Study On Glass Forming Ability For Alloys

Amorphous alloys, as a new widely studied field among material science, have been noticedby scientists of material and physical. The knowledge of the amorphous alloys is perfected dayby day, from the initial one that only a relatively restricted number of materials could be preparedin the form of amorphous solids to the later one that nearly all materials can, if cooled fastenough and far enough, be prepared as amorphous solids. But the condition of big production ofmany amorphous alloys can not be reached, and the output is almost in form of thin ribbon,which limited the big production and widely application.In recent years, through numerous experiments, the observation is that amorphous alloys canbe made with a slow cooling rate in some systems. Therefore, the research of the amorphousalloys goes further. The three empirical rules coming from the phenomena theory depending onthe experiments are as follows:1.multi-component systems consisting of more than three elements2.significant difference in atomic size ratios above about 12% among the three mainconstituent elements3.negative heats of mixing (△H) among the three main constituent elements.The three rules agree with the practice, for they coming from the numerous experiments.However, some fact deviates from the rules appear, as the new amorphous alloys are obtained bycontinuing exploration. For instance, Pd-Cu-Ni-P system having good glass forming ability (GFA)does not follow the above principles where the atomic radius difference among Pd, Cu and Ni issmaller than 12% and the negative △H arises only between these metals and metalloids P. Thusthe rules need be perfected and think over deeply.As we known, a small mass density difference of 0.30 -0.54% between the bulk metallicglass and the corresponding crystal while this value for usual glasses is generally 2%. Anotherexperiment has proved that when the volume difference between a BMG and the correspondingcrystal approaches zero, the BMG has the best GFA. The fact implies that the bulk metallic glassand the corresponding crystal have the similar topological structure, and the energy difference isvery small.Since the empirical rules are based on phenomena theory, they could not reach the essentialof the question that leads to the deficiency for the composition design of new kinds of bulkmetallic glasses. Therefore, in this paper, based on the former work and the theory, by adoptedthe new results, some principles are brought for the composition design of new kinds of bulkmetallic glasses.From the experiments the bulk metallic glass and the corresponding crystal have the similartopological structure with the small energy difference. This is consistent with the Jones' e/apremise. The e/a value controls the structural preference in a family of polytypic structureswhere the structural change from one to another does not alter the coordination number (CN).Since CN does not change significantly from one structure to the other, the major contribution tothe enthalpy and thus to the stability of the phases remains unchanged, which completelydominate smaller changes in electron energies associated with energy band effects. Therefore thedriving force of crystallization is small, and the crystal could not be form, the liquid has goodGFA. In the paper, the liquid is a multi-component system.According to above analyses, liquid has lowest energy with some special value of e/a, whilethe energy difference between the liquid and corresponding crystal is small, the GFA is good.Therefore, the relation between e/a and GFA can be obtained. By adopted the results of Germanstudy on the valences of transition metals, the e/a values of representative glasses are gotten, theconclusion is as follows:1.e/a = 3.5;2.The component number (n) in the alloy is larger than three;3.The component percentage of the i-th component (xi) in the alloy approaches 1/n.Without the temperature effect on, free energy difference (?G) is the result of internalenergy difference (?U) and entropy (S) action. The former discusses the ?U effect on the ?G,namely, when .e/a = 3.5 the alloy has the lowest ?U. Since the effect of ?U on ?G is mainly,sometime the effect of S can be ignored. However, if ?U remains constant ?G decreases furtherwith the mixing entropy difference between the liquid and the crystal ?S = Sl –Sc increasing.Although this ?S is only a part of overall melting entropy ?Sm, it is clear that ?Sm should beproportional to ?S. The ?S can replace the ?Sm.Usually, if Sc ≈ 0 where the solid has a certain crystalline structure with a unique structuralconfiguration, ?S ≈ Sl, which implies that a larger Sl value, which can be achieved by increase ofn of an alloy, is related with a better GFA. In order to get larger ?S, the n should be larger than 3.In this paper, its Sl is decided only by mixing of different atoms with different sizes. Accordingthe results of experiments, when n > 4, GFA of a BMG is not further improved with the ?Sincreasing. Thus, it is reasonable that Sc ≠ 0 is a general phenomenon for BMGs and the effect ofSc on ?S must be considered where Sl and Sc may increase simultaneously when n > 4.In this contribution, through calculating Sc value in light of experimental evidence thatmetallic glasses have a A2B crystalline structure, the components in the crystalline alloys occupyeither the A site or B site. When an alloy consists of only A and B with an atom ratio of 2:1, Sc =0. When other components are added in the alloy, they may occupy one of the two sites of A or Bdepending on the relative atom size and chemical nature determined by their positions in theperiodic table. Based on the definition of the ?S and the atom size effect on, the formula of ?Swith Sc ≠ 0 can be approached.In this contribution, ?S and GFA of Zr base glasses are considered as an example. In lightof the calculated results from the Sc, ?S and Sl as functions of n for the Zr base glasses are shownin figure. As shown in the figure, although Sl increases steadily as n increases, ?S ≈ 5Jg-atom-1K-1 for n > 3. The reason is that Sl and Sc may increase simultaneously. This is accordwith the critical cooling rate.It is found that n = 4-5 is enough to maintain ?S = 5 Jg-atom-1K-1 or to decrease ?G of Zrbase alloys when their e/a ratio has been controlled to be near 3.5, which decides the size of ?U.Thus, the alloy has the best GFA.In a word, based on the former results and the thermodynamics and kinetics, we study theGFA of the alloys and some principles for the composition design of new kinds of bulk metallicglasses are concluded.