ASM Handbook, 3, Diagrams Copyright # 2016 ASM InternationalW H. Okamoto, M.E. Schlesinger and E.M. Mueller, editors All rights reserved asminternational.org

Introduction to Phase Diagrams*

IN , a phase is a a system with varying composition of two com- Nevertheless, phase diagrams are instrumental physically homogeneous state of with a ponents. While other extensive and intensive in predicting phase transformations and their given and arrangement properties influence the phase structure, materi- resulting microstructures. True equilibrium is, of . The simplest examples are the three als scientists typically hold these properties con- of course, rarely attained by and alloys states of matter (, , or ) of a pure stant for practical ease of use and interpretation. in the course of ordinary manufacture and appli- element. The solid, liquid, and gas states of a Phase diagrams are usually constructed with a cation. Rates of heating and cooling are usually pure element obviously have the same chemical constant of one atmosphere. too fast, times of treatment too short, and composition, but each phase is obviously distinct Phase diagrams are useful graphical representa- phase changes too sluggish for the ultimate equi- physically due to differences in the bonding and tions that show the phases in equilibrium present librium state to be reached. However, any change arrangement of atoms. in the system at various specified compositions, that does occur must constitute an adjustment Some pure elements (such as and tita- , and . It should be recog- toward equilibrium. Hence, the direction of nium) are also allotropic, which means that the nized that phase diagrams represent equilibrium change can be ascertained from the phase dia- structure of the solid phase changes with conditions for an alloy, which means that very gram, and a wealth of experience is available to and pressure. For example, iron slow heating and cooling rates are used to gener- indicate the probable degree of attainment of undergoes several distinct solid-state changes ate data for their construction. The equilibrium equilibrium under various circumstances. As of its with temperature. Alloy- states that are represented on phase diagrams are such, alloy phase diagrams are useful to - ing, the formation of a substance with metallic known as heterogeneous equilibria, because they lurgists, materials engineers, and materials properties composed of two or more elements, refer to the coexistence of different states of mat- scientists in four major areas: also affects the occurrence of phase changes. ter (gas, liquid, and/or solid phases with different For example, the temperature for complete melt- crystal structures). When two or more phases are Development of new alloys for specific applications ing (100% liquid phase) of an alloy depends on in mutual equilibrium, each phase must be in the the relative of alloying elements. lowest free-energy state possible under the restric- Fabrication of these alloys into useful configurations Alloying also affects the stable crystalline phase tions imposed by its environment. This equilib- of a solid. Depending on how two or more ele- rium condition means that each phase is in an Design and control of heat treatment proce- ments behave when mixed, the elements may internally homogeneous state with a chemical dures for specific alloys that will produce form different crystalline phases and/or chemical composition that is identical everywhere within the required mechanical, physical, and chemical properties compounds. each phase, and that the molecular and atomic Phase diagrams and the systems they describe species of which the phase is composed (if Solving problems that arise with specific are often classified based on the number of com- more than one) must be present in equilibrium alloys in their performance in commercial ponents (typically elements) in the system proportions. applications (Table 1). A unary plots the phase Because industrial practices almost never changes of one element as a function of tempera- approach equilibrium, phase diagrams should ture and pressure. A binary diagram plots the be used with some degree of caution. Kinetic Unary Systems phase changes as a function of temperature for effects, including surface energies, activation energies, and diffusion and reaction rates, affect A system containing only one pure metal is the time needed to initiate and complete a physi- referred to as a unary system, which can exist cal or chemical phase change. With rapid heat- as a solid, liquid, and/or gas, depending on the Table 1 Terminology for the number of ing, any phase change, such as , occurs specific combination of temperature and pres- elements in an alloy phase diagram at a slightly higher temperature than with slow sure. Assuming a constant , Number of components Name of system or diagram heating. Conversely, with rapid cooling, the a metal melts when heated to a specific tempera- One Unary phase change occurs at a lower temperature than ture and boils with further heating to a specific Two Binary with slow cooling. Therefore, transformations temperature. Through evaporation, metal Three Ternary observed during heating are at higher tempera- atoms leave the container as a . In condi- Four Quarternary Five Quinary tures than the reverse transformations observed tions where matter can enter or leave the system, Six Sexinary during cooling, except in the hypothetical case these systems are known as open systems. Seven Septenary wherein the rates of heating and cooling are infi- To create a closed system where no matter Eight Octanary nitely slow, where the two observations of tem- enters or leaves the system, an airtight cover Nine Nonary Ten Decinary perature would coincide at the equilibrium can be placed on top of the container. If pres- transformation temperature. sure is held constant during boiling, then an

* Adapted from F.C. Campbell, Ed., Phase Diagrams: Understanding the Basics, ASM International, 2012, and H. Baker, Alloy Phase Diagrams and Microstructure, Metals Handbook Desk Edition, ASM International, 1998, p 95–114. 4 / Introduction to Alloy Phase Diagrams increase in volume of the closed system must are 14 kinds of space lattices, derived from all the for the “arrest” in temperature during heating or occur. Conversely, if pressure is increased, the possible varieties of interatomic spacing, lattice cooling through the transformation temperature. vapor phase is compressed and the volume of arrangements, and interaxial angles within the The letter A also is followed by the letter c or r the vapor becomes smaller. The volume of liq- seven crystal systems. to indicate transformation by either heating or uid metal remains fairly constant with the appli- cooling, respectively. The use of letter c for cation of pressure; that is, liquid metals are heating is derived from the French word chauf- essentially incompressible. The same situation Polymorphism and Allotropy fant, meaning warming. If cooling conditions holds if the liquid metal is cooled so that only apply, the critical temperature is designated as a solid remains. Metal atoms in a liquid state Some elements and compounds exhibit more “Ar,” with the letter r being derived from the are almost as closely packed as they are in the than one stable solid crystalline phase depen- French word refroidissant for cooling. solid state, so the effects of pressure and vol- dent on temperature and pressure; these materi- Many other metals and also exhibit ume change can often be approximated as sta- als are called polymorphic,orallotropic for allotropic transformations (Table 2). For exam- tistically negligible. pure elements. ple, , zirconium, and hafnium all exhibit All pure metals have a unique . The most commonly used allotropic element a transition from an hcp structure to bcc on heat- Metals with weak interatomic bonds melt at is iron, which undergoes a series of phase trans- ing. Note that in each case, a close-packed struc- lower temperatures. This includes several soft formations as a function of temperature and pres- ture is stable at room temperature, while a looser metals such as , , and . In contrast, sure (Fig. 3). At room temperature, iron has a bcc packing is stable at elevated temperatures. While the refractory metals have high melting points. structure (also known as ferrite, or a iron). this is not always the case, it is a trend experi- The refractory metals include niobium, tantalum, Ferrite is ferromagnetic at room temperature, enced with many metals. molybdenum, tungsten, and rhenium. A two- but if temperature is slowly increased, a ferrite phase system exists when a container consists becomes paramagnetic at the Curie temperature of both a liquid phase and a vapor phase. In this of 770 C (1418 F). When temperature is increa- Binary Systems example, a liquid is in equilibrium with its vapor sed further, the paramagnetic ferrite changes to phase, when the average rate of atoms leaving an fcc crystal structure referred to as the liquid equals the rate joining the liquid from or g iron. Austenite is completely paramagnetic. A binary phase diagram plots the different the gas. If pressure is increased, the result is At higher temperatures, austenitic iron changes states of matter as a function of temperature for fewer atoms in the gas phase and more in the liq- to a high-temperature bcc structure, referred to a system at constant pressure with varying com- uid phase, and the ratio of atoms in each phase as d iron. Similar phase changes occur during position of two components (or elements). The remains constant once equilibrium (constant cooling. Under equilibrium conditions, the solid- addition of an alloying element represents temperature and pressure) is reestablished. ification of pure iron from the liquid occurs at another degree of freedom (or variable), which 1540 C (2800 F) and forms d iron. Delta iron thus allows two distinct forms (or phases) to is then stable on further cooling until it reaches coexist under equilibrium conditions. The mixed Solid-State Crystal Structures approximately 1390 C (2541 F), where it (i.e., heterogeneous) equilibria phases in binary undergoes a transformation to an fcc structure alloys also can occur in either or when liq- The solid state of a unary system involves of austenite (g iron). On still further cooling, aus- uid and solid phases change during melting or bonding of adjacent atoms in different arrange- tenite changes into the bcc structure of ferrite solidification, as described in the following sec- ments. Atoms in some solids have an amorphous (a iron). This last transformation is extremely tions. (See also the section “The Gibbs Phase structure, which is an arrangement without a pat- important because it forms the basis for the hard- Rule” in this article on the variables affecting tern or periodic structure. Atoms in solids usually ening of . Also, it should be noted (as in the heterogeneous phase equilibria). form a crystal structure, which is a repetitive and previous example of transformation from d iron In many industries, alloy composition is nor- symmetrical arrangement of atoms in three to g iron) that the temperatures of transformation mally expressed in weight percentage, but for dimensions. Metal atoms, for example, tend to during cooling may be slightly lower than the certain types of scientific work the atomic per- form crystal structures due to the nature of metal- temperatures of transformation during heating centage scale may be preferred. If desired, com- lic bonding. Metals are atoms with an outermost (Fig. 3). This temperature differential is known position may also be given in terms of the electron shell that has a spherically shaped elec- as the temperature hysteresis of allotropic phase percentage by volume, but this usage is rare in tron cloud with a . The metallic transformation, and its magnitude increases with the representation of metal systems. bond occurs when valence electrons are shared increases in the cooling rate. Very slow heating The conversion between weight percentage amongst the metal atoms in the matrix, and the or cooling approaches equilibrium conditions, (wt%) and atomic percentage (at.%) may be type of crystal structure is influenced by the in which case transformation temperatures accomplished in a binary system by the use of nature of the electron cloud in the outer shell. would be essentially the same for heating and the formulas: The crystal structure depends on the type of cooling. at:% A at:wt of A bonding (ionic, covalent, metallic) between When any phase transformation occurs during wt% A ¼ 100 atoms in a solid. The Periodic Table of elements slow heating or cooling, there is an arrest or pla- ðÞþat:% A at:wt of A ðÞat:% B at:wt of B in Fig. 1 identifies the crystal structures of the teau in the temperature change during transfor- (Eq 1) various metallic elements. The most common mation. That is, temperature remains constant crystal structures of the metallic elements at the critical temperature for some period until and (Fig. 2) are face-centered cubic (fcc), body- the phase change is complete for all the matter wt% A=at:wt of A centered cubic (bcc), and hexagonal close- in the system. At the transformation temperature, at:% A ¼ 100 packed (hcp). the process of transformation continues with any ðÞat:% A=at:wt of A þ ðÞwt% B=at:wt of B have been classified into seven basic addition or extraction of thermal energy. After (Eq 2) systems (see the appendix “Crystal Structure” transformation is complete, then slow heating in this Volume). Six of the seven crystal systems or cooling results in a temperature change of The equation for converting from atomic per- also can occur with one of five different lattice the system. centages to weight percentages in higher-order arrangements of atoms, such that the same kind In the case of iron (and also steel as an iron- systems is similar to that for binary systems, of (or a group of atoms in a compound) of alloy), the critical temperatures for except that an additional term is added to the identical composition occurs in a periodic array phase transformation are assigned the letter A, denominator for each additional component. of points (lattice points) in space. In effect, there derived from the French word arreˆt that stands For example, for ternary systems: nrdcint hs igas/5 / Diagrams Phase to Introduction

Fig. 1 The periodic table of elements. F, face-centered cubic; B, body-centered cubic; H, hexagonal; O, orthorhombic; T, tetragonal; R, rhombohedral; M, monoclinic; cc, complex cubic. Source: Ref 1 6 / Introduction to Alloy Phase Diagrams

Fig. 2 Arrangement of atoms: (a) face-centered cubic (fcc), (b) hexagonal close-packed (hcp), and (c) body-centered cubic (bcc) crystal structures. Source: Ref 2 Fig. 3 Phase changes of pure iron with very slow (near equilibrium) heating and cooling. When heating, the critical temperatures of phase change are designated as Ac2,Ac3, and Ac4. Because some hysteresis occurs (depending on the rate of heating or cooling), critical temperatures of phase change during cooling are designated as at:% A ¼ Ar2,Ar3, and Ar4. If critical temperatures are determined under equilibrium conditions, the designations of critical wt% A=at:wt of A 100 temperatures are Ae2,Ae3, and Ae4. Source: Ref 2 ðÞat:% A=at:wt of A þðwt% B=at:wt of BÞþðwt% C=at:wt of CÞ (Eq 3) nickel-copper phase diagram (Fig. 4). Pure cop- alloy is referred to as a eutectic alloy (where wt% A¼ per has a lower melting point than nickel. the term eutectic is taken from the Greek word at:% A at:wt of A 100 Therefore, in a nickel-copper alloy, the copper for “easily melted”). ðÞþat:% Aat:wt of A ðÞþat:% Bat:wt of B ðÞat:% Cat:wt of C atoms melt before the regions rich in nickel. Eutectic systems are described in more detail (Eq 4) The mushy zone becomes 100% liquid when in the article “Eutectic Alloy Systems” in this the temperature is raised above the Volume. Eutectic alloys are important because where A, B, and C represent the metals in the alloy. line. As one might expect, the liquidus line con- complete melting occurs at a low temperature. verges to the melting points of the pure metals For example, cast are based on composi- Effect of Alloying on in an alloy phase diagram. The liquidus denotes tions around the iron-carbon eutectic composi- Melting/Solidification for each possible alloy composition the temper- tion of iron with 4.30 wt% C (Fig. 5). This ature at which begins during cooling improves castability by lowering the tempera- Binary systems have two elements of varying or, equivalently, at which melting is completed ture required for melting and also promotes concentration. Unlike pure metals, alloys do not on heating. better solidification during casting. Another necessarily have a unique melting point. The lower curve, called the , indicates industrially significant eutectic is in the lead- Instead, most alloys have a melting range. This the temperatures at which melting begins on tin system for solders. behavior can be seen in the phase diagram for heating or at which freezing is completed on copper-nickel (Fig. 4). The solidus line is where cooling. Above the liquidus every alloy is mol- the solid phase begins to melt, and the new ten, and this region of the diagram is, accord- Effect of Alloying on Solid-State phase consists of a of a solid- ingly, labeled L for the liquid phase or liquid Transformations phase and some liquid phase. This solid-liquid solution. Below the solidus all alloys are solid, mix in a two-phase region is referred to as the and this region is labeled a because it is cus- When small amounts of another element are mushy zone. Like the two-phase system of a liq- tomary to use Greek letters to designate differ- added to a pure metal, some (or all) of the alloy- uid and vapor phase in a closed system, the liq- ent solid phases. At temperatures between the ing element can be accommodated (dissolved) uid phase is in equilibrium with the solid when two curves, the liquid and solid phases both within the solid crystal matrix of the parent metal the average rate of atoms leaving the liquid are present in equilibrium, as is indicated by (see the article “Solid and Phase Trans- equals the rate joining the liquid from the solid the designation L + a. formations” in this Volume). This dissolution of phase. Temperature and alloy concentration However, it also is important to understand the element within the crystal matrix of solid changes the relative amount of liquid and solid that some specific alloy compositions do not metal is referred to as a . The parent phase in the mushy zone (see the section “The have a mushy zone. That is, some alloys have metal is referred to as the , while the ” in this article). a unique composition with a specific melting alloying element is referred to as the solute. The reason for the mushy zone can be under- point that is lower than the melting points of The phase structure of a solid solution is the same stood in qualitative terms by examining the the two pure metals in the alloy! This type of as that of the parent metal. Introduction to Phase Diagrams / 7

There are limits to the solid of one the fcc crystalline structure and the atoms are solidus line (Fig. 4). This type of system is element in another, depending on the similarity of roughly similar size. Thus, the elements are referred to as isomorphous system, which is and differences of the two atoms. For example, completely miscible, and the nickel-copper sys- described in more detail in the article “Isomor- nickel and copper are very similar. Both have tem consists of one solid phase (a) below the phous Alloy Systems” in this Volume. More complex behavior occurs when alloy- ing elements have significantly dissimilar Table 2 Solid-state allotropes in selected metals atomic sizes and different crystalline structures

Temperature range than that of the parent matrix. When the host (solvent) lattice cannot dissolve any more - Element Structure C F ute atoms, the excess solute may group together hcp <1250 <2280 to form a separate distinct phase. Excess solute bcc >1250 >2280 Calcium fcc <450 <840 atoms can also react with solvent atoms to form bcc >450 >840 an intermetallic compound (which is a phase hcp –150 to –10 –238 to –14 with a crystal structure different from that of fcc –10 to 725 –14 to 1340 either pure metal). For example, when the level bcc >725 >1340 hcp <425 <795 of carbon exceeds its solubility limit in iron, the fcc Quenched from >450 Quenched from >840 excess carbon not dissolved in the iron gets tied hcp <1381 <2520 up in separate phases of a metastable interme- bcc >1381 >2520 tallic compound, Fe C, called cementite. hcp <1264 <2310 3 bcc >1264 >2310 The iron-carbon phase diagram (Fig. 5) also Hafnium hcp <1950 <3540 makes a distinction between phase boundaries > > bcc 1950 3540 for iron-carbon and iron carbide (Fe3C). In Holmium hcp <906 <1660 actuality, cementite (Fe C) is not a true equilib- ? >906 >1660 3 Iron bcc <910 <1670 rium phase in the iron-carbon system and is fcc 910 to 1400 1670 to 2550 referred to as metastable. Over a long exposure bcc >1400 >2550 time (depending on temperature), cementite Lanthanum hcp –271 to 310 –455 to 590 eventually decomposes into iron and free car- fcc 310 to 868 590 to 1590 bcc >868 >1590 bon (). However, the rate of decompo- hcp <–202 <–330 sition of iron carbide is extremely slow under bcc >–202 >–330 the most favorable conditions and is usually Lutetium hcp <1400 <2550 imperceptible under ordinary conditions for ? >1400 >2550 ...... Manganese Four cubic phases the applications designed. bct <–194 <–317 Because of its reluctance to decompose, the Rhombohedral >–194 >–317 metastable cementite phase is represented on hcp <868 <1590 the iron-carbon phase diagram as a phase that is bcc >868 >1590 Orthorhombic <280 <535 approximately in stable equilibrium in most Tetragonal 280 to 577 535 to 1070 practical circumstances. The iron-cementite bcc 577 to 637 1070 to 1180 ...... phase diagram is important to industrial metal- Several phases lurgy, especially in the solid-state heat treatment Simple cubic >76 >169 Rhombohedral ...... of steel. However, the development of an iron- hcp <798 <1470 base microstructure with graphite (in the shape bcc >798 >1470 of flakes or nodules) is important in the produc- Rhombohedral <917 <1680 tion of cast irons. These alloys contain bcc >917 >1680 Scandium hcp <1000 to 1300 <1830 to 2370 additions as a catalyzing substance that helps bcc >1335 >2440 hcp <237 <460 bcc >237 >460 fcc <540 <1000 bcc >540 >1000 hcp <1317 <2400 Copper, at% > > bcc 1317 2400 Ni 10 20 30 40 50 60 70 80 90 Cu 1500 Thallium hcp <230 <445 1452 °C 2700 1410 °C bcc >230 >445 2600 1400 L Titanium hcp <885 <1630 Liquid solution 2500 > > 1370 °C 2400 bcc 885 1630 1300 L + α Liquid Curve Orthorhombic <662 <1220 2300 Tetragonal 662 to 774 1220 to 1430 1200 Solidus curve 2200 bcc 774 to 1132 1430 to 2070 2100 1100 2000 fcc RT to 798 RT to 1470 1083 °C 1900 bcc >798 >1470 1000 1800 Yttrium hcp RT to 1460 RT to 2660 Solid solution α 1700 bcc >1460 >2660 900 Temperature, °F Temperature, Temperature, °C Temperature, 1600 Zirconium hcp <865 <1590 1500 > > 800 bcc 865 1590 1400 Summary: 700 1300 1200 In 14 metals, hcp transforms to bcc as temperature increases. 600 1100 In 3 metals, hcp transforms to fcc as temperature increases. 1000 500 In 6 metals, fcc transforms to bcc as temperature increases. NI 10 20 30 40 50 60 70 80 90 Cu In 0 metals, fcc transforms to hcp as temperature increases. In 0 metals, bcc transforms to hcp as temperature increases. Copper, wt% In 1 metal, bcc transforms to fcc as temperature increases. hcp, hexagonal close-packed; bcc, body-centered cubic; fcc, face-centered cubic; bct, body-centered tetragonal; RT, room temperature Fig. 4 The nickel-copper phase diagram. Adapted from Ref 3 8 / Introduction to Alloy Phase Diagrams

Fig. 5 Iron-carbon phase diagram with constituents in the form of cementite (Fe3C) and graphite (as the equilibrium form of carbon). Although cementite is strictly a metastable phase that eventually decomposes into iron and graphite, cementite is sufficiently stable over the time scales of most engineering applications. Introduction to Phase Diagrams / 9 breaks down iron carbides to produce iron and through the solid diagram. See the article “Intro- illustration, component A is placed at the bottom free carbon as graphite during solidification. duction to Ternary Diagrams” in this Volume for left, B at the bottom right, and C at the top. The In addition, the iron-carbon phase diagram more details. amount of constituent A is normally indicated also illustrates the coexistence of two solid- Isothermal Sections. Composition values in from point C to point A, the amount of constitu- state phases in several areas. One area of het- the triangular isothermal sections are read from ent B from point A to point B, and the amount erogeneous phase equilibria is the indicated a triangular grid consisting of three sets of lines of constituent C from point B to point C. This region of ferrite (a iron) and austenite (g iron). parallel to the faces and placed at regular compo- scale arrangement is often modified when only At lower of carbon and high sition intervals (see Fig. 7a). Normally, the point a corner area of the diagram is shown. enough temperatures, a two-phase region of of the triangle is placed at the top of the Vertical sections are often taken through solid a iron and solid g iron in equilibrium one corner (one component) and a congruently forms. Another two-phase region of solid iron melting binary compound that appears on the exists with g iron and d iron at higher tempera- C opposite face; when such a plot can be read like tures. These regions are the solid-state analog any other true binary diagram, it is called a of the two-phase (mushy) zone that occurs on quasi-binary section. An example of such a sec- melting or solidification. tion is illustrated by line 1–2 in the isothermal section shown in Fig. 7(b). A vertical section between a congruently melting binary compound Ternary Diagrams X A on one face and one on a different face might also form a quasi-binary section (see line 2–3). X When more than two components exist in a C All other vertical sections are not true binary system, illustrating equilibrium conditions diagrams, and the term pseudobinary is applied graphically in two dimensions becomes more to them. A common pseudobinary section is complicated. One option is to add a third compo- one where the percentage of one of the compo- sition dimension to the base, forming a solid dia- nents is held constant (the section is parallel to gram having binary diagrams as its vertical sides. one of the faces), as shown by line 4–5 in Fig. 7(b) This can be represented as a modified isometric Another is one where the ratio of two constitu- A X B projection, such as shown in Fig. 6. Here, bound- (a) B ents is held constant, and the amount of the third aries of single-phase fields (liquidus, solidus, and is varied from 0 to 100% (line 1–5). solvus lines in the binary diagrams) become sur- C Projected Views. Liquidus, solidus, and sol- faces; single- and two-phase areas become vus surfaces by their nature are not isothermal. ; three-phase lines become volumes; Therefore, equal-temperature (isothermal) con- and four-phase points, while not shown in tour lines are often added to the projected views

Fig. 6, can exist as an invariant plane. The com- AC2 of these surfaces to indicate the shape of the sur- position of a binary eutectic liquid, which is a faces (see Fig. 7c). In addition to (or instead of) 3 point in a two-component system, becomes a line contour lines, views often show lines indicating in a ternary diagram, as shown in Fig. 6. BC the temperature troughs (also called valleys or While three-dimension projections can be 2 grooves) formed at the intersections of two sur- helpful in understanding the relationships in the faces. Arrowheads are often added to these lines diagram, reading values from them is difficult. 4 5 to indicate the direction of decreasing tempera- In order to represent three dimensions on two- ture in the trough. dimensional diagrams, pressure and temperature 1 are typically fixed. Ternary systems are often represented by views of the binary diagrams that A B Phase Diagrams comprise the faces and two-dimensional projec- (b) tions of the liquidus and solidus surfaces, along Rules and Notation with a series of two-dimensional horizontal sec- C T tions (isotherms) and vertical sections (isopleths) 11 The construction of phase diagrams is greatly T 10 facilitated by certain rules that come from ther- T L 9 modynamics. Foremost among these is the sec- Liquidus T ond law of and the Gibbs surface 8 T , which to rules governing the L + α 7 T construction of more complex phase diagrams L + β 6 Solidus (for more details see the article “Thermody- Temperature T 5 β surface namics and Phase Diagrams” in this Volume). Solidus L α β T surface + + 4 The term phase is based on the work of the T American mathematician Solvus 3 Solvus surface (Ref 4), who first introduced the term in describ- surface T 2 ing the thermodynamics of heterogeneous A B T solids. Solids may have a homogenous structure, 1 α α + β such that the alloying elements (or components) A arrange themselves into a repeatable type of (c) crystal structure. However, solids do not always have a homogenous structure of the constituent C atoms. Many solids are heterogeneous—that is, Fig. 7 Two-dimensional projections from a ternary phase diagram. (a) Isothermal section. (b) the atomic elements can arrange themselves into Fig. 6 Ternary phase diagram showing three-phase Vertical section. (c) Isothermal contours for a liquidus or more than one type of structure under equilib- equilibrium. Adapted from Ref 3 solidus projection rium conditions. 10 / Introduction to Alloy Phase Diagrams

The Gibbs Phase Rule mixture of stable compounds, such as salt and of a binary alloy. Examples of invariant point , the number of components may be taken binary phase diagrams include: The construction of phase diagrams is greatly as two (NaCl + H2O), unless the mixture is car- facilitated by certain rules that come from ther- ried to a degree of temperature and pressure Eutectic points at a specific composition and modynamics. Foremost among these is Gibbs’ where one or both of the compounds decom- temperature, where the liquid phase coexists phase rule, which applies to all states of matter pose, when it becomes necessary to consider with two different solid phases (three phases total in equilibrium) (solid, liquid, and gaseous) under equilibrium four components (Na, Cl, H, and O). conditions. The maximum number of phases For most practical applications of the phase Eutectoid points at a specific composition (P) that can coexist in a chemical system, or rule used in materials science, pressure is kept and temperature, where three different solid alloy, is: constant, and the number of degrees of freedom phases can coexist in equilibrium (three is reduced by 1, so that: phases total in equilibrium) P ¼ C F þ 2 (Eq 5) P ¼ C F ðÞþat constant pressure 1 (Eq 6) The iron-carbon system is the best example of a where F is the number of degrees of freedom, eutectoid system. and C is the number of chemical components In this case, the only variables of the system (usually elements for alloys). The phases are would be the change in chemical concentration Theorem of Le Chaˆtelier and the the homogeneous parts of a system that, having and/or temperature. In a binary alloy (C = 2), Clausius-Clapeyron Equation definite bounding surfaces, are conceivably sep- both composition and temperature could be vari- arable by mechanical means alone, for exam- ables during equilibrium, so that F = 2. In this Although the phase rule tells what lines and ple, a gas, liquid, and solid. The degrees of case, the maximum number of heterogeneous fields should be represented on a phase dia- freedom, F, are those externally controllable phase equilibria would be: P = 2 – 2+ 1 = 1. Just gram, it does not usually define their shapes or conditions of temperature, pressure, and com- one homogeneous phase would be possible, if the directions of the lines. Further guidance in position, which are independently variable and both composition and temperature could vary the latter respect may be explained by several which must be specified in order to completely while the system stays in equilibrium. additional thermodynamic rules. define the equilibrium state of the system. However, if equilibrium in a binary alloy can The theorem of Le Chaˆtelier says that if a sys- The derivation of the Gibbs phase rule is be maintained while either composition or tem- tem in equilibrium is subjected to a constraint by described in various texts on thermodynamics perature remains constant (F = 1), then two het- which the equilibrium is altered, a reaction takes of materials (Ref 2–6). In the simple case of a erogeneous phases can exist in equilibrium: place that opposes the constraint, that is, one by pure metal (where C =1), composition is a con- P = 2 – 1+ 1= 2. This possible coexistence of which its effect is partially annulled. Therefore, stant, while there could be two degrees of free- two phases in a binary system represents phase if an increase in the temperature of an alloy dom if both temperature and pressure are regions such as the mushy zone (liquid + solid results in a phase change, that phase change will allowed to change while staying in equilibrium. phases) or a two-phase mixture in a solid (such be one that proceeds with heat absorption, or if In this case (C = 1 and F = 2), then only one as the ferrite-austenite region of the iron-carbon pressure applied to an alloy system brings about coexisting phase (P =1) would be possible from system, Fig. 5). Phase diagrams of binary alloys a phase change, this phase change must be one the phase rule: P =12 + 2 = 1. However, the also can have an invariant point (where F = 0), that is accompanied by a contraction in volume. phase rule also can define equilibrium condi- such that equilibrium conditions are defined by The usefulness of this rule can be shown by ref- tions for multiple coexisting phases in a pure a specific (constant) composition and tempera- erence to Fig. 8. Consider the line showing freez- metal. For example, if pressure is held constant, ture. In this case, three coexisting phases ing, which represents for a typical pure metal the then temperature is the only variable (F = 1). In (P = 2 – 0 + 1= 3) occur for the invariant point temperature at which melting occurs at various this case, the phase rule predicts two coexisting phases in equilibrium: P =12 +2 = 2. In this case, the two phases occur when a pure metal is in the thermal plateau region that occurs dur- ing melting, solidification, or allotropic trans- formation of a pure metal. The Gibbs phase rule also defines unique and constant conditions (referred to as invariant points in phase diagrams) where the maximum number of phases can coexist while still in equi- librium. Invariant points refer to equilibrium conditions when composition, pressure, and temperature must all remain constant (F = 0). In the case of a pure metal, for example, there is a unique and constant invariant point with three coexisting phases according to the phase rule: P =10 +2 = 3. This occurrence of three coexisting phases of a pure metal refers to the unique and constant of the metal, which occurs at a very specific combination of pressure and temperature unique to a metal. The phase rule also applies to alloys with any number of independent chemical components (C). The components, C, are the smallest num- ber of substances of independently variable composition making up the system. In alloy systems, it is usually sufficient to count the number of elements present. In the case of a Fig. 8 Pressure-temperature diagram for a pure metal. Source: Ref 10 Introduction to Phase Diagrams / 11

pressures. This line slopes upward away from the proportion up to composition c, which, at this W0 ¼ Wa þ Wb (Eq 8) pressure axis. The typical metal contracts on freez- temperature, is the boundary of the single-phase ing. Hence, applying an increased pressure to the b field. Therefore, at temperature b, any alloy This equation can be used to eliminate Wa from liquid can cause the metal to become solid, experi- that contains less than a% of metal B will exist Eq 7, and the resulting equation can be solved encing at the same time an abrupt contraction in at equilibrium as the homogeneous a solid solu- for Wb to give the expression: volume. Had the metal bismuth, which expands tion; and any alloy containing more than c%of  % % on freezing, been selected as an example, the theo- metal B will exist as the b solid solution. How- ¼ A0 Aa Wb W0 % % (Eq 9) rem of Le Chaˆtelier would dictate that the solid-liq- ever, any alloy whose overall composition is Ab Aa uid line be drawn so that the conversion of liquid to between a and c (e.g., at d) will, at the same tem- solid with pressure change would occur only with a perature, contain more metal B than can be dis- Although a similar expression can be obtained reduction in pressure; that is, the line should slope solved by the a and more metal A than can be for the weight of the a phase Wa, the weight upward toward the pressure axis. dissolved by the b. It will therefore exist as a of the a phase is more easily obtained by means A quantitative statement of the theorem of Le mixture of a and b solid solutions. At equilib- of Eq 8. Chaˆtelier is found in the Clausius-Clapeyron rium, both solid solutions will be saturated. The Because the weight of each phase is deter- equation. Referring again to Fig. 8, this equa- composition of the a phase is therefore a%of mined by chemical composition values accord- tion leads to the further conclusion that each metal B and that of the a phase is c% of metal B. ing to Eq 9, the tie-line ac shown in Fig. 9 of the curves representing two-phase equilib- When two phases are present, as at composi- can be used to obtain the weights of the phases. rium must lie at such an angle that on passing tion Y in Fig. 9, their relative amounts are deter- In terms of the lengths in the tie-line, Eq 9 can through the point of three-phase equilibrium, mined by the relation of their chemical be written as: each would project into the region of the third compositions to the composition of the alloy.  length of line da phase. Thus, the sublimination-desublimination This is true because the total weight of one of Wb ¼ W0 (Eq 10) line must project into the liquid field, the vapor- the metals, for example metal A, present in the length of line ca ization curve into the solid field, and the solid- alloy must be divided between the two phases. where the lengths are expressed in terms of the liquid curve into the vapor field. This division can be represented by: numbers used for the concentration axis of the    diagram. The lever rule, or inverse lever rule, %A0 %Aa %Ab W0 ¼ Wa þ Wb can be stated: The relative amount of a given 100 100 100 The Lever Rule  phase is proportional to the length of the tie- weight of metal ¼ line on the opposite side of the alloy point of The lever rule is one of the cornerstones of A in alloy the tie-line. Thus, the weights of the two phases understanding and interpreting phase diagrams. weight of metal weight of metal þ are such that they would balance, as shown in A portion of a binary phase diagram is shown in A in a phase A in b phase (Eq 7) Fig. 10. Fig. 9. In this diagram, all phases present are Using Eq 10, the weight of the b phase at solid phases. There are two single-phase fields where W0, Wa, and Wb are the weights of the composition Y in Fig. 9 is: labeled a and b, separated by a two-phase field alloy, the a phase, and the b phase, respec-  labeled a + b. It indicates that, at a temperature tively, and %A0,%Aa, and %Ab are the respec- 40 25 W ¼ W ¼ W ðÞ0:375 (Eq 11) such as b, pure metal A can dissolve metal B in tive chemical compositions in terms of metal A. b a 65 25 a any proportion up to the limit of the single-phase Because the weight of the alloy is the sum of a field at composition a. At the same tempera- the weight of the a phase and the weight of The percentage of b phase can be determined ture, metal B can dissolve metal A in any the b phase, the following relationship exists: by use of:   W Percentage of b phase ¼ b 100 W 0 %A %A ¼ 0 a 100 %Ab %Aa (Eq 12) At composition Y the percentage of b phase is:  40 25 % b phase ¼ 100 ¼ 37:5% (Eq 13) 65 25 The percentage of a phase is the difference between 100% and 37.5%, or 62.5%.

Lines and Labels

Magnetic transitions (Curie temperature and Ne´el temperature) and uncertain or speculative boundaries are usually shown in phase dia- grams as nonsolid lines of various types. The components of metallic systems, which usually are pure elements, are identified in phase diagrams by their symbols. Allotropes of poly- morphic elements are distinguished by small Fig. 9 Portion of hypothetical phase diagram. Source: Ref 11 (lowercase) Greek letter prefixes. 12 / Introduction to Alloy Phase Diagrams

in a bcc matrix of a iron (Fig. 12), reaching a maximum of only 0.0218 wt% C at 727 C. When carbon exceeds the solubility limit, the excess carbon combines with iron to form cementite (Fe3C). However, if the steel is heated above the upper critical temperature (Ac3), the austenitic (fcc) phase can accommodate (dissolve) more carbon atoms than ferrite (a-phase) iron. The lattice spacing between atoms of g iron is greater than that of a iron (Fig. 13), resulting in a higher solu- bility limit of carbon. Thus, when steel is heated above the Ac3 temperature, the hard cementite phase (Fe3C) dissolves into a solid solution of carbon in g iron. Then by controlled cooling of the austenitic steel back down into the ferritic (a iron) phase, the morphology and distribution of iron-carbide phase (cementite) can be manipu- lated to produce a wide variety of microstruc- tures and mechanical properties, as described in numerous publications. However, it should also be mentioned that the Fig. 10 Visual representation of lever rule. Source: Ref 11 iron-carbon equilibrium phase diagram is not a complete description of phase formation in steel. Terminal solid phases are normally designated Iron-Carbon Phase Diagram When austenitized steel is rapidly cooled by the symbol (in parentheses) for the allotrope (quenched), a different mechanism of phase of the component element, such as (Cr) or Steel and cast iron behavior are described in transformation occurs. During rapid quenching (aTi). Continuous solid solutions are designated large part (but not exclusively) by the iron- from austenite to ferrite, there is not enough time by the names of both elements, such as (Cu, Pd) carbon phase diagram. As noted, at standard for the excess carbon atoms to diffuse and form or (bTi, bY). pressure pure iron has three allotropic forms of cementite along with the bcc ferrite. Therefore, Intermediate phases in phase diagrams are solid iron and a magnetic at some (or all) of the carbon atoms become trapped normally labeled with small (lowercase) Greek 770 C (1420 F). The addition of carbon intro- in the ferrite lattice, causing the composition to letters. However, certain Greek letters are con- duces an additional degree of freedom and rise well above the 0.02% solubility limit of car- ventionally used for certain phases, particularly results in coexisting heterogeneous phases with bon in ferrite. This causes lattice distortion, so disordered solutions: for example, b for disor- several invariant points (Fig. 5) that include: much so that the distorted bcc lattice rapidly dered bcc, z or e for disordered close-packed transforms into a new metastable phase called hexagonal (cph), g for the g-brass-type struc- A eutectic point at 4.30 wt% C and 1148 C . Martensite does not appear as a phase ture, and s for the sCrFe-type structure. (2100 F) where liquid phase coexists with on the iron-carbon equilibrium phase diagram For line compounds, a stoichiometric phase austenite-carbon solid solution and because it is a metastable (nonequilibrium) phase name is used in preference to a Greek letter cementite that occurs from rapid cooling. (for example, A2B3 rather than d). Greek letter A eutectoid point at 0.77 wt% C at 727 C The unit cell of the martensite crystal is a prefixes are used to indicate high- and low- (1340 F) with three coexisting solid phases: body-centered tetragonal (bct) crystal structure, temperature forms of the compound (for exam- austenite-carbon solid solution, ferrite-carbon which is similar to the bcc unit cell, except that ple, aA2B3 for the low-temperature form and solid solution, and cementite one of its edges (called the c axis) is longer than bA2B3 for the high-temperature form). the other two axes (Fig. 14). The distorted form Besides the occurrence of eutectic and eutec- of the bct is a supersaturated phase that accom- toid reactions, the formation of austenite at high modates the excess carbon. The bct structure Application of Phase Diagrams temperatures is instrumental in the solid-state also occupies a larger atomic volume than fer- processing of steel. The fcc crystal structure of rite and austenite, as summarized in Table 3 for different microstructural components as a The use of phase diagrams in the application austenite is much more ductile that the bcc crystal structure of ferrite (because fcc crystals function of carbon content. The of mar- of typical alloy systems is discussed in separate tensite thus is lower than ferrite (and also aus- articles based on the type or category of phase have more active slip systems than bcc crys- tals). Thus, hot working of steel is typically tenite, which is denser than ferrite). The diagram. The different types of liquid-solid resulting expansion gives martensite its high transformations are discussed in articles on: done at temperatures above the upper critical temperature (Ac3) with complete austenitiza- hardness and is the basis for strengthening Isomorphous alloy systems tion (100% g iron) of steel. by heat treatment. Eutectic alloy systems Carbon additions also lower the upper critical Peritectic alloy systems temperature (Ac3) up to the eutectoid point of Binary Titanium Phase Diagrams with a Monotectic alloy systems approximately 0.77 wt% C (Fig. 11). This and b Stabilizing Elements allows for complete dissolution of cementite at Application of phase diagrams also is instru- lower temperatures and the subsequent manipu- Like iron, titanium is an allotropic element mental in understanding solid-state transforma- lation of carbide formation during cooling. Dis- with two solid-state phases: tions for the processing and heat treatment of solution of cementite can occur during heat alloys. The following sections provide brief treatment, because there is a great difference a titanium with a hcp crystal structure at room introductions on various types of alloy systems in the solid solubility of carbon in ferrite and temperature and up to 882 C (1620 F) that involve solid-state transformations. More austenite. The limit on solid solubility can be b titanium with a bcc crystal structure that is details also are given in the article “Solid-State increased by raising the temperature to some stable from 882 C up to the melting Transformations” in this Volume. extent, but carbon has very limited solubility temperature