3
• Dispersants which preferentially attach to the silica
and then in turn promote the adsorption of the col-
lector to the silica
• Dispersants which preferentially attach to the hema-
tite/magnetite or other iron-bearing mineral, and
compete with the collectors for these surface sites –
depressing the valuable minerals
From the viewpoint of flotation reagents, caustic soda
is an example in the first category, calcium and magne-
sium cations could be viewed as a limited implementation
of the second (Parra-Álvarez et al., 2023), and causticized
modified starches could be seen as an implementation of
the third option (Zhang et al., 2021). However, while the
latter two reagents are proven in effect in flotation proper,
their impact in the grind is likely to be minimally effec-
tive for separate reasons: one, Ca and Mg are already
unavoidable in almost all process waters (e.g., as reported in
Haselhuhn et al. 2012 Da Cruz et al., 2023 Parra-Álvarez
et al., 2023) and two, modified starches tend to be more
potent of flocculants than dispersants as they are typically
designed. However, some preliminary results using polysac-
charide-based reagents as grinding aids have been reported
(Chipakwe et al., 2022).
The other major promising development is the utiliza-
tion of improved grinding technology. In general, the use of
stirred media mills or high pressure grinding rolls (HPGR)
have received considerable interest in other mineral cat-
egories, with energy improvements generally reported in
hard rock applications, including a Chilean iron ore mine
(von Michaelis, 2009). The main advantage of an HPGR
is increased energy efficiency over earlier technologies, but
they also tend to have reduced plant footprints and particle
breakage patterns that are more convenient for subsequent
separations.
For processing iron ore feeds, however, the implemen-
tation of HPGR centered grinding circuits remains a rela-
tively long-term process. Interestingly, however, HPGRs
are also finding application in preparation of iron ore con-
centrate for agglomeration, being used to increase the sur-
face area of the material to be agglomerated at relatively low
energy cost. This will be discussed in a later section focusing
on agglomeration.
SEPARATION
Once the iron-bearing minerals, whether hematite, magne-
tite, goethite, siderite, or other iron-bearing materials have
been liberated from their gangue materials, the next step is
separating and concentrating the valuable materials.
There are several technically and economically viable
options for this process. Magnetite is particularly simple
to process via magnetic separation, though the other iron-
bearing minerals can often be separated with higher-inten-
sity magnetic separation as well. All these materials can also
be processed via flotation, though goethite and siderite are
somewhat more involved and newer developments tech-
nologically. For larger liberation sizes it is also potentially
feasible to use gravity separation techniques or visual sepa-
ration of the materials.
In practice, however, most iron ore concentrators end
up utilizing flotation (Zhang et al., 2021), potentially with
magnetic separation steps in-between. There are several
advantages here, but primarily that it can be tuned to work
very well for mixed hematite/magnetite ores and is low cost
in terms of reagent and power consumption for the sepa-
ration capacity achieved. Magnetic separation is even sim-
pler to operate for magnetite ores, but it often rejects the
other iron-bearing minerals in exchange. As the goal is usu-
ally the iron and not the magnetite specifically, and since
most deposits are not purely magnetite, a flotation step is
often indicated regardless. And since in turn, flotation is
usually capable of effectively collecting magnetite anyways,
the presence of a separate magnetic separation step is not
always economically justified.
Since most iron ores require fine grinding to achieve
good liberation, it is often the case that significant quan-
tities of silica slimes are present after the comminution
step. The addition of grinding aids can help mitigate their
impact in the grinding and help prevent excessive forma-
tion of the slimes but are generally not sufficient to entirely
prevent their formation. An excess of these gangue slimes is
exceedingly problematic in flotation:
• In neutral pH values between the iron minerals’
isoelectric points and the isoelectric points of the
gangues, heteroflocculation can occur, and thus the
slimes coat the surfaces of the valuable particles.
Selective separation becomes impossible.
• In alkaline pH values, the available surface area of
the dispersed slimes increases flotation reagent con-
sumption by one or more orders of magnitude, mak-
ing flotation either ineffective economically or just
entirely ineffective in general.
Operation in acidic pH values, performing direct flo-
tation of iron minerals (usually siderite) is possible but
presents other significant challenges and utilizes entirely
different reagent and processing schemes. These are well
summarized in the review by Zhang et al. (2021).
The typical solution to the presence of silica slimes is
desliming. Desliming is accomplished either by a size classi-
fication (such as via hydrocycloning) with an extremely fine
• Dispersants which preferentially attach to the silica
and then in turn promote the adsorption of the col-
lector to the silica
• Dispersants which preferentially attach to the hema-
tite/magnetite or other iron-bearing mineral, and
compete with the collectors for these surface sites –
depressing the valuable minerals
From the viewpoint of flotation reagents, caustic soda
is an example in the first category, calcium and magne-
sium cations could be viewed as a limited implementation
of the second (Parra-Álvarez et al., 2023), and causticized
modified starches could be seen as an implementation of
the third option (Zhang et al., 2021). However, while the
latter two reagents are proven in effect in flotation proper,
their impact in the grind is likely to be minimally effec-
tive for separate reasons: one, Ca and Mg are already
unavoidable in almost all process waters (e.g., as reported in
Haselhuhn et al. 2012 Da Cruz et al., 2023 Parra-Álvarez
et al., 2023) and two, modified starches tend to be more
potent of flocculants than dispersants as they are typically
designed. However, some preliminary results using polysac-
charide-based reagents as grinding aids have been reported
(Chipakwe et al., 2022).
The other major promising development is the utiliza-
tion of improved grinding technology. In general, the use of
stirred media mills or high pressure grinding rolls (HPGR)
have received considerable interest in other mineral cat-
egories, with energy improvements generally reported in
hard rock applications, including a Chilean iron ore mine
(von Michaelis, 2009). The main advantage of an HPGR
is increased energy efficiency over earlier technologies, but
they also tend to have reduced plant footprints and particle
breakage patterns that are more convenient for subsequent
separations.
For processing iron ore feeds, however, the implemen-
tation of HPGR centered grinding circuits remains a rela-
tively long-term process. Interestingly, however, HPGRs
are also finding application in preparation of iron ore con-
centrate for agglomeration, being used to increase the sur-
face area of the material to be agglomerated at relatively low
energy cost. This will be discussed in a later section focusing
on agglomeration.
SEPARATION
Once the iron-bearing minerals, whether hematite, magne-
tite, goethite, siderite, or other iron-bearing materials have
been liberated from their gangue materials, the next step is
separating and concentrating the valuable materials.
There are several technically and economically viable
options for this process. Magnetite is particularly simple
to process via magnetic separation, though the other iron-
bearing minerals can often be separated with higher-inten-
sity magnetic separation as well. All these materials can also
be processed via flotation, though goethite and siderite are
somewhat more involved and newer developments tech-
nologically. For larger liberation sizes it is also potentially
feasible to use gravity separation techniques or visual sepa-
ration of the materials.
In practice, however, most iron ore concentrators end
up utilizing flotation (Zhang et al., 2021), potentially with
magnetic separation steps in-between. There are several
advantages here, but primarily that it can be tuned to work
very well for mixed hematite/magnetite ores and is low cost
in terms of reagent and power consumption for the sepa-
ration capacity achieved. Magnetic separation is even sim-
pler to operate for magnetite ores, but it often rejects the
other iron-bearing minerals in exchange. As the goal is usu-
ally the iron and not the magnetite specifically, and since
most deposits are not purely magnetite, a flotation step is
often indicated regardless. And since in turn, flotation is
usually capable of effectively collecting magnetite anyways,
the presence of a separate magnetic separation step is not
always economically justified.
Since most iron ores require fine grinding to achieve
good liberation, it is often the case that significant quan-
tities of silica slimes are present after the comminution
step. The addition of grinding aids can help mitigate their
impact in the grinding and help prevent excessive forma-
tion of the slimes but are generally not sufficient to entirely
prevent their formation. An excess of these gangue slimes is
exceedingly problematic in flotation:
• In neutral pH values between the iron minerals’
isoelectric points and the isoelectric points of the
gangues, heteroflocculation can occur, and thus the
slimes coat the surfaces of the valuable particles.
Selective separation becomes impossible.
• In alkaline pH values, the available surface area of
the dispersed slimes increases flotation reagent con-
sumption by one or more orders of magnitude, mak-
ing flotation either ineffective economically or just
entirely ineffective in general.
Operation in acidic pH values, performing direct flo-
tation of iron minerals (usually siderite) is possible but
presents other significant challenges and utilizes entirely
different reagent and processing schemes. These are well
summarized in the review by Zhang et al. (2021).
The typical solution to the presence of silica slimes is
desliming. Desliming is accomplished either by a size classi-
fication (such as via hydrocycloning) with an extremely fine