Surf. Coat. Technol. (in press)

[7]     A.P.Gulaiev, Metalloviedenie, Izdatilestwo “Metallurgia” Moscow (1977) (in Russian)

[8]     A.Saker, et al., Mater. Sci. Engin., A 140 (1991)702

[9]     K.Oda, K.Umezu, H.Ino, J. Phys. Condens. Matter. 2(1990)10147

 

1)         Institute of Nuclear Chemistry and Technology, Warsaw, Poland

2)         Institute of Electronic Materials Technology, Warsaw, Poland

3)         Forschungszentrum Rossendorf e.V. Institut für Ionenstrahlphysik und Materialforschung, Dresden, Germany

9.2         Ion Implanted Nanolayers in AIN for Direct Bonding with Coper

by M.Barlak, W.Olesińska1), J.Piekoszewski, M.Chmielewski1), J.Jagielski1), D.Kaliński1), Z.Werner, B.Sartowska2)

 

Direct bonding (DB) of conductors to ceramics (especially AlN) is considered as the most promising technology of packaging electronic devices used in high power density applications. In the DB technique, metal is bonded directly to the ceramics with only a very thin transition layer at their interface. One of the pioneering works [1] proved that satisfactory bonding of Cu to AlN can be achieved if 1-1.5 at% of oxygen is added to the system as an active element – even if the surfaces of the joined components have not been intentionally modified.

In our preliminary experiments with implantation of Ti, Fe and O ions into AlN substrates we tried to replace the conventional process of thermal oxidation. The results [2] suggested that the best shear strength could be expected for relatively low energy of Ti ions.

In the present work, we implanted Ti, Fe and O ions. 1016-1018 Ti and Fe ions/cm2 were implanted in the MEVVA-type TITAN direct beam implanter [3] at 15 or 70 kV into commercial (Goodfellow) AlN substrates of 12x3x0.63 mm3 with roughness of about Ra=0.1 μm. Oxygen was implanted in a home-made semi-industrial implanter with non-mass separated beam. To avoid overheating, samples were clamped onto a water-cooled stainless steel plate mounted in both implanters. The ion current densities were kept below 10 μA/cm2, and the substrate temperature did not exceed 200ºC during the entire run.

Metallic components of the joints, i.e. 30x3x0.3 mm3 strips of oxygen-free Cu were annealed at 600ºC for about 30-40 min in flowing nitrogen containing 1.5 ppm oxygen. The components were subsequently oxidized in air at 380ºC for 3 min and directly bonded to the ceramics in a conventional DB process.

Mechanical properties of the obtained joints were examined. Careful SEM observations were performed on fractured surfaces of both components of the joints to reveal the joint microstructure.

Fig. 1   Comparison of various AlN pre-treatment techniques for Direct Bonding of AlN-Cu joints.

 

Shear strengths between 20 and 70 kG/cm2 have been obtained. Oxygen implantation gave consistently better results than implantation of iron. For all elements implanted at 15 kV shear strength was greater by a factor of about 2 than that obtained at 70 kV. The best shear strength obtained for 5x1016 cm-2 Ti (70 kG/cm2) exceeds by a factor of 5 typical values (14 kG/cm2) obtained using conventional isothermal oxidation AlN pre-treatment process [2].

Microstructure inspections showed that the strongest joints are of adhesive type over the entire surface. Over 90% of the copper surface in contact with the ceramics exhibited neither changes, nor inhomogeneities – its structure was continuous and homogeneous. Insignificant copper grooves observed at grain boundaries can be associated with copper oxidation prior to the joining process.

The ceramic surface had a homogeneous compact grain structure homogeneously covered with nano-precipitates. At higher Ti doses larger numbers of needle-shape precipitates were seen and a new phase of multifaceted crystals appeared. On the other hand, the grain-like phase increases with both energy and dose of Ti ions.

Fe and O implantations did not lead to substantial changes of the microstructures at the joint surfaces of ceramics and copper for different implantation conditions. Such effects were seen only for Ti.

To conclude, ion implantation seems to be ideally suited for Direct Bonding process. Optimized implantation process leads to much better results than a conventional process at a comparable processing time. Advantages of ion implantation include:

·       potentially fast processing time (several minutes at 100 mA beam current instead of tens of minutes in the conventional process)

·       accuracy in creating an appropriate dopant content

·       flexibility in tailoring the desired distribution of the introduced atoms at nanometer depth range

·       ability to form non-equilibrium compounds.

 

[1]     E.Entezarian, R.Drew, Mat. Sci. Egin. A 212 (1996) 206

[2]     J.Piekoszewski, Z.Werner, M.Barlak, W.Szymczyk, Solid State Phenomena 231 (2004)99

[3]     S.Bugaev, et al., Rev. Sci. Instr. 10(1994) 3110

 

1)         Institute of Electronic Materials Technology , Warsaw, Poland

2)         Institute of Nuclear Chemistry and Technology , Warsaw, Poland

9.3         Superconductivity of MgB2 Thin Films Prepared by Ion Implantation and Pulsed Plasma Treatment

by J.Piekoszewski, W.Kempiński1), B.Andrzejewski1), Z.Trybuła1), L.Piekara-Sady1), J.Kaszyński1), J.Stankowski1), Z.Werner, E.Richter2), F.Prokert2), J.Stanisławski, M.Barlak

 

The superconductive phase of inter-metallic MgB2 compound discovered in 2001 by Nagamatsu et al. [1] has attracted a considerable interest due to its relatively high transition temperature TC=39K and potential application on the industrial scale. The new material has been studied along two paths: solid MgB2 and thin films formed on various substrates. We try to form superconducting MgB2 layer grown on substrates with surfaces molten by intense plasma pulses.

In our preliminary experiments [2] Mg substrates were implanted with 5x1018 B ions per cm2 at energy of 100 keV an melted by two ms duration H plasma pulses of energy 1.9 and 3.0 J/cm2. Magnetically Modulated Microwave Absorption (MMMA) and four point probe (FPP) methods were used to detect superconductivity. The highest obtained transition temperature was TC=31 K. However, macroscopic percolation chains did not occur and MMMA signals were very weak.

In the present work, the number of pulses was increased (2-4), and H plasma was replaced by Ar one. Mg substrates were implanted with 3x1018 B ions per cm2 at energy of 80 keV. Samples were characterized by MMMA, FPP, XRD and RBS techniques.

XRD data indicate that the a lattice constant decreased with respect to the as-implanted samples by Da/a = –0.71%, whereas the c lattice constant increased by Dc/c = 0.37%. According to theoretical predictions given by Wan [3] this should lead to an increased density of states near the Fermi level and therefore to an increased TC. On the other hand, the highest value of TC was observed when the c lattice constant had the smallest value with respect to the bulk Dc/c = –0.2% [4]. However, some authors claime that TC rises with lattice expansions [5-7].

Fig. 1   RBS spectra of as-implanted and pulse-treated samples.

 

A pronounced valley centered around channel 400 appeared in the RBS spectrum of as-implanted sample (a). Samples (b) and (c) were treated with 2 H pulses with energy density 2 and 3 J/cm2, respectively. It may be seen that the width of the valley (related to the width of the boron profile) grows with the pulse fluence. For sample (c) the boron profile seems to be spread over the greatest depth. On the other hand, the Mg signal value at the minimum of the spectrum does not change significantly. Rough estimations based on SIMNRA simulations indicate that the Mg content at the peak of boron profile amounts to no more than 15-25%. It means that boron concentration is no less that 75-85% – far in excess of the stoichiometric composition of MgB2. However, this excess of boron does not preclude the existence of MgB2 phase, as evidenced by the XRD spectra (not shown here).

The strongest MMMA signal hysteresis (higher by an order of magnitude with respect to samples (b) and (c)) has been obtained for the sample treated with 4 pulses of Ar plasma at the fluence of about 2 J/cm2. How­ever, TC for that sample had a rather disappointing value of about 12K, probably due to lack of stoichiometry. The highest TC has been obtained for high-energy pulses, but the highest superconducting phase content has been obtained for the largest number of pulses, corresponding to a long time of diffusion in the molten phase.

Still no breakthrough has occurred with respect to the macroscopic percolation of the superconductive regions.

 

[1]     J.Nagamatsu, et al., Nature 410 (2003) 63

[2]     J.Piekoszewski, Z.Werner, et al., Inst. of Nucl. Chem. and Techn. Annual Report 2003, Warsaw 2004, p 118

[3]     X.Wan, et al., cond-mat /0104216v3 (2001)

[4]     H.Yamazaki, et al., Appl. Phys. Lett. 83, 18 (2003)3740

[5]     N.Hur, et al., Appl. Phys. Lett. 79(2001)4180

[6]     J.Tang, et al., Phys. Rev. 64 (2001)132509

[7]     J.B.Neaton, A.Perali, cond-mat/0104098v1 (2001)

 

1)         Institute of Molecular Physics Polish Academy of Sciences, Poznań, Poland

2)         Forschungszentrum Rossendorf e.V. Institut für Ionenstrahlphysik und Materialforschung, Dresden, Germany

9.4         The Effect of Nitrogen Implantation on Lifetime of Cutting Tools Made of SK5M Tool Steel

by J.Narojczyk1), Z.Werner, J.Piekoszewski

 

A number of tools used presently in industry for cutting and forming are made of high speed steels (HSS). To increase wear resistance of such cutting tools, hard coatings of nitrides, borides, carbides or oxides of such elements like Al, Cr, Ta, Ti and Fe are used [1]. Nitrogen ion implantation has for a long time been used as an alternative to coatings [2, 3, 4, 5]. This report describes the results of such application in the ball-bearing industry.