Advanced Ceramics Processing

Chapter 9 – Sintering

See also Chapter 24 in:

“Ceramic Materials-Science & Engineering (C . Barry Carter and M. Grant Norton)

Louis Winnubst

Inorganic Membranes,
University of Twente,
Enschede, The Netherlands
E-mail: a.j.a.winnubst@utwente.nl
 Advanced Ceramics Processing - Lecture Notes

1. Introduction. The process (microstructure, homogeneity)

2. Characteristics of powders. Definitions; particle size
3. Interaction between powder particles. DLVO theory;
supension stability
4. Reaction between solid particles. Solid state diffusion
5. Preparation of powders and coatings. Solid state reactions;
nucleation – crystal growth
6. Treatments of powders. Milling, mixing, granulation
7. Compaction processing. Pressing, slip-casting,etc
8. Thermal processing. Drying, binder burn-out
9. Sintering. Solid state sintering

2
 In this lecture

• Introduction to theory of solid state sintering

• Study of sintering behaviour by using a dilatometer

4
 The Ceramic Fabrication Process
Powder Preparation

Compaction Processing
Microstructure
Sintering

Post Treatment

Sintering as a part of the ceramic fabrication process

5
 Sintering
• Why sintering?
• Which sinter stages are often distinguished?
• What is the driving force for sintering?
• What are the “sources” and “sinks” for matter transport
• Which transport paths are available?
• What is the influence of pore coordination?
• What is the mechanism of grain growth?
• Is pore growth possible? (“de –sintering”)

6
 Why sintering?
 A dense material at Temperature < Tmelt
 A porous system with sufficient strength
• Catalyst support
• Membrane/filter
 Special microstructures
• Small grains (e.g.mechanical properties)
• Large grains (special dielectrics)

7
 Which sinter stages are often distinguished?
Powder Solid state sintering:
 Powder particles fuse by

solid state diffusion

green compact neck formation
Compaction
Processing
Tsinter < 0.3 -0.7 Tm
Three stages:
•Initial initial
Sintering •Intermediate pore closure grain growth
•Final

Post
Treatment intermediate final
8
 Solid-state sintering ---- Solid state diffusion
See chapter 4: Reaction between solids
• Solid state diffusion paths:
 Ds: Surface diffusion (at solid-gas interface)
 Db: Grain boundary diffusion (grain-grain interface)
(in metals also dislocation diffusion)
 Dv: Diffusion through the lattice (= bulk or volume diffusion

• Order of diffusion coefficients:

Ds ≥ Db >> Dv
 What is the driving force of sintering?
 Reduction of the total free energy of the system (∆G ):

∆G = ∆Gv + ∆Gs + ∆Gb Change in free energy associated with volume,

surface and boundary of the grains

∆Gv ~ 0 ⇒ ∆G = ∆Gs + ∆Gb = γ s ∆As + γ b ∆Ab

 Reduction of the interface free energy

γs , γb : Interface energy (surface, boundary)

∆As , ∆Ab: Change in surface area

s: solid-vapour interface = (Surface)

b: solid-solid interface (= Grain Boundary)
 What is the driving force of sintering?
∆Gv ~ 0 ⇒ ∆G = ∆Gs + ∆Gb = γ s ∆As + γ b ∆Ab
 Reduction of the interface free energy

γs , γb : Interface energy (surface, boundary)

∆As , ∆Ab: Change in surface area

The result of sintering is a decrease in surface area by

elimination of solid-vapour or solid-solid interfaces.

 Two solid-vapour interfaces become one solid-solid interface

 Disappearance of solid-solid interfaces (at higher temperature)
 Initial sintering stage: strong decrease in ∆As (solid-vapour): γs > γb
 Sintering and curvature

Stresses exist at curved surfaces if compared with a flat surface

• Convex and concave surfaces

Concave (ρ < 0)
surface

Convex
(R > 0)
surface
 Stresses in materials causes sintering
• ∆P: Stress on the atoms/ions under a curved surface
• Differences in curvature causes vacancy(defect)/matter transport

Pressure difference between inside and outside of curved surface:

2γ
Laplace equation (See also section 13.6 Carter & Norton): ∆P = SV
r
γSV: Specific surface energy of a surface with radius of curvature r
 Convex surface (r > 0): tensile stresses : vacancy sink
 Concave surface (r < 0): compressive : vacancy source

Molar volume (Vmi) and Laplace equation;

 Chemical potential (or free energy) difference of
a curved surface compared to a flat surface: ∆µ i = Vmi ∆P
 Sources and sinks for matter transport
• ∆P: Stress on the atoms/ions under a curved surface
• Differences in curvature causes vacancy(defect)/matter transport
Lower vacancy
chem. potential

Vacancy flux

Convex Higher vacancy

Vacancy transport from
surface chem. potential
concave to convex (r > 0) (r < 0)

Concave
surface

Chemical potential difference with flat surface: ∆µ i = Vmi ∆P

 Sources and sinks for matter transport
Matter transport from convex to concave

Lower vacancy
Higher material chem. potential
chemical
potential
Material flow Vacancy flux

Convex Higher vacancy

Convex surface Lower material surface chem. potential
chemical
Atoms/ions under
potential (r > 0) (r < 0)
compressive stress

Concave Concave
surface surface

Chemical potential difference with flat surface: ∆µ i = Vmi ∆P

 Sources and sinks for matter transport
Atoms at convex surfaces have higher energy than at concave
 Material flux Ji to neck (from convex to flat surface):
 Sources and sinks for matter transport

Matter transport from convex to concave

Concave (ρ < 0)
surface

Convex
(R > 0)
surface
 When does a material shrink during sintering?

See also Carter Fig 24.5

Shrinkage:
• Centre approach
• Which type of matter transport is necessary for shrinkage?
From which source to neck?
 Several transport paths are distinguished
When does a material shrink during sintering?

Book: Carter and Norton

 Which transport paths are available?
Two particles (initial sintering stage)
Mechanism Source
1 vapour transport grain surface
2 viscous flow grain boundary
3 surface diffusion grain surface
4 grain boundary diffusion grain boundary
5 volume diffusion grain boundary
6 volume diffusion grain surface
7 bulk diffusion bulk (dislocations)
Table 24.1 (Book: “Ceramic Materials” Carter & Norton)

Which mechanisms lead to densification?

Which mechanisms dominate at lower temperature?
 Which transport paths are available?
Two particles (initial sintering stage)
Mechanism Source
1 vapour transport grain surface
2 viscous flow grain boundary
3 surface diffusion grain surface
4 grain boundary diffusion grain boundary
5 volume diffusion grain boundary
6 volume diffusion grain surface
7 bulk diffusion bulk (dislocations)
Neck formation

Cross-section of 316L stainless-steel support:

Neck formation between two adjacent particles

S. Abedini et al. / Materials Science and Engineering A 533 (2012) 1– 8

 Transport paths – Summary

In practice:
• More grains Densification: removes
• Pores material from grain boundary
region, leading to shrinkage
Lattice diffusion

Surface
Evaporation diffusion
Boundary
diffusion
 Grain growth Coarsening:
 Pore growth microstructural changes
 Effect inhomogeneities without shrinkage

23
 Intermediate and final stage of sintering

 More particle approach

 Pore elimination
 Grian growth
 Pore growth
 Grains – Grain Boundaries – Grain Growth
Grain growth: Important process during final stage sintering
• Takes also place in initial and intermediate stage
• Driving force: ∆G = ∆Gs + ∆Gb = γ s ∆As + γ b ∆Ab
Reduction of the solid-solid interface ∆Ab

movement of grainboundary
material flow
Matterboundary
Grain transport from
convex 
convex to flat
concave
 Grain Growth
Important process during final stage sintering

Ref: Mater. Sci. Eng., 48 (1981) 53

 Control of grain growth in final sintering stage
Pinning grain boundaries by inclusions
 Inclusions increases the energy for grain-
boundary movement
 Inclusion: pore or second-phase particle
 However: inclusions must have certain size
and must be present in sufficient amounts
Grain growth “retarders”
solid
particle
The process is similar, whether the
“obstacle” is a pore or a particle

Fig. 24.14: Carter and Norton

 Pinning grain boundaries
Few % ZrO2 in Al2O3 matrix

15 wt. % ZrO2 in Al2O3 matrix

Grain growth “retarders”?

solid
particle
Chevalier et al. Nano Letters (2005)

 Percentage 2nd phase

 Difference in grain size
Here: pores show no pinning effect

Book: Carter and Norton

 More grain growth

Zriconia ceramics with different grain sizes

Pores within the grains

 Pores enter in the grain bulk by
fast moving grain boundaries

Abnormal or exaggerated grain growth

Grain Growth
Abnormal grain growth

Abnormal grain growth

Easily occurs in
inhomogeneous compacts!
Abnormal or exaggerated grain growth

Book: Carter and Norton

Pores and abnormal grain growth
Pores enter in the grain bulk
by fast moving grain
boundaries
The final sintering stage – A closed pore in more detail

γsv

Pore
γgb
ψ

γsv
 Influence of pore coordination
More particles approach (two-dimensional model)

(A) Pore surrounded by 9 grains

(B) Pore surrounded by 3 grains
 Pore growth or pore shrinkage?

γb θ
Dihedral angle θ: cos =
s, b: resp. solid-vapour
2 2γ s and solid-solid interface

Identical to wetting behaviour

γ sv − γ sl
Contact angle θ : cos θ =
γ lv
 Influence of pore coordination
More particles approach (two dimensional model)

(A) Pore surrounded by 9 grains

(B) Pore surrounded by 3 grains
 Pore growth or pore shrinkage?

θγb
Dihedral angle θ: cos =
s, b: resp. solid-vapour
2 2γ s and solid-solid interface

Dihedral angle is a materials property

Example: material with θ = 60o

 This results in:

 Sintering
Dihedral angle θ
θ = 60o θ < 60o θθ>>60
60oo

Assume: Triangular pore enclosed by 3 cylindrical

grains
 effect of dihedral angle on sintering

γb = γs√3 γb < γs√3 γb > γs√3

Assume: Boundary is convex Boundary is concave
•pore shrinks Pore grows Pore shrinks
•each boundary by 2X
Relation between pore coordination
Gibbs equation: and shrinkage or growth of pores
 Depends on θ
∆G = ∆Gs + ∆Gb = γ s ∆As + γ b ∆Ab
Influence of pore coordination

 Dihedral angle θ:
A materials property
 Is density decrease possible? (“de –sintering”)

Yes
Yes

Schematic final-stage microstructure of a system containing a

mixture of both small and large pores
 Both are thermodynamically unstable, but behave differently
 Do pores with (too) high coordination always grow?

J. Pan, A. Cocks, J. Rödel, R. Huang, H.N. Ch’ng

“Densification of Powder Compact Containing Large and Small Pores”
J. Amer. Ceram. Soc., 92 [7] (2009) 1414-1418

 The classical sintering theory predicts that a pore with high coordination
number grows instead of shrinks during sintering.
 This has led to the proposition that grain growth may be beneficial to
densification
 Pan and colleagues argue against this theory using computer simulations
 Flinn and colleagues have provided direct experimental evidence showing
that very large pores shrink despite their high coordination number

 The current paper brings the analytical and experimental work together
 Practical aspects of sintering
• Homogeneity of the green compact

General Gibbs equation: ∆G = ∆Gs + ∆Gb = γ s ∆As + γ b ∆Ab

Gibbs energy change per unit dG γ s dAs

Initial stage: dG ≅ γ s dAs = σ=
volume (= compressive stress): dV dV

Assume in a certain (local) volume of the compact

•The number of necks (n) is higher than average (nav) higher density
Shrinkage is stronger in this volume
σt
This results in a local (effective) tensile stress: = ( nav − n ) γ s dAs
 Sintering versus homogeneity
Inhomogeneities in green compact have effect on microstructure after sintering

• Starts as area with lower density in green compact

• Difficult to remove during sintering
Practical aspects of sintering

An inhomogeneous green compact

can easily result in an even more
inhomogeneous sintered material
 Practical aspects of sintering - Summary
Inhomogeneities in pore size distribution and/or density:
 More dense regions densify faster
 A local tensile stress can arise during sintering: =
σt ( nav − n ) γ s dA '
 The faster sintering units brake away
 But: This effect can also easily occur in a compact of
monosized particles (close packed but partly simply stacked)
Particle Rearrangement and Pore
Space Coarsening During Solid-
State Sintering
Hans Eckart Exner and Clemens Müller

Copper spheres, 63 μm diameter, sintered

for 1 min (top) and for 12 h (bottom)

Journal of the American Ceramic Society

Vol. 92, 7 Pages: 1384-1390 (2009)
[1384]
 Practical aspects of sintering
Particle size distribution Particle size

Not too narrow Small particles (why?)

(and certainly not too broad) what can happen if too small?

48
 Grain size and relative density as function of temperature

• Figure, submitted to a journal

• Two curves:
- Grain size
- Relative density
• Why is this figure not clear?
• What is remarkable on these
results?

 Density decreases above a certain sintering temperature

- How is that possible?
- What microstructure do you expect for the green compact
and how does it change during sintering?

49
 Sintering – More practical things

 Sintering 5 pieces to one monolithic body

 Study of sintering behaviour by using a dilatometer
 An example of microstructural changes during initial
sintering (fine-grained zirconia ceramics)
 The use of a sintering additives (MgO to YAG)
 Sintering 5 pieces to one monolithic body

 cylinder-built geometries; so-called 5-piece concept

 monolithic body: compressive joining

• extrusion
 Sintering 5 pieces to one monolithic body
1200oC 1200oC

1300oC 1300oC

2000oC
 Study sintering behaviour by using a dilatometer
Dilatometry A definition:
A technique in which a dimension of a substance under a negligible load
is measured as a function of temperature while the substance is
subjected to a controlled temperature program.
(ICTA, ASTM E 473-85)

Dilatometry – Principle (horizontal)

53
Dilatometer

DIL 402 C (1600 °C)

54
Dilatometer results of commercial and nanocrystalline ZrO2
 Nano-structured zirconia
Metal salt solution
(ZrCl4 + YCl3)
Nucleation
burst
Hydrolysis in excess
ammonia (pH > 12)
Amorphous
metal hydroxide
Water/Ethanol washing
(control agglomeration)

Drying

Calcining
Weakly agglomerated powder

“Chloride derived (CP) nanopowder” Differential Thermal Analysis (DTA)

of the dried Zr,Y-hydroxide gel
 3-mol% Y2O3 doped ZrO2 = 3Y-TZP (TZ3Y)
Influence of microstructure on sintering behaviour

Sintering behaviour of two

powders is compared

Dilatometer study

Typical microstructure of zirconia Commercial Tosoh TZ3Y

after precipitation/calcination of • Identical composition

chloride salts ( “Chloride” powder) • Different microstructure

Crystallite size: 8 nm Crystallite size: 34 nm

Sintering
Pure 3Y-TZP
8 nm (CP) versus 35 nm (Tosoh) crystals
• Chloride derived (CP) nanopowder
(crystallite size 8 nm)
 1130°C; 2 hours
 Density: 96%
 Grain size: 150 nm
• Tosoh powder compact
(crystallite size 35nm)
 1350°C; 3 hours
 Density: 98%
 Grain size: 350 nm

Density and linear shrinkage rate of CP and Tosoh

powder compacts as a function of temperature
Heating rate: 2 °C⋅min-1
 Sintering of yttria-doped zironia ceramics

• Look in detail to “chloride” powder!

 At low temperature already a slight
densification (only a few %)
 Microstructural development
during initial sintering stage
 Initial stage during sintering of nanocrystalline zirconia
• From room temperature till 900 oC
• Grains grow (confirmed by TEM and X-Ray diffraction analyis)
• How does the pore size change in
this temperature range?
 Studied by N2 adsorption/desorption

G.S.A.M. Theunissen, PhD Thesis, University of Twente

“Microstructure, fracture toughness and strength of (ultra) fine-
grained tetragonal zirconia ceramics” (1991)
61
 Pore evolution during initial stage of sintering

After isostatic pressing at 400 MPa the

compact has a relative density ~ 45 %
Pore radius ~4
 Initial stage during sintering of nanocrystalline zirconia
• N2 adsorption/desorption studies
• “Hysteresis” loop tells us
something about:
 Pores size
 Pore shape

Green

E
change in pore shape

G.S.A.M. Theunissen, PhD Thesis, University of Twente A

“Microstructure, fracture toughness and strength of (ultra) fine-
grained tetragonal zirconia ceramics” (1991)
63
 Initial stage during sintering of nanocrystalline zirconia
• Slight densification
• Grains as well as pores grow!
• Pore morphology becomes even more homogeneous
• A good start for further densification

Green

G.S.A.M. Theunissen, PhD Thesis, University of Twente

“Microstructure, fracture toughness and strength of (ultra) fine-
grained tetragonal zirconia ceramics” (1991)

64
 Sintering additive
• MgO used as sintering additive
a b c

a) Nd-YAG ceramics, made with (top) 0.01 wt % MgO, and (bottom) no MgO
b) SEM picture of the Nd-YAG ceramic microstructure made with no MgO
c) SEM picture of the Nd-YAG ceramic microstructure made with 0.01 wt % MgO

Liu, W., et al., Synthesis of Nd:YAG powders leading to transparent ceramics: The effect of
MgO dopant. Journal of the European Ceramic Society, 31 [4] (2011): p. 653-657.
[2283]
 Summary - Solid State Sintering
 Driving force: Decrease of external surface
 Transport of material from convex (r > 0) to concave (r < 0)

Requirements:
 Fine-grained powders (however ….!)
 (not too) uniform grains
 Uniform stacking
 Narrow pore size distribution
 High green density
- Often too high density results in worse densification (sinter stress!?)
 ….

66
 The Ceramic Fabrication Process
Powder Preparation

Compaction Processing
Microstructure
Sintering

Post Treatment

Sintering as a part of the ceramic fabrication process

67
 Discussion on a paper on sintering

• Zirconia – Yttria phase diagram

 Metastable tetragonal zirconia (TZP)
• Sintering behaviour of Y-TZP/CuO composites

68
Microstructure evolution during sintering of
CuO-doped 3Y-TZP nano-powder composites
The importance of several interface reactions

Louis Winnubst, Shen Ran, Emiel A. Speets & Dave H.A. Blank

J. Eur. Ceram. Soc., 29 [12] (2009) 2549-2557

“Analysis of reactions during sintering of CuO-doped 3Y-TZP nano-powder composites”

Sintering Behaviour of 0.8mol%-CuO-doped
3Y-TZP ceramics

Starting from commercial coarse-grained powders

Shen Ran, Louis Winnubst, Wika Wiratha, Dave H.A. Blank

J. Am. Ceram. Soc., 89 [1] (2006) 151-155

“Sintering Behavior of 0.8 mol%-CuO-Doped 3Y-TZP Ceramics”

 Pure undoped: not interesting as a ceramic
 Cooling after sintering: T → M;
Volume expansion Fracture

 Stabilize high temperature C-phase to RT:

 17 mol% YO1.5: stabilized ZrO2; Ionic conductor

 Stabilizing T-phase to RT is interesting (TZP)

 Metastable T-phase: high strength/toughness

Monoclinic Tetragonal Cubic

Undoped ZrO2:

RT 1170ºC 2370ºC
 Phase stability in Zirconia: Masahiro Yoshimura et al.
Thermodynamics and kinetics of phase transformation and phase stability:
• 3 mol% Y2O3: < 1200 ºC monoclinic thermodynamically stable (ΔG)
• < 1200 ºC: Phase transformation is mainly kinetically determined (ΔG*)
• Kinetic energy barrier, ΔG*, increases with:
 decreasing temperature
 decreasing grain size

 For thermodynamic equilibrium: ΔG* =0:

 How to reduce ΔG* ?
 An extra driving force can e.g. be introduced by aZirconia-yttria
chemical reaction
phase diagram
Solid State Ionics 86-88 (1996) 1131-1149
 Phase stability in Zirconia: Masahiro Yoshimura et al.
Thermodynamics and kinetics of phase transformation and phase stability:
• 3 mol% Y2O3: < 1200 ºC monoclinic thermodynamically stable (ΔG)
• < 1200 ºC: Phase transformation is mainly kinetically determined (ΔG*)
• Kinetic energy barrier, ΔG*, increases with:
 decreasing temperature
 decreasing grain size

 For thermodynamic equilibrium: ΔG* =0:

 How to reduce ΔG* ? Zirconia-yttria phase diagram
 An extra driving force can e.g. be introduced by a chemical reaction

Solid State Ionics 86-88 (1996) 1131-1149

 Toughening mechanism in tetragonal zirconia (TZP)

 Stress-induced phase transformation toughening

 A Model !!
Is this the only toughening mechanism??

Load/stress on ceramic:

 Crack propagates

 tensile stress at tip leads to:

 tetragonal → monoclinic

 transformed zone is larger than the constrained t-phase

 Volume expansion gives compression stresses at crack

 Compressive strain screens further crack opening

Y-TZP: Yttria-doped Tetragonal Zirconia Polycrystals

A ceramic with excellent mechanical properties if:

 Grain size below a critical value (600 nm)
 Correct yttria content (5-6 at% Y)
 Dense, uniform, defect-poor microstructure

Sintering of 3Y-TZP / CuO

composites

3Y-TZP
Ceramic roof tile with integrated solar cells