Precipitation Behavior and Properties of Cu-Ti Alloys with Added Nitrogen+1

Jun Ikeda1,+2, Satoshi Semboshi1,2, Akihiro Iwase1, Weilin Gao3 and Akira Sugawara3

1Department of Materials Science, Osaka Prefecture University, Sakai 599-8531, Japan2Kansai Center, Institute for Materials Research, Tohoku University, Sakai 599-8531, Japan3Technology Center, Dowa Metaltech Co. Ltd., Iwata 438-0025, Japan

The effects of adding nitrogen (N) to age-hardenable Cu-Ti alloys on their microstructure, hardness, and conductivity have beeninvestigated. It was found that the aging of Cu-Ti-(0.06­0.6)mol% N alloys resulted in the continuous formation of finely precipitated needle-shaped ¡-Cu4Ti grains and the discontinuous formation of coarse cellular components composed of a stable ¢-Cu4Ti and Cu solid solution at thegrain boundaries, in a manner similar to that in the case of conventional Cu-Ti alloys without any N. Furthermore, a small amount of granularTiN particles was also formed in Cu-Ti-N alloys. The hardening behavior of Cu-Ti-N alloys was similar to that of Cu-Ti alloys without N. Thissimilarity was attributed to the finely dispersed precipitation of ¡-Cu4Ti that was similar between Cu-Ti alloys with and without N. The electricalconductivity of Cu-Ti-N alloys increased more steeply than that of Cu-Ti alloys without N. This is because in the case of Cu-Ti-N alloys, the co-precipitation of ¡-Cu4Ti, ¢-Cu4Ti, and TiN efficiently reduced the Ti concentration in the matrix. Thus, the conductivity of peak-hardened Cu-Ti-N alloys can be improved by optimizing the N concentration and aging temperature. [doi:10.2320/matertrans.M2014216]

(Received August 1, 2014; Accepted December 3, 2014; Published February 25, 2015)

Keywords: copper-titanium alloy, nitrogen, aging, precipitate, hardness, electrical conductivity

1. Introduction

A balance between the electrical conductivity and me-chanical properties of Cu alloys used is required forconnector materials in electrical and electronic components.Among Cu alloys used for connectors, age-hardenable Cu-Tialloys have high tensile strength, good workability, and goodstress relaxation properties.1­3) However, their conductivityis approximately 10­15% IACS (% IACS: a measure ofelectrical conductivity with the International AnnealedCopper Standard at 25°C being 100%IACS) which is lessthan half that of typical alloys such as Cu-Be alloys andCorson alloys.1­4) For this reason, the use of Cu-Ti alloys hasbeen limited to applications that place emphasis on mechani-cal properties.

The Ti concentration in typical age-precipitation-typeCu-Ti alloys is about 3­6mol%. These alloys are quenchedafter a solution treatment at a primary Cu solid solutionrange of 800°C or higher to transform it into a supersaturatedsolid solution, and aging is performed at 400­500°Cthereafter. By this process, a metastable phase, ¡-Cu4Ti(tetragonal; a = 0.5864 nm, c = 0.3649 nm), with a finenanoscale structure is produced in a Cu matrix phase.1­3,5)

In addition, a cellular component having a lamination of a Cusolid solution phase and a stable ¢-Cu4Ti (orthorhombiccrystal; a = 0.4523 nm, b = 0.4341 nm, c = 1.2918 nm)phase at the grain boundaries of the Cu matrix phase underover-aging conditions is generated and grown.5­7) This alloyhas excellent strength, mainly owing to the precipitationstrengthening of fine ¡-Cu4Ti; whereas, over-aging causes thecellular component to become coarse to the micron order,significantly deteriorating the workability and strength of thealloy.7) On the other hand, the conductivity of this alloygradually improves with aging. This is because the amount ofTi solid solution in the Cu matrix phase decreases with the

generation of a cellular component and ¡-Cu4Ti. However,when aging is allowed to proceed so that the conductivityreaches 20% IACS normally, it corresponds to a state of over-aging and the cellular component will develop excessively,thereby deteriorating the mechanical properties. Therefore,the age-precipitation-type Cu-Ti alloys are often used in astate where the hardness has been adjusted to a maximum bypeak-aging (i.e., before over-aging) even if the conductivityis still less than 20% IACS. If it is possible to set theconductivity at 20% IACS or higher while maintaining theexcellent mechanical properties of Cu-Ti alloys, theirapplication can be greatly expanded as they can be used asalternative materials to Cu-Be alloys, which are harmful tohumans despite their high strength and conductivity.

The application of sever plastic deformation process8) orthe addition of a third element9) has been proposed asalternative approaches to improve the balance between themechanical properties and conductivity of Cu-Ti alloys. Inrecent years, Semboshi et al. found that when Cu-Ti alloysare aged in a hydrogen (H) atmosphere, their conductivityincreases rapidly and significantly,10) showing that thebalance between strength and conductivity can be consid-erably improved by optimizing the aging conditions.11,12)

However, safety concerns arise during the manufacturingprocess when utilizing H, so it has not been put into practicaluse yet. Similar to H, nitrogen (N) has been identified as anelement that can improve the conductivity of Cu-Ti alloys.This is because N has high affinity with Ti as in the case withH; thus, it is expected to reduce the Ti solid solution thatremains in the matrix phase. Therefore, this study was aimedat preparing age-hardenable Cu-Ti alloys with added N toinvestigate changes in conductivity and hardness associatedwith changes in its microstructure.

2. Experimental Procedure

Specimens of compositionally controlled homogeneousCu-Ti-N alloys were prepared by the following processes.Pure Ti (99.99%) was subjected to arc melting under an

+1This Paper was Originally Published in Japanese in J. JICu 53 (2014) 62­67.

+2Graduate Student, Osaka Prefecture University

Materials Transactions, Vol. 56, No. 3 (2015) pp. 297 to 302©2015 Journal of Japan Institute of Copper

Ar + 10 vol% N gas atmosphere to prepare alloys having aTi-14.7mol% N (melting point: ³2300°C) chemical compo-sition. After weighing the Ti-14.7mol% N alloy and O-freeCu (purity 99.99%) to obtain a weight ratio of 1 : 1 in orderto lower its melting point so that it would dissolve easily, arcmelting was carried out in an Ar atmosphere to obtain amatrix alloy. This was weighed with O-free Cu and pure Tiand subjected to high-frequency melting to prepare four typesof alloy ingots having the chemical composition listed inTable 1. The four alloys are hereby designated as Cu-Ti, Cu-Ti-0.06N, Cu-Ti-0.3N, and Cu-Ti-0.6N. The alloys werehomogenized at 850°C for 24 h in vacuum; thereafter, theywere cold-rolled and plate-like specimens with the dimen-sions 40­55mm length, 5­7mm width, and 0.5mm thicknesswere cut. In order to obtain a supersaturated solid solution,the specimens were quenched in water after being heated for

3 h at 850°C in vacuum. As an aging treatment, the test pieceswere kept at 420°C and 450°C for 0­450 h in vacuum.

The hardening behavior of the specimens due to agingwas evaluated on the basis of their Vickers hardness. Thehardness of each specimen was measured at 10 points ormore with a load of 300 g and a load time of 10 s. Theirelectrical resistance at room temperature was measured bythe four-probe technique (measured current: 10mA), andtheir conductivity was calculated. The microstructure ofthe specimens was observed with field emission scanningelectron microscopy (FE-SEM) and a transmission electronmicroscopy (TEM). The precipitation phase in the specimenswas collected by a dissolution extraction separation methodcarried out using a nitric acid aqueous solution.13,14) Theaging specimens and precipitate powder acquired by theextraction and separation methods were subjected to X-raydiffraction (XRD) to carry out a structural analysis.

3. Results and Discussion

3.1 MicrostructureFigure 1 shows the FE-SEM images of solution-treated

Cu-Ti, Cu-Ti-0.06N, Cu-Ti-0.3N, and Cu-Ti-0.6N alloys.The Cu-Ti and Cu-Ti-0.06N alloys were single-phase super-saturated solid solutions with a grain size of 50­100 µm(Fig. 1(a), (b)). Massive inclusions of about 1 µm in diameter

Table 1 Chemically analyzed Ti and N concentrations for the specimensprepared in this study.

specimen Ti, mol% N, mol% Cu

Cu-Ti 3.72 <0.001 bal.

Cu-Ti-0.06N 3.51 0.06 bal.

Cu-Ti-0.3N 3.64 0.32 bal.

Cu-Ti-0.6N 3.64 0.60 bal.

Fig. 1 FE-SEM images of solution-treated (a) Cu-Ti, (b) Cu-Ti-0.06N, (c) Cu-Ti-0.3N, and (d) Cu-Ti-0.6N alloys.

J. Ikeda, S. Semboshi, A. Iwase, W. Gao and A. Sugawara298

were partially observed in the matrix phase of the Cu-Ti-0.3Nand Cu-Ti-0.6N alloys (Fig. 1(c), (d), indicated by circles).

Figure 2 shows the FE-SEM and TEM images of the Cu-Ti-0.6N alloy aged at 450°C for 24 h. Cellular componentscan be observed at several locations at the crystal grainboundaries as marked by solid circles in Fig. 2(a). Further-more, the massive inclusions in the grains and needle-shapedcontinuous precipitates about 10 nm in length can also beobserved (Fig. 2(b), (c)). From the selected area electrondiffraction analysis, the needle-shaped precipitates wereidentified as ¡-Cu4Ti, and the massive inclusions wereidentified as TiN (cubical crystal; a = 0.4238 nm) (Fig. 2(d),(e)). Here, it has been confirmed that the cellular componentswere composed of lamination of the ¢-Cu4Ti and Cu solidsolution.5,6,14)

Figure 3 shows the FE-SEM images of the Cu-Ti-0.6Nalloy aged at 450°C for 450 h. The cellular components werefound to occupy almost the entire surface of the specimenaged for 450 h. Needle-shaped ¡-Cu4Ti was also coarsenedowing to aging; however, because of the development of thecellular components during the erosion of ¡-Cu4Ti, almost all¡-Cu4Ti disappeared after 450 h. The size of TiN particlesincreased to several microns, suggesting that they becamecoarser than the solution material and initial aging material(Fig. 1(d) and Fig. 2).

Figure 4 shows the XRD patterns of the Cu-Ti-0.60Nalloys aged at 450°C for 0, 12, and 450 h. In the alloys agedfor 0 h and 12 h, we found only peaks from Cu solid solution,

Fig. 3 FE-SEM images of Cu-Ti-0.6N alloys aged at 450°C for 450 h.Granular particles of TiN are marked by solid circles.

Fig. 2 FE-SEM and TEM images of Cu-Ti-0.6N alloys aged at 450°C for 24 h: (a), (b) low- and high-magnification FE-SEM images,respectively. (c) Bright-field TEM image. (d), (e) Selected area diffraction patterns taken from areas marked as “d” and “e” in (c),respectively. Weak spots marked in (d) are indexed to ¡-Cu4Ti.

Precipitation Behavior and Properties of Cu-Ti Alloys with Added Nitrogen 299

while in the alloys aged for 450 h, peaks from ¢-Cu4Ti aswell as Cu were detected. Extraction and separation werecarried out in order to identify the precipitates in the agedalloys clearly. From the specimen aged for 450 h, cellular andmassive precipitates were collected, and diffraction peaksof ¢-Cu4Ti and TiN were detected in the XRD patterns(Fig. 4(cA)). The structure of the precipitates was thenevaluated by performing the Rietveld analysis. The constit-uent phases and volume fractions in Cu-Ti, Cu-Ti-0.06N, Cu-Ti-0.3N, and Cu-Ti-0.6N alloys aged for 450 h are summa-rized in Table 2. From the results shown, it can be confirmedthat after 450 h of aging, ¡-Cu4Ti was eliminated from everyCu-Ti-N alloy and ¢-Cu4Ti and TiN were formed. Theformation rate of TiN increased with the N concentration inthe Cu-Ti-N alloys.

Massive inclusions about 1 µm in diameter were observedin the solution-treated Cu-Ti-0.3N and Cu-Ti-0.6N alloys(Fig. 1(c), (d)). They were identified as TiN (Fig. 2(c), (e)).Further, since TiN was not observed in the solution-treatedCu-Ti-0.06N alloy, it can be said that the solubility limit ofsolid N in the Cu-3.6mol% Ti alloy was between 0.06 and0.3mol%. In every Cu-Ti-N alloy, similar to the case of theCu-Ti alloys, fine needle-shaped ¡-Cu4Ti was continuouslyprecipitated in the crystal grains during the initial aging, and

simultaneously, a cellular component gradually generated anddeveloped at the crystal grain boundary, especially in over-aging. Further, in Cu-Ti-N alloys, the generation and growthof TiN occurred competitively. It is considered that the Ti andN solid solutions in the Cu matrix phase were graduallyconsumed with the growth of ¡-Cu4Ti, ¢-Cu4Ti, and TiN.

3.2 Hardness and electrical conductivityFigure 5 shows the Vickers hardness and electrical

conductivity of Cu-Ti, Cu-Ti-0.06N, and Cu-Ti-0.6N alloysaged at 450°C and 420°C. Every alloy had a hardness ofabout 120Hv at the time of solution treatment. When thealloys were aged at 450°C, their hardness increased withaging, reaching a maximum at 12­24 h, and decreasedthereafter. When the alloys were aged at 420°C, theirhardness reached the maximum value at 96 h. The hardnessof the alloys decreased as a whole as the N concentration inthe alloys increased at both the aging temperatures.

Regarding electrical conductivity, every alloy had about5% IACS at the time of solution treatment. The conductivityof the all alloys increased with the aging time. In the alloywith higher N concentration, the increasing rate of theconductivity with the aging time became more rapid. Theconductivity of the Cu-Ti-0.6N alloys at the aging time at themaximum hardness was 18% IACS at the aging temperatureof 450°C and 22% IACS at the aging temperature of 420°C.In other words, the conductivity was improved by 4% IACSat the aging temperature of 450°C and 5% IACS at the agingtemperature of 420°C compared to Cu-Ti alloys.

The main factor responsible for the high hardness of Cu-Tialloys is the dispersion of fine needle-shaped ¡-Cu4Ti;1­3,5,14)

the amount, size, and number density of ¡-Cu4Ti are effectivefor the strength. In the Cu-Ti-N alloys, a portion of the Tisolid solution is spent for the generation and growth of TiN,leading to relative decrease in the amount and number densityof ¡-Cu4Ti precipitates, although the size is similar in boththe alloys with and without N. Therefore, the hardnessdecreases slightly with the N concentration of the agedalloys.

Since the effect on the electrical resistance of the Tisolid solution is extremely large with respect to Cu,10) theconductivity of Cu-Ti alloys mainly increases with decreasein the concentration of the Ti solid solution in the Cu matrixphase. In the Cu-Ti-N alloys, the Ti solid solution is spent notonly in the formation of ¡-Cu4Ti and ¢-Cu4Ti but also for theformation of TiN; therefore, the amount of Ti solid solutiondecreases rather efficiently. For this reason, the rate of

Fig. 4 XRD patterns of Cu-Ti-0.6N alloys solution-treated (a), and aged at450°C for 12 h (b) and 450 h (c). XRD pattern of powder extracted fromalloy aged for 450 h is shown in (cA).

Table 2 Constituent phases of precipitates in Cu-Ti and Cu-Ti-N alloys aged at 450°C for 450 h. Their mass fractions and lattice parameters are analyzed byRietveld fitting.

specimen

¡-Cu4Ti TiN ¢-Cu4Ti

mass fraction (%) mass fraction (%)lattice constant (nm)

mass fraction (%)lattice constant (nm)

a a b c

Cu-Ti 0.0 0.0 ® 100.0 0.4525 0.4340 1.2921

Cu-Ti-0.06N 0.0 3.0 0.4229 97.0 0.4525 0.4340 1.2919

Cu-Ti-0.3N 0.0 6.6 0.4229 93.4 0.4525 0.4341 1.2923

Cu-Ti-0.6N 0.0 26.8 0.4229 73.2 0.4524 0.4341 1.2920

J. Ikeda, S. Semboshi, A. Iwase, W. Gao and A. Sugawara300

conductivity increase on aging is higher in the Cu-Ti-Nalloys.

From a practical point of view, the maximum hardness andcorresponding conductivity are important in aged-hardenableCu-Ti alloys. This is because after peak-hardening condition(i.e., under the over-aging condition) cellular componentsdevelop significantly,5,14) which cause to degradation of notonly the strength but also workability and fatigue character-istics. In this study, aging conditions under which hardnesswas highest (250Hv or higher) and the conductivityexceeded 20% IACS were present in both alloys: with orwithout added N, when aging at 420°C: for the Cu-Ti-Nalloys, aging at 420°C resulted in the conductivity exceeding20% IACS (22% IACS for a Cu-Ti-0.6N alloy) at the agingtime corresponding to the maximum hardness. On the otherhand, the conductivity of the Cu-Ti alloys without N did notreach 20% IACS at the aging time corresponding to themaximum hardness, but reached it after over-aging for 240 h.Thus, in the case of Cu-Ti-N alloys, it is possible to realizeaging materials with cellular components not yet developedhaving a Vickers hardness of 250Hv or higher and aconductivity of 20% IACS or higher. Decrease in themaximum hardness due to N addition is an issue that needsto be addressed. However, it can be said that one of theeffective additive elements is N that can improve the balancebetween workability and conductivity, as well as provideoptimum structure control when an appropriate amount of Nis added in the Cu-Ti alloys.

4. Conclusions

In this study, we prepared Cu-Ti-N alloys with constituents

controlled homogenously by using the dissolution method.The precipitation behavior of Cu-Ti-N alloys was evaluatedby investigating the changes in their conductivity andhardness, and observing the texture of aged Cu-Ti-N alloyssubjected to a solution treatment. The following conclusionswere obtained:(1) The solid solubility limit of N for Cu-3.6mol% Ti

alloys was between 0.06 and 0.3mol%.(2) In Cu-Ti-N alloys (similar to Cu-Ti alloys), a

continuous precipitation of fine needle-shaped ¡-Cu4Tiand a discontinuous precipitation of coarse cellular¢-Cu4Ti occurred with aging. In addition, in Cu-Ti-Nalloys, the generation and growth of TiN, with atetragonal structure, occurred.

(3) In Cu-Ti-N alloys, an age hardening curve similar tothat for Cu-Ti alloys was observed. However, theoverall hardness decreased as the amount of added Nincreased. Furthermore, the electrical conductivityimproved as the amount of added N increased.

(4) If the amount of added N and the aging conditions areselected, it is possible to obtain alloys having aconductivity of more than 20% IACS with a Vickershardness of more than 250Hv.

Acknowledgements

The authors are grateful to Prof. T. Takasugi and Prof. Y.Kaneno of Osaka Prefecture University, and Prof. N.Masahashi and Prof. S. Sato of the Institute for MaterialsResearch (IMR) of Tohoku University for their usefuldiscussions and comments. The authors also thank Mr. M.Ishikuro, Mr. E. Aoyagi, and Mr. Y. Hayasaka of the IMR for

150

200

250

300

150

200

250

300

Cu-Ti

Cu-Ti-0.06N

Cu-Ti-0.60N

0

5

10

15

20

25

30

1 10 100 10000

Cu-Ti

Cu-Ti-0.06N

Cu-Ti-0.60N

Aging time, h

Con

duct

ivit

y, %

IA

CS

Vic

kers

har

dnes

s,

(a) 450 °C (b) 420 °C

0

5

10

15

20

25

30

0.1 1 10 100 1000

Cu-Ti

Cu-Ti-0.06N

Cu-Ti-0.60N

Aging time, h

Con

duct

ivit

y, %

IA

CS

Vic

kers

har

dnes

s,

Cu-Ti

Cu-Ti-0.06N

Cu-Ti-0.60N

Fig. 5 Vickers hardness (upper) and electrical conductivity (lower) of Cu-Ti, Cu-Ti-0.06N, and Cu-Ti-0.6N alloys aged at (a) 450°C and(b) 420°C.

Precipitation Behavior and Properties of Cu-Ti Alloys with Added Nitrogen 301

their technical assistance. Financial support provided by theJapan Society for the Promotion of Science (JSPS) via aGrant-in-Aid for Young Research (A) (No. 236860104), andby the Japan Copper and Brass Association is gratefullyacknowledged.

REFERENCES

1) A. Datta and W. A. Soffa: Acta Metall. 24 (1976) 987­1001.2) D. E. Laughlin and J. W. Cahn: Acta Metall. 23 (1975) 329­339.3) W. A. Soffa and D. E. Laughlin: Prog. Mater. Sci. 49 (2004) 347­366.4) S. Semboshi and T. J. Konno: J. Mater. Res. 23 (2008) 473­477.5) S. Semboshi, M. Ishikuro, S. Sato, K. Wagatsuma, A. Iwase and T.

Takasugi: Metall. Mater. Trans. A 45 (2014) 3401­3411.6) R. C. Ecob, J. V. Bee and B. Ralph: Metall. Trans. A 11 (1980) 1407­

1414.7) R. C. Ecob, J. V. Bee and B. Ralph: Phys. Status Solidi (a) 52 (1979)

201­210.8) H. Miura, T. Namiki and M. Kobayashi: J. Japan Inst. Copper. 53

(2014) 50­54.9) R. Markandeya, S. Nagarjuna and D. S. Sarma: Mater. Sci. Eng. A 371

(2004) 291­305.10) S. Semboshi and T. J. Konno: J. Mater. Res. 23 (2008) 473­477.11) S. Semboshi, T. Nishida, H. Numakura, T. Al-Kassaab and R.

Kirchheim: Metall. Mater. Trans. A 42 (2011) 2136­2143.12) S. Semboshi, S. Orimo, H. Suda, W. L. Gao and A. Sugawara: Mater.

Trans. 52 (2011) 2137­2142.13) S. Semboshi, M. Ishikuro, S. Sato, K. Wagatsuma and T. Takasugi:

Mater. Charact. 82 (2013) 23­31.14) S. Semboshi, M. Ishikuro, S. Sato, K. Wagatsuma, A. Iwase and T.

Takasugi: Metall. Mater. Trans. A 45 (2014) 3401­3411.

J. Ikeda, S. Semboshi, A. Iwase, W. Gao and A. Sugawara302