Content uploaded by M. Fernanda Mora
Author content
All content in this area was uploaded by M. Fernanda Mora on Apr 28, 2015
Content may be subject to copyright.
The effects of alkyl sulfates on the analysis of phenolic compounds by
microchip capillary electrophoresis with pulsed amperometric detection
Yongsheng Ding,{ Maria F. Mora, Grant N. Merrill and Carlos D. Garcia*
Received 22nd March 2007, Accepted 12th July 2007
First published as an Advance Article on the web 3rd August 2007
DOI: 10.1039/b704364c
The effects of different surfactants (sodium 2-ethylhexyl sulfate, sodium decyl sulfate, sodium
dodecyl sulfate and sodium tetradecyl sulfate) on the analysis of phenolic compounds by
microchip-CE with pulsed amperometric detection were investigated. Using sodium decyl sulfate
as a model surfactant, the effects of concentration and pH were examined. Under the optimized
conditions, the analysis of six phenolic compounds was performed and compared with control
runs performed without surfactant. When these surfactants were present in the run buffer,
decreases in the migration time and increases in the run-to-run reproducibility were observed.
Systematic improvements in the electrochemical response for the phenolic compounds were also
obtained. According to the results presented, surfactants enhance the analyte–electrode
interaction and facilitate the electron transfer process. These results should allow a more rational
selection of the surfactants based on their electrophoretic and electrochemical effects.
Introduction
Miniaturized analysis systems have great potential in biologi-
cal and environmental analysis due to their low cost, minimal
sample and reagent requirements, short analysis time and
portability.
1,2
Although glass was initially used as a substrate,
the use of polymeric materials has significantly increased of
late.
1,3
Polymeric materials are less expensive than silicon or
glass and involve simpler and less expensive manufacturing
processes. Although different polymers (such as polyester,
4
polyethylene,
5
polycarbonate,
6,7
polymethyl methacrylate,
8–10
polystyrene,
11
polyethyleneterephthalate glycol, and poly-
ethylene
12
) have been used in the fabrication of microfluidic
devices, one of the most convenient is poly(dimethylsiloxane)
(PDMS).
3,13
PDMS devices can be fabricated in hours, do not
require complex instrumentation, and are relatively inexpen-
sive. PDMS can also be sealed to a wide variety of substrate
materials: it is also optically transparent above 230 nm,
flexible, non-toxic, inert, and permeable to gases.
14
One of the
greatest limitations of PDMS is, however, its hydrophobicity
(PDMS/water contact angle of y108u),
13
which allows
hydrophobic analytes to interact with the channel surface,
leading to significant peak tailing and loss of separation
efficiency. Because uncured monomers can migrate from the
bulk to the surface, PDMS can also suffer from unstable
electro-osmotic flow (m
EOF
),
13,15,16
which may compromise the
reproducibility of the analysis. To overcome these two major
drawbacks of PDMS, different strategies have been pro-
posed,
17–24
including coating the surface with surfactants. A
number of reasons exist for this: (1) surfactants spontaneously
adsorb to PDMS, and no instrumentation or elaborated
procedures are therefore required; (2) surfactant concentration
(in solution and adsorbed on the surface) can be controlled
without difficulty; (3) most surfactants can be removed by
rinsing the surface with background electrolyte;
25–31
(4) the
same surfactants can be used to effect the separation;
32
and (5)
different surfactants are commercially available, allowing a
variety of experimental designs.
33
Brij-35,
16,34
Tween 20,
cetyltrimethylammonium bromide,
16
sodium dodecyl sulfate,
35
didodecyldimethylammonium bromide,
25,36,37
1,2-di-lauroyl-
sn-phosphatidylcholine, Triton X-100,
38–40
and palmityl sulfo-
betaine
41
have all been shown to strongly influence both the
m
EOF
and separation efficiency.
42
Different aspects of adsorption of surfactants to the
capillary surfaces
16,30,43–56
have been previously reported. It
is generally accepted that surfactants can adsorb to the
capillary wall by a combination of hydrophobic and electro-
static interactions.
57
As a consequence, the direction, magni-
tude, and stability of the m
EOF
can be controlled. Several
groups have also reported the beneficial effects of surfactants
on electrochemical detection. In this regard, cationic surfac-
tants are used to exploit electrostatic interactions with anionic
analytes.
58–61
It has also been proposed that anionic micelles
could be responsible for the increased stability of the electro-
chemical signal, but the mechanism by which surfactants
improve the electrochemical detection at submicellar concen-
trations remains unclear.
52,62
The improvement in electro-
chemical signal is particularly significant when anionic
surfactants are used in combination with pulsed amperometric
detection (PAD).
30,52–54,56,63,64
The main purpose of our work was to simultaneously study
the electrophoretic and electrochemical effects of the addition
of a series of alkyl sulfates to the background electrolyte.
Experiments were conducted using a microchip-CE device and
pulsed electrochemical detection. As a result, m
EOF
, migration
time, and peak current as function of surfactant type (sodium
2-ethylhexyl sulfate, sodium decyl sulfate, sodium dodecyl
Department of Chemistry, The University of Texas at San Antonio, San
Antonio, TX 78249, USA. E-mail: carlos.garcia@utsa.edu;
Tel: +1 (210) 458-5465
{ Permanent address: The Department of Chemical Biology, Peking
University Health Science Center, Beijing 100083, P. R. China.
PAPER www.rsc.org/analyst | The Analyst
This journal is
ß
The Royal Society of Chemistry 2007
Analyst
, 2007, 132, 997–1004 | 997
sulfate, and tetradecyl sulfate sodium) were analyzed. SD
10
S
as a model surfactant, and the effect of the surfactant
concentration and pH, were also investigated. In all the cases,
six phenolic compounds with different charge distributions,
structure, and pK
a
values were selected as analytes.
Experimental
Chemicals and apparatus
SU-8 2035 photoresist was purchased from Micro Chem Co.
(Newton, MA), and Sylgard 184 silicone elastomer and curing
agent were obtained from Dow Corning (Midland, MI).
Aqueous solutions were prepared using analytical grade
reagents and 18 MV cm
21
resistance water (NANOpure
Diamond, Barnstead, Dubuque, IA). Sodium 2-ethylhexyl
sulfate (SE
8
S, Aldrich, Saint Louis, MO), sodium decyl sulfate
(SD
10
S, Fluka, Buchs, Switzerland), sodium dodecyl sulfate
(SD
12
S, Sigma, Saint Louis, MO), and sodium tetradecyl
sulfate (ST
14
S, Aldrich, Saint Louis, MO) were used as
received. Pentachlorophenol (PCP), 4,6-dinitro-o-cresol
(DNOC), 2-nitrophenol (NP), vanillic acid (VA), and
4-hydroxybenzoic acid (HBA) were purchased form Sigma–
Aldrich (St. Louis, MO). o-Aminophenol (AP) was purchased
from Eastman Kodak Co (Rochester, NY). Sample stock
solutions of analytes were prepared by dissolving the desired
amount in methanol (Fisher Scientific, Fairlawn, NJ).
Electrolytes were prepared by dissolving the desired amount
of sodium borate (Na
2
B
4
O
7
?10 H
2
O) in water and adjusting
the pH with either 0.1 M NaOH or 0.1 M HCl (Fisher
Scientific, Fairlawn, NJ). The pH measurements were
performed with a combined glass electrode and a digital pH
meter (Orion 420A+, Thermo, Waltham, MA). A home-made
high voltage power supply (HVPS)
65
was used for separation
and injection in all the experiments. As previously
reported, the HVPS consists of two positive and one negative
dc–dc converters, a microprocessor controlled timer, and a
110 V–12 V transformer. This arrangement allows for the
control of the applied potentials in the 0 to ¡4000 V range
with noise values in the nA range.
65
Pulsed amperometric detection (PAD) was performed using
a CHI812 electrochemical detector (CH Instruments, Austin,
TX). Unless otherwise noted, the waveform applied to the
electrode consists of a cleaning step (1.2 V for 0.05 s), a
reconstruction step (20.5 V for 0.025 s), and a detection step
(0.6 V for 0.15 s). These values based upon the respective
hydrodynamic voltammograms (data not shown).
66–70
A
25-mm gold wire was used as the working electrode, which
has been previously described.
30,52,53,56,66–69,71
The detection
area was estimated to be 0.0039 mm
2
(25 mm electrode
diameter, 50 mm channel width). Ag wire and Pt electrodes
were used as pseudo-reference and auxiliary electrodes, respec-
tively, and placed in the waste reservoir. In all cases, the
plotted points and error bars represent the average and
standard deviation, respectively, of at least 5 measurements.
Fabrication of the PDMS microchip
PDMS microchips were fabricated according to a previously
described procedure.
72
Briefly, a 100 mm silicon wafer (Silicon
Valley Microelectronics Inc., Santa Clara, CA) was cleaned
and oxidized with piranha solution (1:1 H
2
SO
4
:H
2
O
2
), rinsed
with deionized water, and dried at 200 uC for 15 min. (Piranha
solution is a powerful oxidizing agent that reacts violently with
organic compounds; it should be handled with extreme care.)
The wafer was then coated with SU-8 2035 negative photo-
resist using a spin coater (Laurell Technologies, North Wales,
PA) by dispensing approximately 3 mL of photoresist onto the
wafer. A spread cycle of 500 rpm for 10 s, followed by
1900 rpm for 30 s, was performed. Two pre-exposure baking
steps at 65 and 95 uC for 5 and 10 min, respectively, were then
performed. A digitally produced mask (The Photoplot Store,
Colorado Springs, CO) containing the channel pattern was
placed on the coated wafer, exposed to light via a near-UV
flood source (365 nm–13 mW cm
22
, Optical Associates Inc.,
San Jose, CA) for 30 s, and then baked at 95 uC for 1 min. The
positive relief was developed by placing the wafer in propylene
glycol methyl ether acetate for 15 min, rinsed with methanol,
and dried under a N
2
stream. The height of the positive patterns
on the molding masters, which is equal to the channel depth
created in the PDMS layer, was 50 mm as measured with a pro-
filometer (Dektak IIA, SloanTechnology, Santa Barbara, CA).
Two PDMS layers were fabricated by pouring a de-gassed
mixture of Sylgard 184 silicone elastomer and curing agent
(10:1) onto either a molding master or a blank wafer, followed
by curing for at least 2 h at 65 uC. The cured PDMS was
separated from the mold, and reservoirs were made at the end
of each channel using a 6 mm circular punch. As previously
reported, the detection electrode was then aligned at the end of
the separation channel in an ad hoc perpendicular chan-
nel.
52,56,67,68,71
The two PDMS layers were then placed in an
air plasma cleaner (Harrick, PDC-32G, Ithaca, NY), oxidized
for 20 s and immediately brought into conformal contact to
form an irreversible seal. The extremities of the electrode
channel were sealed with two drops of super glue. Finally, an
electrical connection to the working electrode was made using
silver paint and a copper wire. A double-T injector with a
500 mm gap between side channels was used for all experi-
ments, thus defining a 1.2 nL sample plug. The separation
channel was 52 mm long. The standard amount of buffer
dispensed into each reservoir (buffer, sample waste, and waste)
was 50 mL, except for the sample reservoir, in which only 45 mL
were dispensed to avoid sample leakage into the separation
channel. During the sample injection, potentials of +500 V,
250 V, and +500 V were applied to the sample, sample waste,
and buffer reservoirs, respectively. In order to avoid Joule
heating in the sample-to-buffer channel during the injection, a
resistor (1 MV) was included in series with the chip. During the
separation step, the potential applied to the buffer reservoir
was raised to +1200 V (or to the corresponding separation
potential) while the potential applied to the sample waste
reservoir was changed to +500 V. The waste reservoir was
always grounded. Electrical connections to the microfluidic
devices were made with platinum electrodes placed into the
reservoirs at the ends of each channel. In order to minimize
noise, a separate ground electrode, positioned in the waste
reservoir, was used for the separation system.
To avoid the formation of micelles, all the surfactants were
used at concentrations below their respective critical micellar
998 |
Analyst
, 2007, 132, 997–1004 This journal is
ß
The Royal Society of Chemistry 2007
concentrations (CMC). Before each analysis, the microchips
were preconditioned by sequentially rinsing the channels for
5 min with 0.01 M NaOH solution and water. The PDMS
channels were then hydrodynamically rinsed for 10 min with a
solution containing the surfactant. Preliminary experiments
had shown that this rinsing time ensured equilibrium
conditions (adsorbed surfactant u surfactant in solution).
Because not all the surfactant can be completely removed
from the surface, a new chip was used for each different
surfactant.
Computational methods
All structures were fully optimized at the Hartree–Fock (HF)
level of theory.
73,74
Structures were considered converged
when the root-mean-square average and the largest component
of the gradients fell below 0.004 and 0.012 kcal mol
21
A
˚
21
,
respectively. To verify that all optimized structures corre-
sponded to minima, analytic Hessians were computed at the
HF level. Minima are characterized by the absence of negative
eigenvalues for the Hessian matrices. Optimizations and
Hessians were performed with the 6-31G double split-valence
basis set of Pople and co-workers,
75–77
to which were added
sets of d-orbital polarization and sp-orbital diffuse func-
tions
78–80
to all atoms except hydrogen: 6-31+G(d) ;
6-31+G*. Charges were derived from Mulliken population
analyses,
81–84
and they were centered on the nuclei. All
calculations were carried out with the GAMESS program,
85
which was run on small LINUX clusters of personal
computers in a Beowulf configuration. The resulting wave
functions for these structures permitted the corresponding
dipole moments (m) to be calculated. It was also possible to
derive partial charge distributions from these wave functions.
While dipole moments are observable quantities, partial
charges are not, and there is, therefore, a certain degree of
arbitrariness to them. The results are expressed in electrons
(e
2
) and dipoles in debye (D). The partial charges have been
summed together for functional groups (e.g., phenyl). The
following convention for dipole moments was used: the
arrow’s tail is positively charged, while the arrow’s head is
negatively charged. The well-established scheme of Mulliken
for assigning partial charges was used, and the charges
themselves were centered on nuclei. The foregoing results are
for species in the gas phase, i.e., solvation effects were ignored.
Results and discussion
It is generally accepted that the adsorption of negatively
charged surfactants to the capillary wall produces a self-
adsorbing, non-permanent coating that increases the m
EOF
.
16,29
To evaluate the effects of the selected surfactants on the
separation using microchips, experiments were performed
under different analytical conditions. It is worth noting that,
because surfactants will adsorb to both the PDMS surface and
the detection electrode, their effects on the separation and the
detection of different phenolic compounds were analyzed
simultaneously. For these experiments, six phenolic com-
pounds (PCP, DNOC, NP, VA, HBA, and AP) were selected
considering their different structures, pK
a
values, and mole-
cular weights.
Effect of surfactant structure on separation and detection
Surfactants adsorb to surfaces by a combination of electro-
static and hydrophobic interactions.
44,46,48,86,87
In the case
of PDMS, hydrophobic interactions overcome electrostatic
repulsions and allow different anionic surfactants to be
spontaneously adsorbed to the channel wall.
30,52
To evaluate
the effects of different surfactants (at the same concentration),
the electrophoretic separation and PAD of six phenolic com-
pounds was investigated. As can be observed in Fig. 1, all the
surfactants produced increases in the m
EOF
with respect to the
bare PDMS channel (m
EOF PDMS
= 3.58 6 10
24
cm
2
V
21
s
21
)
under the same experimental conditions. Typical m
EOF
values
of 3.61 6 10
24
cm
2
V
21
s
21
, 3.64 6 10
24
cm
2
V
21
s
21
, 3.86 6
10
24
cm
2
V
21
s
21
, and 4.35 6 10
24
cm
2
V
21
s
21
were obtained
for SE
8
S, SD
10
S, SD
12
S, and ST
14
S, respectively. The
differences in m
EOF
can be attributed to different channel
surface charges, as a function of the adsorbed amount of
surfactant. The addition of alkyl surfactants to the back-
ground electrolyte also resulted in smaller (2.4%) relative
standard deviations in migration time with respect to PDMS
(4.9%), probably due to the stabilization of the surface charge.
Because submicellar concentrations were selected for these
experiments, the migration order of the phenolic compounds
was preserved.
The addition of surfactants to the background electrolyte
(BGE) can affect not only migration time (t
M
) but also
electrochemical response.
52,54
It has been reported that the
addition of SD
12
S produced significant increases in peak
current for carbohydrates, phenols and other small mole-
cules.
30,52–54,63,64
The exact mechanism associated with these
signal increases is, however, not fully understood. To
investigate the role of surfactant structure on the observed
signal enhancement, the peak current of six selected phenolic
compounds was analyzed. As can be seen in Fig. 1, the
addition of surfactants to the background electrolyte produced
relative improvements in peak current. To summarize the
effects of the addition of surfactants on the electrochemical
Fig. 1 Electropherograms obtained in the presence of different
surfactants (at 0.5 mM each). Conditions: 5 mM borate, pH = 11.0,
E
SEP
= 1200 V, t
INJ
=10s,E
DET
= 0.6 V. Phenols concentrations and
migration order: 20 mM AP, 13 mM PCP, 15 mM DNOC, 37 mM NP,
25 mM VA, and 37 mM HBA.
This journal is
ß
The Royal Society of Chemistry 2007
Analyst
, 2007, 132, 997–1004 | 999
response, the results (I
p
versus surfactant) are included in
Table 1. Although the relative signal magnitude depends on
the structure of both the surfactant and the phenolic com-
pound, a significant signal enhancement was obtained for all
the compounds (44 ¡ 8%) when 0.5 mM SE
8
S was used. A
smaller increase was obtained when 0.5 mM SD
10
S (27 ¡
15%) or SD
12
S (27 ¡ 26%) was used. After the addition of
0.5 mM of ST
14
S, signal increases of 47% and 36% were only
obtained for AP and NP, respectively, while signal decreases
were observed for PCP, DNOC, VA, and HBA. This behavior
is most likely due to the aggregation of surfactants on the
electrode surface CMC ST
14
S=2mM
88
), precluding the
interaction with some analytes, even before the oxidation
process. It is also worth noting that the largest signal enhance-
ments were observed for 4-aminophenol (SE
8
S = 40%, SD
10
S=
47%, SD
12
S = 73%, and ST
14
S = 47%) and the smallest
changes were seen for vanillic acid (SE
8
S = 50%, SD
10
S = 8%,
SD
12
S = 8%, and ST
14
S=233%). SD
10
S was selected as a
model surfactant and used in the remaining experiments because
it provided a significant increase in m
EOF
and signal magnitude,
while its CMC value (SD
10
SCMC=24mM)
89
would allow
its use in micellar electrokinetic chromatography (MEKC).
Effect of surfactant concentration on the migration time and
peak current
The effect of surfactant concentration was studied in the range
of 0 (no surfactant added) to 2 mM using SD
10
S. This range
ensures that micelles will not be formed (SD
10
S CMC =
24 mM)
89
and that all the changes in migration time can only
be attributed to changes in m
EOF
. As shown in Fig. 2(a), the
addition of SD
10
S produced significant increases in m
EOF
without affecting the electrophoretic mobility (m
EP
) or the
migration order of the compounds. The immediate conse-
quence of increasing the m
EOF
without affecting the m
EP
is a
significant decrease in analysis time. As can be also observed,
the change in m
EOF
is more significant at lower concentrations
(,0.1 mM). No significant improvements were obtained when
surfactant concentrations greater than 1 mM were used,
suggesting that the surface is saturated around this concentra-
tion. These results are in good agreement with those previously
published for other surfactants under similar condi-
tions.
30,31,52,54
They also suggest that at concentrations below
the CMC, surfactants may adsorb to PDMS by (mainly)
hydrophobic interactions.
31
The conformations adopted on
the surface cannot, however, be determined from the present
experiments.
A characteristic feature of surfactant adsorption at con-
centrations below the CMC is the formation of local
aggregates on electrode surfaces.
90
Because surfactant con-
centration is one of the most important parameters affecting
the aggregation process (and hence the oxidation of phenolic
compounds), the effect of SD
10
S concentration on the
electrochemical signal was investigated. As can be observed
in Fig. 2(b), the addition of increasing amounts of SD
10
S
produced two significantly different effects, which depended
Table 1 Peak current (nA) of the selected phenolic compounds as a function of the surfactant structure (at 0.5 mM each). Conditions: 5 mM
borate, pH = 11.0, E
SEP
= 1200 V, t
INJ
=10s,E
DET
= 0.6 V, 20 mM AP, 13 mM PCP, 15 mM DNOC, 37 mM NP, 25 mM VA, 37 mM HBA. Results
expressed as average and standard deviation of at least 5 measurements
AP PCP DNOC NP VA HBA
BGE 1.5 ¡ 0.1 1.2 ¡ 0.3 1.9 ¡ 0.2 1.4 ¡ 0.1 2.4 ¡ 0.1 3.0 ¡ 0.1
SE
8
S 2.1 ¡ 0.1 1.8 ¡ 0.4 2.8 ¡ 0.4 1.8 ¡ 0.2 3.6 ¡ 0.8 4.4 ¡ 0.1
SD
10
S 2.2 ¡ 0.1 1.7 ¡ 0.4 2.3 ¡ 0.4 1.8 ¡ 0.1 2.6 ¡ 0.2 3.4 ¡ 0.1
SD
12
S 2.6 ¡ 0.2 1.5 ¡ 0.3 2.0 ¡ 0.1 2.0 ¡ 0.1 2.6 ¡ 0.4 3.3 ¡ 0.2
ST
14
S 2.2 ¡ 0.3 1.0 ¡ 0.2 1.1 ¡ 0.1 1.9 ¡ 0.1 1.6 ¡ 0.3 2.3 ¡ 0.3
Fig. 2 (a): Effect of the SD
10
S concentration on the electro-osmotic
(–w–) and electrophoretic mobility of 20 mMAP(–&–), 13 mM PCP
(–$–), 15 mM DNOC (–m–), 37 mMNP(–.–), 25 mMVA(–b–), and
37 mM HBA (–c–). Conditions: 5 mM borate, pH = 11.0, E
SEP
=
1200 V, t
INJ
=10s,E
DET
= 0.6 V. Plotted points and error bars
represent the average and standard deviation, respectively, of at least
5 measurements. Lines included to guide the eye. (b): Effect of the
SD
10
S concentration on the peak current of 20 mMAP(–&–), 13 mM
PCP (–$–), 15 mM DNOC (–m–), 37 mMNP(–.–), 25 mMVA(–b–),
and 37 mM HBA (–c–). Conditions: 5 mM borate, pH = 11.0, E
SEP
=
1200 V, t
INJ
=10s,E
DET
= 0.6 V. Plotted points and error bars
represent the average and standard deviation, respectively, of at least
5 measurements. Lines included to guide the eye.
1000 |
Analyst
, 2007, 132, 997–1004 This journal is
ß
The Royal Society of Chemistry 2007
on the phenolic compound. Signal increases were observed for
AP, PCP, and 2NP, while a signal decrease was observed for
DNOC, VA and HBA. In order to explain these results, a
series of contributions have to be considered. First, the
electrode potential strongly affects the adsorption to (and
conformation on) the electrode surface.
90–93
Second, phenolic
compounds must interact with electrode surface before the
electron-transfer reaction. Third, it is accepted that the
primary one-electron oxidation of phenol (and phenolates) to
its corresponding phenoxy radical is the rate-determining step
in the oxidation process.
94
And fourth, adsorption of
surfactants on electrodes may stabilize the oxidation products
and hence increase the electron transfer rate.
95
The increases in
signal observed for AP, 2NP, and PCP could be explained by
assuming that the PAD waveform is facilitating the formation
of surface aggregates (hydrophobic core and negative shell),
which enhances the interaction of the analytes with the
electrode surface. We believe that this interaction is more
favorable if the analytes have a surfactant-like conformation.
The higher the concentration of surfactants, the higher the
concentration of these aggregates and, therefore, the higher the
electrochemical response. For the same reason, because
DNOC, VA and HBA have substituents spread around the
phenolic ring, their interaction with the surface aggregates is
less favorable than with the bare gold surface, leading to a
decrease in peak current (in the presence of negatively-charged
surface aggregates). To prove this hypothesis, partial charges
and dipole moments were calculated for the selected phenolate
anions. As shown in Fig. 3, for the singly charged anions,
dipole moments increase in the following series: DNOC , AP
, PCP , NP. The partial charge distributions readily explain
the direction and magnitude of these dipole moments. Two of
these anions deserve comment. The dipole moment for DNOC
is attenuated somewhat due to rotation about the C–N bond
between the phenyl ring and the nitro group in the 6-position,
which disrupted conjugation of the p-system. The planar C
2v
structure, which is a transition state, is slightly higher in energy
(DE = +0.1 kcal mol
21
) and has a larger dipole moment (m =
3.74 D). In a similar manner, hydrogen bond formation
between the amino group and the oxygen atom in AP also
leads to a reduction in the phenolate’s dipole moment. From
these results, it is clear that NP, PCP, and AP bear a close
resemblance in terms of their charge distributions to the alkyl
sulfates: both possess (ionic) polar and non-polar groups, each
with substantial dipole moments. While DNOC also has a
fairly sizable dipole moment, the negatively charged substi-
tuents are arranged such that polar and non-polar regions are
not clearly delineated. The free energy of concentrating DNOC
within the alkyl sulfate aggregates should, therefore, be greater
than those for NP, PCP, and AP. In many respects, the charge
distributions for the dianions resemble that of DNOC. While
the negatively charged substituents on DNOC are arrayed in
the 1-,4-, and 6-positions, those on VA are found in the 1-, 2-,
and 5-positions. This substitution pattern makes the incor-
poration of VA within the alkyl sulfate aggregates more
difficult. HBA is the least polar of the six phenolates (m =
1.73 D). This is the result of the 1-,4-substitution pattern in
HBA, in which the local dipole moments created by the oxygen
Fig. 3 Calculated Mulliken partial charges (e
2
) and dipole moments (D) for six phenolates. Partial charges have been summed for functional
groups. Dipole moments in grey: arrow tail = positive; arrowhead = negative. Level of theory: HF/6-31+G(d). See text for details.
This journal is
ß
The Royal Society of Chemistry 2007
Analyst
, 2007, 132, 997–1004 | 1001
atom and carboxylate group largely cancel one another. This,
in turn, leads to a charge distribution unlike that found in the
alkyl sulfate aggregates. Upon consideration of the theoretical
charge densities for the six phenolates, a plausible explanation
for the experimentally observed correlation between peak
current and surfactant concentration may be proffered: a pre-
concentration effect is seen for those phenolates (NP, PCP,
and AP) capable of insinuating themselves into the alkyl
sulfate aggregates. For those species (HBA, VA, and DNOC)
whose charge distribution precludes such partitioning, a pre-
concentration effect is not observed. Experimental and com-
putational studies are currently being designed to rigorously
test this explanation, and the results from this work will be
reported in due course.
The fact that high electrode potentials allow the adsorp-
tion of greater amounts of surfactants
91,93
may also explain
why the addition of surfactants yields better results in
combination with PAD than with dc amperometry.
55,56,67,69
It is possible that, during the reconstruction step (20.5 V
for 0.025 s), all the aggregates (and oxidation products)
are removed from the surface, allowing the detection step
(0.6 V for 0.15 s) to be performed on a clean surface (surface–
surfactant without oxidation products). These experiments
91,93
could also explain why even low amounts of surfactant
(below CMC values) added to the background electrolyte
can produce a significant improvement in electrochemical
signal.
30,52,54,63
Effect of pH on the migration time and peak current
The pH of a 5 mM borate buffer with 0.5 mM SD
10
S was
adjusted from 9.2 to 12 in order to analyze the effect of pH on
the separation and response of phenols. As shown in Fig. 4(a),
as pH (and the dissociation of phenols) increases a significant
increase in m
EP
(toward more negative values) for the six
phenolic compounds was obtained. These results are in good
agreement with the corresponding pK
a
values (see below). As
was also expected, the m
EOF
was not significantly affected
because the surface charge is dominated by the number of
silanol groups (from the PDMS) and exposed sulfate ions
(from the SD
10
S).
49,50
Of still greater interest are changes in the
solution pH that produce significant changes in electrochemi-
cal response of the phenols. As can be observed in Fig. 4(b),
higher pH values produced increases in the peak current of
PCP (pK
a
= 4.74),
96
DNOC (pK
a
= 4.3),
96
and NP (pK
a
=
7.08).
96
Under the same conditions, the peak current of AP
(pK
a1
= 4.74, pK
a2
= 9.76),
97
VA (pK
a1
= 4.5, pK
a2
= 9.0),
98
and HBA (pK
a1
= 4.47, pK
a2
= 9.17)
99
reached a maximum
value at pH = 11.0. These results can be explained by
considering that alkaline conditions facilitate the oxidation of
both the analyte and the electrode surface.
52,67,69
It should also
be observed that the ionization of two different functional
groups (e.g., –OH and –COOH in VA and HBA) offers a
significant electrostatic barrier, which diminishes the analyte–
electrode interaction and leads to significant current decreases
(Fig. 4(b)). These experiments not only support the hypothesis
that the charge distribution of the analyte has a significant
influence on the electrochemical response when surfactants
are present; these experiments also allow a rational selection of
the pH to maximize the response, considering the pK
a
value of
the analytes.
Effect of surfactants in analytical performance
To demonstrate the advantages of using alkyl sulfate surfac-
tants on PDMS microchips, the calibration curves, limits of
detection and reproducibility of both migration time and peak
current for the six phenols were obtained. In these experi-
ments, results obtained with 0.5 mM SD
10
S were compared
with control experiments without surfactant. The results are
summarized in Table 2. The measured limit of detection
represents the lowest point in the calibration curve and was
defined as the lowest measured concentration with a signal-to-
noise ratio of at least 3. As can be observed, under the
optimum solution conditions (5 mM borate buffer, pH = 11.0,
0.5 mM SD
10
S), significant improvements in sensitivity and
limits of detection were obtained. The addition of SD
10
S also
Fig. 4 (a): Effect of the solution pH on the electro-osmotic (–w–) and
electrophoretic mobility of 20 mMAP(–&–), 13 mM PCP (–$–),
15 mM DNOC (–m–), 37 mMNP(–.–), 25 mMVA(–b–), and 37 mM
HBA (–c–). Conditions: 5 mM borate, 0.5 mM SD
10
S, E
SEP
= 1200 V,
t
INJ
=10s,E
DET
= 0.6 V. Plotted points and error bars represent the
average and standard deviation, respectively, of at least 5 measure-
ments. Lines included to guide the eye. (b): Effect of the solution pH
on the peak current of 20 mMAP(–&–), 13 mM PCP (–$–), 15 mM
DNOC (–m–), 37 mMNP(–.–), 25 mMVA(–b–), and 37 mM HBA
(–c–). Conditions: 5 mM borate, 0.5 mM SD
10
S, E
SEP
= 1200 V, t
INJ
=
10 s, E
DET
= 0.6 V. Plotted points and error bars represent the average
and standard deviation, respectively, of at least 5 measurements. Lines
included to guide the eye.
1002 |
Analyst
, 2007, 132, 997–1004 This journal is
ß
The Royal Society of Chemistry 2007
improved the reproducibility of the migration times (t
M
) and
peak currents (I
p
) of the phenolic compounds.
Conclusions
The analysis of phenolic compounds by microchip-CE and
PAD can be significantly improved upon addition of alkyl
sulfates. These surfactants can adsorb to the microchip
substrate (PDMS), increase the surface charge, and therefore
decrease the analysis time. Surfactants with longer alkyl chains
(at 0.5 mM) can, however, produce excessive increases in m
EOF
,
precluding the separation of the phenolic compounds on the
microchips. The second effect of the addition of surfactants to
the BGE is the improvement in the electrochemical detection
of phenols (under optimized conditions). Our results, which
are in accord with the formation of aggregates on the gold
electrode surface,
90,91,93
show that surfactants can enhance
the analyte–electrode interaction and, therefore, facilitate the
electron-transfer process. The current results should allow a
more rational selection of surfactants, via consideration of
their electrophoretic and electrochemical impacts.
Acknowledgements
Financial support to this project was provided by The
University of Texas at San Antonio.
References
1 J.-J. Xu, A.-J. Wang and H.-Y. Chen, TrAC, Trends Anal. Chem.,
2007, 26, 125–132.
2 V. Dolnik and S. Liu, J. Sep. Sci., 2005, 28, 1994–2009.
3 H. Becker and L. Locascio, Talanta, 2002, 56, 267–287.
4 G. S. Fiorini, G. D. M. Jeffries, D. S. W. Lim, C. L. Kuyper and
D. T. Chiu, Lab Chip, 2003, 3, 158–163.
5 A. Han, O. Wang, M. Graff, S. K. Mohanty, T. L. Edwards,
K.-H. Han and B. Frazier, Lab Chip, 2003, 150–157.
6 Y. Liu, D. Ganser, A. Schneider, R. Liu, P. Grodzinski and
N. Kroutchinina, Anal. Chem., 2001, 73, 4196–4201.
7 V. Dolnik, Electrophoresis, 2004, 25, 3589–3601.
8 O. Vogt, M. Pfister, U. Marggraf, A. Neyer, R. Hergenroder and
P. Jacob, Lab Chip, 2005, 5, 205–211.
9 M. Castano-Alvarez, M. T. Fernandez-Abedul and A. Costa-
Garcı´a, Electrophoresis, 2005, 26, 3160–3168.
10 J. Wang, M. Pumera, M. P. Chatrathi, A. Escarpa, R. Konrad,
A. Griebel, W. Dorner and H. Lowe, Electrophoresis, 2002, 23,
596–601.
11 N. Bogdanski, H. Schulz, M. Wissen, H. C. Scheer, J. Zajadacz and
K. Zimmer, Microelectron. Eng., 2004, 73, 190–195.
12 G. H. Seong, W. Zhan and R. M. Crooks, Anal. Chem., 2002, 74,
3372–3377.
13 D. C. Duffy, J. C. McDonald, O. J. A. Schueller and
G. M. Whitesides, Anal. Chem., 1998, 70, 4974–4984.
14 J. M. Ng, I. Gitlin, A. D. Stroock and G. M. Whitesides,
Electrophoresis, 2002, 23, 3461–3473.
15 G. Ocvirk, M. Munroe, T. Tang, R. Oleschuk, K. Westra and
D. J. Harrison, Electrophoresis, 2000, 21, 107–115.
16 M. Y. Badal, M. Wong, N. Chiem, H. Salimi-Moosavi and
D. J. Harrison, J. Chromatogr., A, 2002, 947, 277–286.
17 G. M. Janini, H. J. Issaq and G. M. Muschik, J. Chromatogr., A,
1997, 792, 125–141.
18 Y. Liu, J. C. Fanguy, J. M. Bledsoe and C. S. Henry, Anal. Chem.,
2000, 72, 5939–5944.
19 B. C. Giordano, E. R. Copeland and J. P. Landers, Electrophoresis,
2001, 22, 334–340.
20 C. Fujimoto, Electrophoresis, 2002, 23, 2929–2937.
21S.Hu,X.Ren,M.Bachman,C.E.Sims,G.P.Liand
N. Allbritton, Electrophoresis, 2003, 24, 3679–3688.
22S.Hu,X.Ren,M.Bachman,C.E.Sims,G.P.Liand
N. L. Allbritton, Langmuir, 2004, 20, 5569–5574.
23 G. T. Roman, T. Hlaus, K. J. Bass, T. G. Seelhammer and
C. T. Culbertson, Anal. Chem., 2005, 77, 1414–1422.
24 M. F. Mora and C. D. Garcia, Electrophoresis, 2007, 28,
1197–1203.
25 J. E. Melanson, N. E. Baryla and C. A. Lucy, Anal. Chem., 2000,
72, 4110–4114.
26 C. Wang and C. Lucy, Anal. Chem., 2005, 77, 2015–2021.
27 C. Wang and C. A. Lucy, Electrophoresis, 2004, 25, 825–832.
28 M. M. Yassine and C. A. Lucy, Anal. Chem., 2004, 76, 2983–2990.
29 M. M. Yassine and C. A. Lucy, Anal. Chem., 2005, 77, 620–625.
30 C. D. Garcia, B. M. Dressen, A. Henderson and C. S. Henry,
Electrophoresis, 2005, 26, 703–709.
31 M. F. Mora, C. E. Giacomelli and C. D. Garcia, Anal. Chem.,
2007, DOI: 10.1021/ac070953g.
32 S. Terabe, Anal. Chem., 2004, 76, 240A–246A.
33 A. Mohanty and J. Dey, J. Chromatogr., A, 2005, 1070, 185–192.
34 A. V. Jager, F. G. Tonin and M. F. Tavares, J. Sep. Sci., 2005, 28,
957–965.
35 Z. Deyl, I. Miksik and A. Eckhardt, J. Chromatogr., A, 2003, 990,
153–158.
36 N. E. Baryla, J. E. Melanson, M. T. McDermott and C. A. Lucy,
Anal. Chem., 2001, 73, 4558–4565.
37 A. E. Nassar, S. V. Lucas, W. R. Jones and L. D. Hoffland, Anal.
Chem., 1998, 70, 1085–1091.
38 E. Orejuela and M. Silva, Analyst, 2002, 127, 1433–1439.
39 T. E. Rosso and P. C. Bossle, J. Chromatogr., A, 1998, 824,
125–134.
40 J. Seo and L. P. Lee, Sens. Actuators, B, 2006, 119, 192–198.
41 W. Wei and H. Ju, Electrophoresis, 2005, 26, 586–592.
42 E. A. Doherty, R. J. Meagher, M. N. Albarghouthi and
A. E. Barron, Electrophoresis, 2003, 24, 34–54.
Table 2 Analytical parameters (linear range, sensitivity, and R
2
) obtained in the absence of surfactants and the presence of 0.5 mM SD
10
S.
Reproducibility in migration time and signal magnitude are informed as the RSD% of 10 consecutive injections. Conditions: 5 mM borate, pH =
11.0, E
SEP
= 1200 V, t
INJ
=10s,E
DET
= 0.6 V
Compound BGE t
M
RSD% Ip RSD% Linear range (mM) Sensitivity (nA/mM) R
2
AP No surfactant 3.2 8.4 2.5–10.1 128.3 0.95
0.5 mM SD
10
S 2.0 5.9 1.0–16.1 204.7 0.99
PCP No surfactant 4.2 3.6 1.6–6.5 211.2 0.93
0.5 mM SD
10
S 1.9 8.4 0.6–10.4 210.8 0.99
DNOC No surfactant 4.3 2.7 1.9–7.5 364.7 0.98
0.5 mM SD
10
S 2.0 3.9 0.7–12.0 390.7 0.99
NP No surfactant 4.9 4.6 4.6–18.6 64.3 0.95
0.5 mM SD
10
S 2.3 1.8 1.8–29.7 93.4 0.99
VA No surfactant 6.2 4.2 3.1–12.5 228.5 0.99
0.5 mM SD
10
S 2.9 4.1 1.2–20.0 255.2 0.99
HBA No surfactant 6.9 4.4 4.6–18.5 200.2 0.99
0.5 mM SD
10
S 3.2 3.2 1.8–29.6 251.7 0.98
This journal is
ß
The Royal Society of Chemistry 2007
Analyst
, 2007, 132, 997–1004 | 1003
43 H. Motschmann, R. Teppner, S. Bae, K. Haage and D. Wantke,
Colloid Polym. Sci., 2000, 278, 425–433.
44 R. Atkin, V. S. Craig, E. J. Wanless and S. Biggs, J. Colloid
Interface Sci., 2003, 266, 236–244.
45 O. J. Rojas, P. M. Claesson, K. D. Berglund and R. D. Tilton,
Langmuir, 2004, 20, 3221–3230.
46 E. Sabah, M. Turan and M. S. Celik, Water Res., 2002, 36,
3957–3964.
47 M. Chorro, C. Chorro, O. Dolladille, S. Partyka and R. Zana,
J. Colloid Interface Sci., 1999, 210, 134–143.
48 L. H. Torn, L. K. Koopal, A. d. Keizer and J. Lyklema, Langmuir,
2005, 21, 7768–7775.
49 B. J. Kirby and E. F. Hasselbrink, Jr., Electrophoresis, 2004, 25,
187–202.
50 B. J. Kirby and E. F. Hasselbrink, Jr., Electrophoresis, 2004, 25,
203–213.
51 A.-J. Wang, J.-J. Xu and H.-Y. Chen, Anal. Chim. Acta, 2006, 569,
188–194.
52 C. D. Garcia and C. S. Henry, Anal. Chim. Acta, 2004, 508, 1–9.
53 C. D. Garcia and C. S. Henry, Electroanalysis, 2005, 17, 223–229.
54 C. D. Garcia and C. S. Henry, Analyst, 2004, 129, 579–584.
55 R. W. Hompesch, C. D. Garcia, D. J. Weiss, J. M. Vivanco and
C. S. Henry, Analyst, 2005, 130, 694–700.
56 Y. Ding, M. F. Mora and C. D. Garcia, Anal. Chim. Acta, 2006,
561, 126–132.
57 P. G. Righetti, C. Gelfi, B. Verzola and L. Castelletti,
Electrophoresis, 2001, 22, 603–611.
58 M. Kawakami, K. Tanaka, N. Uriuda and S. Gondo,
Bioelectrochemistry, 2000, 52, 51.
59 X.-G. Wang, Q.-S. Wu, W.-Z. Liu and Y.-P. Ding, Electrochim.
Acta, 2006, 52, 589–594.
60 M. Ye, H. Zou, Z. Liu, J. Ni and Y. Zhang, J. Chromatogr., A,
1999, 855, 137–145.
61 R. Vittal, H. Gomathi and K.-J. Kim,
Adv. Colloid Interface Sci.,
2006, 119, 55–68.
62 R. A. Wallingford and A. G. Ewing, Anal. Chem., 1988, 60,
258–263.
63 C. D. Garcı´a, G. Engling, P. Herckes, J. L. Collett, Jr. and
C. S. Henry, Environ. Sci. Technol., 2005, 39, 618–623.
64 C. D. Garcia and C. S. Henry, Electroanalysis, 2005, 17,
1125–1131.
65 C. D. Garcia, Y. Liu, P. Anderson and C. S. Henry, Lab Chip,
2003, 3, 331–335.
66 Y. Ding, A. Ayon and C. D. Garcia, Anal. Chim. Acta, 2007, 584,
244–251.
67 Y. Ding and C. D. Garcia, Analyst, 2006, 131, 208–214.
68 Y. Ding and C. D. Garcia, Electrophoresis, 2006, 27, 5119–5127.
69 Y. Ding and C. D. Garcı´a, Electroanalysis, 2006, 18, 2202–2209.
70 M. F. Mora, Y. Ding, E. Mejia and C. D. Garcı´a, J. Capillary
Electrophor. Microchip Technol., 2007, 10, 7–18.
71 C. D. Garcia and C. S. Henry, Anal. Chem., 2003, 75, 4778–4783.
72 C. D. Garcı´a and C. S. Henry, in Microchip Capillary Electro-
phoresis: Methods and Protocols, ed. C. S. Henry, Humana Press,
Totowa, 2006, vol. 339, pp. 27–36.
73 C. C. J. Roothaan, Rev. Mod. Phys., 1951, 23, 69–89.
74 G. G. Hall, Proc. R. Soc. London, Ser. A, 1951, 205, 541–552.
75 J. S. Binkley, J. A. Pople and W. J. Hehre, J. Am. Chem. Soc.,
1980, 102, 939–947.
76 R. Ditchfield, W. J. Hehre and J. A. Pople, J. Chem. Phys., 1971,
54, 724–728.
77 W. J. Hehre, R. Ditchfield and J. A. Pople, J. Chem. Phys., 1972,
56, 2257–2261.
78 P. C. Hariharan and J. A. Pople, Theor. Chim. Acta, 1973, V28,
213–222.
79 G. W. Spitznagel, Diplomarbeit, Erlangen, 1982.
80 T. Clark, J. Chandrasekhar, G. W. Spitznagel and P. v. R. Schleyer,
J. Comput. Chem., 1983, 4, 294.
81 R. S. Mulliken, J. Chem. Phys., 1955, 23, 1833–1840.
82 R. S. Mulliken, J. Chem. Phys., 1955, 23, 1841–1846.
83 R. S. Mulliken, J. Chem. Phys., 1955, 23, 2338–2342.
84 R. S. Mulliken, J. Chem. Phys., 1955, 23, 2343–2346.
85 M. W. Schmidt, K. K. Baldridge, J. A. Boatz, S. T. Elbert, M. S.
Gordon, J. H. Jensen, S. Koseki, N. Matsunaga, K. A. Nguyen,
S. Su, T. L. Windus, M. Dupuis and J. J. A. Montgomery,
J. Comput. Chem., 1993, 14, 1347–1363.
86 R. Atkin, V. S. J. Craig, E. J. Wanless and S. Biggs, Adv. Colloid
Interface Sci., 2003, 103, 219–304.
87 S. Paria and K. C. Khilar, Adv. Colloid Interface Sci., 2004, 110,
75–95.
88 H. Gharibi and A. A. Rafati, Langmuir, 1998, 14, 2191–2196.
89 P. Sehgal, J. E. Mogensen and D. E. Otzen, Biochim. Biophys.
Acta, 2005, 1716, 59–68.
90 K. Hu and A. J. Bard, Langmuir, 1997, 13, 5418–5425.
91 I. Burgess, V. Zamlynny, G. Szymanski, J. Lipkowski, J. Majewski,
G. Smith, S. Satija and R. Ivkov, Langmuir, 2001, 17, 3355–3367.
92 P. Chandar, P. Somasundaran and N. J. Turro, J. Colloid Interface
Sci., 1987, 117, 31–46.
93 V. Zamlynny, I. Burgess, G. Szymanski, J. Lipkowski, J. Majewski,
G. Smith, S. Satija and R. Ivkov, Langmuir, 2000, 16, 9861–9870.
94 P. I. Iotov and S. V. Kalcheva, J. Electroanal. Chem., 1998, 442,
19–26.
95 X.-L. Wen, Y.-H. Jia and Z.-L. Liu, Talanta, 1999, 50, 1027–1033.
96 M. Cledera-Castro, A. Santos-Montes and R. Izquierdo-Hornillos,
J. Chromatogr., A, 2005, 1087, 57–63.
97 P. Forlay-Frick, D. Mangelings, T. Ivanyi, I. Lazar, K. Heberger
and Y. Van der Heyden, J. Pharm. Biomed. Anal., 2006, 41,
1164–1170.
98 T. Stalin and N. Rajendiran,
J. Mol. Struct., 2006, 794, 35–45.
99 H. Miklautz, D. Keller, F. L. Holguin and R. Woloszczak, Anal.
Bioanal. Chem., 2006, V384, 1191–1195.
1004 |
Analyst
, 2007, 132, 997–1004 This journal is
ß
The Royal Society of Chemistry 2007
