Field of the invention The present invention relates to an alloy with the features of the preamble of claim 1 or 19, to the use of such an alloy, and to articles manufactured from such an alloy, in particular implants such as endoprostheses.

Background of the invention A number of alloys may be brought into a glassy state, i.e., an amorphous, non- crystalline structure, by splat cooling at very high cooling rates, e.g., 106 K/s. However, most of these alloys cannot be cast into a bulk glassy structure at much lower cooling rates achievable with casting.

In recent years, many bulk metallic glass-forming liquids have been discovered for which cooling rates of less than 1000 K/s are sufficient for vitrification. For the purposes of this document, a "bulk metallic glass" is to be understood as an alloy which develops an at least partially amorphous structure when cooled from a temperature above the melting point to a temperature below the glass- transition temperature of the amorphous phase with a cooling rate of 1000 K/s or less, preferably with a cooling rate of 100 K/s or less. Cooling rates in this range are typically experienced in bulk casting operations.

Bulk metallic glasses generally have mechanical properties that are superior to their crystalline counterparts. Due to the absence of a dislocation mechanism for plastic deformation, they often have a high yield strength and elastic limit. Furthermore, many bulk metallic glasses show good fracture toughness, corro¬ sion resistance, and fatigue characteristics. For an overview of the properties and areas of application of such materials see, for example, Johnson WL, MRS Bull. 24, 42 (1999) and Lόffler JF, lntermetallics 11 , 529 (2003). Reference is made explicitly to the disclosure of these documents and the references cited therein for teaching properties of glass-forming metallic alloys and methods for the determination of such properties. Commercial applications of bulk metallic glasses are described, e.g., in Buchanan O, MRS Bull. 27, 850 (2002).

Currently, only Zr-based bulk metallic glasses (and some Pt-based glasses for jewelry) have found their way into applications. The following documents of the prior art deal with Zr-based glass-forming alloys: US Patent No. 5,740,854 discloses an alloy of composition

US Patent No. 5,288,344 discloses alloys of general composition Zr-Ti- Cu-Ni-Be. Specifically, the alloy Zr4i.2Tii3.8Cui2.sNiioBe22.s, which has be- come known under the trade name Vitreloy 1 ™ or Vit1 ™, and Zr46.75Ti8.8NiioCu7.5Be27.5, which is known under the trade name Vitreloy 4™ or Vit4™, are disclosed in that document. US Patent No. 5,737,975 discloses alloys of the general composition Zr- Cu-Ni-Al-Nb. Specifically, an alloy of composition Zr57CUi54NiI2 GAIiONb5, which is known under the trade name Vitreloy 106™ or as Vit106™, is dis¬ closed in this document. Lin X H, Johnson W L, Rhim W K, Mater. Trans. JIM 38, 473 (1997)) dis¬ closes the alloy Zr52 5Ti5Cui7.9Nii4.6Alio, also known as Vit105™. Lόffler JF, Bossuyt S, Glade SC, Johnson WL, Wagner W, Thiyagarajan P, Appl. Phys. Lett. 77, 525 (2000) and Lόffler JF, Johnson WL, Appl. Phys. Lett. 76, 3394 (2000) describe comparative investigations of Vit1 ™, Vit105TM and Vit106™. Kϋndig AA, Lόffler JF, Johnson WL, Uggowitzer PJ, Thiyagarajan P, Scr. mater. 44, 1269 (2001 ) describes alloys of the general formula Zr52.5Cui7.9Nii4.6Alio-χTi5+χ, i.e., alloy compositions which have been varied in the vicinity of the composition of Vit105™. lnoue A, Shibata T. and Zhang T., Mater. Trans. JIM 36, 1426 (1995) dis- closes alloys of composition Zrβs-xTixAlioCuisNiio. Zhang T, lnoue A, Mater. Trans. JIM 39, 1230 (1998) discloses alloys of composition Zr7o-x-yTiχAlyCu2oNiio. Xing LQ, Ochin P, Harmelin M et al, Mat. Sci. Eng. A220, 155 (1996) dis- closes, inter alia, an alloy of composition Zr57Cu2OAIiONJeTi5, as well as other Zr-Cu-Al-Ni-Ti alloys. Lόffler JF, Thiyagarajan P, Johnson WL1 J. Appl. Cryst 33, 500 (2000) describes Zr-Ti-Cu-Ni-Be alloys whose (Zr, Ti) and (Cu, Be) contents were varied between the compositions of Vit1 ™ and Vit4™. - lnoue A, Zhang T, Nishiyama N, Ohba K, Masumoto T, Mater. Trans. JIM 34, 1234 (1993) discloses an alloy of composition Zr65AI7-5CUi75Ni1O.

According to the following documents, the addition of Fe to an Zr-Al-Ni-Cu alloy was believed not to improve or to even decrease the glass-forming ability: - lnoue A, Shibata T, Zhang T, Mater. Trans. JIM 36, 1420 (1995). Eckert J, Kubler A, Reger-Leonhard A et al, Mater. Trans. JIM 41 , 1415 (2000). Mattern N, Roth S, Kuhn U et al, Mater. Trans. JIM 42, 1509 (2001 ).

Due to their favorable mechanical properties, bulk metallic glasses are interest¬ ing candidate materials for biomedical applications. However, most known glass-forming alloys, especially Zr-based alloys, contain a considerable propor¬ tion of nickel (Ni). Exposure to nickel is known to possibly cause allergies. Therefore these alloys are not well suited for medical applications, in which the alloy can come into contact with body fluids, with the skin, with tissue or other body parts. Specifically, these alloys may cause allergic reactions because they tend to release small amounts of nickel when they come into a prolonged con¬ tact with the body. Copper (Cu) may also be problematic, albeit to a lesser ex¬ tent.

Fan C, lnoue A, Mater. Trans. JIM 38, 1040 (1997) describes the improvement of mechanical properties by precipitation of nanoscale compound particles in Zr-Cu-Pd-Al amorphous alloys. However, these alloys are not bulk metallic glasses; they are only amorphous when using melt spinning or splat quenching.

Summary of the invention It is therefore an object of the present invention to provide an alloy which has good glass-forming ability and an improved biocompatibility, in particular, an alloy which does not release nickel in contact with body liquids.

This object is achieved by an alloy with the features of claim 1.

It is another object of the present invention to provide an alloy which has good glass-forming ability and an improved biocompatibility, in particular, an alloy which is essentially free of both copper and nickel.

This object is achieved by an alloy with the features of claim 19.

Thus, an alloy is provided which contains at least four components A, D, E and G. Optionally, a fifth component Z may be present. The alloy preferably has a bulk structure containing at least one amorphous phase, i.e., a volume fraction of at least 10%, preferably at least 50% of the alloy is amorphous. In the context of this document, a structure is considered to be fully amorphous if the material having this structure does not exhibit significant Bragg peaks in an X-ray diffrac¬ tion pattern. Accordingly, the volume fraction of the amorphous phase in a mixed-phase material may be estimated by integrating the intensity of Bragg peaks and comparing with the intensity of non-Bragg features.

Preferably, the amorphous phase can be obtained by cooling from a tempera¬ ture above the melting point to a temperature below the glass-transition tem- perature of the amorphous phase with a cooling rate of 1000 K/s or less, i.e., preferably the alloy is a bulk metallic glass. More preferably, the amorphous phase can be obtained by cooling with a cooling rate of 100 K/s or less. This enables the material to be formed by casting, in particular copper-mold casting. In other words, preferably the alloy with at least one amorphous phase can be obtained in a shape with dimensions of at least 0.1 mm, preferably at least 0.5 mm, more preferred at least 1 mm in any spatial direction. This is not possible for alloys which adopt an amorphous structure only at cooling rates as achiev¬ able by splat cooling or melt spinning.

Component A consists of at least one element selected from the group consist¬ ing of Zr (zirconium), Hf (hafnium), Ti (titanium), Nb (niobium), La (lanthanum), Pd (palladium) and Pt (platinum). The other components D, E, G and, option¬ ally, Z are all different from each other and from component A. Each of these components may consist of more than one element, as long as all elements of all components are different. Preferably, however, components D, E and G each consist of a single element. The alloy composition follows an "80:20 scheme", i.e., the ratio of the combined atomic content of components A and D to the combined atomic content of components E and G is approximately 80 to 20, within a band of plus or minus 10, preferably a band of plus or minus 5, in particular a band of plus or minus 2.

Expressed as a chemical formula, the alloy composition is [(AxDi 00-x)a( EyG 1 oo-y)i 00-a] 100-bZb, where x, y, a and b are independent numbers selected from zero and the posi¬ tive real numbers and denote atomic percentages, with 70 < a < 90 , preferably 75 < a < 85 , more preferred 78 < a < 82. The following example is meant to H- lustrate the meaning of the term "atomic percentage": Before multiplying indices outside and inside of brackets, the indices inside the brackets should be divided by 100, e.g., (Zr72 5Cu275)So(Fe40AI6O)2O = Zr58Cu22Fe8AI12. After all brackets have been removed, each index indicates the number of atoms contributing to a for¬ mula unit of the alloy. In the present example, 58 atoms of Zr would be com- bined with 22 atoms of Cu, 8 atoms of Fe and 12 atoms of Al in order to arrive at one formula unit. In other words, if a number is an "atomic percentage", this means that the number, when divided by 100, indicates the stoichiometry in the sense as it is usually understood in chemistry.

Component A is the main component of the alloy, in the sense that x > 50 . In order to have a significant content of component D, preferably x < 95 and more preferably x < 90 . Advantageously, the content of component G relative to component E is not too small, preferably y > 5 , more preferred y > 10. On the other hand, the content should not be too large. Preferably y < 95 , more pre¬ ferred y < 90. If a fifth component Z is present at all, then it is present in a com¬ paratively small proportion only. In numbers, 0 < b < 6 , preferably 0 < b < 4 , more preferably 0 < b < 2. The numbers x, y, a and b are generally independent of each other.

Importantly, the alloy is substantially free of nickel. In the context of this docu¬ ment, "substantially free of nickel" means that the total nickel content of the al- loy is less than 1 atomic percent, preferably less than 0.1 atomic percent. It may even be required that the nickel content is below 10 atomic ppm, e.g., in medi¬ cal applications. In particular, none of the components A, D, E, G or Z should comprise nickel.

Preferably, components A and E are miscible in a wide composition and tem¬ perature range. The term "wide composition and temperature range" is to be understood as a range extending over a temperature range of at least 600 K and over a range of compositions spanning at least 60 at.% of either component in the liquid state and below the liquidus temperature in the A-E phase diagram. In the present example, a wide composition range would, e.g., be the range from 20 at.% to 80 at.% of component A in the binary mixture A-E.

More preferably, components A and E are capable of forming a deep eutectic composition in the absence of other components. The term ..capable of forming a deep eutectic composition" is to be understood as meaning that, if A and E are mixed in the melt in the absence of other components, there is a composi¬ tion for which A and E are miscible down to the liquidus temperature, and the liquidus temperature of the mixture for that composition has a local minimum as a function of composition. In other words, when varying the composition in a small vicinity of a deep eutectic, the liquidus temperature is higher than at the composition of the deep eutectic itself. Often, the liquidus temperature of the binary mixture at the deep eutectic will additionally be lower than the melting point of each of the components taken alone. As an example for a very deep eutectic, for A = Zr, the melting temperature is rm(Zr) = 2128 K, for E = Fe, it is Tm(Fe) = 1811 K; an eutectic occurs at 1201 K = 0.66 Tm(Fe); likewise, for 7m(Au) = 1337 K, 7m(Si) = 1687 K, and an eutectic is at 636 K = 0.47 Tm(Au).

Preferably, the components are chosen such that a deep eutectic composition of the A-E mixture occurs at a composition A9 EiOO a' with 70 < a'< 90 , preferably 75 < a'< 85. Then the number a is preferably chosen such that the absolute value of the difference between a and a' is smaller or equal to 10 (i.e., |a - a'| < 10 ), preferably |a - a'| < 5.

Preferably, also components A and D are miscible over a wide temperature and composition range. More preferably, they are capable of forming a deep eutec¬ tic composition when mixed in a binary mixture. If components A and D form a deep eutectic composition at AxDiOo-x', then x is preferably chosen such that |x - x'l < 10 , more preferably |x - x'| < 5.

Preferably, component G is miscible with component E over a wide temperature and composition range, in particular if E is at least one element selected from the group consisting of the transition metals, in particular the group consisting of Fe and Co. It is then preferred that G is capable of forming a deep eutectic composition with component A.

More preferably, components G and E are capable of forming a deep eutectic composition at EyG10O-V1- Tnen y is preferably chosen such that |y — y'j < 10 ,

more preferably jy — y'j < 5. Alternatively or additionally, A and G are preferably capable of forming a deep eutectic composition.

Preferably, the atomic Goldschmidt radius of each element in component A is relatively large, at least 0.137 nm, preferably at least 0.147 nm, more preferred at least 0.159 nm. In particular, if the atomic Goldschmidt radius of each ele¬ ment in component A is at least 0.159 nm, then preferably 70 < a < 90 , if this radius is at least 0.147 nm, then preferably 75 < a < 85 , and if this radius is at least 0.137 nm, then preferably 78 < a < 82. In particular, this means that for Zr-, Hf-, and La-based alloys, preferably 70 < a < 90 ; for Ti- and Nb-based al¬ loys, preferably 75 < a < 85 ; and for Pt- and Pd-based alloys, preferably 78 < a < 82.

The components A, D, E and G may have similar atomic radii and atomic prop¬ erties. However, it is preferred that the atomic radius of each element in com- ponent E is smaller than the atomic radius of each element in component A.

The atomic (Goldschmidt) radii of the elements can be found tabulated in stan¬ dard textbooks or in the 2004 Goodfellow Catalog, available from Goodfellow Inc., Huntingdon, U.K. In particular, for selected elements, reference is made to Table 1 below.

Table 1: Atomic Goldschmidt radii of selected elements

In general terms, component D is preferably at least one element selected from the group consisting of Cu (copper), Be (beryllium), Ag (silver) and Au (gold). Specifically, if component A is at least one element selected from the group consisting of La (lanthanum), Pd (palladium) and Pt (platinum), component D is preferably Cu (copper). If A is at least one element selected from the group consisting of Zr (zirconium), Hf (hafnium) and Ti (titanium), then D is preferably Cu (copper) or Be (beryllium). Both copper and beryllium have deep eutectics with Zr, Hf and Ti.

In general terms, component E is preferably at least one metal selected from the group consisting of the transition metals except Ni (nickel); particularly Sc (scandium), Ti (titanium), V (vanadium), Cr (chromium), Mn (manganese), Fe (iron), Co (cobalt), Zn (zinc), Y (yttrium), Mo (molybdenum), Ta (tantalum), and W (tungsten). A transition metal is defined as any of the thirty chemical ele¬ ments with atomic number 21 through 30, 39 through 48, and 71 through 80. These metals are preferred because of their tendency to form deep eutectics with component A and because of their specific electronic properties. In particu¬ lar, component E is preferably at least one metal selected from Fe (iron) and Co (cobalt). These metals have empirically been found to be preferred.

Component G is preferably at least one element selected from the group con- sisting of Al (aluminum), Zr (zirconium), P (phosphorus), C (carbon), Ga (gal¬ lium), In (indium) and the metalloids, particularly B (boron), Si (silicon), and Ge (germanium). The known metalloids are B (boron), Si (silicon), Ge (germanium), As (arsenic), Sb (antimony), Te (tellurium), and Po (polonium). It is believed that the specific electronic properties of these elements favorably influence the glass-forming ability. Furthermore, the elements B, P, C, and Si have particu- larly small atomic sizes (< 0.117 nm), which contributes to a large size differ¬ ence between the components A and G. In particular, if component E is Fe (iron), component G is preferably selected from the group consisting of Al (aluminum), Zr (zirconium), P (phosphorus), B (boron), Si (silicon) and C (car¬ bon). More preferred, if component E is Fe (iron), then component G is Al (alu- minum). Then y is advantageously chosen to be in the range from about 30 to about 50, in particular approximately 40. Alternatively, if component E is Co (cobalt), component G is preferably at least one element selected from the group consisting of Zr (zirconium), Al (aluminum), B (boron), Si (silicon), Ge (germanium), Ga (gallium) and In (indium).

In a preferred embodiment, component A is Zr (zirconium) or a mixture of Zr (zirconium) with either Hf (hafnium) or Ti (titanium) or both wherein at least 80 atomic percent of component A is Zr (zirconium). It is then preferred that com¬ ponent D is Cu (copper). It has been found empirically that this combination leads to alloys with superior glass-forming ability.

If component A is Zr and component D is Cu, it is preferred that x is chosen be¬ tween 62 and 83 (i.e., 62 < x < 83 ), preferably 68 < x < 77 , in particular that x is approximately 72.5. If component A is Zr and component D is Cu, it is further preferred that component E is Fe (iron) and component G is Al (aluminum). Then y is advantageously chosen to be in the range from about 30 to about 50, in particular approximately 40. Alloys of this composition, specifically, the alloy compositions in the vicinity of Zr58Cu22FeSAh2, have been found by the inven¬ tors to belong to the best glass formers known to date.

If a fifth component Z is present, this component is preferably at least one ele¬ ment selected from the group consisting of Ti, Nb, Hf. Alternatively, component Z may preferably be at least one element selected from the group consisting of the transition metals, or component Z may preferably be at least one element selected from the group consisting of Be (beryllium), Y (yttrium), Pd (palladium), Ag (silver), Pt (platinum), and Sn (tin). In general terms, component Z is pref- erably capable of forming a deep eutectic composition with component A.

The alloy may have a structure comprising at least one amorphous phase and at least one crystalline phase. The volume fraction of the amorphous phase preferably is at least 10%. The amorphous and crystalline phases should not be macroscopically separated. Such a structure can be generated by different means. In one approach, a composite comprising crystals embedded in an amorphous matrix is produced by subjecting the alloy to heat treatment at a temperature above the glass transition temperature. For details, see the de¬ scription of the preferred embodiments below. In another approach, the alloy is subjected to electric currents, as described, e.g., in (Holland TB, Lόffler JF, Mu- nir ZA, J. Appl. Phys. 95, 2896 (2004)), who describe the crystallization of me¬ tallic glasses under the influence of high density DC currents. In still another approach, the alloy composition in the melt is chosen to be initially outside the glass-forming region. During cooling, crystals start forming in the melt. This al- ters the composition of the mixture remaining in the melt, which is shifted into the glass-forming region. Upon further cooling, a glassy matrix with embedded crystals is formed. For details, see (Hays CC, Kim CP, Johnson WL, Phys Rev. Lett. 84, 2901 (2000)). In yet another approach, development of crystals in the amorphous matrix is fostered by a suitable choice of the fifth component Z. Suitable components Z are preferably at least one element selected from the group consisting of Ti, Nb, Ta, or at least one element selected from the group consisting of the transition metals, or at least one element selected from the group consisting of Be and Pd. For details, see, e.g., (He G, Eckert J, Loser W, Schultz L, Nature Materials 2, 33 (2003)).

In a preferred embodiment, A is Zr (zirconium) and D is selected from the group consisting of Cu (copper) and Fe (iron). Specifically, it is preferred that A is Zr (zirconium), D is Cu (copper), and E is selected from the group consisting of Fe (iron) and Co (cobalt). Then G is pref¬ erably at least one element selected from the group consisting of Al (aluminum) and the metalloids. A particularly preferred system is the Zr-Cu-Fe-Al system, i.e., A is Zr (zirconium), D is Cu (copper), E is Fe (iron) and G is Al (aluminum). It has been found that alloys of this composition, when following the 80:20 con¬ cept, have favorable glass-forming properties.

If A is Zr (zirconium) and D is Cu (copper), it is preferred that the ratio of these is chosen according to 62 < x < 83 if E jS Fe (iron) and F is Al (aluminum), it is preferred that their ratio is chosen according to 30 < y < 50 jne combination of these ranges, together with the general 80:20 concept, defines a region of quaternary compounds with exceptionally good glass-forming properties.

In particular, the alloy may substantially be represented by the formula (ZrxCuioo-x)8o(Fe4oAl6o)2o with 62 < x < 83 , in particular, with x = 62, 64, 66, 68, 72.5, 77, 79, 81 or 83, or by one of the formulas (Zr95Ti5)72Cui3Fei3AI2, Zr7OCUi3FeI3AI3Sn1, Zr7OCUi3FeI3AI2Cr2, Zr70CUi3FeI3AI2Nb2, Zr70CUi3FeI3AI2Zn2, (Zr72Cu13FeI3Ab)98Mo2, (Zr72Cui3Fei3AI2)98P2, (Z95Hf5)72Cui3Fe13AI2, Zr70CUn Fen Al8, Zr71Cu1 IFe10AI8, (Zr74Cu13Fe1 S)90AI10, Zr72Cu13Fe13AI2, (Zr74Cu13Fei 3)98AI2) Zr73Cu13FeI3AIi, Zr72CUi3Fe13AI2, Zr71Cu13Fe13AI3, Zr72Cu12Fe12AI4, Zr70CUi3Fe13AI4, Zr72CUi1Fe11AI6, Zr72Cu11 5Fe1 IAI5 5, Zr73Cu11Fe11AI5, Zr71Cu11Fe11AI7, Zr69Cu11Fe11AI9, Zr70Cu10 5Fe10 5AI9, Zr70Cu10FeH Al9, Zr70Cu11Fe10AI9, Zr69CUi0Fe10AI11, Zr69CUi0FeI 1AI10, Zr70Cu13FeI3AI2Sn2, Zr72Cu13Fe13Sn2, (Zr74Cu13Fe13)98Sn2, (Zr79Cu2I )βo( Fe40AI60)20) (Zr8iCui9)80(Fe40AI60)20) (Zr83Cui7)80(Fe4oAI6o)20, (Zr66Cu34)8O(Fe4oAI6o)2o, (Zr64Cu36)8o(Fe4OAI6o)2O, and (Zr62Cu38)80(Fe40AI60)20.

Another system having excellent glass-forming properties if following the 80:20 concept is the Zr-Fe-AI-(Pd/Pt) system. This system has the additional advan¬ tage that it is free of copper. In other words, preferably A is Zr (zirconium), D is Fe (iron), E is Al (aluminum), and G is one or both elements selected from Pd (palladium) and Pt (platinum). Specifically, excellent glass formers have been found if G is palladium, while a slightly improved biocompatibility may result by partially or fully replacing Pd by Pt. In this connection, it is to be noted that Pd and Pt are known to occupy the same group of the periodic system of elements, and have a similar (outer-shell) electronic structure, almost the same GoId- schmidt radius and a similar chemical behaviour. It is therefore to be expected that Pd may be replaced by Pt without dramatic changes in the glass-forming properties of the alloys. In these systems, it has been found to be advantageous if the atomic percentages of Fe and Al are substantially equal. A range of good glass formers was found for 68 < x < 89 and 73 < a < 87 . Particularly good re¬ sults were achieved for 81 < x < 85 , 80 < a < 83 , and 65 < y ≤ 80 , in particular if G was Pd. The ratio of Al to Pd/Pt is favourably chosen according to 40< y < 82.

Generally, it is preferred that only small amounts of additional elements are pre¬ sent, i.e., 0 < b ≤ 2. In particular, it is preferred that b=0, i.e., that there are sub¬ stantially at most trace amounts of additional elements present. If such ele¬ ments are present, i.e., if b > 0 , then Z is preferably at least one element se- lected from the group consisting of Ti, Hf, V, Nb, Y, Cr, Mo, Fe, Co, Sn, Zn, P, Pd, Ag, Au and Pt.

Expressed in another way, Zr-Fe-AI-Pd/Pt system has been found to have good glass-forming properties if conforming to the general formula

wherein X is one or both elements selected from Pd and Pt, a, b, c and ε are zero or real positive numbers signifying atomic percentages, and ε < 10 , i ≥ 50 , j ≥ \9 , £ > 0.5 and i + j + k = 100. Excellent glass-forming abilities were achieved in examples where X was Pd, while a slightly improved biocompatibil- ity may be expected by partially or fully replacing Pd by Pt, which has very simi¬ lar properties as Pd. Preferred ranges are (independently or in combination) βl ≤ i ≤ ll , 19 < y < 34 , and ε < 2. Preferably, ε is substantially zero, i.e., the atomic percentages of Fe and Al are approximately equal. For the best glass formers which have been found in this system, ε is substantially zero, 66 < i < 70 , 25 < j ≤ 29 and 4 < k ≤ 7 . The best glass formers of this system also conform to the 80:20 concept as described above.

In particular, alloys being substantially represented by one of the following for¬ mulas were found to be good glass formers: An alloy represented by one of the formulas Zr67Fei3.2Ali3.2Pd6.6, Zr69-7Fe12-95AI^ 95Pd4-41 Zr66-7FeI4-45AI14-45Pd4-4, Zr68.3Fei3.4Ali3.4Pd4.9, Zr65-4Fe14-85AI14-85Pd4-9, Zr62.3Fe-|6.7Ali6.7Pd4.3, Zr59.2Feiβ.3Ali8.3Pd4.2. Zr72Feii.SAlii.5Pd5, Zr73.4Fβio.9Alio.9Pcl4.8, Zr75.2Feio.2Alio.2Pd4.3, Zr77Fe9-5AIa5Pd41 Zr65Fe1 i.4Alii.4Pdi2.2, Zr62.5Feio.7sAlio.7sPdi6, by the formula Zη(Fe5oAI5O)3oPd7o-i with 62 < / < 69.5 , in particular by one of the formulas Zr69 5Fe15AI15Pd0 5, Zr69Fe15AI15Pd0 5, Zr68Fe15AI15Pd2, Zr67Fe15AI15Pd3, Zr66Fe15AI15Pd4, Zr65Fe15AI15Pd5, Zr64Fe15AI15Pd6, Zr63Fe15AI15Pd7, Zr62Fe15AI15Pd8, or by one of the formulas Zr71Fe12AI12Pd5, Zr69Fe12 85AI12 85Pd5 3, Zr66-8Fei3-7Ali3-7Pd5-8, Zr65Fe14-5AI14-5Pd6, Zr61 -9Fe16-2AI16-2Pd5-7, Zr50Fe12AI12Pd26, Zr53-2Fe12-6AI12-6Pd21 -6, Zr57 6Fe13 95AI13 95Pd14-5, Zr60Fe14-3AI14-3Pd1 I4.

Preferably, the alloy has a structure comprising at least one amorphous phase and at least one crystalline phase. The at least one amorphous phase is pref¬ erably obtainable by cooling from a temperature above the melting point of the alloy to a temperature below the glass-transition temperature of the amorphous phase at a cooling rate of 1000 K/s or less, i.e., the alloy is preferably a bulk metallic glass.

The present invention is further directed at a method of manufacture of the in- ventive alloys. The method comprises preparing a melt of aliquots of A, D, E, G, and optionally Z, and cooling the melt from a temperature above the melting point to a tern- perature below the glass-transition temperature of the amorphous phase with a cooling rate of 1000 K/s or less to obtain a solidified material. Preferably, the method comprises casting of the melt into a mold, in particular, a copper mold.

Alternatively, the inventive alloys may be produced by mechanical alloying, as described, e.g., in (Eckert J, Mater. Sci. Eng. A 226-228, 364 (1997): Mechani¬ cal alloying of highly processable glassy alloys). Mechanical alloying means mechanical processing of the alloy or its constituents in the solid state, without passing through the liquid state. In particular, by mechanical alloying of, e.g., a crystalline powder, an amorphous metallic alloy may be obtained. Suitable me¬ chanical alloying methods include, but are not restricted to, ball milling. For de¬ tails, explicit reference is made to the teachings of the above-mentioned Eckert paper.

The method may additionally comprise a step of processing the alloy above the glass transition temperature, e.g., for obtaining a mixed-phase material. In par¬ ticular, the method may comprise a step of heat-treating the solidified material for a few minutes up to 15 hours at a temperature below the first crystallization temperature or for a few seconds up to 2 hours at a temperature above the first crystallization temperature. The first crystallization temperature is the tempera¬ ture of the first exothermic feature in a DTA scan of the amorphous alloy when the temperature is raised from the glass transition temperature. Heat treatment at relatively low temperatures results in slow kinetics, which is believed to lead to the formation of small crystals. For details, see the description of the pre- ferred embodiments below.

For obtaining material with specific surface properties, the alloy may be sub¬ jected to a microstructuring process as described, e.g., in (Kundig AA, Cucinelli M, Uggowitzer PJ, Dommann A, Microelectr. Eng. 67, 405 (2003): Preparation of high aspect ratio surface microstructures out of a Zr-based bulk metallic glass) or in the patent application PCT/CH 2004/000401. The content of these documents is incorporated herein by reference in its entirety. Microstructuring may be achieved by casting the liquid alloy into a mold having itself a micro- structured surface. For details, reference is made to the teachings of the above- mentioned Kundig et al. paper and to PCT/CH 2004/000401. In a different em¬ bodiment, an already solidified alloy is brought into a superplastic state, i.e, into a state in which it can be easily shaped, by heating the alloy to a temperature above the glass-transition temperature, and is pressed onto a microstructured matrix. For details, reference is made to PCT/CH 2004/000401. In an advanta¬ geous embodiment, the microstructured mold resp. matrix is a silicon wafer which has been structured by etching, as it is well known in the art. In yet an- other embodiment, the liquid alloy is drawn into a system of capillaries by the capillary effect and rapidly solidified within the capillaries. For details, reference is made to the teachings of the application PCT/CH 2004/000401.

The invention is also directed at the use of an inventive alloy for the manufac- ture of an article destined to be brought into contact with the human or animal body. In particular, the invention is directed at the use of such an alloy for the manufacture of a surgical instrument, a jewelry item, in particular a watch case, or a prosthesis, in particular an endoprosthesis, specifically, a so-called stent. A stent is an endoprosthesis for insertion into a blood vessel, lining the inner sur- face of the vessel. Stents are used in particular for ensuring sufficient blood flow through the vessel, or for stabilizing the blood vessel to prevent aneurisms. Other implants for which the inventive alloys can be used are in the field of os¬ teosynthesis, e.g., hip implants, artificial knees, etc. The present invention is also directed at an endoprosthesis, in particular a stent, manufactured from an inventive alloy.

The inventive alloys are particularly suited for such biomedical applications due to their good biocompatibility, high strength and high elasticity. In particular, the inventive alloys of general composition Zr-Cu-Fe-Al or Zr-Fe-Al-Pd are well suited for these purposes. Brief description of the drawings The invention will be described in more detail in connection with an exemplary embodiment illustrated in the drawings, in which Fig. 1 shows a strongly simplified, schematic phase diagram of a binary Zr-Fe alloy; Fig. 2 shows a strongly simplified, schematic phase diagram of a binary Cu-Zr alloy; Fig. 3 shows a strongly simplified, schematic phase diagram of a binary Fe-Al alloy together with the £^phase; Fig. 4 shows XRD patterns of as-cast 1 mm x 1 cm2 alloys of composi¬ tion Zr54.4Cu25.6Fe8Ali2, Zr58Cu22Fe8AIi2, and Zr6I eCUi84Fe8AIi2; Fig. 5 shows SANS intensity data of as-cast 1 mm x 1 cm2 alloys of composition Zr54-4Cu256Fe8AIi2, Zr58Cu22Fe8AIi2, and Zr6LeCUi8-4Fe8AIi2 (wave number Q = 4π s\r\θ/λ, with θ = half the scattering angle and λ = wavelength of neutrons); Fig. 6 shows DTA scans on samples of composition Zr544Cu25-6Fe8AIi2, Zr58Cu22Fe8AIi2, Zr6L6CUi84Fe8AIi2, and Zr65AI75NiI0CUi7 5, per¬ formed with a heating rate of 20 K/min (T9 = glass transition, 7χi = first crystallization temperature); Fig. 7 shows a DTA scan of Zr58Cu22Fe8AIi2, performed with a heating rate of 20 K/min; Fig. 8 shows a photograph of cast samples of composition Zr58Cu22Fe8AIi2 together with a ruler illustrating their actual size; Fig. 9 shows XRD patterns of Zr58Cu22Fe8AIi2 cast to cylindrical rods of diameters 5, 7 and 8 mm, and to a plate of 1 mm thickness (inset); Fig. 10 shows DTA scans of Zr58Cu22Fe8AI12 cast to cylindrical rods of diameters 5, 7 and 8 mm (heating rate 20 K/min); Fig. 11 shows XRD patterns of Zr544Cu256Fe8AIi2 cast to a cone with outer diameter 6 mm; Fig. 12 shows a DTA scan of Zr6LeCUi84Fe8AIi2, performed with a heating rate of 20 K/min; Fig. 13 shows a SEM image showing the fracture surface of glassy Fig. 14 shows a room-temperature tensile stress-strain curve of an as- cast cylindrical Zr58Cu22Fe8AI12 sample with a diameter of 5 mm; Fig. 15 shows XRD patterns of Zr58Cu22Fe8AI12 in the as-prepared state and after annealing for several hours at different temperatures; Fig. 16 shows an XRD pattern (72 hours scan) of Zr58Cu22Fe8AI12 after annealing at 708 K for 12 h. The indexing shows an icosahedral phase with a lattice constant of 4.76 A; Fig. 17 shows DTA scans of Zr58Cu22Fe8AI12 in the as-prepared state and after annealing for several hours at different temperatures, as indi¬ cated in the figure (heating rate 20 K/min); Fig. 18 shows SANS intensity data of Zr58Cu22Fe8AI12 obtained from in- situ SANS measurements performed at a temperature of 708 K at different times, as indicated in the figure; Fig. 19 shows the time evolution of the particle size, Φ, of Zr58Cu22Fe8AI12 using the Guinier approximation; Fig. 20 shows a pseudoternary mixing diagram; Fig. 21 shows a DTA scan of the alloy Zr68^ Fβo.sAlo.sWβPd-i.θ cast to a thickness of 1 mm; and Fig. 22 shows an X-ray diffraction pattern of the alloy Zr68.3(Feo.5Alo.5)26.8Pd4.9 cast to a thickness of 1 mm.

Detailed description of the invention Before describing specific examples of inventive alloys and their characteriza- tion, the concept which led to the development of the inventive alloys shall be described and exemplified.

Many binary alloys which form metallic glasses when splat-cooled have the composition A8oX2o, where the atomic radius of A is significantly larger than that of X. The good glass-forming ability of such alloys with large size ratio has been explained by topological effects. In the present invention, this "80-20 concept" has been generalized to quaternary or higher-component alloys and has been successfully applied for developing Ni-free bulk metallic glasses. It has surpris¬ ingly been found that alloys with exceptionally good glass-forming ability result when following the principles laid down in claim 1. While it is generally believed in the art that the presence of nickel improves the glass-forming abilities of an alloy, making nickel an essential component of many quaternary bulk glass- forming alloys, and especially of Zr-based alloys, it has been found by the in¬ ventors that nickel can be dispensed with by following the principles of the pre¬ sent invention, while still alloys with excellent glass-forming abilities are ob¬ tained.

While the invention is not limited to the particular compositions described here¬ after, the underlying principles of the invention will in the following be exempli¬ fied for an alloy with general composition Zr-Cu-Fe-Al. Of the four components present in such an alloy, Zr is the element with the largest atomic size (r = 0.160 nm). With Fe (r = 0.128 nm), it forms a deep eutectic composition near 20 atomic percent (at.%) Fe. This is illustrated in Fig. 1 , which shows, in a highly schematic manner, part of the phase diagram of a binary Zr-Fe alloy. The tran¬ sitions between the various solid phases have been omitted from the diagram for clarity, such that the diagram shows only the expected liquidus line, i.e., the liquidus temperature as a function of composition (S = solid, L = liquid). A deep eutectic feature at 24 at.% Fe is clearly visible. This deep eutectic can be quali¬ tatively explained by topological considerations.

Also Zr and Cu have eutectic compositions, one of which occurs at 72.5% Zr, as illustrated in Fig. 2. This diagram shows, again in a highly schematic fashion, the liquidus line. At various compositions between 38.2 at.% and 72.5 at.%, several other eutectics are expected.

The fourth component in the above-mentioned general composition is Al. Fig. 3 shows, again in a highly schematic fashion, part of the phase diagram of a bi¬ nary Al-Fe alloy. Several solid-solid transitions have been included in this dia¬ gram. In particular, a high-temperature phase, the so-called £>-phase 301 , is present around the composition AIeFe4. This phase prevents a deep eutectic to be present at around 60 at.% in the Al-Fe phase diagram, which would other¬ wise be expected by extrapolation, as indicated by the dotted line in Fig. 3. However, since the eutectics of Zr76Fe24 and Zr72 SCu27 S are already below 1000°C, it is likely that the high-temperature ^phase, which spans a tempera¬ ture range between 1102 and 1232 0C, will not form any more in the quaternary alloy.

These considerations led to the development of the composition (Zr72.5Cu27.5)8o(Fe4oAl6o)2o as a starting point for further investigations as detailed below. It was found that this alloy, even without any further refinement of the composition, exhibits excellent glass-forming ability. In addition, the composition of the alloy was varied, and it was found that the alloy retained its good glass- forming properties in a rather wide range of compositions.

This shows that the "80-20 concept" can be successfully generalized to quater¬ nary alloys. The concept is believed to be generally applicable and not to be restricted to the particular Zr-Cu-Fe-Al system described above. In particular, the same considerations may be applied to alloys which are based on Ti, Hf, Nb, La, Pd or Pt as a main component. Instead of Cu, other elements having a deep eutectic with the main component may be employed. Particularly good candidates are Be, Ag and Au. The Fe component may be replaced by one or more of the transition metals except Ni, e.g. by Co. The Al component may be replaced by, e.g., Zr or one or more of the metalloids.

In the following, examples of the manufacture and characterization of inventive alloys will be given.

Example 1: Preparation and characterization of amorphous (ZrxCuioo-x)8o(Fe4oAI6o)2o samples Several Zr-based Ni-free alloys with composition (ZrxCuioo-x)8o(Fe4oAI6o)2o were prepared, where x = 60, 62, 64, 66, 68, 72.5, 77, 79, 81 , 83 and 85. Ingots were prepared by arc melting the constituents (purity > 99.9%) in a titanium-gettered argon atmosphere (99.9999% purity). Using an induction-heating coil, the ingots were remelted in a quartz tube (vacuum « 10"5 mbar) and injection cast into a copper mold with high-purity argon. Samples were cast into plates with a thick¬ ness of 0.5 mm, width of 5 mm and length of 10 mm. To determine the critical casting thickness, some samples were additionally or alternatively cast into various rod- and cone-like shapes with diameters ranging up to 10 mm. Fur¬ thermore, several samples were made with a thickness of 1 mm and cross sec- tion 1 cm x 4 cm. The samples were then, where appropriate, cut into various pieces of length 1 cm and investigated by X-ray diffraction (XRD), small-angle neutron scattering (SANS), differential thermal analysis (DTA) and/or hardness measurements. XRD was performed with a Scintag XDS-2000 X-ray diffracto- meter, using a collimated monochromatic Cu K01 x-ray source. The thermo- physical properties were investigated with a Netzsch Proteus C550 DTA and SANS was performed at Paul Scherrer Institute, Switzerland, using a wave¬ length of λ = 6 A and sample-detector distances of 1.8 m, 6 m, and 20 m.

Fig. 4 shows XRD patterns of as-cast alloys of composition Zr544Cu2SeFe8AIi2, Zr58Cu22Fe8AIi2, and Zr61.6CUi814Fe8AIi2, i.e., (ZrxCuioo-x)8o(Fe4OAI6o)2o with x = 68, 72.5, and 77. All samples show a typical XRD pattern of an amorphous structure without any Bragg peaks. The amorphicity is also confirmed by SANS. As can be seen in Fig. 5, the same samples do not show any small-angle scat¬ tering over a wide Q-range, giving evidence for a homogeneous, amorphous structure.

The DTA scans in Fig. 6, performed with a heating rate of 20 K/min, reveal for all three alloys a clear glass transition, followed by an extended undercooled liquid region and an exothermic crystallization peak. For comparison, the Ni- bearing alloy Zr65AI75NiIoCUi75 was also investigated by DTA. This result is also shown in Fig. 6 for comparison. Additionally, the DTA scan in Fig. 7, which was performed over an extended temperature range, shows the endothermic melting peak of Zr58Cu22Fe8Ali2.

Table 2 gives the characteristic values extracted from DTA scans like those of Figs. 6 and 7. The glass transition temperatures T9 were extracted from the on¬ set of the endothermic events in Fig. 6 (arrows pointing up) and the first crystal¬ lization temperatures Tx1 were obtained from the onset of the exothermic peaks (arrows pointing down). The onset of melting Tm and the offset of melting 7Ϊ were obtained from scans like that in Fig. 7. The new Ni-free alloys show an undercooled liquid region ΔTX = TxI-T9 of 78 to 86 K and a reduced glass transi¬ tion temperature T9ITi between 0.56 and 0.57. Table 2 lists the ratios of rg/7m also, since in many publications this ratio has been used as the reduced glass transition temperature. The value of TgITm is 0.59 to 0.62 for the new Ni-free alloys and thus significantly larger than that Of Zr65AI75Ni10Cu175.

Table 2. Glass transition temperature T9, first crystallization temperature 7"x1,

undercooled liquid region ATx = Tx1- T9, liquidus temperature (offset of melting)

7|, reduced glass transition temperature Tg/Ti, onset of melting Tm, and ratio 7g/rm for three Ni-free alloys and for the Ni-bearing alloy Zr65AI75Ni10Cu175, ob¬ tained by DTA using a heating rate of 20K/min.

Table 3 shows the Vickers hardness HV of the Ni-free alloys that was measured with a load of 500 g. From these measurements, one obtains an estimated yield strength of 1.56 to 1.68 GPa, using the scaling relation σy = 3 HV. Indeed, de¬ tailed tensile tests show a yield strength of σy = 1.71 GPa and an elastic limit of 2.25% for the alloy Zr58Cu22Fe8AI12.

Table 3. Vickers hardness HV (measured with a load of 500 g) and estimated

yield strength σy of the Ni-free alloys.

Detailed casting experiments were performed on these Ni-free alloys, and these were compared with the critical casting thicknesses of Zr65AI75Ni10Cu175 and Zr5215Ti5Cu179Ni146AIiO (Vit105™) under equal experimental conditions. The al¬ loy Zr58Cu22Fe8AI12 (x = 72.5) could be cast into a fully amorphous state up to a rod-diameter of 7 mm. Fig. 8 shows some examples of such cast samples. These examples prove that indeed articles to be used in real-life applications can be manufactured from the inventive alloys. The wedge-shaped sample is fully amorphous up to a diameter of 7 mm.

Fig. 9 shows X-ray diffraction patterns Of Zr58Cu22Fe8AI12 cast to cylindrical rods of diameters 5, 7 and 8 mm, and to a plate of 1 mm thickness (inset). No Bragg peaks are apparent either in the 5 mm rod sample or in the 1 mm plate, while only very weak Bragg peaks seem to arise in the 7 mm rod sample. In contrast, a clear crystalline component is present in the 8 mm rod sample, as apparent from the strong Bragg peaks from that sample. These findings are consistent with the DTA scans shown in Fig. 10, which were performed on the 5 mm, 7mm and 8 mm rod samples. Clear exothermic crystal¬ lization peaks are visible for the 5 mm and 7 mm samples, while no such peak is observed for the 8 mm sample.

Likewise, the alloys with x = 68, 77 could be cast in rod shape with a diameter of at least 5 mm with an amorphous structure.

Fig. 11 shows XRD patterns of Zr544Cu2SeFe8AI12 (x = 68) cast to a cone with a maximum outer diameter of 6 mm. The XRD scans were performed on 0.5 mm thick plates cut perpendicularly to the longitudinal axis of the cone. The average diameter of the corresponding plates is given in the figure. The XRD patterns of the plates with diameters of 5 mm or less show typical amorphous structures, while the plate with 6 mm diameter appears to show some Bragg peaks indicat¬ ing a small volume fraction of crystals in the amorphous matrix. This is perfectly consistent with the findings for rods with uniform diameter.

Fig. 12 shows a DTA scan of Zr61 6CUi84Fe8AI12 (x = 77) performed with a heat- ing rate of 20 K/min. Clear glass-transition, crystallization and melting features are observed. Fig. 13 shows a SEM image, showing the fracture surface of glassy Zr61 ^Cu184Fe8AI12 (x = 77) which is typical for an amorphous glass. These findings demonstrate that also Zr61 6Cu184Fe8AI12 (x = 77) is an excellent bulk metallic glass-former.

In summary, of the three alloys with x = 68, 72.5 and 77, the alloy Zr58Cu22Fe8AIi2 (x = 72.5) has the greatest glass-forming ability, comparable to that of Vit105™, followed by Zr6I-6Cu184Fe8AIi2 and Zr5414Cu25 6Fe8AIi2, followed by the prior-art alloy Zr65AI75Ni10Cu17-5. These experimental results agree well with the Turnbull theory (D. Tumbull, Contemp. Phys. 10, 473 (1969), F. Spa- epen and D. Turnbull, Proc. Sec. Int. Conf. on Rapidly Quenched Metals (Cam¬ bridge, Mass.: M. IT. Press, 1976), pp. 205-229), which predicts that the best glass-forming ability is obtained for the alloy with the highest ratio of T9IT] (see Table 2).

Fig. 14 shows the tensile stress-strain curves of an as-cast cylindrical Zr58Cu22 Fe8A112 (x = 72.5) sample with a diameter of 5 mm. Hooke's law is well fulfilled for strain up to 2.25%. The excellent elasticity and high tensile strength as visible from this diagram are just one example of the excellent mechanical properties of the inventive alloys.

The alloys with x = 60, 62, 64, 66, 79, 81 , 83 and 85 were also investigated by selected similar methods. It was found that the alloys with x between 62 and 81 were amorphous when cast to a thickness of 0.5 mm, the alloy with x = 60 was crystalline, the alloy with x = 83 was partially amorphous, and the alloy with x = 85 was crystalline when cast to a thickness of 0.5 mm.

It is apparent from this example that the composition of the material can be var¬ ied within rather broad limits without losing the good glass-forming properties. Specifically, it may be expected that a variation in the composition with respect to the other constituent elements, in particular a moderate variation of the num- bers a and y, will not alter the glass-forming ability dramatically. Furthermore, it is expected that addition of a small amount of an additional component will not negatively affect the glass-forming ability or even possibly improve the glass- forming ability of the inventive materials, while possibly improving certain de¬ sired properties.

Example 2: Preparation of mixed-phase samples Samples with a mixed-phase structure were prepared as follows: Fully amor¬ phous samples of Zr58Cu22Fe8AI12 were prepared as in Example 1. The samples were subjected to heat treatment (annealing) at various temperatures for 12 hours. XRD patterns and DTA scans were recorded for the heat-treated sam¬ ples. Fig. 15 shows XRD patterns of the samples in the as-prepared state (bot¬ tom trace) and after annealing. The XRD patterns show typical amorphous structures up to an annealing temperature of 683 K. At higher annealing tem¬ peratures, however, clear Bragg peaks arising from an icosahedral phase (I. P.) can be observed. At still higher temperatures, peaks which are typical for a Zr2Fe structure are observed. Fig. 16 shows the XRD pattern of the sample an- nealed at 708 K for 12 hours in more detail. The indexing indicates the pres¬ ence of an icosahedral phase with a lattice constant of 0.476 nm. Fig. 17 shows DTA scans of the same samples as in Fig. 15, which are consistent with the development of a structure with both glassy and crystalline components.

In order to better characterize the structure after annealing, in-situ small-angle neutron scattering (SANS) experiments were performed during annealing at a temperature of 708 K of a Zr58Cu22Fe8AI12 sample which was initially fully amor¬ phous. The results are shown in Fig. 18, for total annealing times as indicated. The results show that crystalline regions develop in the initially fully amorphous sample, with typical sizes on the order of only nanometers. These data were analyzed by applying the Guinier approximation. Fig. 19 shows the time evolu¬ tion of the particle size, Φ, in this approximation. This clearly demonstrates the emergence of nanocrystals within the glassy matrix. It is believed that the gen¬ eration of such nanocrystals is fostered by keeping the annealing temperature only slightly above the laboratory glass transition temperature, in particular, in a range between 0 and 150 K above the laboratory glass transition temperature. The laboratory glass transition temperature is to be understood as the glass transition temperature as determined by DSC (differential scanning calorimetry) with a typical heating rate of 20 K/min. Higher annealing temperatures often lead to the precipitation of larger crystals; for example in the range of 0.1 - 20 μm.

Such mixed-phase materials exhibit somewhat different mechanical properties than a fully glassy material. In particular, ductility is often improved, which can be rationalized by the fact that shear bands which develop as a result of shear forces during forming and which might lead to breaking of the material are dis¬ rupted by the crystals. These properties may be particularly beneficial in appli- cations where the material must be shaped or deformed during manufacture of the end product.

Example 3: Variations of composition Samples in a widely varying range of compositions were prepared and investi¬ gated. The compositions of the following Tables proved to be at least partially amorphous when cast to a plate with thickness of 1 mm (Table 4), 0.5 mm (ta¬ ble 5), or 0.2 mm (Table 6):

Table 4: Alloys having a partially or fully amorphous structure when cast to a thickness of 1 mm.

(Zr95Ti5)72 Cu13Fe13AI2 Zr72 Cu12Fe12AI4 Zr7O Cu-I3Fe13AI3 Sn1 Zr70 Cu13Fe13AI4 Zr70 Cu13Fe13AI2Cr2 Zr72 Cu11Fe11AI6 Zr70 Cu13Fe13AI2Nb2 Zr72 Zr7O Cu13Fe13AI2Zn2 Zr73 Cu11Fe11AI5 (Zr72 Cu13Fe13AI2)98Mo2 Zr71 Cu11Fe11AI7 (Zr72 Cu13Fe13AI2)98P2 Zr69 Cu11Fe-I 1AI9 (Zr95Hf5)72 Cu13Fe13AI2 Zr70 Cu1C5Fe1C5AI9 Zr70 Cu11Fe11AI8 Zr70 Cu10Fe11AI9 Zr71 Cu11Fe10AI8 Zr70 Cu11Fe1OAIg (Zr74 Cu13Fe13)90AI10 Zr69 Cu10Fe10AI11 Zr72Cu13Fe13AI2 Zr69 CUi0Fe11AIi0 (Zr74Cu15Fe13)98AI2 Zr70 Cu13Fe13AI2Sn2 Zr73 Cu13Fei3Ali Zr72 Cu13Fe13Sn2 Zr72Cu13 Fe13AI2 (Zr74Cu13Fe13)98Sn2 Zr71 Cu13Fe13AI3 Table 5: Alloys with a partially or fully amorphous structure when cast to a thickness of 0.5 mm.

(Zr79Cu2I )8o(Fe4OAI6o)2o (Zr66Cu34)8o(Fe4oAl6o)2o (Zr8iCui9)8o(Fe4oAl6o)2o (Zr64Cu36)8o(Fe4oAl6o)2o (Zr83Cui7)8o(Fe4oAl6o)2o (Zr62Cu38)8o(Fe4oAl6o)2o

Table 6: Alloys with a partially or fully amorphous structure when cast to a thickness of 0.2 mm.

Zr72Cu13Fe13AI2 (Zr74 Cu13Fe13J98Ge2 Zr72 Cu-I3Fe13Sn2 (Zr74Cu13Fe13)98Sn2

For comparison, the alloys in Table 7, while being binary, ternary or Ni- containing alloys, were also investigated and developed an at least partially amoφhous structure when cast to a thickness of 0.2 mm.

Table 7: Comparative listing of other alloys with a partially or fully amorphous structure when cast to a thickness of 0.2 mm.

Zr70 Cu13Fe13AI2Ni2 Zr76Fe2oAI4 Zr70Cu6 5 Fe13AI2Ni6 5 Zr70Fe27Nb3 (Zr74 Cu13Fe13)98Ni2 Zr68Fe27Nb5 (Zr74 Cu13Fe13)96Ni4 Zr66Fe28Nb6 Zr76Fe24 Zr68Fe25Nb7 Zr75Fe23Sn2 Zr75Fe24Ni1 Zr7oFe28Nb2 Zr75-5Fe23 5Ge1 Zr76Fe22Sn2 Zr70Fe28NbiSn-ι Zr76Fe23Sn1 Zr75 5Fe23-5Si1 Zr75Fe24Sn1 Zr77Fe23 Zr74Fe24Sn2 Zr69Fe30Nb1 Zr73.72Fe23.2βSn3 Zr68Fe3INb1

Zr76Fθ2iSn3 ZreeFe26Nb6 Zr69Fe29NbiSni Zr69Fe27Nb4 Zr75.5Fe23.5AI1 Zr6eFe2eNb4 Zr76Fe23Ali Zr7IFe26Nb3

Zr74Fe2B Zr70Fe26Nb4 Zr7oFe29Nbi Zr74FeI3Cu13 Zr72Fe27Nbi Zr71Fei6Cui3 Zr74Fe25Nbi Zr74FeI3Cu13 Zr73Fe2SNb2 Zr76Fe23Cu1 Zr76Ni24 Zr76Fe12Cu12 Zr6oFe2oNi2o Zr73.5Fe2i.5Cu5 Zr75 5Fe23 5SJi Zr72Fe14Cu14

Specifically, this list shows that also ternary, nickel-free alloys can be reasona¬ bly good glass-formers, especially if composed according to the "80:20 scheme". Specifically, the list shows that ternary alloys of composition (ZrxD100-x)aFe100-a, where the number a is in the range from about 70 to about 90, in particular approximately 80, are good glass formers. Here D is advanta¬ geously Cu, Nb, Al or Sn.

The alloys in Table 8 have also been prepared and were found to be fully amor- phous when subjected to splat cooling to a thickness of 20 micrometers at high cooling rates of approximately 106 K/s. These alloys may be regarded as candi¬ date materials for bulk metallic glasses, while casting experiments will be nec¬ essary to verify which of these are indeed bulk metallic glasses. Table 8: Alloys having a fully amorphous structure when splat-cooled. All num¬ bers are atomic percentages.

Zr58Cu22Fei8AI2 (Zr58Cu22FeβAI12)98Nb2 Zr58Cu22FeI6AU (Zr58Cu22FeδAI12)98Ta2 Zr58Cu22FeI4AI6 (Zr58Cu22Fe8AI12)98Cr2 Zr58Cu22FeI2AI8 (Zr58Cu22Fe8AI12)98Co2 Zr58Cu22FeioAho (Zr58Cu22Fe8AI12)98Mo2 Zr58Cu22Fe6AIi4 (Zr58Cu22FeβAI12)98Sn2 Zr58Cu22Fe4AI16 Zr58Cu22Fe6AI12Nb2 Zr58Cu22Fe2AIi8 (Zr72-5Cu27-S)7BFe8AI12Nb4 Zr62-4Co17.6Fe8AI12 Zr58Cu22Fe4AI12Nb4 Zr65 AI15Fe15Nb5 Zr58Cu22Fe8AI10Nb2 Zr58Cu22Co8AI12 (Zr72-5Cu27 S)78Fe8AI12Co2 Zr68 AI15Fe15Nb2 (Zr72 5Cu27-S)78Fe8AI12Cr2 (Zr72.5Cu27.5)78FesAI12Nb2 (Zr72-5Cu27-S)78Fe8AI12Ta2 (Zr72 5Cu271S)78Fe8AI12Sn2 (Zr72-5Cu27 S)78Fe8AI12Mo2 (Zr72 5Cu27 5)SoFe6AI12Nb2 (Zr72-5Cu27-S)76Fe8AI12Sn4

Also the ternary and binary alloys in Table 9 were found to be fully amorphous when splat-cooled. These are listed for comparative purposes.

Table 9: Ternary and binary alloys having a fully amorphous structure when splat-cooled.

Zr60Fe15AI15 Zr58Cu22Fe20 Zr75Fe^Sn2 Zr58Cu22AI20 Zr70Fe28Nb2 Nb60Co40

The wide range of alloys according to the present invention which were investi¬ gated in these experiments clearly demonstrate that wide variations of composi¬ tion are possible without losing the glass-forming properties of the alloys. Example 4: Biocompatibility tests As an example of the newly developed Ni-free alloys, the cytotoxicity of the al¬ loy Zr58Cu22Fe8Ali2 was determined. The effect of surface modification by pas- sivation in diluted nitric acid was also investigated.

Surface analysis using XPS showed that a natural oxide layer, composed al¬ most exclusively of zirconium oxide, forms on the surface on this glass and that it has a thickness of 7-8 nm. This layer protects mouse fibroblasts used in the study from the toxic metals, especially Cu, present in the bulk, allowing for good cell growth on the alloy. The results of indirect tests demonstrate that this layer is stable in PBS (phosphate-buffered solution) for many weeks, and that no toxic effects due to high ion concentrations diffusing into the medium occur.

The thickness of the zirconia layer is only slightly increased by passivation with nitric acid. However, this treatment clearly improves the quality of the surface layer, which leads to increased corrosion resistance and lower diffusion of bulk elements into the medium, and thus to improved biocompatibility. After this pas¬ sivation treatment, the alloy shows cell growth comparable to that on polysty- rene, which is used here as a negative control.

In conclusion, the cytotoxic properties of the metallic glasses of the present in¬ vention are very promising and thus indicate a very good biocompatibility.

Example 5: Cu- and Ni-free alloys As Cu may nevertheless be problematic in many medical applications, a search for Cu-free alloys was conducted. Starting from the Zr-Cu-Fe-Al bulk metallic glasses of the previous examples, Pd (palladium) was found to be promising in replacing Cu in such alloys. For a systematic search for bulk metallic glasses, alloys belonging to the pseudoternary Zr-(Feo.5Alo.5)-Pd system were screened. Initially, the amount of Pd was varied between 0% and approximately 22% in a pseudoternary Zr-(Feo.5Alo.5)-Pd system along the (Fβo.sAlo.stao line, while choosing the ratio of the sums of the atomic percentages of Zr and Fe on the one hand and Al and Pd on the other hand roughly according to the 80:20 con¬ cept. In this manner, a number of initial alloy compositions with favorable glass- forming properties were identified. The composition was then varied around these initial compositions in an iterative manner within the range of pseudoter- nary Zr-(Feo.5AIO-5)-Pd compositions.

The following tables summarize the results found in these investigations.

Table 10: Cu-free Zr-Fe-Al-Pd alloys having a partially or fully amorphous struc¬ ture when cast to a thickness of 3 mm

Zr69.7Fei2.95 AI12-95Pd4-4

Zr66.7Fe-|4.45AI-|4.45Pd4.4

Table 11 : Cu-free Zr-Fe-Al-Pd alloys having a partially or fully amorphous struc¬ ture when cast to a thickness of 1 mm

Zr68-3Fe13-4AI13-4Pd4-9

Zr65.4Fei4.85Ali4.85Pd4.9 Zr62^Fe167AI16-7Pd4-3

Zr59-2Fe-18.3Ali8.3Pd4.2

Zr73.4Feio.9Alio.gPd4.8 Zr75.2Feio.2AI-ιo.2Pd4.3 Zr77Fe9 5AI9-5Pd4Zr67-9

Zr65Fe11 4AI11 4Pd12-2 Zr62-5Fe10-7SAI10J5Pd16

Table 12: Cu-free Zr-Fe-Al-Pd alloys having a partially or fully amorphous struc- ture when cast to a thickness of 0.5 mm

Zr695Fe15AI15Pd0 5 Zr69FeI5AI15Pd1 Zr68Fe15AI15Pd2 Zr67Fe15AI15Pd3 Zr66Fe15AI15Pd4 Zr65Fe15AI15Pd5 Zr64Fe15AI15Pd6 Zr63Fe15AI15Pd7 Zr62Fe15AI15Pd8 Zr71Fe12AI12Pd5 Zr69FeI2 85AI12 85Pd5-3 Zr668Fe13.7Ali3.7Pd5.8 Zr65Fe14.5AI14.5Pd6 Zr61.9Fe16.2AI16.2Pd5.7 Zr5OFe-I2AI12Pd26 Zr532Fei2 6Ali2.6Pd21.e Zr57.6Fe13.95Ali3.95Pd14.5 Zr60FeI413AI1413PdIi4

The examples of Tables 10, 11 and 12 are indicated by black squares in the pseudoternary mixing diagram of Fig. 20. From this diagram, it may be appreci¬ ated that alloys containing at least 50 at.-% Zr, at least 0.5 at.-% Pd and at least 19 at.-% of a mixture of Fe and Al in approximately equal atomic proportions are expected to be good glass formers. This is even more true for alloys of this type containing at least approximately 59 at.-% of Zr, up to approximately 36 at.-% of the Fe-Al mixture and/or at least approximately 4 at.-% Pd. In particular, all al¬ loys in the trapezoidal area indicated in Fig. 20 may reasonably be expected to be good glass formers. Small variations of the relative proportions between Fe and Al within a few percent, say, between 60:40 and 40:60 or better between 55:45 and 45:55, are not expected to strongly affect the glass-forming ability.

Notably, all alloys in Tables 10 and 11 and most of the alloys in Table 12 corre¬ spond to the 80:20 principle in the following sense: The ratio of the sum of the atomic percentages of Zr and Fe to the sum of the atomic percentages of Al and Pd is approximately 80:20. In the examples of Tables 10 and 11 , the ratio of the atomic content of Zr+Fe to that of Al+Pd varies between approximately 73:27 and approximately 87:13. The 80:20 principle is fulfilled to an excellent degree for the alloys in Table 10, i.e., for those alloy compositions which have been found to have the highest critical casting thickness. There, the corresponding ratio varies between approximately 80:20 and approximately 83:17.

Concerning the variations within the Zr-Fe subsystem, in the preferred composi¬ tions of Tables 10 and 11 , the ratio of the atomic percentage of Zr to the atomic percentage of Fe is in the range between approximately 76:24 and approxi¬ mately 89:11. It appears that this is a preferred range. In particular, in the ex¬ amples of Table 10, this ratio varies between approximately 81 :19 and ap¬ proximately 85:15. In contrast, the ratio between Al and Pd may apparently vary in a wider range without detrimental effects on the glass-forming ability of the alloy. In the examples of Tables 10 and 11 , the ratio of the atomic percentage of Al to the atomic percentage of Pd varies between approximately 40:60 and ap¬ proximately 82:18. In particular, in the examples of Table 10, this ratio varies between approximately 65:35 and approximately 78:22.

An even improved biocompatibility may be achieved by replacing Pd partly or fully by Pt (platinum) in the above examples. Pt (platinum) has very similar properties as Pd, such as outer electronic structure, in consequence, similar chemical properties, and almost the same Goldschmidt radius. Therefore, a par¬ tial or full replacement of Pd by Pt will not strongly alter the mechanical proper- ties of the alloy or its glass-forming ability.

As an example of measurements performed on the Cu-free alloys, Fig. 21 shows a DTA scan and Fig. 22 shows an X-ray diffraction pattern, using a Co- K0 X-ray source, of the alloy cast to a thickness of 1 mm. The DTA scan exhibits a clear glass transition and a second crystallization event, while the X-ray diffraction pattern exhibits the broad hump indicative of an amorphous material.

Also the following Cu-free alloys were found to be at least partially amorphous when cast to a thickness of 0.5 mm: Zr69FeI5AI15Yi, Zr685FeI5AI15Yi 5. In these examples, Pd has been replaced by Y (yttrium).

A further example of an alloy found to be at least partially amorphous when cast to a thickness to 0.2 mm is Zr70Fe28NbiSni.

It is to be understood that the above examples are only provided for illustrative purposes and that the invention is in no way limited to these examples.

List of abbreviations, symbols and reference signs at.% atomic percent XRD X-ray diffraction SEM scanning electron microscopy SANS small-angle neutron scattering DTA differential thermal analysis DSC differential scanning calorimetry Tg glass transition temperature 7χi first crystallization temperature ATx undercooled liquid region T\ offset of melting (liquidus temperature) Tm onset of melting T temperature σy yield strength HV Vickers hardness S solid L liquid 2Θ scattering angle lnt intensity a. u. arbitrary units Q wave number S(Q) scattering intensity q heat transfer cps counts per second σ tensile stress ε strain IP. icosahedral phase ann. annealed Φ particle size