Chapter 8
What is sintering?
Sintering is the process by which a compacted powder is converted into a solid body
by heating below the melting point of the main constituents, so that the object essentially
remains a solid throughout the process. There is (in principle) no composition change
involved – all that happens is that the degree of agglomeration of the material alters. The
first stage that occurs during sintering is an initial reduction in the surface roughness of the
individual particles, followed by a period in which the particles start to join together.
Usually this involves shrinkage, change of shape and the formation of pores. Finally, the
solid becomes denser by the elimination of internal pores and voids, resulting in a compact
grain structure. Sintering is widely used in the ceramics and powder metallurgy industries
to form small components such as valves and in the polymer industry to fabricate porous
polymer blocks used in filtration.
As sintering does not usually involve chemical reactions the driving force is a reduction in
the total surface area and the associated reduction in surface energy.
What is a first-order phase transition?
Phase transitions are changes that occur without chemical reaction, generally at a
definite temperature, the transition temperature and a definite pressure, the transition
pressure. Melting, as when ice turns to water, is typical. From a thermodynamic
perspective, a phase transition will occur if the change results in a lowering of the Gibbs
energy of the system. This change can manifest itself as one of two general categories,
first-order or second-order transitions. First-order phase transitions are those in which the
molar Gibbs energy of the system changes slope at the transition temperature. They are
characterised by an enthalpy change called the latent heat of transition.
What is a martensitic transformation?
Martensitic transformations are essentially solid-state transitions that take place at a
constant temperature and composition due to synchronized atomic displacements over
distances smaller than the normal interatomic distances in the parent phase. The name is
derived from a characteristic microstructure of hard steels called martensite. Martensitic
transitions take place very rapidly, because significant atomic diffusion does not occur and
for this reason are frequently called diffusionless transitions. Martensitic transitions are
important in steel making, when rapid cooling the homogeneous alloy austenite to room
temperature leads to the formation of metastable martensite. Martensitic transformations
take place in many materials other than steel, including zirconia, ZrO2 and isostructural
HfO2, and the titanium – nickel alloy TiNi, called Nitinol, used in shape-memory alloys.
Martensitic transitions are extremely rapid because only a slight displacement of atoms
takes place and progress through the reacting solid at speeds approaching the speed of
sound in the material.
Quick quiz
1. c; 2. c; 3. a; 4. b; 5. b; 6. a; 7. c; 8. c; 9. b; 10. a; 11. c; 12. a; 13. b; 14. a;
15. b; 16. b; 17. b; 18. c; 19. a; 20. c; 21. b.
Calculations and questions
1. 1.69.
2. 0.930.
4. (a) 6.79 x 10-8 m diameter; (b) 2.15 x 10-7 m diameter; (c) 6.80 x 10-5 m diameter.
5. 1.386 x 10-5 m diameter.
6. 5.95 x 10-5 m diameter.
7. (a) gold.
8. (a) rhenium.
9. (a). Cr2O3.
13. Cu, protective; Fe, protective; K, non-protective; Ti, protective; Al, protective; Na, non-
protective.
14. ~6.22 x 10-17 m2 s-1.
15. (a) 0.0044 m; (b) 26.41 kg; (c) 2.95 kg m-2.
16. 7.15 x 10-18 m2 s-1.
Solutions
1 The surface energy of solid Al2O3, at 1850C is 0.905 J m-2 and the density is 3970 kg
mol-1. Calculate the relative pressure, p/p0 over a hemispherical “hill”, of diameter 1 x 10-7
m, on the surface of a flat plate of corundum, at this temperature.
The vapour pressure of a curved surface is given by:
ln (p / p0) = (2V) / (r RT)
The molar volume, V = molar mass / density
For Al2O3, V = 0.10961 / 3970 m3 mol-1 = 2.57 x 10-5 m3 mol-1
ln (p / p0) = (2 x 2.57 x 10-5 x 0.905) / (0.5 x 10-8 x 8.31451 x 2123)
= 0.527
(p / p0) = 1.694
2 The surface energy of solid MgO, at 1500C is 1.2 J m-2 and the density is 3580 kg mol-
1
. Estimate the pressure differential between a spherical pit of diameter of 5 x 10 -7 m and
that over the surface of a flat plate of magnesia, at this temperature.
The vapour pressure of a curved surface is given by:
ln (p / p0) = (2V) / (r RT)
The molar volume, V = molar mass / density
For MgO, V = 0.04304 / 3580 m3 mol-1 = 1.120 x 10-5 m3 mol-1
ln (p / p0) = (2 x 1.12 x 10-5 x 1.2) / (-2.5 x 10-8 x 8.31451 x 1773)
= -0.0729
(p / p0) = 0.930
3 Assuming that the ideal gas law (pV = nRT) holds, deduce the formula
ri 2
rf pi ri
where ri and rf are the initial and final radii of the pore in the material, p i is the initial gas
pressure in the pore, and is the surface energy of the material.
The gas law gives: (piVi = RTi) = (pfVf = RTf)
where i and f stand for the initial and final states of the system.
Substituting for the volumes and pressures:
(Vi = 4 ri3); (Vf = 4 rf3)
pi = 2 / ri ; pf = 2 / rf
Noting that Ti = Tf = T
pi x 4 ri3 / T = (2 / rf) x (4 rf3 / T)
pi x ri3 = (2 rf2)
ri3 / rf2 = (2 / pi)
ri2 / rf2 = (2 / pi ri)
ri / rf = (2 / pi ri)
4 A spherical pore in a soda-lime glass at 1200 K contains trapped gas. Assuming the
gas trapped in the pore is at atmospheric pressure, and the surface energy of the glass at
1200 K is 0.350 J m-2 calculate the equilibrium pore size when the initial pore is of diameter
(a) 0.4 m; (b) 4 m; (c) 40 m.
ri / rf = (2 / pi ri)
Inserting values:
(a) ri / rf = (2 x 0.350) / (1.01 x 105 x 0.2 x 10-6)½ = 5.89
rf = ri / 5.89 = 3.40 x 10-8 m
pore diameter = 6.79 x 10-8 m, pore shrinks
(b) ri / rf = (2 x 0.350) / (1.01 x 105 x 2 x 10-6)½ = 1.862
rf = ri / 1.862 = 1.07 x 10-7 m
pore diameter = 2.15 x 10-7 m, pore shrinks slightly
(c) ri / rf = (2 x 0.350) / (1.01 x 105 x 20 x 10-6)½ = 0.589
rf = ri / 0.589 = 3.40 x 10-5 m
pore diameter = 6.80 x 10-5 m, pore grows
5 Using the data in question 4, what is the equilibrium size of a pore that will neither shrink
nor expand?
At equilibrium ri / rf = 1; (ri = rf)
i.e 1 = (2 x 0.350) / (1.01 x 105 x ri)½
1.01 x 105 x ri = 2 x 0.350
ri = 6.93 x 10-6 m
pore diameter = 13.86 x 10-6 m = 13.86 m
6 Solid titanium nitride, TiN, has a surface energy of 1.19 J m -2 at 1200C. If the voids in
this material contain nitrogen gas at the same partial pressure as found in the atmosphere,
approximately 8 x 104 Pa, estimate the maximum void size to ensure that voids in a
sintered ceramic will shrink.
At equilibrium ri / rf = 1; (ri = rf)
i.e 1 = (2 x 1.19) / (8 x 104 x ri)½
For no shrinkage or growth: ri = (2 x 1.19) / (8 x 104)
= 2.975 x 10-5 m
Maximum void size = 2 ri = 5.95 x 10-5 m
7 The gold – silver system forms a complete solid solution. The melting point of gold is
1064.43C and that of silver is 961.93C. (a) During rapid cooling of a 50 at % gold: 50 at %
silver mixture, which phase will be richest in the core of a grain? (b) Sketch the
microstructure of the solid at a temperature of 800C if the melt is cooled very slowly. (c)
Sketch the microstructure of the solid at a temperature of 800C if the melt is cooled
quickly.
(a) The highest melting point phase will solidify first and so be richest at the grain centre.
In this case it is Au.
(b) homogeneous AuAg
(c) inhomogeneous, core richer in Au.
8 The ruthenium – rhenium system forms a complete solid solution (see Chapter 4 Q 13
for the phase diagram). The melting point of ruthenium is 2334C and that of rhenium is
3186C. (a) During rapid cooling of a 45 at % Re: 55 at % Ru mixture, which phase will be
richest in the core of a grain? (b) Sketch the microstructure of the solid at a temperature of
2000C if the melt is cooled very slowly. (c) Sketch the microstructure of the solid at a
temperature of 2000C if the melt is cooled quickly.
(a) The highest melting point phase will solidify first and so be richest at the grain centre.
In this case it is Re.
(b) homogeneous RuRe
(c) inhomogeneous, core richer in Re
9 The Al2O3 – Cr2O3 system forms a complete solid solution (see Chapter 4 Q 9 for the
phase diagram). The melting point of Al2O3 is 2035C and that of Cr2O3 is 2330C. (a)
During rapid cooling of a 33 mol % Al2O3: 67 mol % Cr2O3 mixture, which phase will be
richest in the core of a grain? (b) Sketch the microstructure of the solid at a temperature of
2000C if the melt is cooled very slowly. (c) Sketch the microstructure of the solid at a
temperature of 2000C if the melt is cooled quickly.
(a) The highest melting point phase will solidify first and so be richest at the grain centre.
In this case it is Cr2O3.
(b) homogeneous (Al, Cr)2O3
(c) inhomogeneous, (Al, Cr)2O3, core richer in Cr
10 With respect to the phase diagram of the copper – silver system, Figure 8.37: (a)
Sketch and label the microstructure of a solid containing 2.5 at % Ag (about half way across
the phase field at 800C), when the melt is slowly cooled to 800C, and then to 400C. (b)
Sketch and label the microstructure of a solid containing 30 at % Ag when the melt is slowly
cooled to 700C. (c) Sketch and label the microstructure of a solid formed by slowly cooling
the eutectic composition containing 60 at % Ag to 700C.
(a) homogeneous CuAg
(b) primary , eutectic , eutectic
(c) eutectic , eutectic
11 With respect to the iron - carbon phase diagram, Figure 4.15: (a) Sketch and label the
microstructure of a solid containing 3.44 wt % C when the melt is slowly cooled to 800C.
What phases will occur if the sample is further cooled to 600C? (b) Sketch and label the
microstructure of a solid containing 6.2 wt % C when the melt is slowly cooled to 800C. (c)
The alloy in (b) is further cooled to 600C. What phases will now be present? How much
of each is present?
(a) At 800C a composition of 3.44 wt % C would contain (austenite) plus Fe3C,
(cementite). On cooling crystals of primary would form first in the liquid. As cooling
continues the amount of would increase and the composition of the liquid move towards
4.3 wt % C. At this point it will solidify to eutectic plus Fe3C.
On further cooling to 600C the phase would transform to pearlite ( + Fe3C). Rapid
cooling would produce martensite.
(b) The composition 6.2 wt % C is very close to that of Fe 3C (6.7 wt % C) so that the
microstructure would consist of polycrystalline Fe 3C with small amounts of between the
grains.
(c) The small grains of will transform to pearlite. The composition of the Fe 3C is fixed.
To find the quantities draw tie-lines.
Amount of Fe3C (6.2 – 1.0) / (6.69 – 1.0) = 81.4 %
Amount of = amount of pearlite (6.69 - 6.2) / (6.69 – 1.0) = 8.6 %
The composition of the -phase will be 0.1 wt % C.
12 With respect to the iron - carbon phase diagram, Figure 4.15: (a) Sketch and label the
microstructure of a solid containing 1.0 wt % C when the melt is slowly cooled to 1000C,
and then to 750C and finally to 700C. (b) Sketch and label the microstructure of a solid
containing 0.3 wt % C when the melt is slowly cooled to 1000C, and then to 750C and
finally to 700C. (c) Sketch and label the microstructure of a solid containing 0.76 wt % C
when the melt is slowly cooled to 1000C, and then to 750C and finally to 700C.
(a) At 1000C the solid will consist of homogeneous grains of (austenite). Cooling to
750C will cause some to transform to Fe3C. Cooling to 700C will cause the remaining
to transform to pearlite.
(a) At 1000C the solid will consist of homogeneous grains of (austenite). Cooling to
750C will cause some to transform to (ferrite). Cooling to 700C will cause the
remaining to transform to pearlite.
(c) At 1000C the solid will consist of homogeneous grains of (austenite). This will be
unchanged at 750C. Cooling to 700C will cause the to transform to pearlite.
13 Which of the metals in the table are likely to be protected from corrosion by the
formation of a protective oxide film.
Metal / metal Density of Density of
oxide metal / kg m-3 oxide / kg m-3
Cu / CuO 8933 6315
Fe / Fe2O3 7873 5240
K / K2O 862 2320
Ti / TiO2 4508 4260
Al / Al2O3 2698 3970
Na / Na2O 966 2270
For protection the Pilling-Bedworth ratio must be greater than 1.
XPB = (molar mass of oxide x density of metal) / (m x molar mass of metal x
density of oxide)
Note: this method is semiquantitative and great accuracy in the numerical values used is
not really necessary.
Cu/CuO: XPB = (79.55 x 8933) / (63.55 x 6315) = 1.77; protective
Fe/Fe2O3: XPB = (159.7 x 7873) / (2 x 55.85 x 5240) = 2.15; protective
K/K2O: XPB = (79.1 x 862) / (2 x 39.1 x 2320) = 0.45; non-protective
Ti/TiO2: XPB = (79.88 x 4508) / (47.88 x 4260) = 1.77; protective
Al/Al2O3: XPB = (101.96 x 2698) / (2 x 26.98 x 3970) = 1.28; protective
Na/Na2O: XPB = (61.98 x 966) / (2 x 22.99 x 2270) = 0.57; non-protective
Note that all of the alkali metals have non-protective coatings:
Li/Li2O: XPB = (29.88 x 533) / (2 x 6.94 x 2013) = 0.57; non-protective
Rbi/Rbi2O: XPB = (186.94 x 1533) / (2 x 85.47 x 3720) = 0.45; non-protective
14 A nickel foil is oxidised at 1000C. The film thickness as a function of time is given in
the table. Confirm that the rate is parabolic and calculate the parabolic rate constant.
Heating time / hr Film thickness / nm
1 4.74 x 10-7
2 0.67 x 10-6
3 0.82 x 10-6
4 0.95 x 10-6
5 1.06 x 10-6
6 1.16 x 10-6
The parabolic rate law is: x2 = kt
Hence, for agreement with the parabolic rate law the plot of x2 versus kt will be linear with a
slope of k.
x/m x2 / m2 t / hr t / s t / min
4.74 x 10-7 2.25 x 10-13 1 3.6 x 104 60
0.67 x 10-6 4.49 x 10-13 2 7.2 x 104 120
0.82 x 10-6 6.72 x 10-13 3 10.8 x 104 180
0.95 x 10-6 9.02 x 10-13 4 14.4 x 104 240
1.06 x 10-6 11.24 x 10-13 5 18.0 x 104 300
1.16 x 10-6 13.46 x 10-13 6 21.6 x 104 360
The slope 11.2 x 10-13 / 300 = 3.733 x 10-15 m2 min-1 = 6.22 x 10-17 m2 s-1
15 When copper is oxidised under low partial pressures of oxygen gas, it forms Cu 2O via a
parabolic rate law. The rate constant is 5.38 x 10 -10 m2 s-1 at 0.05 atm pressure, and
900C. (a) What will the film thickness be after oxidation of copper foil for 10 hours at this
temperature? (b) What will the weight of the copper oxide film be? (c) Experimentally, it is
easier to measure the weight gain as a function of time rather than film thickness. What
will the weight gain of the film be?
The parabolic rate law is: x2 = kt
(a) x = (kt) = (5.38 x 10-10 x 10 x 60 x 60) = 0.0044 m
(b) The density of Cu2O is 6000 kg m-3.
The film weight is: 0.0044 x 1 x 1 x 6000 = 26.41 kg m -2
(c) The weight of Cu is unchanged, and the weight gain is due solely to the added oxygen.
The fraction of the film weight due to oxygen is:
molar mass of oxygen added / molar mass of oxide
= 16 / 142.7 = 0.1118
Hence the weight gain is 26.41 x 0.1118 = 2.95 kg m-2
16 The thickness of the layer of the spinel NiAl2O4 formed between NiO and Al2O3 when
reacted at 1350C is given in the table. Check that the reaction is diffusion controlled and
calculate the rate constant. The reaction equation is:
NiO + Al2O3 NiAl2O4
Layer thickness / m Time / h
1.0 20
1.4 40
1.8 60
2.0 80
2.3 100
If the reaction is diffusion controlled the parabolic rate law will be followed. The parabolic
rate law is: x2 = kt. Hence, for agreement with the parabolic rate law the plot of x2
versus kt will be linear with a slope of k.
x/m x2 / m2 t / hr t / s
1.0 x 10-6 1.0 x 10-12 20 7.2 x 104
1.4 x 10-6 1.96 x 10-12 40 14.4 x 104
1.8 x 10-6 3.24 x 10-12 60 21.6 x 104
2.0 x 10-6 4.0 x 10-12 80 28.8 x 104
2.7 x 10-6 5.29 x 10-12 100 36.0 x 104
The plot is linear so the parabolic rate law is followed.
From th graph k 1.5 x 10-17 m2 s-1
