This article appeared in a journal published by Elsevier. The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article （e.g. in Word or Tex form） to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/copyright
Author's personal copy
Colloids and Surfaces A: Physicochem. Eng. Aspects 368 （2010） 75–83
Contents lists available at ScienceDirect
Colloids and Surfaces A: Physicochemical and
Engineering Aspects
journal homepage: www.elsevier.com/locate/colsurfa
A study of adsorption of dodecylamine on quartz surface using quartz crystal
microbalance with dissipation
J. Koua,b, D. Taoa,∗, G. Xua
a Department of Mining Engineering, University of Kentucky, Lexington, KY 40506, USA
b School of Civil and Environment Engineering, University of Science and Technology Beijing, 30 Xueyuan Road, Haidian District, Beijing 100083, PR China
a r t i c l e i n f o
Article history:
Received 23 April 2010
Received in revised form 13 July 2010
Accepted 15 July 2010
Available online 23 July 2010
Keywords:
Adsorption
Quartz
QCM-D
Dodecylamine hydrochloride
Zeta-potential
FTIR
a b s t r a c t
In this study the adsorption characteristics of dodecylamine hydrochloride （DACL） on quartz surface have
been investigated using a high sensitivity surface characterization technique referred to as quartz crystal
microbalance with dissipation （QCM-D） technique in conjunction with zeta-potential and FTIR analyses.
The experimental results have demonstrated the versatility and accuracy of the QCM-D for surface
adsorption characterization and for the first time revealed the changes in structure and orientation of
the amine adsorption film on quartz surface during the adsorption process by real-time measurements
of frequency and dissipation shifts with quartz coated sensor. Five distinct adsorption behaviors were
identified from the different slopes of D−f plots at different concentrations of DACL at pH 6 and 9.5.
The physisorption, coadsorption of dodecylammonium and dodecylamine, and surface precipitation of
neutral amine molecules with variable conformation and orientation were revealed on the quartz surface
by FTIR and QCM-D. Physisorption of ammonium ion and coadsorption of dodecylammonium and dodecylamine
dominated at concentrations <0.11mMatpH6 and 9.5 forming a rigid and thin adsorption layer.
A compaction stage was present at pH 9.5 at concentrations lower than 1.13mM. Surface precipitation
of neutral molecules dominated at higher concentrations at pH 6 and 9.5 to form a thick but dissipated
adsorption layer. The adsorption density was calculated with Sauerbrey equation and Voigt model and
the results indicated the existence of a critical concentration of 0.45mM at pH 6 and 1.13mM at pH 9.5
which led to a significant increase in adsorption density and a structural change in adsorption layer.
© 2010 Elsevier B.V. All rights reserved.
1. Introduction
Long-chain primary alkylammonium salts are employed in the
reverse flotation of silicates as well as in the flotation of quartz,
dolomite and calcite from phosphate minerals [1]. Since theamount
of surfactant adsorbed, the orientation of the adsorbed molecules,
and the nature of the adsorbed layer have amajor influence on the
hydrophobicity of mineral surface, a good understanding of adsorption
mechanisms of long-chain primary alkylammonium salt on
quartz surface and properties of the adsorbed layer is crucial for a
variety of industrial applications [2].
The mechanism of amine-silicate interaction has been studied
extensively by indirect methods and ex situ measurements such as
flotation recovery response [3], adsorption experiments [4,5], AFM
[6], contact angle [7], zeta-potential [8,9] and FTIR spectroscopy
[10–13] in the last several decades. Many concepts have been proposed
based on these studies.
∗ Corresponding author. .: +1 859 257 2953; +1 859 323 1962.
address: dtao@engr.uky.edu （D. Tao）.
Bijsterbosch [14] investigated the mechanism of cationic surfactant
adsorption on silica surfaces by electrophoresis experiments.
The results supported the theory of monolayer formation at low
surfactant concentrations due to interaction of opposite charges on
the silica surface and the surfactant ions. The formation of bilayers
at high surfactant concentration on quartz surface was demonstrated
by Menezes et al. [15] using contact angle and zeta-potential
measurements, which indicated that the decrease in contact angles
at concentration above the CMC can be attributed to the formation
of bilayers.
The adsorption mechanism of cationic surfactant dodecylpyridinium
chloride on quartz surface was also delineated through
measurements of adsorption isotherms, zeta-potentials, suspension
stability, contact angles, induction times, and flotation
response by Fuerstenau and Jia [9]. The results showed that
dodecylpyridinium adsorption on quartz occurred in four distinct
regions as the concentration of surfactant was increased. The
adsorption was controlled primarily by electrostatic interactions at
low concentrations, but at higher concentrations, adsorbed surfactant
ions began to associate at the interface forming hemimicelles.
At the CMC, the adsorbed surfactant ions existed as a bilayer or
its equivalent, with the charged head groups oriented towards
0927-7757/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.colsurfa.2010.07.017
Author's personal copy
76 J. Kou et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 368 （2010） 75–83
the solution phase. Churaev et al. [16] developed two models of
Langmuir-type adsorption of cationic surfactant on quartz surfaces,
including the adsorption of cations on both charged and neutral
sites which formed the monolayer and the consecutive adsorption
that first occurred on charged sites only, followed by adsorption
on hydrophobic tails of pre-adsorbed ions, forming two layers or
hemimicelles.
Novich and Ring [3] investigated the flotation behavior and
adsorption mechanism for the alkylamine-quartz flotation system.
They demonstrated the existence of three adsorption regions in
the adsorption isotherm. In the lower concentration range single
ion adsorbed on the quartz surface by electrostatic effect and the
adsorption density increased linearly with equilibrium concentration.
At the equilibrium concentration of approximay 0.1mM
（neutral pH） the surface micelle formed with the flattened adsorption
curves, which then abruptly steepened at an equilibrium
concentration closed to the CMC and continued to rise linearly with
the increase of concentration. Multilayer adsorption occurred at
high surfactant concentrations. These results are consistent with
G–F model [7,17] which also postulated the formation of surface
micelles and hemimicelle prior to monolayer coverage at HMC
（hemimicelle concentration）.
The adsorption mechanisms of long-chain alkylamines and their
acetate salts on quartz and mixed cationic and anionic collectors
on feldspar were investigated by Vidyadhar and Rao [18,19] using
Hallimond flotation, zeta-potential, FTIR, and X-ray photoelectron
spectroscopy （XPS） at neutral and acidic pH. In differentiation
to the electrostatic adsorption theory of G–F model, their FTIR
results exhibited the strong hydrogen bonds between the amine
cations and surface silanol groups. The acetate counterions were
found to influence the amine adsorption. The presence of neutral
amine molecules together with protonated ammonium ions was
also revealed by the XPS spectra above the critical hemimicelle
concentration （CHC）. Vidyadhar and Rao [18] also proved spectroscopically
that long-chain alcohols were coadsorbed along with
amine cations, which led to formation of a closely packed surface
layer, as compared to the case of adsorption of pure amine alone at
the same concentration.
The correlation between solid surface free energy and flotation
activity was verified by Chibowski and Holysz [20] with quartz conditioned
with DACL dissolved in methanol. They concluded that
at 0.25 statistical monolayer of DACL the sample of quartz floats
in 90%, which can be interpreted by the drastic reduction in the
polar component of quartz surface free energy, and the maximum
flotation activity appears at one statistical monolayer （20 ?A2 was
assumed for the molecule） of the amine.
The hydrophobic attractive forces between mica surfaces in
dodecylammonium chloride solution were reported by Yoon and
Ravishankar [21]. At concentration of 0.003mM and pH 5.7 where
the surfactant is in ionized form （dodecylammonium）, only shortrange
hydrophobic forces were observed, which can be attributed
to the difficulty in forming close-packed monolayers on the mica
surface. At pH 9.5 where neutral molecule of dodecylamine was
formed as a result of hydrolysis long-range hydrophobic forces
were observed, which may be attributed to the coadsorption of
both dodecylammonium and dodecylamine, increasing the packing
density of hydrocarbon chains on the surface. They also concluded
that the pH at which the long-range hydrophobic force appeared
corresponded to the maximum quartz flotation.
McNamee et al. [22] investigated the adsorption of the cationic
surfactant octadecyl trimethylammoniumchloride （C18TAC） at low
concentrations to negatively charged silica surfaces in water using
atomic force microscopy （AFM）. From the AFM images of silica surface,
C18TAC was seen to form islands of bilayers and patches on
mica at a concentration of 0.03mM, which was identified as partial
surfactant bilayers or hemimicelles. The electrostatic attraction
between the bilayer islands of surfactant and bare areas of the
substrate was observed at low surfactant concentrations.
Most of these concepts about the adsorption of amines on
silicate minerals were based on ex situ studies and indirect methods
such as measurement of contact angle, zeta-potential, surface
forces, and recovery response, which unfortunay cannot monitor
the formation process of the adsorbed layer. The objective of
the present study was to perform in situ investigation of the behavior
of the adsorbed layer of dodecylamine on quartz surface by
use of the quartz crystal microbalance with dissipation technique
（QCM-D）, which is capable of providing real-time information on
the adsorption layer density, strength, and structural change under
various process conditions. To our knowledge, this technique has
not been employed previously for studying the adsorption of amine
on quartz.
Quartz crystal microbalance （QCM） measures the mass change
per unit area by monitoring the change in frequency of a quartz
crystal resonator. It offers an opportunity to study molecular interactions
and conformation changes of adsorbed layer on many
different types of surfaces with acute sensitivity [23]. QCM-D is the
second generation of QCM, which cannot only determine the mass
of surface bound layers, but also simultaneously give information
about their structural （viscoelastic） properties based on the data of
dissipation factor shift. A brief discussion of this technique is provided
later in this paper.QCM-Dhas recently been used to study the
viscoelastic properties of protein adsorbed on biosensors surface
[24] and the adsorption behavior of surfactants and polymers from
aqueous solutions [25–28]. In this work, in situ adsorption behavior
of dodecylammonium onto the quartz coated sensor at various
concentrations and pH was studied for the first time using QCMD.
The obtained data was fitted with Voigt model to get physical
properties and mechanical properties of the adsorbed layer.
2. Materials and methods
2.1. Materials
Dodecylamine hydrochloride （CH3（CH2）11NH2·HCl） and anhydrous
ethanol （C2H5OH） with 99% and 99.5% purity, respectively
were acquired from Acros Organics. Sodium hydroxide, which was
purchased from Fisher Scientific, contained 99.8% NaOH. Deionized
water was used throughout the experiments.
The quartz coated sensor used in QCM-D analysis was an ATcut
quartz disc （14mm in diameter and 0.3mm in thickness with
an active sensor crystal area of 0.2cm2） with 5 nm Cr, 100nm Au,
50nm Ti and 50nm SiO2 sputter-coated onto the crystal surface
successively. The sensors and Q-sense E4 system were supplied by
Q-Sense Co.
Zeta-potential and FTIR tests were conducted with quartz powder
（50% <16.93m, 90% <85.90m） made of pulverized pure
quartz crystals purchased from Ward’s Natural Science. XRD analyses
with the sample did not show any impurity.
2.2. Quartz crystal microbalance with dissipation
All QCM-D measurements in this study were conducted at 25? C
（±0.02 ?C）. The stock solution was prepared by dissolving dodecylamine
hydrochloride （DACL） in deionized water in the presence of
ethanol. The solutionpHwas adjusted to 6.00±0.05 and 9.50±0.05
using 1% NaOH and 1% HCl solution. To ensure dissolution and
degassing, the solutions were left in an ultrasonic bath for 5–10 min.
For each experiment, the data generated with the solvent only
（deionized water） was accepted as baseline when it became stable.
Following this, the DACL solutions were injected into the flow
module by a chemical feeding pump capable of precise control of
Author's personal copy
J. Kou et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 368 （2010） 75–83 77
flow rate. The flow rate in the experiment was kept at 0.50 mL/min.
Prior to each test, the flow modules, crystals and tubings were
cleaned with 2% Hellmanex II solution （an alkaline cleaning liquid
from Hellma GmbH & Co. KG, Germany） for 0.5–1 h and deionized
water for 0.5 h. Then the crystal was removed and rinsed with more
water or ethanol. After drying with nitrogen gas, the crystal was
remounted into the module and deionized water was injected into
the system.
Software QTools 3.0 was used for data modeling and analysis.
For a rigid, thin, uniform film the change in dissipation factor
D<1×10−6 for 10 Hz frequency change [29] and the Sauerbrey
equation （Eq. （1）） was used for the mass calculation:
m = −qtqf
f0n
= −qvqf
2f 2
0 n
= −Cf
n
（1）
where the constant C has a value of 17.8 ngcm−2 Hz−1 and n is
the harmonic number （when n=1, f0 = 5MHz）. If the adsorbed film
is “soft” （viscoelastic）, it will not fully couple to the oscillation of
the crystal [30], causing energy dissipation of the system. The dissipation
factor D is proportional to the power dissipation in the
oscillatory system （Eq. （2）） and can give valuable information about
the rigidity of the adsorbed film [30]:
D =
Edissipated
2Estored
（2）
where Edissipated is the energy dissipated during one oscillation and
Estored is the energy stored in the oscillating system [30]. Therefore,
the changes in the property of adsorbed layer will result in the
measured change in D. During the tests, QCM-D simultaneously
measures the changes in resonance frequency f and dissipation
D as a result of adsorption on a crystal surface. The quantitative
information on the adsorbed layer on the mineral surface can be
obtained by data modeling.
When the adsorption caused greater shift in D value, i.e.,
D>1×10−6, as a result of the viscous and soft layer, Voigt modeling
（Eqs. （3） and （4）） [28] was used:
f ≈ − 1
20h0
??
?
3
?3
+

j=1,2

hjjω−2hj

3
?3
2 jω2
2j
+ 2j
ω2
 ??
? （3）
D ≈ 1
2f0h0
??
?
3
?3
+

j=1,2

2hj

3
?3
2 jω2
2j
+ 2j
ω2
 ??
? （4）
According to Voigt model for viscous adsorbed layer, f and D
depend on the density （）, thickness （h）, elastic shear modulus （）
and shear viscosity （） of adsorption layer （j: number of adsorbed
layer）. The Sauerbrey equation and Voigt model are the theoretical
basis for data modeling using software QTools 3.0 （Q-Sense Co.）.
2.3. Zeta-potential measurements
The zeta-potential measurements were made with Zeta-plus
analyzer of Brook Haven Instruments Corporation. All experiments
were conducted with DACL solutions with 1.0mMKCl under ambient
conditions. 1.0 g quartz powder was conditioned in 20mL stock
solution with amagnetic stirrer for 1 h. The mineral suspension was
filtered using Whatman filter paper （pore size 25m） and then
poured into the rectangular cell for zeta-potential measurements.
2.4. FTIR analysis
The infrared transmission spectra were recorded on a Thermo
Nicolet Nexus 470 FTIR spectrometer. The quartz powders were
conditioned with 20mL DACL solution with ethanol at different
Fig. 1. Real-time experimental data of frequency shifts （a） and dissipation shifts （b）
for the third overtone （15 MHz） of QCM-D resonator for different concentrations of
dodecylamine hydrochloride adsorption on quartz surface atpH6. Arrow 1 indicates
the injection of dodecylamine hydrochloride solution.
pH’s and different concentrations while being agitated with a magnetic
stirrer for 0.5 h to make 1% suspension. The suspension was
then filtered with Whatman filter paper （pore size 25m） and the
solids were air-dried overnight at the room temperature. The samples
were prepared by dispersing 0.025 g air-dried powder in 5 g
KBr followed by pressing into a transparent tablet for scanning.
The untreated （initial） quartz powder was used as reference. Each
spectrum is the average of 250 scans.
3. Results
3.1. Adsorption behavior of DACL on quartz surface
Fig. 1 displays the real-time experimental data of the frequency
shiftf （Fig. 1（a）） and dissipation shiftD （Fig. 1（b）） from the third
overtone （15 MHz） associated with DACL adsorption onto quartz
surface at different concentrations at pH 6 （the same trend was
observed at 0.11 and 6.76mM as at 0.07 and 2.25mM, respectively,
and thus not shown in Fig. 1）. Immediay after the injection
of DACL at arrow 1, a gradual decrease in f （adsorption density
increase） and a slight increase in D appeared with 0.07 and
0.11mM DACL. At the steady state D was <0.3×10−6 and f
was −2.5 Hz at 0.07 and 0.11mM, which indicated the formation of
thin and rigid adsorption film. After the injection of 0.45mMDACL,
f decreased gradually until it reached the steady state at about
−5HzwhileD remained <0.4×10−6, suggesting the formation of
a thicker adsorption layer than at 0.11 and 0.07mM DACL.
At concentrations of 1.13, 2.25 and 6.76mM, the injection of
DACL caused an initial rapid decrease in f and a significant
Author's personal copy
78 J. Kou et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 368 （2010） 75–83
Fig. 2. Real-time experimental data of frequency shifts （a） and dissipation shifts （b）
for the third overtone （15 MHz） of QCM-D resonator for different concentrations
of dodecylamine hydrochloride adsorption on quartz surface at pH 9.5. Arrow 1
indicates the injection of dodecylamine hydrochloride solution.
increase inDfollowed by a gradual decrease inf and an increase
inD which became stable after 1.7 h. At the steady state,f of−9,
−12 and −17 Hz of was achieved with 1.13, 2.25 and 6.76mMDACL
solution, respectively and theD was higher than 2×10−6 at these
concentrations, revealing the formation of a thick but less rigid
adsorption layer with adsorbed molecules not densely compacted.
Fig. 2 shows the f and D as a function of time when the
quartz surface was exposed to different dosages of DACL at pH 9.5
（the same trend was observed at 0.07mMas at 0.11mMand therefore
is not shown in Fig. 2）. The first observation was the gradual
decrease in f and increase in D after the injection of ≤0.45mM
solutions （Fig. 2, arrow 1）. The steady states of f and D were
reached with≤0.45mMpH9.5 solutions 1.2 hafter the solutionwas
injected, or 0.5 h earlier than with the same concentration at pH 6.
The value of D smaller than 1×10−6 indicates that the adsorbed
layer was non-dissipative. The stable D values were almost the
same as those observed with the adsorption at the same concentration
of ≤0.45mM at pH 6 but the f was much lower at pH 9.5,
which suggests an increase in adsorption density with increasing
solution pH value.
After the injection of 1.13, 2.25 and 6.76mM DACL solutions
（Fig. 2a, arrow 1）, a sharp decrease in f was observed to a steady
state value of−13,−16 and−24 Hz, respectively, which were much
lower and appeared much earlier than at pH 6. It should be noted
that the steady state value of 0.8×10−6 for D at 1.13mM was
much lower at pH 9.5 than 2.6×10−6 produced at pH 6, suggesting
that a thicker and more rigid adsorption layer was formed in
1.13mM DACL solutions at pH 9.5 than at pH 6. The D’s at 2.25
and 6.76mMwere also lower at pH 9.5 than at pH 6, but still higher
than 1×10−6.
Fig. 3. Real-time experimental data of frequency and dissipation shifts for the third
overtone （15 MHz） of QCM-D resonator for DACL adsorption and rinsing with deionized
water from quartz surface. The straight lines are the f and D of 2.25mM
DACL at pH 6; the dash lines are the f and D of 1.13mM DACL at pH 6; the
dash-dotted curves are the f and D of 6.76mM DACL at pH 9.5; the dotted
curves are the f and D of 0.45mM DACL at pH 9.5. Arrows 1 and 2 indicate
the injection of dodecylamine hydrochloride solution and reagent-free water,
respectively.
Fig. 3 shows the f and D as a function of time when the
quartz surface was exposed to 1.13 and 2.25mM DACL solutions
at pH 6 and 0.45 and 6.76mM DACL solutions at pH 9.5 at
arrow 1 followed by rinsing with water of the same pH value but
without DACL at arrow 2. It can be observed that after the injection
of water at arrow 2, f and D at all these concentrations
changed significantly. With a 0.45mM solution at pH 9.5, the D
and f were immediay back to zero after rinsing with water,
implying that the desorption occurred readily at lower concentrations
at pH 9.5. At concentrations of 1.13 and 2.25mM at pH
6 and of 6.76mM at pH 9.5, D changed to about 0–0.5×10−6
with f increased to about −2 Hz, which corresponded to
0.5×10−10 mol/cm2 adsorption density. This adsorption density
was much lower than the statistical monolayer, which was
6.64×10−10 mol/cm2 on the crystal sensor （the active sensor crystal
area was 0.2cm2 and 25 ?A2 was assumed for the molecule） [3].
It can be concluded that even at higher pH or higher concentration,
almost all adsorbed DACL can be washed off from the quartz
surface.
3.2. Adsorption density of DACL
The measured data of f and D from different overtone
（from 3rd to 13th） was fitted using the Sauerbrey equation for
D<10−6 or Voigt model for D>10−6. The calculated adsorption
density at concentrations from 0.07 to 6.76mM is shown
in Fig. 4. It can be seen that the adsorption density increased
with increasing concentration at both pH values. In pH 6 solutions
the adsorption density was lower than 5×10−10 mol/cm2 at
concentrations ranging from 0.07 to 0.45mM. The adsorption density
jumped to 19.4×10−10 mol/cm2 at 1.13mM and continued to
increase to about 56.7×10−10 mol/cm2 with increasing the concentration
to 6.76mM. A thicker adsorption layer appeared at all
concentrations at pH 9.5 than at pH 6. For example, the adsorption
density was 19.4×10−10 mol/cm2 and 47.3×10−10 mol/cm2
at pH 6 and 9.5, respectively in 1.13mM solutions. The dissipation
shift D was only 0.8×10−6 with 1.13mM DACL at pH 9.5,
which was <2.5×10−6 observed at pH 6. It can be concluded that
the adsorption layer was thicker and less dissipated at higher
pH.
Author's personal copy
J. Kou et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 368 （2010） 75–83 79
Fig. 4. Adsorption density for different concentrations of dodecylamine hydrochloride
on quartz surface at pH 6 and 9.5.
3.3. Kinetics of DACL adsorption
To characterize the adsorption behavior of DACL on quartz, the
measured data of f and D from third overtone （15 MHz） in
Figs. 1 and 2 were correlated in D−f plots in Fig. 5. D−f
plots serve as the fingerprint of an adsorption process and provide
the information on the energy dissipation per unit mass added
to the crystal sensor [29]. The slope of the plots is defined as K
（K =D/f, absolute value） which is indicative of kinetic and structural
alternation during adsorption process [29]. Small value of K
indicates the formation of a compact and rigid layer, while a high
value indicates a soft and dissipated layer [23,29].
Fig. 5 shows the D−f plots for the adsorption of DACL on
the quartz surface at （a） 0.11mM at pH 6, （b） 2.25mM at pH 6, （c）
0.45mM at pH 9.5, and （d） 2.25mM at pH 9.5. The data in Fig. 5
（a） shows only one phase （one slope） and Fig. 5（b） and （d） shows
two distinct phases （two slopes） of kinetics whereas Fig. 5（c） shows
three phases （three slopes）. According to the previous studies conducted
by Rodahl et al. [23], Hook et al. [31], and Paul et al. [29], a
morerigid and compact adsorption mass is expected to yield a small
K value and a soft and dissipated layer is associated with a higher
K value. One slope only indicates a direct adhesion on the surface.
The presence of multiple slopes suggests direct adhesion and orientational
changes associated with hydrodynamically coupled water
[29,23]. Data shown in Fig. 5 reveals that DACL adsorption took
Fig. 5. D−f plots for DACL adsorbed on the quartz surface at （a） 0.11mM and pH 6; （b） 2.25mM and pH 6; （c） 0.45mM and pH 9.5; （d） 2.25mM and pH 9.5.
Author's personal copy
80 J. Kou et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 368 （2010） 75–83
Table 1
Slopes in D−f plots for adsorption of different concentrations of dodecylamine hydrochloride at pH 6 and 9.5 on quartz.
Concentration of DACL （mM） 0.07 0.11 0.45 1.13 2.25 6.76
pH 6 slope K （×10−6）Hz−1
0.0313 0.102
K1 = 0.2495 K1 = 0.3609 K1 = 0.4904 K1 = 0.2516
K2 = 0.0288 K2 = 0.093 K2 = 0.1409 K2 = 0.0026
pH 9.5 slope K （×10−6）Hz−1 K1 = 0.1038 K1 = 0.0438 K1 = 0.1417
K1 = 0.1549 K1 = 0.0467 K1 = 0.0387
K2 = 0.0051 K2 = 0.0027 K2 = 0.0072
K3 = 0.1035 K3 = 0.0314 K3 = 0.0672 K2 = 0.006 K2 = 0.1544 K2 = 0.4146
place with constant kinetics at 0.11mM in pH 6 solution. However,
the adsorption in pH 9.5 solutions or in pH 6 solutions with
DACL concentration higher than 0.11mM resulted in changes in
adsorption characteristics.
Table 1 shows calculated K values for adsorption processes at
various DACL concentrations and pH’s represented in Fig. 4. Four
distinct adsorption processes can be identified: （1） the adsorption
layer was thin and rigid at lower DACL concentrations （0.07
and 0.11mM） with only one K value of <0.15 at pH 6, which indicates
the direct adhesion with no kinetic changes; （2） two slopes
（K1 > 0.15 > K2） were observed with DACL concentrations of greater
than 0.45mM at pH 6 and 1.13mM at pH 9.5, which indicates the
formation of an initial flexible and water rich adsorption layer, followed
by a structural and orientational changes as evidenced by a
decrease in energy dissipation per unit adsorption mass K1 values
increased with increasing concentration of DACL at pH 6, as shown
in Table 1, which indicates that the adsorption layer became more
and more dissipated. （3） Three slopes （0.15 > K1 ≈K3 > 0.01 > K2）
were observed with lower DACL concentrations （0.07, 0.11 and
0.45mM） at pH 9.5, which shows the same slopes at the beginning
and the end of the adsorption process separated by a stage with
a much lower slope of <0.01. Small values of three slopes （<0.15）
indicate the formation of compact and rigid layers as evidenced by
D<10−6 and orientational changes during the adsorption. （4） At
higher DACL concentrations of 2.25 and 6.67mM and pH 9.5, two
slopes （K2 > 0.15 > K1） were observed, indicating the rapid adhesion
followed by conformational changes or water molecules trapped in
the adsorption layer with low structural stability. Comparing the K
values in Table 1 also shows that a smaller K1 value was observed
at pH 9.5 than at pH 6 with the same DACL concentrations, which
indicates that although the initial adsorption of DACL on quartz was
rapid at pH 9.5, as shown in Figs. 1 and 2, the adsorbed layer was
more rigid and less dissipated than at pH 6. When the adsorption
became stable, the adsorbed layer of DACL had higher adsorption
density and lower D at pH 9.5 than at pH 6, which indicated better
organized structure of DACL on the quartz surface at higher pH.
These adsorption results can be used to explain why quartz flotation
at pH 9.5 can achieve better performance than at pH 6 when
amine is used as collector. According to Mielczarski et al. [32] the
hydrophobicity of mineral is closely related to the structure of the
adsorbed surfactant on the surface. The highly packed adsorption
layer of surfactant imparts strong hydrophobicity to solid surface
whereas a poorly organized structure does not.
3.4. Zeta-potential measurement
To better understand the adsorption mechanism of DACL in
weakly acidic solutions, the relationship between the initial DACL
concentration and the zeta-potential of quartz particles at pH 6 is
shown in Fig. 6. It is evident that increasing the concentration to
about 0.36mMsharply increased the surface charge of quartz from
−110 to 0mV. Novich and Ring [3] suggested that the quartz surface
will be negatively charged if the suspension pH is above the point of
zero charge （pH 2.0） and cationic amine surfactant will absorbwhen
it is placed in solution with the negatively charged quartz surface.
According to G–F model, alkylammonium ions specifically adsorb
at negative sites on the quartz surface in the low-concentration
region, oriented with the charged head toward the surface and the
hydrocarbon tail into the solution, which accounted for the increase
in zeta-potential. The QCM-D results also showed a low adsorption
density of 1.8×10−10 mol/cm2 and only one slope of K≤0.15 on
the D−f plot at 0.07 and 0.11mM at pH 6. It can be concluded
that a rigid and well-ordered monolayer was present on the quartz
surface at a concentration lower than 0.36mM at pH 6.
A further increase in DACL concentration from 0.36 to 6.76mM
led to positively charged phase although the rate of increase in surface
charge slowed with increasing DACL concentration. It can be
postulated based on the G–F model that after the monolayer coverage,
a second layer forms with the charged amine heads toward
the solution and the hydrocarbon tails oriented toward the surface.
This is responsible for the positive zeta-potential at the beginning
of the higher concentration region and agrees well with the QCMD
results that showed two slopes （K1 > 0.15 > K2） on the D−f
plot and significantly increased adsorption density （decreased f）
at higher DACL concentrations at pH 6.
3.5. FTIR analysis
Infrared transmission spectra of quartz after adsorption in DACL
solution at different pH’s and concentrations were investigated to
confirm the adsorption mechanism of amine on the quartz surface.
Fig. 7 shows the infrared transmission spectra of dodecylamine
hydrochloride （DACL） as reference. The bands characteristic of
alkyl chains for dodecylamine are identified at 2950, 2920, and
2850cm−1, which were assigned to asymmetric stretching vibration
in theCH3 （as （CH3）） andCH2 radical （as（CH2）） and symmetric
stretching vibration in the CH2 radical （s（CH2）） [18], respectively.
The broad band at 3000cm−1 was observed, which was assigned to
as（NH3
+） and s（NH3
+） [18].
Fig. 6. Zeta-potential of quartz conditioned with DACL at constant ionic strength as
a function of concentration.
Author's personal copy
J. Kou et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 368 （2010） 75–83 81
Fig. 7. Infrared transmission spectra of DACL.
Curves a and b in Fig. 8A show the absorption bands ranging
from 2750 to 3050cm−1 in the FTIR spectra of the species adsorbed
on quartz surface from the solutions of 4.5mMDACL at pH 6 and 9.5
and curves c and d show the same absorption bands with 0.45mM
DACL solution at pH 6 and 9.5, respectively. Curve e shows the
adsorption bands with quartz powder conditioned with ethanol
Fig. 8. Infrared transmission spectra of quartz after adsorption of DACL at different
concentrations and pHs for （A） alkyl chain region 2750–3050cm−1 and （B） （OH）
group and amine molecule region 3050–3500cm−1 （a） 4.5mM, pH 9.5; （b） 4.5mM,
pH 6; （c） 0.45mM, pH 9.5; （d） 0.45mM, pH 6; （e） quartz powder with ethanol.
Fig. 9. Infrared transmission spectra of quartz after adsorption of DACL at different
concentrations and pHs for the alkyl chain region 2750–3100cm−1 （a） 4.5mM, pH
9.5; （b） 4.5mM, pH 6; （c） 0.45mM, pH 9.5; （d） 0.45mM, pH 6; （e） quartz powder
with ethanol; （f） DACL.
solution as reference. Curves a–d show the absorption peaks at
2950, 2920, and 2850cm−1, which are undoubtedly caused by the
adsorption of alkyl chain on the quartz surface since they are consistent
with DACL spectra shown in Fig. 7. In comparison with curve
e, the intensities of these bands in Fig. 8 were found to increase with
increasing DACL concentration, which is in agreement with QCM-D
measurements and zeta-potential results.
According to Vidyadhar et al. [2], the amine molecule in the
bulk phase in neutral and protonated forms can be identified with
the appearance of the band at 3333cm−1 and the alcohol spectrum
shows a broad band around 3318cm−1 characteristic of （OH）
group. Fig. 8B （curves a–e） shows the spectra in the region from
3300 to 3350cm−1 of quartz treated with DACL solution under different
conditions and with ethanol solution. The broad band on the
spectra indicated the existence of H-bond of alcohol （OH） group
on quartz surface at all concentrations and pH’s as well as at the
higher concentration of amine.
The DACL spectrum in Fig. 7 also displays several bands in the
region of 1700–1000cm−1. The band at 1460–1480cm−1 is often
assigned to the methylene scissoring band ?（CH2） on the quartz surface.
But quartz also exhibits strong absorption in the same region
and thus it is hard to identify these bands undermono-or submonolayer
adsorption [18]. Fig. 9 shows the infrared transmission spectra
for the adsorption bands in the region from 1400 to 1800cm−1. It
was observed that the absorption of DACL gave rise to characteristic
absorption bands in the region of 1400–1700cm−1 on quartz
surface and the intensity of these absorption peaks increased with
increasing concentration.
The FTIR results did not show significant effects of pH on the
intensities of absorption peaks, which is inconsistent with flotation
performance of amine at different pH’s. In contrast the QCM-D
results clearly showed differences in adsorption behavior at different
pH’s. For example, theD−f plot shows K1 > K2 at high DACL
concentrations at pH 6 and K1 < K2 at pH 9.5. According to Sirkeci
[33] the coadsorption of ionic and molecular species is responsible
for quartz flotation with dodecylamine at high pHs. Sirkeci [33]
further claimed that pH 9.3 represents the equilibrium between
ionized and molecular amine species. Vieira and Peres [34] also
reported that the most favorable pH for quartz flotation was pH 9.0
Author's personal copy
82 J. Kou et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 368 （2010） 75–83
for both ether monoamine and ether diamine. In good agreement
with previous studies QCM-D results have shown the significant
increase in adsorption density and adsorption layer stability as a
result of surface precipitation of molecular amine at pH 9.5 on the
quartz.
4. Discussion
The previous studies on the adsorption of dodecylamine on
quartz surface have revealed three adsorption mechanisms: （1）
single ion adsorption at lower concentration; （2） surface micelle
formation at intermediate concentration; and （3） multilayer
adsorption with molecular amine precipitation at higher concentration
[3]. The present QCM-D study provided for the first time the
real-time adsorption data and the kinetic D−f plots displayed
five distinguishable adsorption stages:
（1） At lower concentrations （0.07 and 0.11mM） at pH 6, a slight
decrease in f and a small D value of <1×10−6 with only
one slope K < 0.15 was observed, which indicates the adsorption
process formed a thin but rigid adsorption layer. The species
distribution diagram developed by Novich and Ring [3] for
0.04mM dodecylamine solution at different pH’s suggests that
cationic amine （RNH3
+） is the dominant species in pH 6 solution.
According to Vidyadhar and Rao [18], at pH from 2 to
7, where the surface potential of silicate is negative, ammonium
ions undergo physisorption and are electrostatically held
at the silicate–water interface much below the critical micelle
concentration （CMC, 10mM at pH 6）, which is in agreement
with zeta-potential measurement results. Novich and Ring
[3] reported an adsorption density of 1.24×10−10 mol/cm2 at
first adsorption step, in agreement with QCM-D results which
showed 1.8×10−10 mol/cm2 adsorption density on quartz surface
at concentrations of 0.07 and 0.11mM.
（2） Three slopes （0.15 > K1 ≈K3 > 0.01 > K2） were observed at lower
DACL concentrations at pH 9.5, which revealed different
adsorption processes from at pH 6. According to the species distribution
diagram, the concentration of cationic amine starts
to decrease and a small amount of RNH2 is observed in the
0.04mM solution at pH 9.5. Yoon and Ravishankar [21] indicated
that neutral molecules of dodecylamine were coadsorbed
with dodecylammonium at low concentrations atpH9.5, which
may increase the packing density of hydrocarbon chains on the
surface and form a higher adsorption density but lower dissipative
adsorption layer. The existence of one small slope between
two higher slopes may indicate that the orientational changes
and layer compression step resulted in more rigid adsorption
layer.
（3） It should be mentioned that the QCM-D results show different
adsorption behavior at concentrations of 0.45mM at pH 6
and 1.13mM at pH 9.5. Although the D−f plot displayed
two slopes and higher adsorption density at 0.45mM and pH 6
and 1.13mM at pH 9.5, the dissipation shift D was <1×10−6.
These two adsorption conditions produced higher adsorption
density than at concentration <0.45mM, but lower dissipation
shift than at concentration >1.13mM. After the monolayer
coverage the formation of second layer with the hydrocarbon
tails oriented toward the surface and the charged amine heads
toward the solution accounted for the charge reversal of zetapotential
（0.36mM at pH 6 as shown in Fig. 6）. According to
QCM-D results, the critical concentration for hemimicelle formation
increased with increasing pH from 6 to 9.5 and the
hemimicelle formed at this critical concentration was rigid and
well-ordered.
（4） At DACL concentrations higher than 0.45mM at pH 6, a significant
decrease in f happened simultaneously with a sharp
increase in D. Two slopes （K1 > 0.15 > K2） were observed on
D−f plots, suggesting the formation of a thick, dissipated
and less rigid structure followed by a compaction stage on the
quartz surface caused by surface precipitation. According to
Chernyshova et al. [8] the adsorbed film on quartz includes
both neutral and protonated amine H-bonded to the surface
silanols at higher concentration, which is in agreement with
FTIR results shown in Fig. 8 that also showed adsorption peak
intensity higher than at lower concentration and without DACL.
（5） At concentrations higher than 1.13mM and pH 9.5, a dramatic
decrease in f and a significant increase in D were observed
along with two slopes （K2 > 0.15 > K1） on D−f plots, which
suggests the formation of a rigid structure followed by a surface
precipitation that resulted in higher adsorption density. This
result is consistent with the theory proposed by Novich and
Ring [3] that there is a multilayer of adsorbates held together
by Van de Waals force associated with hydrocarbon chains.
McNamee et al. [35] also reported the Van der Waals attraction
between the cationic surfactant and negatively charged
silica surface based on AFM and zeta-potential measurements.
The poorly organized multilayer structure is reflected in high
dissipation shift D and low resonance frequency change f
observed in the present study. The DACL adsorption density in
2.25mM solution was estimated to be 47.3×10−10 mol/cm2 at
pH 9.5, which is very close to the value of 39.6×10−10 mol/cm2
Novich and Ring [3] reported as a result of multilayer adsorption.
5. Conclusions
In situ measurements of dodecylamine adsorption on quartz
surface at different concentrations and pH’s were conducted using
the QCM-D technique to investigate the adsorption density and
kinetics and the structural properties of the adsorption layer. The
following major conclusions were derived from this study:
（1） Three adsorption mechanisms, i.e., physisorption, formation
of hemimicelle and surface precipitation of neutral molecule,
were revealed by the QCM-D technique combined with Zetapotential
measurements and FTIR analysis for the adsorption
of dodecylamine on quartz surface. At low concentrations and
pH 6, physisorption of ammonium ion formed a rigid and thin
adsorption layer, resulting in a small change in both f and
D. The coadsorption of both dodecylammonium and dodecylamine
were observed in pH 9.5 solutions, which formed a
thicker but more rigid layer than at pH 6. Surface hemimicelle
with a higher density and a lower dissipation was formed
at intermediate concentrations, and the critical concentration
for hemimicelle formation increased from 0.45 to 1.13mM
with increasing pH from 6 to 9.5. Surface precipitation of neutral
amine molecules dominated at higher concentrations and
caused a rapid decrease in f and increase in D, which indicated
the formation of a thick but soft and dissipated adsorption
layer.
（2） Different adsorption behaviors can be readily derived from
the different slopes of D−f plots. At lower concentrations
the adsorption layer was thin and rigid with only
one small slope （K < 0.15） at pH 6 and three small slopes
（0.15 > K1 ≈K3 > 0.01 > K2） at pH 9.5. At higher concentrations
two slopes were observed with K1 > 0.15 > K2 at pH 6 and
K2 > 0.15 > K1 at pH 9.5, which indicated the formation of higher
adsorption density and more dissipated multilayer structure. A
critical concentration, i.e., 0.45mMat pH 6, and 1.13mMat pH
Author's personal copy
J. Kou et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 368 （2010） 75–83 83
9.5, was observed which indicated the occurrence of surface
hemimicelle with higher adsorption density and lower dissipation.
（3） Adsorption densities under different solution conditions were
calculated with Sauerbrey equation and Voigt model. At concentrations
≤0.45mM at pH 6 and ≤0.11mM at pH 9.5
the adsorption density was close to the theoretical monolayer
coverage of 6.64×10−10 mol/cm2; the adsorption density
increased significantly at concentrations higher than 0.45mM
at both pH’s.
（4） Zeta-potential and FTIR studies showed that the DACL adsorption
on quartz involved both physisorption as a result of the
electrostatic interaction at lower concentrations and surface
precipitation of amine molecules on quartz surface at higher
concentrations, which helped explain the different adsorption
processes observed in QCM-D studies.
References
[1] H. Sis, S. Chander, Reagents used in the flotation of phosphate ores: a critical
review, Miner. Eng. 16 （2003） 577–585.
[2] A. Vidyadhar, K.H. Rao, I.V. Chernyshova, Mechanisms of amine-feldspar interaction
in the absence and presence of alcohols studies by spectroscopic
methods, Colloids Surf. A: Physicochem. Eng. Aspects 214 （2003） 127–142.
[3] B.E. Novich, T.A. Ring, A predictive model for the alkylamine–quartz flotation
system, Langmuir 1 （1985） 701–708.
[4] H. Matsubara, T. Eguchi, H. Takumi, K. Tsuchiya, T. Takiue, M. Aratono, Surface
adsorption and aggregate formation of cationic gemini surfactant and longchain
alcohol mixtures, J. Phys. Chem. B 113 （2009） 8847–8853.
[5] Q. Zhou, P. Somasundaran, Synergistic adsorption of mixtures of cationic gemini
and nonionic sugar-based surfactant on silica, J. Colloid Interface Sci. 331 （2009）
288–294.
[6] J.H. Zhang, R.H. Yoon, J.C. Eriksson, AFM surface force measurements concluded
with silica in CnTACL solutions: effect of chain length on hydrophobic force,
Colloids Surf. A: Physicochem. Eng. Aspects 300 （2007） 335–345.
[7] D.W. Fuerstenau, H.M. Jang, On the nature of alkylsulfonate adsorption at the
rutile/water interface, Langmuir 7 （1991） 3138–3143.
[8] I.V. Chernyshova, K.H. Rao, A. Vidyadhar, Mechanism of adsorption of longchain
alkylamines on silicates. A spectroscopic study. 1. Quartz, Langmuir 16
（2000） 8071–8084.
[9] D.W. Fuerstenau, R. Jia, The adsorption of alkylpyridinium chlorides and their
effect on the interfacial behavior of quartz, Colloids surf. A: Physicochem. Eng.
Aspects 250 （2004） 223–231.
[10] I.V. Chernyshova, K.H. Rao, A. Vidyadhar, Mechanism of adsorption of longchain
alkylamines on silicates: a spectroscopic study. 2. Albite, Langmuir 17
（2001） 775–785.
[11] R.M.F. Lima, P.R.G. Brandao, A.E.C. Peres, The infrared spectra of amine collectors
used in the flotation of iron ores, Miner. Eng. 18 （2004） 267–273.
[12] I.V. Chernyshova, K.H. Rao, A new approach to the IR spectroscopic study of
molecular orientation and packing in adsorbed monolayers. Orientation and
packing of long-chain primary amines and alcohols on quartz, J. Phys. Chem. B
105 （2001） 810–820.
[13] I.V. Chernyshova, K.H. Rao, Mechanism of coadsorption of long-chain alkylamines
and alcohols on silicates: Fourier transform spectroscopy and X-ray
photoelectron spectroscopy studies, Longuire 17 （2001） 2711–2719.
[14] B.H. Bijsterbosch, Characteriztion of silica surfaces by adsorption from solution:
investigation into themechanismof adsorption of cationic surfactants, J. Colloid
Interface Sci. 47 （1973） 186–198.
[15] J.L. Menezes, J. Yan, M.M. Sharma, The mechanism of alteration of macroscopic
contact angles by the adsorption of surfactants, Colloids Surf. 38 （1989）
365–390.
[16] N.V. Churaev, I.P. Sergeeva, V.D. Sobolev, H.J. Jacobasch, P. Weidenhammer, F.J.
Schmitt, Modification of quartz surfaces using cationic surfactant solutions,
Colloids Surf. A: Physicochem. Eng. Aspects 164 （2000） 121–129.
[17] A.M. Gaudin, D.W. Fuerstenau, Quartz flotation with cationic collectors, Trans.
AIME 202 （1955） 958–962.
[18] A. Vidyadhar, K.H. Rao, Adsorption mechanism of mixed cationic/anionic collectors
in feldspar-quartz flotation system, J. Colloid Interface Sci. 305 （2007）
195–204.
[19] A. Vidyadhar, K.H. Rao, I.V. Chernyshova, K.S.E. Forssberg, Mechanisms of
amine-quartz interaction in the absence and presence of alcohols studied by
spectroscopic methods, J. Colloid Interface Sci. 256 （2002） 59–72.
[20] E. Chibowski, L. Holysz, Correlation of surface free energy changes and floatability
of quartz, J. Colloid Interface Sci. 112 （1985） 15–23.
[21] R.H. Yoon, S.A. Ravishankar, Long-range hydrophobic forces between mica surface
in alkaline dodecylammonium chloride solutions, J. Colloid Interface Sci.
179 （1996） 403–411.
[22] C.E. McNamee, H.J. Butt, K. Higashitani, I.U. Vakarelski, M. Kappl, Interaction of
cationic hydrophobic surfactants at negatively charged surfaces investigated
by atomic force microscopy, Longmuir 25 （2009） 11509–11515.
[23] M. Rodahl, F. Hook, C. Fredriksson, C.A. Keller, A. Krozer, P. Brzezinski, M.
Voinova, B. Kasemo, Simultaneous frequency and dissipation factor QCM measurements
of biomolecular adsorption and cell adhesion, Faraday Discuss. 107
（1997） 229–246.
[24] X. Turon, O.J. Rojas, R.S. Deinhammer, Enzymatic kinetics of cellulose hydrolysis:
a QCM-D study, Langmuir 24 （2008） 3880–3887.
[25] J.J.R. Stalgren, J. Eriksson, K. Boschkova, A comparative study of surfactant
adsorption on model surfaces using the quartz crystal microbalance and the
ellipsometer, J. Colloid Interface Sci. 253 （2002） 190–195.
[26] M.A. Plunkett, P.M. Claesson, M.W. Rutland, Adsorption of cationic polyelectrolyte
followed by surfactant-induced swelling, studied with a quartz crystal
microbalance, Langmuir 18 （2002） 1274–1280.
[27] B.D. Vogt, C.L. Soles, H.J. Lee, W.L. Wu, Moisture adsorption into ultrathin
hydrophilic polymer films on different substrate surface, Polymer 46 （2005）
1635–1642.
[28] M.V. Voinova, M. Rodahl, M. Jonson, B. Kasemo, Viscoelastic acoustic response
of layered polymer films at fluid–solid interfaces: continuum mechanics
approach, Phys. Scr. 59 （1999） 391–396.
[29] S. Paul, D. Paul, B. Tamara, A.K. Ray, Studies of adsorption and viscoelastic
properties of proteins onto liquid crystal phthalocyanine surface using quartz
crystal microbalance with dissipation technique, J. Phys. Chem. C 112 （2008）
11822–11830.
[30] P. Ekholm, E. Blomberg, P. Claesson, I.H. Auflem, J. Sjoblom, A. Kornfeldt, A
quartz crystal microbalance study of the adsorption of asphaltenes and resins
onto a hydrophilic surface, J. Colloid Interface Sci. 247 （2002） 342–350.
[31] F. Hook, M. Rodahl, P. Brzezinski, B. Kasemo, Energy dissipation kinetics for
protein and antibody-antigen adsorption under shear oscillation on a quartz
crystal microbalance, Langmuir 14 （1998） 729–734.
[32] J.A. Mielczarski, J.M. Cases, E. Bouquet, O. Barres, J.F. Delon, Nature and structure
of adsorption layer on apatite contacted with oleate solution. 1. Adsorption and
Fourier transform infrared reflection studies, Langmuir 9 （1993） 2370–2382.
[33] A.A. Sirkeci, Electrokinetic properties of pyrite, arsenopyrite and quartz in the
absence and presence of cationic collectors and their flotation behavior, Miner.
Eng. 13 （2000） 1037–1048.
[34] A.M. Vieira, A.E.C. Peres, The effect of amine type, pH, and size range in the
flotation of quartz, Miner. Eng. 20 （2007） 1008–1013.
[35] C.E. McNamee, M. Matsumoto, P.G. Hartley, P. Mulvaney, Y. Tsujii, M. Nakahara,
Interaction forces and zeta potentials of cationic polyelectrolyte coated silica
surfaces in water and in ethanol: effects of chain length and concentration of
perfluorinated anionic surfactants on their binding to the surface, Langmuir 17
（2001） 6220–6227.