Chemical Engineering Science 144 (2016) 260–266

Contents lists available at ScienceDirect

Chemical Engineering Science

journal homepage: www.elsevier.com/locate/ces

Review

Alkyl aminated nanocelluloses in selective flotation of aluminium

oxide and quartz
Ossi Laitinen a,n, Robert Hartmann a, Juho A. Sirviö a, Henrikki Liimatainen a,
Martin Rudolph b, Ari Ämmälä a, Mirja Illikainen a
a
University of Oulu, Fibre and Particle Engineering, P.O. Box 4300, FI-90014 Oulu, Finland
b
Helmholtz Institute Freiberg for Resource Technology, Division of Processing, Halsbrücker Str. 34, 09599 Freiberg, Germany

art ic l e i nf o a b s t r a c t

Article history: There are economic and ecological incentives for developing novel green chemicals from renewable
Received 23 November 2015 resources in order to reduce the environmental impact of mineral processing. Cellulose, the most
Received in revised form abundant natural polymeric source, is a promising green alternative that could replace the synthetic
7 January 2016
chemicals currently used. In this study, linear alkyl aminated nanocelluloses with increasing chain
Accepted 25 January 2016
Available online 5 February 2016
lengths were used for the selective flotation of aluminium oxide and quartz. Methylamine, ethylamine,
n-propylamine, n-butylamine, n-pentylamine and n-hexylamine were introduced into a cellulose back-
Keywords: bone using combined periodate oxidation and reductive amination in an aqueous environment. The
Alumina hydrophobicity of the nanocelluloses was found to be increased by extending the alkyl chain length of
Collector
the amino groups. Flotation experiments proved that alkyl aminated nanocelluloses can be both effective
Flotation
and selective collectors for quartz in a flotation system if they are sufficiently hydrophobic to allow the
Nanocellulose
Quartz particles to effectively attach to the air bubbles. In the case of flotation with a known quartz and alumina
Selectivity mixture, the successful separation solution used aminated nanocelluloses at a pH of around 7.5.
& 2016 Elsevier Ltd. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260
2. Materials and methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
2.1. Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
2.2. Synthesis of alkyl aminated nanocelluloses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262
2.3. Transmission Electron Microscope (TEM). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262
2.4. Microflotation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262
2.5. Hydrophobicity of alkyl aminated nanocelluloses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262
2.6. Zeta potential measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262
2.7. Chemical analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
3. Results and discussion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
3.1. Morphology and hydrophobicity of alkyl aminated nanocelluloses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
3.2. Quartz flotation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
3.3. Aluminium oxide flotation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264
3.4. Zeta potential measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264
3.5. Selectivity of nanocellulose collectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
4. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266

n
Corresponding author. Tel.: þ 358 503504960.
E-mail address: ossi.laitinen@oulu.fi (O. Laitinen).

http://dx.doi.org/10.1016/j.ces.2016.01.052
0009-2509/& 2016 Elsevier Ltd. All rights reserved.
 O. Laitinen et al. / Chemical Engineering Science 144 (2016) 260–266 261

1. Introduction assumed to be beneficial for several water-treatment applications

(Hokkanen et al., 2013; Liimatainen et al., 2012; Suopajärvi et al.,
Aluminium is the third most abundant element in the Earth's 2013; Yu et al., 2013). For example, by attaching hydrophobic alkyl
crust and is one of the most commonly used metallic elements. chains on the nanocellulose backbone (Laitinen et al., 2014;
The Bayer process is the most common commercial concentration Visanko et al., 2014), bio-based collectors with selective perfor-
process for the production of alumina from bauxite (Xu et al., mances could potentially be attained for several mineral oxides, as
2004). In the Bayer process, bauxite is digested in a caustic soda shown in our previous study (Hartmann et al., 2016).
solution at elevated temperatures under pressure, after which the In the present study, aminated nanocelluloses with various
solution contains the dissolved aluminium oxide in the form of chain lengths were synthesised, and their ability to serve as quartz
sodium aluminate and the residue from the bauxite. This insoluble collectors in the mineral flotation process was addressed. Fur-
residue, called “red mud”, consists predominantly of iron oxide, thermore, the aim was to investigate the aminated nanocelluloses
titania and silica, but also contains many trace metals, such as in a mixture of aluminium oxide and quartz, and to compare their
magnesium, chromium, cobalt, nickel, copper and zinc (Collins et flotation performances and selectivities with commercial quartz
al., 2014). The red mud is then separated from the aluminium collectors (i.e. ether amine collectors).
oxide-rich solution, for instance with the aid of synthetic floccu-
lants. The clarified liquor is further purified (mud particle removal)
via filtration. Alumina trihydrate is then precipitated from the 2. Materials and methods
liquor, filtered and washed before the calcination is conducted at
extremely high temperatures. 2.1. Materials
Ores with an Al2O3/SiO2 mass ratio greater than 10 can be
directly processed with the Bayer process (Massola et al., 2009). Aluminium oxide (purity  99% Al2O3), also known as cor-
However, poorer ores must be processed beforehand in order to undum powder, was obtained from Sigma-Aldrich (USA). Quartz
increase the Al2O3/SiO2 ratio. For example, for a low-grade ore
(purity  98% SiO2) crystals were obtained from Sibelco Nordic Ltd.
with an Al2O3/SiO2 mass ratio below 8, a combination of sintering
(Finland). Original quartz particles (median size approximately
and the Bayer process has been used (Zhao et al., 2004). However,
400 mm) were ground in a laboratory ball mill (Retsch PM 200,
the very high energy consumption and negative environmental
Germany) for 5 min to obtain a suitable grain size fraction for
impacts associated with sintering processes reduce its feasibility
flotation. The size distributions of aluminium oxide and quartz
and economic viability in the concentration process. Therefore,
(after comminution), determined with a Beckman Coulter LS 320
low-cost physical separation processes, for instance with selective
particle size analyser, are presented in Fig. 1. The chemical com-
grinding and flotation (Ou et al., 2007) or selective flotation (Liu et
positions of minerals were characterised with an X-ray fluores-
al., 2007; Wang et al., 2004; Xia et al., 2010; Xu et al., 2004), which
cence spectrometer (XRF, PANalytical AxiosmAX). The content of
enables direct application of the Bayer process, are highly desir-
the major components is shown in Table 1.
able for increasing the Al2O3/SiO2 mass ratio of ores.
Alkyl ether monoamines (EDA and EDA-C) were provided by
Flotation is a physicochemical separation technique based on
Clariant (Finland). Fresh solutions (1% ww  1) of these reagents
chemical differences in the surface wettability characteristics of
were prepared each day. Hydrochloric acid and sodium hydroxide
the mineral species present in a pulp. The process utilises
were used to adjust each solution's pH to the desired level.
selectively-adsorbing chemical additives to modify the minerals'
Dowfroth 250C (Dow Chemical Company, USA) was used as a
surface wettability properties, in order to maximise the selective
frother in all flotation tests to modify the surface tension of the
separation of mineral species through the buoyance action of gas
water and to achieve a suitable bubble-size distribution for
bubbles (Araujo et al., 2005; Huang et al., 2014). In most flotation
systems, the particles of the valuable mineral are removed from mineral flotation. Different aminocellulose modifications were
the gangue (i.e. commercially worthless material that surrounds, synthesised in the laboratory, and their production is described
or is closely mixed with, a wanted mineral in an ore deposit) via step by step in the following chapter.
their attachment to gas bubbles and enrichment in the froth, Chemicals for the periodate oxidation of cellulose (NaIO4 and LiCl,
which characterises direct flotation. In reverse or inverse flotation, USA) were obtained from Sigma-Aldrich. 2-picolineborane (Sigma-
gangue particles are removed by the froth while the concentrate Aldrich, USA (95%)) and six primary, open chain amines were used for
remains in suspension (Filippov et al., 2010; Ma et al., 2011; Sahoo the reductive amination. These included: methylamine, n-propyla-
et al., 2015; Vieira and Peres, 2007). Reverse flotation can be useful mine, n-butylamine, and n-hexylamine hydrochloride; n-pentylamine
for cases in which gangue comprised of silica or silicates is floated
with an amine-type collector (i.e. surfactant chemical, which is
imparted a hydrophobicity of minerals in order to float them in
flotation overflow), while aluminium oxide particles are hydrated
and enriched in the underflow fraction.
Future challenges in aluminium processing are related to the
quality of the ores and the fact that the industry must be able to
benefit from deposits of lower grades due to the increasing
demand for sustainability, including the minimisation of water
and energy consumption and of environmental emissions. Fur-
thermore, there are economic and ecological incentives to develop
novel green chemicals from renewable resources, in order to
reduce the environmental impact of processing. For this purpose,
cellulose, which is the most abundant natural polymeric source,
represents a promising green alternative that could replace the
present synthetic chemicals. In particular, cellulose nanoparticles,
which have a very large specific surface area that can be covered
with cationic, anionic and hydrophobic functional groups, are Fig. 1. Particle size distribution of aluminium oxide and quartz.
262 O. Laitinen et al. / Chemical Engineering Science 144 (2016) 260–266
Table 1
XRF component analysis of aluminium oxide (Al2O3) and quartz (SiO2).
Component (%)
SiO2 MgO Al2O3 P2O5 SO3
Quartz 98.28 0.068 0.563 0.017 0.048
Alumina 0.038 0.154 99.37 0.003 0.018
(TCI, Germany); and ethylamine hydrochloride (Sigma-Aldrich, USA).
Ethanol (96%; VWR, Finland) was used to wash the oxidised pulp.
2.2. Synthesis of alkyl aminated nanocelluloses
Functionalised nanocelluloses were synthesised from birch
kraft pulp using periodate oxidation as the first reaction step for
producing dialdehyde cellulose (DAC). The properties of the cel-
lulose pulp are presented elsewhere (Liimatainen et al., 2011). As a
second step, DAC was aminatedusing reductive amination with
alkyl amines followed by cellulose nanofibrillation with a high-
pressure homogeniser.
The pulp was first converted into DAC using lithium chloride
(LiCl)-assisted sodium metaperiodate (NaIO4) oxidation, as pre-
viously reported (Sirviö et al., 2011). Six different (various chain
lengths) aminated nanocelluloses were synthesised from DAC
using reductive amination with methyl-, ethyl-, n-propyl-, n-butyl-
, n-pentyl- and n-hexylamine (Fig. 2). First, a tenfold excess of
amine in relation to the aldehyde groups of DAC was mixed with
deionised water, and the pH of the solution was set to 4.5 with
diluted HCl. Non-dried DAC and a twofold excess of 2-picoline
borane, based on the assessed amount of the aldehyde groups,
were added to the suspension, and the reaction was continued for
72 h with magnetic stirring in a closed container at room tem-
perature. The reaction was stopped by removing the reactive
chemicals from the solution using vacuum filtration. The modified
celluloses were stored in a non-dried state at 4 °C. Finally, all
materials were individualised to alkyl aminated nanofibres using a
two-chamber high-pressure homogeniser (APV  2000, Denmark).
The suspensions were passed through the homogeniser five times
with an increasing pressure sequence (200, 400, 600, 800 and
1000 bar). The methyl-, ethyl-, n-propyl-, n-butyl-, n-pentyl-, and
n-hexylamine modified nanocellulose samples were designated as
MAC, EAC, PRAC, BAC, PEAC and HAC, respectively.
2.3. Transmission Electron Microscope (TEM)
The morphological features of the functionalized nanocellu-
loses were analysed with a Tecnai G2 Spirit transmission electron
microscope (FEI Europe, Eindhoven, the Netherlands). Samples
were prepared by diluting each nanocellulose suspension with
Milli-Q water. A small droplet of the dilution was dosed on top of a
Butvar-coated and glow-discharged copper grid, and excess sam-
ple was removed from the grid by touching the droplet with the
corner of a filter paper. The samples were negatively stained by
placing a droplet of uranyl acetate (2% w/v) on top of each speci-
men. The excess uranyl acetate was removed with filter paper
as described earlier. The grids were dried at room temperature
and analysed at 100 kV under standard conditions. Images were
captured with a Quemesa CCD camera, and iTEM image analysis
software (Olympus Soft Imaging Solutions GMBH, Munster,
Germany) was used to measure the width and length of the
individual particles.
Cl K2O CaO TiO2 Fe2O3 CuO ZnO BaO Na2O
0.021 0.091 0.612 0.017 0.162 0.058 0.002 0.007 –
0.014 0.009 0.046 0.011 0.048 0.006 0.023 0.009 0.436
Fig. 2. Production of alkyl aminated celluloses through sequential periodate
oxidation and reductive amination with a primary amine.
2.4. Microflotation
Flotation responses were measured using a microflotation cell
(400 ml), which was coupled with suspension circulation through
a peristaltic pump. A 2.0 g sample (aluminium oxide, quartz or a
mixture of both) was mixed with 150 ml of distilled water in a
250 ml beaker, the pH was adjusted to the desired level, and the
freshly prepared suspension was mixed with a magnetic stirrer for
5 min. After that, the required quantity of flotation chemicals
(alkylaminated nanocelluloses or ether amines) was added to the
suspension and conditioned for 5 min. Next, 20 ml of frother
(Dowfroth 250C solution) was added, then the suspension was
introduced into the microflotation cell and adjusted to the desired
water level. Airflow during flotation was 25 ml/min, and the speed
of the peristaltic pump was 1400 ml/min. The floated products
were collected after 20 min of flotation. The samples were filtered,
dried, weighed and analysed. Tailings were also collected, dried,
weighed and analysed in order to determine recoveries. The che-
mical compositions of the artificial mixtures (containing 2.0 g each
of aluminium oxide and quartz) after flotation were determined
by XRF in order to calculate the selectivity of each aminated
nanocellulose and alkyl ether monoamine. The EDA, EDA-C and
aminated nanocellulose dosages were 0.1% ww  1 in the single-
oxide experiments. In the case of flotation of the particle mixture
(containing equal amounts of aluminium oxide and quartz), the
EDA, EDA-C and nanocellulose dosages were 0.05% ww  1.
2.5. Hydrophobicity of alkyl aminated nanocelluloses
The hydrophobicity of the fabricated nanocelluloses was eval-
uated by applying a static sessile-drop contact-angle measure-
ment, for which Milli-Q water was used as a probe liquid at room
temperature. For the measurements, self-standing films were
prepared from diluted cellulose suspensions. The suspensions
were dispersed using an Ultra-Turrax mixer (IKA T25, Germany) at
a 10,000 rpm rotational speed for 3 min, after which the suspen-
sions were vacuum-filtrated in a glass filter funnel (7.2 cm dia-
meter) using a filter membrane (polyvinylidene fluoride, 0.65 μm,
Millipore Durapore, France). Prepared films were dried in a drying
cabin. The measurements were carried out with a Krü ss DSA100
(Germany) system. For each sample, three droplets in different
locations were studied, the results were averaged, and the stan-
dard deviations were calculated.
2.6. Zeta potential measurements
The surface (zeta) potential of each mineral and nanocellulose
was measured as a function of pH value with a Zetasizer Nano ZS
 O. Laitinen et al. / Chemical Engineering Science 144 (2016) 260–266 263

(Malvern Instruments Ltd., England). The samples were diluted in
an electrolytic solution of 5 mM KCl. The pH levels of the solutions
were adjusted using 0.1 M HCl or NaOH, and the solutions were
stirred for 10 min and then allowed to settle. Finally, approxi-
mately 2 mL of the suspension was transferred to the measuring
cells. Each measurement was repeated three times, and the aver-
aged values were reported.
2.7. Chemical analysis
The chemical compositions of the flotation mixture samples
were characterised with XRF (PANalytical AxiosmAX). Flotated
overflow and underflow samples were filtrated and dried, and the
amounts of different metals were determined in order to calculate
the recovery of different minerals (especially aluminium oxide and
quartz) after microflotation.
contact angles of over 100°, so they can be characterised as
moderately hydrophobic. The hydrophobicity results indicate, in
practice, that the ability of the alkyl aminated nanocelluloses to
serve as quartz collectors (since amines are utilised as quartz
collectors) is increased when the alkyl chain of the functional
group was longer, due to higher hydrophobicity.
3.2. Quartz flotation
The results of the pure quartz flotation are presented in Fig. 4,
where the performances of different alkyl aminated nanocelluloses
Table 2
Alkyl aminated nanocelluloses and contact angles including standard deviations
(Sirviö et al., 2016).
Alkyl aminated nanocellulose Abbreviation Contact angle [°]

methyl-aminated cellulose MAC 78.2 7 2.77

ethyl-aminated cellulose EAC 64.07 3.40
3. Results and discussion n-propyl-aminated cellulose PRAC 85.5 7 4.47
n-butyl-aminated cellulose BAC 88.5 7 3.37
3.1. Morphology and hydrophobicity of alkyl aminated n-pentyl-aminated cellulose PEAC 101.6 7 2.16
nanocelluloses n-hexyl-aminated cellulose HAC 109.2 7 2.64

According to TEM images (Fig. 3), all of the alkyl aminated

nanocelluloses existed as rod-like individualised cellulose nano-
crystals. The width of individual crystals typically varied from 5 to
6 nm, while their length ranged from 70 to 130 nm, according to
the image analysis. The size distributions were in line with our
earlier studies (Laitinen et al., 2014; Visanko et al., 2014). Alkyl
aminated nanocrystals were well-individualised from each other
without the formation of larger aggregates, and the suspensions
could be considered stable during usage and storage.
Static sessile-drop contact-angle measurements were taken of
the self-standing films to indicate the surface hydrophobicity of
the aminated nanocelluloses, as reported in (Sirviö et al., 2016).
Table 2 shows the aminated nanocelluloses and three parallel
measurements of each of the contact angles of the nanocellulose
films. Based on these results, the chain lengths of the attached
alkyl groups were assumed to influence the hydrophobicity of the
Fig. 4. Flotation recovery curves of quartz as a function of pH with alkyl aminated
aminated nanocellulose surface. HAC had the highest hydro- nanocelluloses, EDA/EDA-C and without reagents. In these tests, the dosages of
phobicity, while MAC and EAC had the lowest. HAC and PEAC had alkyl aminated nanocelluloses and EDA/EDA-C were 0.1% ww  1.

Fig. 3. Example of TEM images of (a) MAC and (b) BAC nanocelluloses. The white dots in (a) are artefacts associated with sample preparation.
264 O. Laitinen et al. / Chemical Engineering Science 144 (2016) 260–266

and ether amines (EDA/EDA-C) are compared as a function of pH. In

these tests, the dosages of alkyl aminated nanocelluloses and ether
amines were 0.1% ww  1, based on our previous study (Laitinen et al.,
2014).
All of the alkyl aminated nanocelluloses served as collectors for
pure quartz, as indicated by the enhanced quartz recovery in the
presence of nanocelluloses. Furthermore, it was revealed that the
nanocelluloses with the longest alkyl chains (BAC, PEAC and HAC)
were much more effective collectors than those with shorter chain
lengths (MAC, EAC and PRAC). Thus, it seems that quartz flotation
efficiency was directly related to the hydrophobicity of alkyl
aminated nanocelluloses (Table 2). All of the nanocelluloses
(except HAC, at a pH of 10) maintained their efficiency throughout
the entire studied pH range of 6–10, showing their stability under
variable chemical conditions. It is assumed that when the amino
groups were attached to quartz particles, the hydrophobic nano-
cellulose side-chains increased the hydrophobicity of the quartz
Fig. 6. Zeta potentials of alkyl aminated nanocelluloses (error bar represents
particles, which in turn improved the interactions between the
standard deviations).
quartz and the air bubbles. According to visual observations during
the experiments, HAC started to flocculate and form agglomerates
at a pH level of around 10. Flocculation was ensured also with the
LS 13320 analyser (Beckmann Coulter). Volumetric particle size
distribution was measured with HAC in pH 7.5 and pH 10.
According to particle size results, median size was increased from
0.070 mm (pH 7.5) to 35.5 mm (pH 10). Flocculation may explain its
significantly decreased flotation efficiency under highly alkaline
conditions due to insufficient adsorption interaction between HAC
and the quartz mineral. The same phenomenon was observed in
the case of aluminium oxide flotation, for which HAC was used
(Fig. 5). Of the reference chemicals, the efficiency of the EDA was
notably affected by pH levels, with its optimum performance
obtained at a pH of 8–9, as stated by the manufacturer.

3.3. Aluminium oxide flotation

The results of the aluminium oxide flotation are presented in

Fig. 5, in which the functions of different aminated nanocelluloses
and ether amines are given as functions of pH. Similar to quartz,
the flotation recovery of aluminium oxide increased as a function
of the alkyl chain length of the nanocelluloses. However, the
nanocelluloses with the shortest alkyl chain lengths (MAC and
EAC) essentially did not float any aluminium oxide particles. When
the hydrophobicity of alkyl aminated nanocelluloses increased,
aluminium oxide particles began to float to some extent, but the
flotation efficiency was poor compared to quartz particles. Addi-
tionally, it was observed that ether amines also floated aluminium
Fig. 7. Zeta potentials of quartz and aluminium oxide (error bar represents stan-
dard deviations).
oxide particles to some extent. Based on these flotation results
with pure oxides, the aminated nanocelluloses had stronger
attractive interactions with quartz particles than with aluminium
oxide particles.
3.4. Zeta potential measurements

A prerequisite for the successful action of alkyl aminated

Fig. 5. Flotation recovery curves of aluminium oxide as a function of pH with alkyl
aminated nanocelluloses and EDA/EDA-C, and without reagents. In these tests, the
dosages of alkyl aminated nanocelluloses and EDA/EDA-C were 0.1% ww  1.
nanocelluloses as flotation reagents is the interaction and attach-
ment of nanocelluloses onto mineral surfaces. Electrokinetic
measurements can be used to delineate interfacial phenomena, in
which electrical double-layer effects are of relevance to the flota-
tion environment (Fuerstenau and Pradip, 2005). In order to
investigate these interactions, surface (zeta) potentials were
measured to determine the mechanisms of interaction between
aminated nanocelluloses and oxide particles. The zeta potentials of
pure alkyl aminated nanocelluloses, in deionised water with a
5 mM KCl background electrolyte solution, were measured in the
pH range of 3–11 (Fig. 6).
The results showed that all of the investigated alkyl aminated
nanocelluloses were positively charged when pH was below 8.5.
This was probably due to protonation of the amine groups of the
nanocelluloses. The alumina used in this research had an iso-
electric point (IEP) at a pH value of 7.7, and it was positively
 O. Laitinen et al. / Chemical Engineering Science 144 (2016) 260–266
charged below that pH level (Fig. 7). At a higher pH, the alumina
charge became negative.
Fig. 7 also shows that the zeta potential of quartz was negative
in the entire measurement range (pH 3–11). Consequently, it is
likely that at a pH value of approximately 7.5, the surface charges
of the two minerals are sufficiently different from each another to
enable the selective adsorption of charged alkyl aminated nano-
cellulose onto the quartz surfaces. Furthermore, it must be noted
that at a pH value of 7.5, quartz and alumina do not carry strong
opposite charges. This means that in practice, they will not form
aggregations due to electrostatic forces if they are placed together
in an aqueous suspension. Xia et al. (2009) observed that the
interactions between the minerals and the ionic collectors were
affected both by electrostatic effects and by hydrogen bonds
(Xia et al., 2009), and the present results support this finding.
3.5. Selectivity of nanocellulose collectors
Since adsorbed alkyl aminated nanocelluloses can promote
mineral particle–bubble interactions, their effect on overall particle
floatability was tested at a pH value of 7.5, at which alumina parti-
cles are slightly cationic and quartz particles are highly anionic.
Based on electrokinetic measurements, a pH of 7.5 was identified as
promising, and was chosen for selectivity experiments with a mix-
ture of alumina and quartz (50/50%-w/w). Only one alkyl aminated
nanocellulose type was used at a time, to investigate the selectivity
of each during flotation of the alumina/quartz mixture. Fig. 8 shows
the grades of alumina and quartz in the flotation overflow with
different alkyl aminated nanocelluloses and ether amines as flota-
tion chemicals. It can be seen that quartz grades in the flotation
overflow decreased slightly when the alkyl chain lengths of the
nanocellulose increased. However, notable separation selectivity was
achieved with all of the chemicals used. In the best experiments, the
quartz grades in the flotation overflow were almost five times
higher than the alumina grades.
The Al2O3/SiO2 ratio in the flotation underflow was increased
from 1.0 (in the feed) to 3.5 during one flotation stage in the case
of PEAC, while with ether amines, an increase from 1.0 to 2.0 was
noted. Thus the flotation efficiency when using PEAC, for example,
is likely sufficiently high that after several flotation cells, the
Al2O3/SiO2 ratio should increase to over 10. This is typically
necessary for the Bayer process (Massola et al., 2009), which is the
only commercial alumina extraction method (Ou et al., 2007).
Fig. 9 shows the recoveries of alumina and quartz when flot-
ation overflow was performed in the presence of different flotation
Fig. 8. Grades of the alumina and quartz in microflotation overflow. The chemical
dosage was 0.05% ww  1 and pH was 7.5.
265
Fig. 9. Flotation overflow recoveries with different alkyl aminated nanocelluloses
and EDA/EDA-C. Flotation experiments were performed with the alumina and
quartz mixture. The chemical dosage used was 0.05% ww  1 and pH value was 7.5.
reagents. The results showed that alkyl aminated nanocelluloses
floated quartz more efficiently than alumina, which was also the
case with the artificial oxide mixtures. Furthermore, the flotation
recoveries of quartz were significantly increased when nano-
celluloses with long chain lengths (PEAC and HAC) were used. It
was observed that commercial quartz collectors (i.e. ether amines)
led to lower quartz recoveries compared to PEAC and HAC, when
the same chemical dosages (0.05% ww  1) were used in the flota-
tion. The results clarify that similar flotation selectivities for a
quartz/alumina mixture are attained with commercial EDA solu-
tions and modified alkyl aminated nanocelluloses. The flotation
recovery results of quartz with aminated nanocelluloses presented
in this study are comparable to the temperature-sensitive polymer,
poly(N-isopropylacrylamide), presented earlier by Forbes et al.
(2011).
Despite promising results, it must be mentioned that the flo-
tation experiments were performed with a dual simple-oxide
mixture rather than a complex multi-mineral system. Systems of
mixed minerals are subject to further complexities, such as
entrainment, when dealing with the flotation of selective particle
aggregates (Grano, 2009; Miettinen et al., 2010). However, these
results provide a clear indication that selective flotation in a mixed
mineral system using alkyl aminated nanocelluloses is possible,
and the results are similar to those typically obtained with con-
ventional collectors, such as dodecyl amines (DDAs) or EDAs.
Moreover, the experiments were performed in only one stage,
while in industrial- or pilot-scale flotation processes, several
sequential stages (such as rougher flotation, scavenger cells and
cleaner cells) are typically used, consequently improving the
beneficiation efficiency. Additionally, only collectors for quartz,
and no additional depressants (such as starch) for alumina, were
employed. Using suitable amounts of both the collectors and the
depressants at the same time could also significantly improve the
overall flotation efficiency. This will be the aim of future research
that will explore more complex flotation systems and provide
more thorough comparisons between alkyl aminated nanocellu-
loses and conventional reagents.
4. Conclusions
The present study shows that alkyl aminated nanocelluloses
can be utilised as novel biochemicals in the flotation of pure alu-
mina and quartz. The alkyl chain length of the amino group affects
266 O. Laitinen et al. / Chemical Engineering Science 144 (2016) 260–266
the hydrophobicity of the modified nanocelluloses, and increased
hydrophobicity can be achieved by extending this length. The
increased particle–bubble attachment probability leads to an
overall increase of selective quartz particle floatability. The results
show that alkyl aminated nanocelluloses can be both effective and
selective collectors in a flotation system, if they are sufficiently
hydrophobic to effectively attach to air bubbles. Based on this
work, a pH of around 7.5 and an alkyl aminated cellulose dosage of
0.05% ww  1 enable the successful flotation of known alumina and
quartz mixtures. With the investigated alkyl aminated nanocellu-
loses, selectivity comparable to that of a commercial ether amine
collector can be achieved.
Acknowledgements
Sylvain Goenvic, Risto Aho and Anna Tenhunen are recognised
for their contribution to the experimental part of the study. This
work was conducted as part of the ERA-MIN CELMIN-project
supported by the Finnish Agency for Technology and Innovation
(TEKES); the Portuguese National Funding Agency for Science,
Research, and Technology (FCT); the Executive Agency for Higher
Education, Research, Development and Innovation Funding,
Romania (UEFISCDI); and several companies (Agnico-Eagle, Haarla,
Sojitz Beralt Tin & Wolfram and Sibelco).
References
Araujo, A.C., Viana, P.R.M., Peres, A.E.C., 2005. Reagents in iron ores flotation. Miner.
Eng. 18, 219–224. http://dx.doi.org/10.1016/j.mineng.2004.08.023.
Collins, R.N., Clark, M.W., Payne, T.E., 2014. Solid phases responsible for MnII, CrIII,
CoII, Ni, CuII and Zn immobilization by a modified bauxite refinery residue (red
mud) at pH 7.5. Chem. Eng. J. 236, 419–429. http://dx.doi.org/10.1016/j.
cej.2013.09.101.
Filippov, L.O., Filippova, I.V., Severov, V.V., 2010. The use of collectors mixture in the
reverse cationic flotation of magnetite ore: The role of Fe-bearing silicates.
Miner. Eng. 23, 91–98. http://dx.doi.org/10.1016/j.mineng.2009.10.007.
Forbes, E., Bradshaw, D.J., Franks, G.V., 2011. Temperature sensitive polymers as
efficient and selective flotation collectors. Miner. Eng. 24, 772–777. http://dx.
doi.org/10.1016/j.mineng.2011.02.003.
Fuerstenau, D.W., Pradip, 2005. Zeta potentials in the flotation of oxide and silicate
minerals. Adv. Colloid Interface Sci. 114  115, 9–26. http://dx.doi.org/10.1016/j.
cis.2004.08.006.
Grano, S., 2009. The critical importance of the grinding environment on fine par-
ticle recovery in flotation. Miner. Eng. 22, 386–394. http://dx.doi.org/10.1016/j.
mineng.2008.10.008.
Hartmann, R., Sirviö, J.A., Sliz, R., Laitinen, O., Liimatainen, H., Ämmälä, A., Fabritius,
T., Illikainen, M., 2016. Interactions between aminated cellulose nanocrystals
and quartz: adsorption and wettability studies. Colloids Surf. Physicochem. Eng.
Asp. 489, 207–215. http://dx.doi.org/10.1016/j.colsurfa.2015.10.022.
Hokkanen, S., Repo, E., Sillanpää, M., 2013. Removal of heavy metals from aqueous
solutions by succinic anhydride modified mercerized nanocellulose. Chem. Eng.
J. 223, 40–47. http://dx.doi.org/10.1016/j.cej.2013.02.054.
Huang, Z., Zhong, H., Wang, S., Xia, L., Zou, W., Liu, G., 2014. Investigations on
reverse cationic flotation of iron ore by using a Gemini surfactant: ethane  1,2-
bis(dimethyl-dodecyl-ammonium bromide). Chem. Eng. J. 257, 218–228. http:
//dx.doi.org/10.1016/j.cej.2014.07.057.
Laitinen, O., Kemppainen, K., Ämmälä, A., Sirviö, J.A., Liimatainen, H., Niinimäki, J.,
2014. Use of chemically modified nanocelluloses in flotation of hematite and
quartz. Ind. Eng. Chem. Res. 53, 20092–20098. http://dx.doi.org/10.1021/
ie503415t.
Liimatainen, H., Sirviö, J., Haapala, A., Hormi, O., Niinimäki, J., 2011. Characteriza-
tion of highly accessible cellulose microfibers generated by wet stirred media
milling. Carbohydr. Polym. 83, 2005–2010. http://dx.doi.org/10.1016/j.
carbpol.2010.11.007.
Liimatainen, H., Sirviö, J., Sundman, O., Hormi, O., Niinimäki, J., 2012. Use of
nanoparticular and soluble anionic celluloses in coagulation–flocculation
treatment of kaolin suspension. Water Res. 46, 2159–2166. http://dx.doi.org/
10.1016/j.watres.2012.01.035.
Liu, G., Zhong, H., Hu, Y., Zhao, S., Xia, L., 2007. The role of cationic polyacrylamide
in the reverse flotation of diasporic bauxite. Miner. Eng. 20, 1191–1199. http:
//dx.doi.org/10.1016/j.mineng.2007.05.003.
Massola, C.P., Chaves, A.P., Lima, J.R.B., Andrade, C.F., 2009. Separation of silica from
bauxite via froth flotation. Miner. Eng. 22, 315–318. http://dx.doi.org/10.1016/j.
mineng.2008.09.001.
Ma, X., Marques, M., Gontijo, C., 2011. Comparative studies of reverse cationic/
anionic flotation of Vale iron ore. Int. J. Miner. Process. 100, 179–183. http://dx.
doi.org/10.1016/j.minpro.2011.07.001.
Miettinen, T., Ralston, J., Fornasiero, D., 2010. The limits of fine particle flotation.
Miner. Eng. 23, 420–437. http://dx.doi.org/10.1016/j.mineng.2009.12.006.
Ou, L., Feng, Q., Chen, Y., Lu, Y., Zhang, G., 2007. Disintegration mode of bauxite and
selective separation of Al and Si. Miner. Eng. 20, 200–203. http://dx.doi.org/
10.1016/j.mineng.2006.09.009.
Sahoo, H., Sinha, N., Rath, S.S., Das, B., 2015. Ionic liquids as novel quartz collectors:
insights from experiments and theory. Chem. Eng. J. 273, 46–54. http://dx.doi.
org/10.1016/j.cej.2015.03.050.
Sirviö, J.A., Visanko, M., Laitinen, O., Ämmälä, A., Liimatainen, H., 2016. Amino-
modified cellulose nanocrystals with adjustable hydrophobicity from combined
regioselective oxidation and reductive amination. Carbohydr. Polym. 136,
581–587. http://dx.doi.org/10.1016/j.carbpol.2015.09.089.
Sirviö, J., Hyvakko, U., Liimatainen, H., Niinimaki, J., Hormi, O., 2011. Periodate
oxidation of cellulose at elevated temperatures using metal salts as cellulose
activators. Carbohydr. Polym. 83, 1293–1297. http://dx.doi.org/10.1016/j.
carbpol.2010.09.036.
Suopajärvi, T., Liimatainen, H., Hormi, O., Niinimäki, J., 2013. Coagulation–floccu-
lation treatment of municipal wastewater based on anionized nanocelluloses.
Chem. Eng. J. 231, 59–67. http://dx.doi.org/10.1016/j.cej.2013.07.010.
Vieira, A.M., Peres, A.E.C., 2007. The effect of amine type, pH, and size range in the
flotation of quartz. Miner. Eng. 20, 1008–1013. http://dx.doi.org/10.1016/j.
mineng.2007.03.013.
Visanko, M., Liimatainen, H., Sirviö, J.A., Heiskanen, J.P., Niinimäki, J., Hormi, O.,
2014. Amphiphilic cellulose nanocrystals from acid-free oxidative treatment:
physicochemical characteristics and use as an oil–water stabilizer. Biomacro-
molecules . http://dx.doi.org/10.1021/bm500628g.
Wang, Y., Hu, Y., He, P., Gu, G., 2004. Reverse flotation for removal of silicates from
diasporic-bauxite. Miner. Eng. 17, 63–68. http://dx.doi.org/10.1016/j.
mineng.2003.09.010.
Xia, L., Zhong, H., Liu, G., 2010. Flotation techniques for separation of diaspore from
bauxite using Gemini collector and starch depressant. Trans. Nonferrous Met.
Soc. China 20, 495–501. http://dx.doi.org/10.1016/S1003  6326(09)60168-0.
Xia, L., Zhong, H., Liu, G., Wang, S., 2009. Utilization of soluble starch as a depres-
sant for the reverse flotation of diaspore from kaolinite. Miner. Eng. 22,
560–565. http://dx.doi.org/10.1016/j.mineng.2009.01.007.
Xu, Z., Plitt, V., Liu, Q., 2004. Recent advances in reverse flotation of diasporic ores––
a Chinese experience. Miner. Eng. 17, 1007–1015. http://dx.doi.org/10.1016/j.
mineng.2004.04.011.
Yu, X., Tong, S., Ge, M., Wu, L., Zuo, J., Cao, C., Song, W., 2013. Adsorption of heavy
metal ions from aqueous solution by carboxylated cellulose nanocrystals. J.
Environ. Sci. 25, 933–943. http://dx.doi.org/10.1016/S1001  0742(12)60145-4.
Zhao, Q., Chen, Q., Yang, Q., 2004. The trends of Chinese alumina production with
combined process. In: Proceedings of the Light Metals-Warrendale, TMS,
pp. 127–130.