Reduction of Lead-Rich Slags with Coke in the Lead Blast Furnace
181
surface of the coke, as shown in Fig. 3. These cover a large section of the coke surface,
but some regions remain uncovered. The uncovered zones did most likely not come
in contact with any dripping phases throughout the experiment. The particles appear
lighter in colour than the coke substrate and range in size from 5 to 200 µm. EDX
analysis was performed to determine the composition and it was found to be a leadrich phase, presumably the reduced metallic lead. This indicates that initially, the
direct reduction reaction (Eq. 2) has taken place, since no air was present throughout
the experiment, and hence no CO could have been generated by incomplete coke
combustion. The generated CO gas from this direct reduction reaction, however,
could in its turn cause further reduction of the slag, by means of the indirect reduction
reaction (Eq. 1).
Wetting between the particles and coke is extremely poor, as observed by the
spherical, droplet-like shape. This can also be seen in Fig. 4, where the contact angle
can be observed to be approximately 180°. When looking more closely at the surface
of these spheres, one can observe a ripple-like structure of ledges, as demonstrated
in Fig. 5. This structure is likely caused by crystallisation of the lead droplets during
cooling, developing facets, as is observed in other studies [15]. Besides the spherical
droplets, some other phenomena can be observed. Another type of spherical particle
is found, for example, as shown in Fig. 6 on the upper right corner, slightly darker
in colour, and smaller in size. This phase proved prone to slight electron charging,
despite the gold nano-coating. Its surface is completely smooth and no evidence of
crystallisation is found. All of these factors point to a glassy slag phase, which is
also confirmed by EDX. EDX also showed that the lead content in this phase was
very low, indicating that the reduction of the slag has taken place to a large extent.
In Fig. 7, we see one such particle surrounded by two of the lead particles discussed
previously. The element mapping supports the hypothesis that the darker particles
are lead-depleted slag and the lighter ones are metallic lead.
Additionally, in Fig. 6, one can also see a complex aggregate of different phases.
Upon closer inspection, we can see that it is one of the slag particles surrounded by
four lead droplets. The lead droplets also have another, polygonally shaped phase dispersed over its surface. These particles are positioned on the top of the lead droplets,
indicating they are lighter and hence float. EDX analysis confirmed these to be an
iron-rich phase with oxygen. However, EDX analysis cannot reliably quantify oxygen content, so to determine the phases that potentially formed; a thermodynamic
calculation was performed using FactSage. This was done by using the actual slag
composition, as listed in Table 3, and subjecting it to decreasing oxygen partial pressures, to simulate the reducing environment produced by the coke. This is illustrated
in Fig. 8. It shows the production of a spinel phase (almost purely Fe 3 O 4 , i.e. magnetite), but with decreasing oxygen partial pressure this is reduced to Fe 2 O 3 , which
redissolves into the slag. As the oxygen partial pressure is lowered, the Fe 2 O 3 in
the slag reduced to FeO before the reduction of liquid lead occurs. Finally, after all
the lead is reduced, the FeO starts to reduce to form solid iron. This implies that the
phase that was detected is either a spinel phase or solid iron crystals. Since EDX
analysis confirms that the slag was fully depleted in lead and its iron content was
181
surface of the coke, as shown in Fig. 3. These cover a large section of the coke surface,
but some regions remain uncovered. The uncovered zones did most likely not come
in contact with any dripping phases throughout the experiment. The particles appear
lighter in colour than the coke substrate and range in size from 5 to 200 µm. EDX
analysis was performed to determine the composition and it was found to be a leadrich phase, presumably the reduced metallic lead. This indicates that initially, the
direct reduction reaction (Eq. 2) has taken place, since no air was present throughout
the experiment, and hence no CO could have been generated by incomplete coke
combustion. The generated CO gas from this direct reduction reaction, however,
could in its turn cause further reduction of the slag, by means of the indirect reduction
reaction (Eq. 1).
Wetting between the particles and coke is extremely poor, as observed by the
spherical, droplet-like shape. This can also be seen in Fig. 4, where the contact angle
can be observed to be approximately 180°. When looking more closely at the surface
of these spheres, one can observe a ripple-like structure of ledges, as demonstrated
in Fig. 5. This structure is likely caused by crystallisation of the lead droplets during
cooling, developing facets, as is observed in other studies [15]. Besides the spherical
droplets, some other phenomena can be observed. Another type of spherical particle
is found, for example, as shown in Fig. 6 on the upper right corner, slightly darker
in colour, and smaller in size. This phase proved prone to slight electron charging,
despite the gold nano-coating. Its surface is completely smooth and no evidence of
crystallisation is found. All of these factors point to a glassy slag phase, which is
also confirmed by EDX. EDX also showed that the lead content in this phase was
very low, indicating that the reduction of the slag has taken place to a large extent.
In Fig. 7, we see one such particle surrounded by two of the lead particles discussed
previously. The element mapping supports the hypothesis that the darker particles
are lead-depleted slag and the lighter ones are metallic lead.
Additionally, in Fig. 6, one can also see a complex aggregate of different phases.
Upon closer inspection, we can see that it is one of the slag particles surrounded by
four lead droplets. The lead droplets also have another, polygonally shaped phase dispersed over its surface. These particles are positioned on the top of the lead droplets,
indicating they are lighter and hence float. EDX analysis confirmed these to be an
iron-rich phase with oxygen. However, EDX analysis cannot reliably quantify oxygen content, so to determine the phases that potentially formed; a thermodynamic
calculation was performed using FactSage. This was done by using the actual slag
composition, as listed in Table 3, and subjecting it to decreasing oxygen partial pressures, to simulate the reducing environment produced by the coke. This is illustrated
in Fig. 8. It shows the production of a spinel phase (almost purely Fe 3 O 4 , i.e. magnetite), but with decreasing oxygen partial pressure this is reduced to Fe 2 O 3 , which
redissolves into the slag. As the oxygen partial pressure is lowered, the Fe 2 O 3 in
the slag reduced to FeO before the reduction of liquid lead occurs. Finally, after all
the lead is reduced, the FeO starts to reduce to form solid iron. This implies that the
phase that was detected is either a spinel phase or solid iron crystals. Since EDX
analysis confirms that the slag was fully depleted in lead and its iron content was