New Ternary Fe-Ni-Cu Invar Alloys Preparation

New Ternary Fe-Ni-Cu Invar Alloys Preparation Preparation and Characterization of New Ternary Fe-Ni-Cu Invar alloys S Ahmada, A B Ziya[1], a, A Ibrahimb, S Atiqb, N Ahmada and F Bashirc Abstract. Six alloys of Fe65Ni35-xCux(x= 0, 0.2, 0.6, 1, 1.4, 1.8 at.%) have been prepared by conventional arc-melting technique and characterized by utilizing in-situ X-ray diffraction (XRD) technique and differential scanning calorimetry (DSC) at a range from room temperature to 773 K for determination of phase. The studies show that these alloys form face centered cubic (FCC) throughout the investigated temperature range. The X-ray integrated intensities of various reflections were used to determine the coefficient of thermal expansion ÃŽ ±(T), mean square amplitude of vibrations and characteristic Debye temperature ÃŽËœD. The ternary substitution of copper has a minor effect on the lattice parameter but the Debye temperature ÃŽËœD is found to decrease with the increase of copper content in the alloy. The coefficients of thermal expansion ÃŽ ±(T) were found to be comparable to those for conventional Fe-Ni invar alloys. Keywords:  Invar alloys; lattice parameters; thermal expansion; X-ray diffraction Introduction Iron rich invar alloys have been of keen interest for researchers and developers, for their own reasons and interests, since their discovery in 1898 Guillaume and Hebd (1987) because of their unique set of properties labeled as invar anomaly or invar effect. A number of theories and models have been postulated to explain these deviations in the behavior of these alloys from other materials but still there are many queries unresolved Sanyal and Bose (2000); Iwase et al (2003); Matsushima et al (2006); Goria et al (2010); Yichun et al (2009); Tabakovic et al (2010); Pepperhoff et al (2001); Duffaut et al (1990); Matsushita et al (2008). One of the most important property of these alloys that made them most sought for material for applications in especially the electrical/ electronic precision instruments is their very low coefficient of thermal expansion around room temperature as compared to other metals and alloys. But, these materials also have their limitations and to overcome them , the researchers have either made ternary additions to the basic alloy or have turned their focus onto other combinations of elements termed as invar type Ono et al (2007); Matsushita et al (2004); Gorria et al (2006); Zhichao et al (2002); Rongjin et al (2010); Kaji et al (2004); Matsushita et al (2009); Matsushita et al (2007). For example, in some electrical/electronic applications another important property required in candidate material is good electrical conductivity. Iron based invar alloys cannot be grouped as good electrical conductors. Consequently, to develop invar alloys that exhibit inherent low coefficient of expansion and comparatively better electrical conductivity, ternary additions of elements like copper have been studied Stolk et al (1999); Bernhard et al. (1987). Not to mention such addition is expected to decrease the manufacturing cost. Many research groups have undertaken the study of effect of addition of copper onto invar properties of binary iron nickel a lloys but lacked correlation between the copper addition to change or no change in invar properties. This study has been carried out to correlate the invar effect to ternary addition of copper to base iron nickel invar alloy by replacing nickel with copper and to determine thermal properties of the newly developed alloys for comparison with same properties of binary invar alloys. Experimental methods For this study, one binary Fe65Ni35 (subscript indicates atomic percent of the element) and five ternary Fe65Ni35-xCux where x was selected to be equal to 0.2, 0.6, 1, 1.4 and 1.8 were prepared. High purity elements (>99.9%) were weighed and combined on water cooled hearth of a vacuum arc melter. The process was carried in 600 mbar argon atmosphere created after evacuating the chamber to 10-5 mbar pressure. The alloys were melted several times to ensure thorough mixing of the ingredients. To ensure homogeneity, the samples were then heated under vacuum in a Nebertherm furnace at 1273 K approximately for one hundred and seventy hours. Homogenized samples were then weighed as well as chemically analyzed and found to be well within the selected range of set composition. Each sample was then cold rolled to about 0.2 mm thickness and then heated at 1273 K for four hours to remove rolling stresses. Samples of suitable dimensions were then cut from each strip for characterization through X-ray diffraction (XRD) and differential scanning calorimetry (DSC). XRD was carried out in a Bruker D8 Advance diffractometer equipped with MRI high temperature chamber fitted with PtRh heater element. Operating conditions for the X-ray tube were set at 40 kV and 40 mA. The diffraction patterns were recorded in the step scan mode in the 2ÃŽ ¸-range from 20 to 120o with a step of 0.01o. The in-situ high temperature X-ray diffraction of all samples was carried out in 10-6 mbar vacuum with Ni-filtered CuK radiation from room temperature to 473 K with a step of 20 K and thereon with a step of 50 K till 773 K. DSC of all samples was carried out on SBT-Q600 differential scanning calorimeter from room temperature to 1473 K at a heating rate of 20 K/minute under argon atmosphere. 3. Results and discussion 3.1. Structure and lattice parameters DSC scans of the six selected invar alloys were measured (not shown here). No sharp exothermal or endothermal peak was observed in the investigated temperature range, it is thus assumed that the samples were single phase. Room temperature XRD patterns of binary classical invar alloy of Fe65Ni35 and ternary alloys of Fe65Ni35-xCux (x=0.2, 0.6, 1, 1.4 and 1.8) are shown in Figure 1. It can be seen that all alloys are single phase and possess face centered cubic (FCC) lattice structure in confirmation to already published data on similar alloy systems Ono et al. (2007). The lattice parameters of the samples under study were determined by the extrapolation of lattice parameters for all reflections against Nelson-Riley function to minimize the random errors Ziya et al (2006). The values of calculated lattice parameters are given in Table 1. It can be seen that copper addition to the binary composition causes marginal decrease in the lattice parameter as expected because the copper with sm aller atomic radii replaced nickel atoms in the structure of relatively larger radius. 3.2 Thermal parameters To investigate invar effect in the newly developed alloys, it was planned to measure / calculate three major thermal properties / parameters vis-à  -vis temperature; namely, coefficient of thermal expansion, Debye temperature and mean square amplitude of vibration. The results obtained for each of them are discussed in succeeding sub sections. 3.2.1 Thermal expansion To investigate invar effect in these newly developed alloys, high temperature XRD technique was employed. A common observation from the scans of all the samples was that these samples are single phase alloys and no phase change occurred in any of the alloy up to scan temperature (773 K). This observation is consistent with the results of DSC measurements. One of the major parameter relating to invar effect is coefficient of thermal expansion which is primarily a reflection of change in lattice parameter with temperature. Temperature dependence of lattice parameter was calculated for each sample from the high temperature XRD data collected during this study. Scan at smaller step, 20 K up to 473 K and then larger step of 50 K to the maximum temperature, 773 K was set based upon the results published in literature for similar type of invar alloys. For calculation purpose data pertaining to (311) peak of binary alloy, (220) peak of Fe65Ni34.8Cu0.2 and (400) peak for all other composition was used. Selection of these peaks was solely made due to their better temperature dependence over the entire temperature range. It can be seen that in all the samples the lattice parameter almost remains unchanged up to about 473 K and there onward, the lattice parameter increases negligibly to a maximum of about 0.004 Aà Ã‚ ¦ at the maximum test tempe rature. However, the effect of increase in temperature on increase in lattice parameter in binary alloy is gradual and almost linear whereas, in ternary alloys, the increase in lattice parameter up to 473 K is insignificant but beyond this temperature it is visible and becomes steep with increase in copper content. Coefficient of thermal expansion ÃŽ ±(T) was then calculated by least square fitting the calculated lattice parameter data to second degree polynomial: ÃŽ ±(T) = A + BT + CT2 Where constant A represents lattice parameter of alloy at absolute zero, while B is the linear term coefficient and C represents the nonlinear term. The calculated values of ÃŽ ±(T) and these constants are tabulated in Table 2 whereas ÃŽ ±(T) versus temperature is plotted in figure 2. It was found that no appreciable change occurs in the thermal coefficient (ÃŽ ±) with temperature which is in line with the conclusion from the lattice parameter calculations. Further, the values of thermal coefficient ÃŽ ±(T) calculated in this study match very well to the values reported earlier for Fe-Cu alloys by other researchers such as (Goria et al. 2004 ). He (Goria et al. 2004) has reported ÃŽ ± (T) for said alloys in the range of 3Ãâ€"10-6K-1 at a temperature of 350 K whereas in the present study same value of ÃŽ ±(T) has been found up to the temperature of 450 K. Based upon above presented results and their analysis it can be concluded that these ternary alloys possess invar characteristics up to test temperature range. 3.2.2  The Debye temperatures and the mean square amplitudes of vibration Debye temperature is usually determined from the slope of ln(Iobs/Ical) versus temperature curves which is then subsequently used to find mean square amplitude of vibrations. Detailed procedure is already presented elsewhere [30]. Accordingly, the ratio of the observed and calculated intensities for each composition over the investigated temperature range was determined for selected Bragg reflections after stripping KÃŽ ±2-components from peak intensity. The peaks selected were (200) for binary, 0.2 at.% Cu and 1.4 at.% Cu containing alloys, (220) for 0.6 at.% Cu and (400) for 1 at.% and 1.8 at.% Cu containing alloys. Again the reflection lines were selected based on their relatively better dependence on temperature and integrated intensities were then determined from selected data by employing a line profile fit software. The results are presented in figure 3. It may be noted that for alloy containing 1.8 at.% Cu, the intensity data below 350 K has not been included because of exces sive scatter. Apart from this exception, for all other compositions and temperatures the points lie well along the fitted line. Debye temperature(ÃŽËœD) was then determined and plotted for all samples over the test temperature range in figure 4. First of all, these values have been found to be in close concurrence to those reported in literature (Gorria et al. 2009). In addition, from the comparison of these curves with each other two major facts can be deduced; firstly, the value of ÃŽËœD decreases as the amount of copper in the alloy increases, secondly up to the temperature of 473 K, ÃŽËœD for each composition remains almost unaffected by the increase in temperature. However, beyond this temperature and up to the maximum increased temperature, the value of ÃŽËœD decreases. These observations are in line with earlier findings that in these alloys invar effect is present up to 473 K because increase in length due to anharmonicity is compensated with magnetostricion. Furthermore, de crease in ÃŽËœD value both with increase in Cu contents as well as increase in test temperature indicates softening of the material. Mean square amplitude of vibrations ( was then calculated from the ÃŽËœD values as explained in reference (Ziya and Ohshima 2006). The result is tabulated in Table 3. Again the results indicate that there is very slight variation in with increase in temperature for every alloy composition. 4.  Conclusions Effect of copper addition in different percentages to binary iron nickel invar alloy has been investigated through in-situ XRD over a temperature range of 298 to 773 K. Thermal properties, i.e. the coefficient of thermal expansion, Debye temperature and mean square amplitude of vibrations of each of the ternary alloy has been determined and compared to the binary invar alloy prepared for this study as well as with the results published by other researchers for similar alloys. The results indicate that the newly developed ternary alloys exhibit Invar effect up to added copper contents although the temperature range is marginally decreased with the increase in copper contents. References Bernhard H, Volker B, and Jurgen H 1987 J. Mag and Mag. Mat 70 423 Duffaut F and Tiers J-F 1990, Industerial application of Invar, J. Written(Ed), The Invar Effects, TMS, Palo Alto, CA, P. 238 Guillaume Ch. E and Hebd C R 1987 Seances Acad. Sci. 125 235 Goria P , Martinez-Blanco D, Jesus A B, Ronald I S 2010 J. Alloys and Compounds 495 495 Gorria P, Martinez-Blanco D, Iglesias R, Palacios S L, Perez M J, Blanco J A, Barquin F, Hernanddo A, Gonzalez M A 2006 J. Mag and Mag. Mat 300 229 Gorria P , Martinez-Blanco D, Blanco J A, Hernando A, Garitaonandia J S, Barquin L F ,Campo J and Ronald I S 2004 Phys Rev B 69 214421 Gorria P, Martinez-Blanco D, Blanco J A, Maria J P, Hernando A, Maria A L, Daniel H, Souza-Neto N, Ronald I S, Marshall W G, Garbarino G, Mezouar M, Fernandez-Martinez A, Chaboy J, Barquin L F, Rodriguez Castrillon J A, Moldovan M, Garcia A J, Zhang J, Liobet A and Jiang J S 2009 Phys Rev B 80 06442 Iwase A, Hamatani Y, Mukomoto Y, Ishikawa N, Chimi Y, Kambara T, C.Muller C, R. Neumann, Ono F 2003 Nuclear Instruments and Methods in Physics Research B 209 323 Kaji S, Chiyoda S, Saito R, Oomi G, Yoshimura M, Tokunaga A and Kagayama T 2004 J. Mag and Mag. Mat 272-276 792 Liu Y, Lei L, Jiake L, Shen B, Wenbin Hu 2009 J. Alloys and Compounds 478 750 Lu Z, Dern L, and Junyi L 2002 J. Mag and Mag. Mat 239 502 Matsushima Y, Sun N Q, Kanamitsu H, Matsushita M , Iwase A, Chimi Y, Ishikawa N, Kambara T and Ono F 2006 J. of Mag and Mag. Mat 298 14 Matsushita M, Inoue T, Yoshimi I, Kawamura T, Kono Y, Irifun T, Kikgaw T, Ono F 2008 Phy Rev B 77 064429 Matsushita M, Endo S, Miura K, and Ono F 2004 J. Mag and Mag. Mat 269 393 Matsushita M, Ogiyama H and Ono F 2009 J of Mag and Mag. Mat 321 595 Matsushita M, Endo S and Ono F 2007 J of Mag and Mag. Mat 310 1861 Ono F, Matsushima Y, Chimi Y, Ishikawa N, Kambara T, Iwase A 2007 J. Mag and Mag. Mat 310 1864 Ono F, Chimi Y, Ishikawa N, Kanamitsu H, Matsushita Y, Iwase A, and Kambara T 2007 Nuclear Instruments and Methods in Physics Research B 257 402 Pepperhoff W and Acet M 2001 Constitution and Magnetism of Iron and its Alloys, Springer, Berlin p. 106 Rongjin, Huang, Zhixiong W , Xinxin C, Huihui Y, Zhen C and Laifeng L 2010 Solid stat sci 12 1977 Sanyal S and Bose S K 2000 Phy. Rev. B 62 12730 Stolk J and Manthiram A 1999 Mat. Sci and Eng. B 60 112 Tabakovic I, Inturi V, Thurn J and Kief M 2010 electchem. Acta. 55 6749 Ziya A B and Ohshima K 2006 J. alloys and compound 425 123 [1] To whom all correspondence should be addressed; Email: [emailprotected] Tel. No.: +92-61-9239942; Fax: +92-61-9210068