Beruflich Dokumente
Kultur Dokumente
art ic l e i nf o a b s t r a c t
Article history: Modelling of (solid-state) precipitation kinetics in terms of particle nucleation and particle growth re-
Received 29 January 2015 quires evaluation of the thermodynamic relations pertaining to these mechanisms, i.e. evaluation of the
Received in revised form nucleation barrier and of the GibbsThomson effect. In the present work, frequently occurring problems
9 April 2015
and misconceptions of the thermodynamic evaluation are identied and a practical approach with regard
Accepted 30 April 2015
to kinetic modelling is proposed for combined and unied analysis of the thermodynamics of nucleation
Available online 4 May 2015
and growth, based on the fundamental thermodynamic equilibrium consideration in a particlematrix
Keywords: system. A computationally efcient method for numerical determination of the thermodynamic relations
GibbsThomson effect is presented which allows an easy and exible implementation into kinetic modelling.
Thermodynamics
& 2015 Elsevier Ltd. All rights reserved.
CALPHAD
Precipitation kinetics
Modelling
n 1
Corresponding author. Fax: 49 711 689 3312. Within the scope of this work, only the case of a, in the Gibbsian sense,
E-mail addresses: b.rheingans@is.mpg.de (B. Rheingans), sharp interface, i.e. an interface with a width small compared to the size of the
e.j.mittemeijer@is.mpg.de (E.J. Mittemeijer). particle will be considered.
http://dx.doi.org/10.1016/j.calphad.2015.04.013
0364-5916/& 2015 Elsevier Ltd. All rights reserved.
50 B. Rheingans, E.J. Mittemeijer / CALPHAD: Computer Coupling of Phase Diagrams and Thermochemistry 50 (2015) 4958
nucleation and the GibbsThomson effect, affecting the growth 2). Analytical expressions for gc (x ,m , x,p), x ,int (r ) and x,int (r ),
(rate) of a particle. In the classical theory of nucleation [8,9], the based on simple thermodynamic solution models for the chemical
rate of nucleation N is dominated by an energy barrier G* for Gibbs energies of the and the phase, are an often used, nu-
formation of a particle of critical size rn above which the particle is merically efcient way to implement thermodynamic data into the
stable: numerical kinetic modelling. For instance, the GibbsThomson
effect in a binary system is often accounted for by application of
the equation [14]
G*
N exp ,
kT (1) 2 V 1
x ,int (r ) = x r exp mol
where k and T denote the Boltzmann constant and the absolute RT r (3)
temperature, respectively.2 For the case of a precipitation reaction,
for the composition of the matrix at the particlematrix interface,
G* and rn are functions of the change in chemical Gibbs energy
where x (r ) is the solute concentration of the phase in the
gc (x ,m , x,p) upon nucleation (with gc (x ,m , x,p) being the
reference state of equilibrium between the phase and the
chemical driving force for nucleation) for given compositions
phase with r , i.e. between the bulk phases in the absence of
x ,m and x,p of the -phase matrix and the -phase precipitate,
the interface. Vmol is the mean molar volume of the -phase and R
respectively, and of the interface energy per unit area, i.e.
denotes the gas constant. Eq. (3) is based on the assumption that
G* = G* (gc (x ,m , x,p) , ) and r * = r * (gc (x ,m , x,p) , ), thus re- the thermodynamic behaviour of the -matrix phase can be
ecting the two competing energy contributions. Growth of a described with the regular solution model and that the -pre-
solute-rich particle leads to solute depletion of the surrounding
matrix; particle growth can then (in any case eventually) become
(
cipitate phase is a pure phase, i.e. x,int (r ) = x r = 1. The )
applicability of such analytical expressions can thus be severely
rate-controlled by solute diffusion through the solute-depleted
limited by the limited capability of the underlying simple solution
matrix towards the particle. The growth rate of a spherical particle
models to adequately describe the actual thermodynamic beha-
of radius r in a binary3 system AB is then often described by
viour of the phase and the phase.
[12,13]
Hence, in recent years, direct numerical derivation of the re-
dr x ,m x ,int D lations gc (x ,m , x,p), x ,int (r ) and x,int (r ) from a comprehensive
= ,
dt kx,int x ,int r (2) thermodynamic assessment of the alloy system (which is typically
based on more complex solid solution models for the chemical
with the diffusion coefcient D of the solute component in the
Gibbs energies) has become more frequently applied, especially for
matrix and the atom fractions4 of solute x ,m in the -phase matrix
multi-component systems (see, e.g., [5,6]). This trend is facilitated
remote from the particle, and x ,int and x,int in the -phase matrix
by the increasing availability of such thermodynamic assessments,
and in the -phase particle at the particlematrix interface,
e.g. in form of CALPHAD data, and commercial software for ther-
respectively; the factor k accounts for the difference in molar
modynamic analysis (e.g. [15]). On the one hand, the numerical
volume of the phase and the phase. For x ,m > x ,int , i.e. for a
determination of the chemical driving force for nucleation,
positive growth rate (considering precipitation of a solute-rich
gc (x ,m , x,p), for a given composition x ,m of the matrix phase,
phase, kx,int x ,int is generally positive), the particle is stable and
can be performed straightforwardly, for instance by application of
grows; for x ,m < x ,int , the particle is unstable and shrinks. x ,int
the parallel tangent/maximum chemical driving force approach
and x,int are often taken according to local establishment of
[16] (see Section 2). On the other hand, the numerical evaluation
thermodynamic equilibrium at the interface.5 For a small particle
of the GibbsThomson effect, i.e. the determination of the com-
size, i.e. for a large ratio of interface area to particle volume, the
positions x ,int (r ) and x,int (r ), is much more elaborate [1719],
state of equilibrium between the -phase matrix and the -phase
since it requires the evaluation of a thermodynamic equilibrium
precipitate can strongly deviate from the state of equilibrium
state including the energy contribution of the interface, e.g. by
between the - and -bulk phases, i.e. the phase and the
minimisation of the total Gibbs energy [15]. In view of the corre-
phase in the absence of the interface. This is the so-called Gibbs
spondingly larger complexity and computational effort, direct
Thomson effect, which, in compliance with the two counteracting
numerical evaluation of the GibbsThomson effect is in practice
contributions of composition-dependent chemical Gibbs energy
often avoided and simple analytical expressions such as Eq. (3),
and interface energy, can be expressed by functions x ,int and x,int
based on generally invalid solid solution models, are adopted in-
depending on particle size (i.e. interface area) and interface energy
stead. Obviously, problems of inconsistency arise when the rela-
per unit area, i.e. x ,int = x ,int r, and x,int = x,int r, .
( ) ( ) tions gc (x ,m , x,p) for nucleation and x ,int (r ) and x,int (r ) for
Kinetic modelling of nucleation and growth thus requires growth are derived, analytically or numerically, by adoption of
evaluation of the thermodynamics of the system dened by che- differing, incompatible thermodynamic solution models for nu-
mical energy and interface energy. Usually, the interface energy cleation and for growth. This is a common shortcoming in kinetic
per area is taken as being constant within a certain range of models of precipitation kinetics based on the KWN-approach (e.g.
composition, particle size and morphology. This assumption ef- [2,3]).6 The problem becomes even more aggravated when an
fectively allows to reduce the evaluation of the thermodynamics elastic strain energy contribution due to a precipitate/matrix mist
for the kinetic modelling to determination of (i) the nucleation is taken into consideration only for nucleation but not for growth
barrier as a function of the composition-dependent chemical (or vice versa) without more ado (e.g. [22,23]). As a consequence,
driving force for nucleation gc (x ,m , x,p) and (ii) the composi- the kinetic model predictions may be strongly biased or com-
tions x ,int (r ) and x,int (r ) as a function of particle size r (cf. Section pletely corrupted by the incongruent thermodynamic descriptions
for nucleation and for growth.7
2
For a full expression of N according to classical nucleation theory, see e.g.
6
[10]. Naturally, this problem does not appear in kinetic models without dis-
3
For multinary systems, see e.g. [11]. crimination of nucleation kinetics and growth kinetics, as in cluster dynamics
4
For binary systems, the convention x j = xBj will be used. models, cf. e.g. [20]. Also, in kinetic models involving a consideration of the total
5
cf. footnote at the end of Section 2.3. (Gibbs) energy of the system [21], such problems are more readily avoided.
B. Rheingans, E.J. Mittemeijer / CALPHAD: Computer Coupling of Phase Diagrams and Thermochemistry 50 (2015) 4958 51
Against the above background, in the present work a generally in the supersaturated -phase matrix of composition x ,m eval-
valid method for combined, inherently consistent, numerical uated at the composition x,p of the nucleus (see Fig. 1, note that all
evaluation of nucleation barrier and GibbsThomson effect is data presented for the binary case in Figs. 1(b) and 3 pertain to the
proposed which is founded on the Gibbsian treatment of nuclea- example given in Section 4).
tion and growth by fundamental consideration of thermodynamic ( )
Introducing the chemical potentials c,ji x j for the components
equilibrium in a particlematrix system [24,25]: starting from ty- i in the (bulk) phase j = , , the chemical Gibbs energy change per
pically applied approaches for separate thermodynamic evalua-
mol, gc,mol = gc,mol
gc,mol , can be expressed as
tions of nucleation and growth, it is rst shown how these ap-
proaches can be retraced to the same evaluation of thermo-
gc,mol
gc,mol = gc,mol (x ,m , x,p)
dynamic equilibrium. Introducing assumptions typically made = xi,p c,i (x,p) xi,p c,i (x ,m).
upon coupling thermodynamics to kinetics (e.g. a constant inter- i i (7a)
face energy per unit area), a computationally efcient method for
For the case of nucleation, it thus follows with Eq. (5):
numerical evaluation of nucleation barrier, critical radius and
GibbsThomson effect is presented, which can be easily adapted to gc,mol (x ,m , x,p) = xi,p c,i (x,p) xi,p c,i (x ,m)
practical cases of modelling transformation kinetics. i i
2
= + gel V mol
r* (7b)
2. Theoretical background
For a binary system, Eq. (7b) can be visualised in a diagram of
2.1. Nucleation barrier j
molar chemical Gibbs energy Gc,mol (x j ) vs. composition xj (see
Fig. 1). gc,mol (x ,m , x,p) is represented by the difference of the
Evaluation of nucleation thermodynamics, i.e. determination of
ordinate value of the tangent of the Gc,mol (x )-curve drawn at x ,m
the nucleation barrier G* and the critical radius rn, classically
(the given composition of the matrix phase) and evaluated at x,p
departs from a consideration of the change of Gibbs energy G (r )
of the system with size of the precipitate particle formed, i.e. in
(the composition of the nucleus), and the Gc,mol
x -curve at ( )
case of formation of a spherical particle with radius r x = x,p . For a given matrix composition x ,m , Eq. (7b) does not
uniquely dene gc,mol (x ,m , x,p) for nucleation, since the compo-
4 3
G (r ) = r + 4 r 2 , sition of the nucleus x,p is not known a priori. In principle, for
3 (4)
given x ,m , the composition x,p of the precipitate nucleus may
adopting the sharp interface model (see footnote in Section 1) assume any value for which gc,mol (x ,m , x,p) is negative. If x,p is
with a constant (i.e. independent of r and composition), isotropic
interface energy per unit area. denotes the contributions to the
chosen such that the tangent of the Gc,mol
( )
x -curve drawn at x,p is
parallel to the tangent of the Gc,mol (x )-curve (xed by the given
Gibbs energy change which scale with the volume of the particle.
matrix composition x ,m ), the chemical driving force
For precipitation in the solid state, two contributions to (per unit
gc,mol (x ,m , x,p) assumes a maximum value and is solely dened
volume of particle) are typically taken into consideration: the
(negative) change in chemical Gibbs energy gc upon formation of by the composition of the matrix phase, i.e.
a particle and an additional (positive) contribution of strain energy ( ) max
max gc,mol (x ,m , x,p) = gc,mol ( )
(x ,m ) and x,p = x,p x ,m . This
gel , arising due to elastic accommodation of the particle/matrix is the so-called maximum chemical driving force approach or
mist (with = gc + gel < 0). gc and gel are usually treated as parallel tangent approach [16,7], an originally graphical method to
constants for a given matrix composition. The Gibbs energy determine the maximum chemical driving force for a binary
change then shows a maximum for dG (r ) /dr = 0 for r equal to system which can be straightforwardly generalised to the multin-
max
the critical radius: ary case to numerically determine gc,mol (x ,m ) and
r* =
2
=
2
,
( )
x,p = x,p x ,m (see Fig. 2). Other methods for describing nuclea-
gc + gel (5) tion depart from setting x,p equal to the composition of the
phase corresponding with equilibrium between the - and -bulk
and with the nucleation barrier: phases, thus inherently employing a lower chemical driving force;
in the following, x,p will always be identied with the composi-
4 2 16 3
G* = G (r *) = r * = . tion of the phase pertaining to the maximum chemical driving
3 3 (gc + gel )2 (6) force. Use of the maximum chemical driving force is justied
For evaluating the change gc in chemical Gibbs energy upon recognising the strong dependency of the nucleation rate N on
particle formation, it is assumed that particle formation occurs gc,mol (x ,m , x,p) (Eqs. (1) and (6)), allowing to neglect nucleation
within an innitely large matrix phase (or that the particle is of particles of composition different from that corresponding with
negligibly small), implying a constant matrix composition x ,m the maximum chemical driving force [10].
(and thus constant gc for certain composition x,p of the particle).
The change in chemical Gibbs energy per unit volume of the 2.2. GibbsThomson effect
particle is dened as the difference gc = gc gc of the chemical
Gibbs energy with gc being the chemical Gibbs energy of the Calculation of the growth rate requires evaluation of the Gibbs
particle-forming components in the phase of composition x,p Thomson effect, i.e. determination of the compositions x ,int (r ) and
and gc being the chemical Gibbs energy of the same components x,int (r ) as a function of particle size (cf. Eq. (2)). This involves
analysis of the equilibrium state for the -phase particle em-
7
bedded in the -phase matrix, accounting for the contributions of
Such type of inconsistencies can in an extreme case lead to the unphysical
the chemical energies, of the interface energy and of the elastic
scenario that a particle of certain size generated by nucleation immediately ex-
periences a negative growth rate due to (x ,m x ,int ) < 0 (cf. Eq. (2)), i.e. it would mist-strain energy. Thermodynamic equilibrium, here at constant
be instantaneously, intrinsically unstable. p and T, corresponds to equality of the (total8) chemical potentials
52 B. Rheingans, E.J. Mittemeijer / CALPHAD: Computer Coupling of Phase Diagrams and Thermochemistry 50 (2015) 4958
Therefore, without restricting the generality of the following dis- x ,m (r ) = x ,int (r ) and x,p (r ) = x,int (r ) as a function of particle ra-
cussion, several simplifying assumptions and boundary conditions dius r employing suitable functions for the chemical potentials
are introduced here: c,ji (x j ) (i.e. employing suitable thermodynamic solution models
8 9
Derived from the total energy function, i.e. in the present case the total Gibbs For r , the elastic energy contribution, in contrast to the interface energy
energy incorporating the contributions of chemical Gibbs energy, interface energy contribution, does not vanish, since both the chemical Gibbs energy and the elastic
and elastic energy, see e.g. [26]. energy contribution scale with the volume of the particle.
B. Rheingans, E.J. Mittemeijer / CALPHAD: Computer Coupling of Phase Diagrams and Thermochemistry 50 (2015) 4958 53
Gc,mol
(x ) and Gc,mol (x ) for the and phases). as presented above may be elucidated in the following way: from
In case of binary systems AB, closed, analytical expressions for Eqs. (9) and (10), the total molar Gibbs energy
,p
x ,int (r ) and x,int (r ) can be obtained if very simple mathematical gmol = i i,p (x,p, r ) x i,p of a precipitate particle with radius r and
expressions hold for the chemical potentials c,ji (x j ), i.e. if very composition x,p is given by (cf. Eq. (9))
simple solid solution models are applicable (cf. Eq. (3)). For ex-
ample, adopting the ideal solution model for the phase and
i i,p (x,p , r ) xi,p = i,m (x ,m , r ) xi,p
i (12)
assuming a pure phase of the solute component B, the chemical
j ,0 and (cf. Eq. (10))
potential c,B (x j ) is given by c,B (x ) = c,B + RT ln (x ), and
c,B ,0
(x ) = c,B , since x = x,p = 1 (the corresponding chemical po- 2 ,p
tential
of component A in the phase is not dened). Then, i c,i (x,p) + + gel V mol,
i xi = c,i (x ,m) xi,p
c,A r i (13a)
,0
c,B ( (
+ RT ln x r )) = ,0
c,B holds for equilibrium between the
bulk phases, i.e. in the absence of interface and mist strain. Ap-
c,i (x,p) xi,p + 2 + gel V mol
c,i (x ,m) xi,p ,
plying Eq. (10) for the equilibrium between the -phase matrix i r
=
(13b)
and the -phase precipitate in this case then leads to a Gibbs i
Thomson-type equation:
with i Vmol, ,p
and i x i,p = 1. In a diagram of molar
i xi = Vmol
V chemical Gibbs energies vs. composition of a binary system, Eq.
2
x ,m (r ) = x ,int (r ) = x (r ) exp
mol,B
+ gel .
(13b), under the constraint of satisfying Eq. (10),10 for each
r RT (11) component i, represents a construction of two parallel tangents,
one of the Gc,mol (x ) curve and one of the Gc,mol (x ) curve, with a
As compared to Eq. (3), this expression includes an additional
correction for the elastic energy contribution. Similar analytical (
vertical offset of 2 /r + gel Vmol )
, since this additional term to
expressions for x ,int (r ) and x,int (r ), based on other (simplifying) each c,i is independent of composition. Eq. (13b), under the
assumptions, can be found in numerous works (see e.g. [29,4] for constraint of satisfying Eq. (10) for each component, thus is
the case of gel = 0, and [28] including a consideration of elastic identical with Eq. (7) derived specically for nucleation, provided
mist-strain and interface stress for the special case of ideal solid the latter is evaluated according to the parallel tangent/maximum
solutions). chemical driving force approach (see discussion below Eq. (7)).
For describing the GibbsThomson effect in the general case, Hence, evaluation of nucleation thermodynamics by combination
i.e. without resorting to specic simple solution models for the of Eq. (7) with the maximum chemical driving force approach
and phases and without restriction to binary systems, Eq. (10) represents an evaluation of the thermodynamic equilibrium of the
has in principle to be evaluated numerically to obtain the (nu- system particlematrix as a function of the variables x ,m , x,p and r
merical) relationships x ,int (r ) and x,int (r ), see e.g. [17,18] or [19]. and thus comprises both nucleation thermodynamics and growth
thermodynamics.11
2.3. The equivalence of nucleation thermodynamics and growth Irrespective of the simplifying assumptions introduced above,
thermodynamics equivalence of nucleation thermodynamics and growth thermo-
dynamics generally holds as long as nucleation is treated as oc-
Now considering the thermodynamics of both nucleation and curring via a critical state for which the particle is in an unstable
growth, as in case of kinetic modelling, separate analysis of nu- equilibrium with the matrix and growth is treated as involving
cleation thermodynamics (Eq. (7b)) and of growth thermo- local establishment of thermodynamic equilibrium at the particle
dynamics (the GibbsThomson effect, Eq. (10)) is unnecessary: matrix interface (see Section 1).12 Separate evaluation of nuclea-
both nucleation and growth thermodynamics are based on the tion thermodynamics and growth thermodynamics is then gen-
same thermodynamic stability consideration. In case of nucleation, erally unnecessary, as will be illustrated in the next section.
the critical size r * = r (g (x ,m , x,p (x ,m )) , , g ) for stability of Despite the fundamental nature of the equivalence of nuclea-
c,mol el
tion thermodynamics and growth thermodynamics [24], it is fre-
a nucleus is determined, which, for given and gel , is dened by a
quently disregarded for analysis of the thermodynamics of nu-
difference in chemical Gibbs energy (Eq. (7b)) and thus by the
cleation and growth with respect to kinetic modelling (but, cf. e.g.,
compositions x ,m and x,p = x,p (x ,m )) of the -matrix phase and
the -precipitate phase, respectively. In case of growth, the com- [27,28,16,7,4]) a possible reason may lie in typical textbook
treatments of nucleation thermodynamics in terms of an energy
positions ( ) ( ) of the -phase matrix and
x ,int r , = x ,m r ,
change (cf. Eq. (4)), and of growth in terms of compositions only
x,int (r, ) = x,p(r, ) of the -phase precipitate particle, as dened (instead of, e.g., activities; cf. Eq. (2)) [14]. Moreover, its implica-
by a difference in chemical Gibbs energy (c.f. Eq. (10)), are eval- tions for thermodynamic evaluation upon kinetic modelling based
uated for a particle of certain size r stable in an -phase matrix. on a modelling approach as outlined in Section 1 have, to the
Thermodynamic analysis in both cases thus breaks down to an
evaluation of the equilibrium state for an -phase matrix of certain 10
Eq. (13b) is only a necessary condition, while Eq. (10) is a sufcient
composition containing a -phase particle of certain composition condition.
and certain size r, associated with a certain interface energy and 11
Eq. (7) results from a maximisation of the system energy change as a function
elastic strain energy. As such, this is a consideration of thermo- of the radius r. Combined with a maximisation of the chemical driving force as a
function of the composition x at given x ,m , this corresponds, under the present
dynamic equilibrium in a heterogeneous system with counter-
assumptions, to the determination of equilibrium as a function of r and x (the
acting contributions to the total energy scaling differently with function G (r , gc,mol (x = x ,m, x )) shows a saddle-point at r = r*, x = x,p ). In-
size of the second-phase domain (see [24], p. 252 ff., for a dis- deed, the maximum chemical driving force/parallel tangent method is derived from
cussion of particle formation/nucleation and stability in a one- a consideration of thermodynamic equilibrium [16].
12
component system ensuing from a consideration of thermo- In case these assumptions do not hold, e.g. in the (very) early stages of
particle growth before particle-growth kinetics become rate-controlled by long-
dynamic equilibrium, there in terms of the variables size and range volume diffusion of solute only (i.e. for an, in comparison, low interface
pressure). mobility), an approach must be chosen which does not use this point of departure;
The equivalence of the nucleation and growth thermodynamics see here, e.g., [21] and, more generally, [30] and [31].
54 B. Rheingans, E.J. Mittemeijer / CALPHAD: Computer Coupling of Phase Diagrams and Thermochemistry 50 (2015) 4958
max
Fig. 4. Maximum change in chemical Gibbs energy gc,mol (x ,m ) as a function of
the matrix concentration x ,m for a ternary system ABC with phases and (a Fig. 5. The relation r (x ,m, , gel ) as a function of the matrix composition x ,m in a
surface in this gure; cf. Fig. 2). The corresponding composition relation x,p (x ,m ) ternary ABC system with phases and , calculated from the maximum change
between the -precipitate phase and the -matrix phase is here represented as a in chemical Gibbs energy gc,mol max
(x ,m ) as shown in Fig. 4 by use of Eq. (5), with
plot of the surface gc,mol (x,p (x ,m )) . Each composition pair (x,p, x ,m ) is related by arbitrarily chosen values for and gel (for gel = 0 , this gure corresponds to the
a tie-line (the dashed red line represents the tie-line for the case depicted in Fig. 2). so-called GibbsThomson phase diagram as presented in [18]). The correspond-
For gc,mol (x ,m ) = 0 , i.e. zero chemical driving force, the ternary equilibrium be- ing composition relation x,p (x ,m ) is represented by a plot of the surface
tween the - and -bulk phases is obtained (cf. Fig. 3). (For interpretation of the r (x,p (x ,m ), , gel ). Corresponding relations pertaining to different values of and
references to colour in this gure caption, the reader is referred to the web version gel can be easily re-obtained from the gc,mol max
(x ,m )-data via Eq. (5). When con-
of this paper.)
sidering nucleation, the critical radius r = r (x ,m, , gel ) and the composition of
the nucleus x,p (x ,m ) can directly be obtained from r (x ,m, , gel ) , equivalent to
the binary case (cf. Fig. 3). In case of growth, direct inverse evaluation of
corresponds to the existence of a single tie-line at given p and T r (x ,m, , gel ) in order to determine x ,int (r ) and x,int (r ) is not possible (in contrast
to the binary case, Fig. 3 (c)): For a given value of r, an innite number of com-
relating one pair of compositions (x ,m , x,p), analogous to equili-
position pairs (x,p, x ,m ) exists (equivalent to equilibrium between bulk phases, i.e.
brium between bulk phases in a binary system, cf. Fig. 3). Starting for r and gel = 0 ). For modelling of particle growth in a multi-component
from a composition x,p (x ,m ) = x,int of the particle dened by the system under assumption of equilibrium at the particlematrix interface, particle
composition of the matrix x ,m = x ,int at the moment of its nu- radius and as well as compositions must therefore be tracked. Size and composi-
tions then follow coupled trajectories on the r (x ,m, , gel ) and
cleation, the compositions x ,int and x,int then evolve upon particle r (x,p (x ,m ), , gel )-surfaces dened by the diffusion kinetics of the different
growth (see, e.g., Eq. (2)) according to x ,int = x ,m (r , , gel ) and components in the matrix phase.
(
x,int = x,p r , , gel .)
In a multinary system, however, equilibrium is not uniquely
then follow a path, which is governed by the particle-growth ki-
dened by the value of the particle radius r (the system of Eqs. (10)
netics in the multi-component system [11], on the size-composi-
would then be an underdetermined system), i.e. the composition
tion surfaces determined by thermodynamics (i.e. the r (x ,m )and
of the phases cannot be determined directly via the inverse eva-
r (x,p) surfaces in Fig. 5; cf. [5,6]).
luation method (this corresponds to the existence of multiple tie-
The equivalence of nucleation and growth thermodynamics in
lines at given p and T relating multiple pairs of compositions
principle also allows application of a reverse procedure to obtain
(x ,m , x,p), analogous to phase equilibrium between bulk phases in
the thermodynamic relations between particle size and phase
a multinary system; cf. Figs. 4 and 5). Starting from the, also for the
compositions: in a rst step, the evaluation of the growth ther-
multinary case, known compositions x ,int = x ,m (r , , gel ) and
modynamics via Eq. (10) could be performed for chosen values of r
( )
x,int = x,p r , , gel = x,p (x ,m ) of matrix and particle at the
to determine x ,int = x ,m (r , , gel ) and x,int = x,p (r , , gel ), for
moment of nucleation, upon particle growth the generally differ- given interface energy and elastic energy contribution gel . Then,
ing diffusivities of the components in the matrix can then lead to a
continuous change of the compositions in addition to the size ef- in a second step, the corresponding relation r * = r (x ,m , , gel ),
fect. Modelling of the growth kinetics in a multi-component sys- required for description of the nucleation kinetics, could be ob-
tem considering the GibbsThomson effect therefore requires tained by inverse evaluation. However, upon assumption of con-
treatment of the coupled evolutions of particle radius and of phase stant and gel , determination of the thermodynamic relations as
compositions under the equilibrium constraint provided by Eq. presented above is much more convenient that this reverse pro-
(10) (see e.g. [11]). This constraint corresponds to subjecting the cedure: evaluation of the thermodynamics of growth as a rst step
size-composition trajectory of the particle to the relations would require a complete evaluation of thermodynamic equili-
r (x ,m , , gel ) and x,p (x ,m ) obtained via the inverse evaluation brium via Eq. (10) for each value of , gel and r [1719]. The ap-
method (Fig. 5): particle size and the compositions at the interface proach via nucleation thermodynamics, by contrast, allows to split
56 B. Rheingans, E.J. Mittemeijer / CALPHAD: Computer Coupling of Phase Diagrams and Thermochemistry 50 (2015) 4958
5. Conclusions
References
[16] M. Hillert, Application of Gibbs energycomposition diagrams, in: H. [30] E. Gamsjaeger, A note on the contact conditions at migrating interfaces, Acta
I. Aaronson (Ed.), Lectures on the Theory of Phase Transformations (1975), 2nd Mater. 55 (2007) 48234833.
Edition,The Minerals, Metals and Materials Society, Warrendale, 1999, [31] F.D. Fischer, J. Svoboda, H. Petryk, Thermodynamic extremal principles for
pp. 133. irreversible processes in materials science, Acta Mater. 67 (2014) 120.
[17] S. Shahandeh, S. Nategh, A computational thermodynamics approach to the [32] B. Rheingans, E.J. Mittemeijer, Analysis of precipitation kinetics on the basis of
GibbsThomson effect, Mater. Sci. Eng. A 443 (2007) 178184. particle-size distributions, Metall. Trans. A 2015, http://dx.doi.org/10.1007/
[18] Q. Du, M. Perez, W.J. Poole, M. Wells, Numerical integration of the Gibbs s11661-015-2937-x, in press.
Thomson equation for multicomponent systems, Scr. Mater. 66 (2012) [33] H. Aaronson, F. Legoues, An assessment of studies on homogeneous diffusional
419422. nucleation kinetics on binary metallic alloys, Metall. Trans. A 23 (1992)
[19] Thermo-Calc Software AB, Stockholm, Sweden, TC-PRISMA User's Guide and 19151945.
Examples, Version 2.0, 2013. [34] T. Ebel, R. Kampmann, R. Wagner, Nucleation in a Cu0.8at-percent-Co alloy-
[20] G. Martin, Reconciling the classical nucleation theory and atomic scale ob- polarized neutron-scattering investigation and clusterdynamics description, J.
servations and modeling, Adv. Eng. Mater. 8 (2006) 12311236. Phys. IV 3 (1993) 295298.
[21] J. Svoboda, F. Fischer, P. Fratzl, E. Kozeschnik, Modelling of kinetics in multi- [35] R. Hattenhauer, P. Haasen, The decomposition kinetics of Cu-1 at.percentCo
component multi-phase systems with spherical precipitatesI: theory, Mater. at 823 K, studied by bright-eld zone-axis-incidence transmission electron-
Sci. Eng. A 385 (2004) 166174. microscopy, Philos. Mag. A 68 (1993) 11951213.
[22] D. den Ouden, F.J. Vermolen, L. Zhao, C. Vuik, J. Sietsma, Modelling of particle [36] M. Palumbo, S. Curiotto, L. Battezzati, Thermodynamic analysis of the stable
nucleation and growth in binary alloys under elastic deformation: an appli- and metastable CoCu and CoCuFe phase diagrams, CALPHAD: Comput.
cation to a Cu0.95 wt% Co alloy, Comput. Mater. Sci. 50 (2011) 23972410. Coupling Phase Diagrams Thermochem. 30 (2006) 171178.
[23] Z. Liu, V. Mohles, O. Engler, G. Gottstein, Thermodynamics based modelling of [37] X. Lu, M. Selleby, B. Sundman, Assessments of molar volume and thermal
the precipitation kinetics in commercial aluminium alloys, Comput. Mater. Sci. expansion for selected bcc, fcc and hcp metallic elements, CALPHAD: Comput.
81 (2014) 410417. Coupling Phase Diagrams Thermochem. 29 (2005) 6889.
[24] J.W. Gibbs, The Collected Works of J. Willard Gibbs, Longmans, Green and Co., [38] F. Legoues, H. Aaronson, Y. Lee, Inuence of crystallography upon critical
New York, 1906. nucleus shapes and kinetics of homogeneous fccfcc nucleation.3. The inu-
[25] M. Hillert, J. gren, Effect of surface free energy and surface stress on phase ence of elastic strainenergy, Acta Metall. 32 (1984) 18451853.
equilibria, Acta Mater. 50 (2002) 24292441. [39] J. Eshelby, The continuum theory of lattice defects, Solid State Phys. 3 (1956)
[26] W.W. Mullins, Thermodynamic equilibrium of a crystalline sphere in a uid, J. 79144.
Chem. Phys. 81 (1984) 14361442. [40] R. Dhl, M. Macht, V. Naundorf, Measurement of the diffusion-coefcient of
[27] J. Cahn, F. Larch, Surface stress and the chemical equilibrium of small crystals cobalt in copper, Phys. Status Solidi A 86 (1984) 603612.
II. Solid particles embedded in a solid matrix, Acta Metall. Mater. 30 (1982) [41] Y. Chang, L. Himmel, Temperature dependence of elastic constants of Cu Ag
5156. and Au above room temperature, J. Appl. Phys. 37 (1966) 35673572.
[28] C. Rottman, P.W. Voorhees, W. Johnson, The GibbsThomson equation for a [42] B. Strauss, F. Frey, W. Petry, J. Trampenau, K. Nicolaus, S. Shapiro, J. Bossy,
spherical coherent precipitate with applications to nucleation, Scr. Metall. Martensitic phase transformation and lattice dynamics of fcc cobalt, Phys. Rev.
Mater. 22 (1988) 293298. B 54 (1996) 60356038.
[29] M. Qian, The GibbsThomson effect in dilute binary systems, Metall. Mater.
Trans. A 33 (2002) 12831287.