Carbohydrate Polymers 233 (2020) 115865

Contents lists available at ScienceDirect

Carbohydrate Polymers
journal homepage: www.elsevier.com/locate/carbpol

A surface functionalized with per-(6-amino-6-deoxy)-β-cyclodextrin for T

potential organic pollutant removal from water
Bárbara A. Herreraa,b,*, Tamara C. Brunaa, Rodrigo A. Sierpec, Erika P. Langd, Marcela Urzúaa,
Marcos I. Florese, Paul S. Jaraa,*, Nicolás I. Yutronica,*
a
Departamento de Química, Facultad de Ciencias, Universidad de Chile, Las Palmeras No 3425, Ñuñoa, CA. 7800003, Santiago, Chile
b
Departamento de Química, Facultad de Ciencias Naturales, Matemática y Medio Ambiente, Universidad Tecnológica Metropolitana, Las Palmeras No 3360, Ñuñoa, CA.
7800003, Santiago, Chile
c
Departamento de Toxicología y Farmacología, Facultad de Ciencias Químicas y Farmacéuticas, Universidad de Chile, Santos Dumont No 964, Independencia, CA.
8380494, Santiago, Chile
d
Departamento de Biología, Centro de Equipamiento Mayor (CEM), Facultad de Ciencias, Universidad de Chile, Las Palmeras No 3425, Ñuñoa, CA. 7800003, Santiago,
Chile
e
Departamento de Física, Facultad de Ciencias Físicas y Matemáticas, Universidad de Chile, Av. Blanco Encalada No 2008, CA. 8370415 Santiago, Chile

A R T I C LE I N FO A B S T R A C T

Keywords: In this work, we present a solid silicon substrate functionalized with modified β-cyclodextrin monolayers as an
Aminated β-cyclodextrin optimal surface for organic contaminant uptake. The inclusion and capture of three potential pollutants, 4-
Emerging pollutants chlorophenoxyacetic acid, 4-aminobenzoic acid and phenylethylamine, were studied. 1H-NMR and ROESY
4-Chlorophenoxyacetic acid studies revealed the complete inclusion and details of the conformational orientation of the three guests in the
Phenylethylamine
per-(6-amino-6-deoxy)-β-cyclodextrin matrix, forming three new inclusion complexes that have not yet been
4-Aminobenzoic acid
Pollutant removal
reported. Capture assays for the guests were carried out by immersing the substrates in an aqueous pollutant
solution and by measuring the UV–vis spectra. This substrate showed a high sorption capacity at equilibrium,
between 2.5 × 10−5 and 6.0 × 10−5 mmol/substrate, for the studied pollutants. In addition, this surface can be
reused four times with an efficiency equal to the initial use. Therefore, it could be a versatile platform that could
be applied for the capture of other organic pollutants from water.

1. Introduction Blachnio, Marczewski, Swiatkowski, & Buczek, 2017; Leopold, van

Schaik, & Neal, 1960).
Global pollutant accumulation, due to excessive use and/or in- Additionally, 4-aminobenzoic acid (4ABA) has been used for dia-
correct handling, creates critical environmental damage (Awad et al., zotization reactions in the synthesis of dyes and related applications
2019; Ivshina, Tyumina, & Vikhareva, 2018). In this way, the pre- such as dye degradation, which means that it may be an emerging
paration and evaluation of substrates capable of removing pollutants pollutant (Naik et al., 2018; Wang, Zhao, Qian, & Huang, 2018). Phe-
(e.g., organic herbicides, dyes, and drugs) is an area of interest for nylethylamine (PhEA) is an antidepressant and psychoactive stimulant
environmental research (Ahmad et al., 2010; Lu et al., 2019). that produces effects such as wakefulness, excitability and tachycardia.
The pollutants present in water that can be classified according to This drug can also be considered an emerging pollutant (Halberstadt,
their application or chemical nature. These include species that, be- van der Zee, Chatha, Geyer, & Powell, 2019; Sabelli, Fink, Fawcett, &
cause of their accumulation or recent unregulated use, are called Tom, 1996; Sierpe et al., 2015).
emerging pollutants (Geissen et al., 2015). Chlorophenoxy pesticides An interesting strategy for capturing organic pollutants in water is
are organic pollutants that can persist in the environment, con- the use of materials based on β-cyclodextrin (βCD) (Morin-Crini &
taminating surface water and groundwater due to their widespread use, Crini, 2013). Cyclodextrins are cyclic sugars obtained from enzymatic
resistance to biodegradation and moderate water solubility. The her- starch degradation (Del Valle, 2004). The hydrophobic character of the
bicide 4-chlorophenoxyacetic acid (4CPA) has higher water solubility cavity and the hydrophilic character of the outside allow for the in-
compared to similar compounds in its class (Derylo-Marczewska, clusion of organic and inorganic guest molecules within the cavity

⁎
Corresponding authors.
E-mail addresses: bahi@ug.uchile.cl, herrera@utem.cl (B.A. Herrera), pjara@uchile.cl (P.S. Jara), nyutroni@uchile.cl (N.I. Yutronic).

https://doi.org/10.1016/j.carbpol.2020.115865
Received 4 June 2019; Received in revised form 16 December 2019; Accepted 12 January 2020
Available online 13 January 2020
0144-8617/ © 2020 Elsevier Ltd. All rights reserved.
B.A. Herrera, et al. Carbohydrate Polymers 233 (2020) 115865

through intermolecular interactions (Crini, 2014; Del Valle, 2004). The 2.4. Determination of the association constant
internal diameter of native βCD is 7.5 Å (Crini, 2014), which is an
adequate dimension for the inclusion of organic pollutant molecules The association constants (Ka values) of βCDNH2@4CPA,
that contain aromatic rings as well as aliphatic chains. βCD-based βCDNH2@4ABA and βCDNH2@PhEA complexes in water were ob-
materials have been applied to remediate different organic pollutant tained using the phase solubility method (Connors, 1997; Higuchi &
molecules (Salazar, Guerra, Yutronic, & Jara, 2018; Schofield, Bain, & Connors, 1965). The measurements were conducted using a V-760
Badyal, 2012; Taka, Pillay, & Mbianda, 2017). JASCO UV spectrophotometer (see Supplementary Material, S3).
Assembled βCD monolayers (βCDMs) on surfaces represent an in-
teresting research topic in areas such as analyte detection because they 2.5. Preparation of substrates modified with the per-(6-amino-6-deoxy)-β-
offer a unique way to confine molecules in two dimensions by con- CD monolayer
trolling binding events at the interface (Lagrost, Alcaraz, Bergamini,
Fabre, & Serbanescu, 2007; Lu et al., 2019; Méndez-Torres, Sandoval- Silicon wafers (100) were cut to 1.5 cm × 1.7 cm. The small silicon
Altamirano, Sánchez-Arenillas, Marco, & Yáñez, 2018; Pan et al., 2015). wafers were immersed in a boiling solution of NH3:H2O2:H2O (1:1:5) at
The arrangement of βCDMs can act as a template for molecular re- 70 °C for 45 min to obtain an external oxide layer with a thickness of
cognition where the matrix-guest intermolecular interactions are re- 1 nm and an additional native oxide layer. The silicon wafers were
versible, which means that βCDMs are able to include and release the rinsed 3 times with water and dried under a N2 stream.
guest molecules (Ling, Reinhoudt, & Huskens, 2008; Méndez-Torres Substrate functionalization was performed following the method
et al., 2018; Niu et al., 2015; Onclin, Mulder, Huskens, Ravoo, & described by Onclin et al. (2004), with modifications. After forming the
Reinhoudt, 2004). oxidized surfaces, the silicon wafers were immersed in a 2 % APTES
Given the excellent properties of βCD in organic pollutant inclusion, solution in toluene at 70 °C for 20 s. Finally, the substrates were rinsed
in this work, we aimed to contribute to the study of materials useful in copiously with toluene and dried under a stream of N2.
removing contaminating agents from wastewater. We modified a silicon The incorporation of isothiocyanate-terminated layers using a DITC
surface with per-(6-amino-6-deoxy)-β-cyclodextrin (βCDNH2) to study solution and the subsequent incorporation of the βCDNH2 solution for
and evaluate its ability to capture organic pollutants from water, spe- the formation of βCDNH2M was performed as previously described
cifically an herbicide (4CPA), a dye precursor (4ABA) and a drug (Onclin et al., 2004).
(PhEA). βCDNH2 was synthesized following the procedure described by
Ashton, Königer, Stoddart, Alker, and Harding (1996)) (see Supple-
2. Materials and methods mentary Material, S1).

2.1. Reagents and solvents 2.6. Characterization of substrates modified with the per-(6-amino-6-
deoxy)-β-CD monolayer
For preparation of the inclusion complexes, per-(6-amino-6-deoxy)-
β-CD hydrochloride (βCDNH2, synthesized by our group, ≥98 % w/w; 2.6.1. Ellipsometric measurements
see Supplementary Material, S1), 4-chlorophenoxyacetic acid (4CPA, Using a vertical computer-controlled DRE-EL02 system (Ratzeburg,
PESTANAL®, ≥98.0 % w/w; Sigma-Aldrich), phenylethylamine (PhEA, Germany) in a room conditioned at 24 ± 1 °C, ellipsometric mea-
99.5 % w/v; Sigma-Aldrich) and 4-aminobenzoic acid (4ABA > 99 % surements were performed. The incidence angle (φ) was set to 70.0°
w/w; Merck) were used. The preparation of substrates modified with using a laser at a wavelength of 632.8 nm. A multilayer model based on
βCDNH2 was performed using 3-aminopropyl triethoxysilane (APTES, the Jones matrix calculation for the interpretation of ellipsometric data
≥99 % w/v; Sigma-Aldrich), 1,4-phenylene diisothiocyanate (DITC, was used (Silva, Urzúa, Petri, & Dubin, 2010). The expected thickness
≥98 % w/w; Sigma-Aldrich), toluene (LC grade; Merck), 25 % am- for each monolayer was determined using ChemDraw software, version
monia in water v/v solution (Merck), 30 % v/v hydrogen peroxide 15 (Perkin Elmer).
(Merck), and ultra-high purity water, which was purified using an SG- The thickness of the SiO2 layers was determined considering a Si
LaboStar 4-DI ultra-pure water system with a conductivity of 0.055 μS/ refractive index of n = 3.858 ± 0.0180 and infinite thickness in con-
cm. Silicon wafers (100) were purchased from Santa Clara, CA, USA. tact with air (n = 1.00). The refractive index used for each subsequent
Deuterated water (D2O, 99.9 % D atoms; Sigma-Aldrich) was used as stage was n = 1.462 for SiO2, n = 1.424 for APTES and n = 1.5 for
the solvent for the NMR studies. Absolute ethanol (≥99.8 % w/v, GC DITC and βCDNH2 (Palik, 1985). The results correspond to an average
grade; Sigma-Aldrich) was used as the solvent for the regeneration tests. of the measurements from 100 to 200 samples.

2.2. Formation of the inclusion complexes 2.6.2. Contact angle

The measurements were performed using a home-built apparatus
The preparation of the inclusion complexes was performed in that included a digital camera connected to a computer. Water droplets
deuterated water using βCDNH2 (7.0 mmolL−1) as the matrix molecule with volumes of 8 and 4 μL were used for the advancing angle (θa) and
and 4CPA, 4ABA and PhEA (7.0 mmolL−1) as guests. Each inclusion receding angle (θr), respectively. The reported data are the average of
complex was defined as βCDNH2@4CPA, βCDNH2@4ABA and fifteen samples for each stage.
βCDNH2@PhEA, respectively.
2.6.3. X-ray photoelectronic spectroscopy
2.3. Characterization by NMR A Perkin Elmer PHI 1257 scanning Auger microscope with a UHV
chamber, a hemispherical electron energy analyser, and an X-ray source
Inclusion complexes were characterized using proton nuclear mag- from the Kα radiation of an Al anode (hν = 1486.6 eV) was used.
netic resonance (1H-NMR) spectroscopy and rotating-frame nuclear During data acquisition, the pressure in the main chamber was main-
Overhauser effect (ROESY) spectroscopy (see Supplementary Material, tained at 10−7 Pa. The link energy scale was calibrated using a carbon
S2). 1H-NMR measurements were carried out at 300 K in D2O using a signal at 284.6 eV, with an accuracy of 70.1 eV.
Bruker Avance 400 MHz NMR spectrometer. ROESY spectra were ac-
quired using pulsed field gradient selection methods over a mixing time 2.6.4. Atomic force microscopy
of 12 h. The resonance signal at 4.700 ppm (TMS) was used as an in- Topographic images and surface roughness measurements were
ternal reference. obtained by atomic force microscopy (AFM, Omicron) in tapping mode

2
B.A. Herrera, et al. Carbohydrate Polymers 233 (2020) 115865

at room temperature under high-vacuum conditions. Silicon tips molecules are included in the matrix, generating an effect of shielding
(AC240TS-R3, Asylum Research) were used with a resonance frequency of their atomic nuclei. For the βCDNH2@PhEA complex, the protons of
of 70 kHz and a constant spring of 2 N/m (see Supplementary Material, the aromatic ring and ethyl chain are shifted towards lower fields,
S4) probably because the PhEA molecule is displaced towards the narrow
matrix edge where the NH2 groups are located, which are electron
density acceptors. This phenomenon is observed with higher intensity
2.7. Sorption of pollutants and reusability of the modified substrate
for the protons of the ethyl chain.
For the three inclusion complexes, there is a noticeable chemical
Pollutant sorption assays were performed at room temperature and
shift in the signals from the H3 and H5 protons of βCDNH2, mainly due
neutral pH. These studies were conducted by immersing six modified
to the inclusion of the contaminant molecules. The displacement to a
substrates in a 0.1 mmol/L initial solution. The variation in their con-
high or low field is due to the degree of interaction between the matrix
centration was recorded in the supernatant solution from 0 to 168 h
and the respective guest and the orientation they adopt within the
using a V-760 JASCO UV spectrophotometer (see Supplementary
matrix cavity. This observation is better clarified in two-dimensional
Material, S5).
resonance studies (see Fig. 2).
Reusability tests of the modified substrate were performed by im-
When analysing the integrals of the signals for the βCDNH2@4CPA,
mersion in a 0.1 mmol/L 4CPA solution. After each assay, the substrates
βCDNH2@4ABA and βCDNH2@PhEA complexes, a 1:1 stoichiometry
were rinsed with ethanol and left in an ultrasonic bath for 5 min to
(matrix:guest) was established (see Supplementary Material, S2.1).
eliminate the included contaminant.
Through ROESY, it was possible to determine the preferential or-
ientation that the guest molecules adopt inside the βCDNH2 cavity.
3. Results and discussion Fig. 2(a)-(c) shows the cross-peaks for the proton signals of greatest
interest for studying the interactions between each guest molecule and
3.1. Characterization of the new inclusion complexes in solution βCDNH2. A strong correlation between the aromatic proton signals of
the guest molecules with the H3 and H5 protons of βCDNH2 demon-
The βCDNH2@4CPA, βCDNH2@4ABA and βCDNH2@PhEA com- strates the inclusion of the aromatic rings. Complementary cross-peak
plexes were characterized in solution using 1H-NMR and ROESY spec- sections in the ROESY spectra are shown in the Supplementary Material
troscopy. 1H-NMR was used to measure the complexation shifts, i.e., the S2.2, which supports the proposed inclusion geometry for the
difference between the free and complexed resonance frequency (in βCDNH2@4CPA, βCDNH2@4ABA and βCDNH2@PhEA complexes.
ppm) for the same nucleus. ROESY was used to elucidate the arrange- βCD inclusion complexes with an average Ka of 490 Lmol−1 have
ment of guest molecules within the matrix (Schneider, Hacket, Rüdiger, been reported (Rao & Stella, 2003). Similar values for some pollutants
& Ikeda, 1998). and molecules containing aromatic rings have been observed (2017,
Fig. 1 shows the full 1H-NMR spectra of the βCDNH2 inclusion Asela et al., 2017; Flaherty et al., 2013; Sierpe et al., 2015). In this
complexes with organic pollutants and the respective proton assign- study, the values of Ka for the βCDNH2@4CPA, βCDNH2@PhEA and
ments. βCDNH2@4ABA complexes were 601, 491 and 40 Lmol−1, respectively
Proton chemical shifts for the individual species and in the inclusion (see Supplementary Material, S3). Higher values of Ka could indicate
complexes are shown below. Table 1 corresponds to the pollutants, and that the pollutants would be more difficult to remove from the sub-
Table 2 corresponds to βCDNH2. strate. However, the values reported for the three complexes support
The NMR study of complexes βCDNH2@4CPA and βCDNH2@4ABA the idea that matrices with βCDNH2M could be used as a strategy of
shows chemical shifts of the aromatic protons of the both guest mole- environmental remediation.
cules to higher fields, probably due to the spatial restriction when these

Fig. 1. 1H-NMR spectra (7.0 mmol/L in D2O) of (a) βCDNH2, (b) βCDNH2@4CPA, (c) βCDNH2@4ABA and (d) βCDNH2@PhEA complexes with the proton as-
signments for all species.

3
B.A. Herrera, et al. Carbohydrate Polymers 233 (2020) 115865

Table 1
1
H-NMR chemical shifts of the individual pollutants and in their respective inclusion complexes.
βCDNH2@4CPA complex βCDNH2@4ABA complex βCDNH2@PhEA complex

H assignments δ 4CPA δ complex Δδ (ppm) H assignments δ 4ABA δ complex Δδ (ppm) H assignments δ PhEA δ complex Δδ (ppm)
(ppm) (ppm) (ppm) (ppm) (ppm) (ppm)

H’2 4.607 4.531 −0.076 H’3/H’7 7.788 7.844 0.144 H’1 2.800 3.165 0.365
H’4/H’8 6.900 6.881 −0.019 H’4/H’6 6.848 6.922 0.056 H’2 2.686 2.942 0.256
H’5/H’7 7.299 7.292 −0.007 H’4/H’8 7.238 7.274 0.036
H’5/H’7 7.320 7.364 0.044
H’6 7.214 7.312 0.098

3.2. Step-by-step characterization of the solid substrate modified with per- each stage, XPS measurements were performed. Fig. 4(a) shows the
(6-amino-6-deoxy)-β-CD acquired full spectra, which were normalized according to the Si2p
peak. The intensity increases in the characteristic C1s, O1s, S2p and N1s
The surface modification stages were i) amino-terminated mono- peaks indicate the stacking of each molecular monolayer on the surface
layer formation, ii) isothiocyanate-terminated monolayer formation, up to the final formation of βCDNH2M. The analysis performed on the
and iii) βCDNH2 incorporation (see Fig. 3). Each stage was monitored oxide surface detected the presence of silicon, oxygen and carbon,
by ellipsometry, contact angle, XPS, and AFM. The results obtained by where the presence of carbon can be attributed to the XPS source. In the
ellipsometry and contact angle are shown in Table 3. following stages, the appearance of C, N and S signals is concordant
The SiO2 average thickness was 1.0 ± 0.1 nm for more than 200 with the expected composition for the incorporated molecular mono-
samples measured. For APTES monolayers, the average thickness was layers.
0.80 ± 0.05 nm, which corresponds to the expected thickness of the Fig. 4(b) corresponds to the curve fittings of each characteristic
APTES monolayer in a stretched and perpendicular conformation on the peak. In the first stage of the silicon wafer oxidation process, the
surface. The contact angles of 61° for the advance angle and 28° for the characteristic Si2p peaks are observed. The lower binding energy at
receding angle are in agreement with previously reported values, but 99.2 eV is associated with Si° from the bulk silicon wafer, the binding
they are lower than expected for an -NH2 terminal surface. This could energy at 100.4 eV corresponds to Si1+, and the higher binding energy
be related to the hydration of the surface caused by environmental at 103.2 eV corresponds to Si4+ from silicon dioxide (Araya et al., 2012;
humidity, transforming the amino-terminal groups into -NH3+ and in- Flores, Fuenzalida, & Häberle, 2005).
creasing the hydrophilic properties of the surface (Pasternack, Amy, & In the XPS spectrum for the C1s peak, a change in the characteristic
Chabal, 2008). peak during modification was observed (Fig. 4). The most intense peak
For the DITC monolayer, the optimum thickness should be ap- corresponded to the CeC bonds, which appeared at a binding energy of
proximately 1.1 nm, assuming an extended conformation. However, an 284.6 eV when the APTES monolayer was incorporated; this peak in-
average thickness of 0.60 nm was obtained. Therefore, a molecular in- creased to 284.8 eV in the subsequent substrate modifications. A second
clination angle of 33° was predicted. The increase in the advance and peak was identified at 285.8 eV, which was assigned to CeN bonds, and
receding angles after the incorporation of the DITC monolayer corre- their binding energies increased to 286.0 eV and 286.4 eV when
lates with the findings described by Manning, Leigh, Ramos, Preece, monolayers of DITC and βCDNH2 were incorporated, respectively. A
and Eritja (2010)), which attribute this increase to greater surface hy- third peak of lower intensity appeared at 288.0 eV in the APTES
drophobicity. monolayer formation and was attributed to CeO bonds (Thangaraju,
Finally, βCDNH2 was incorporated for βCDM formation. The pH Karthikeyan, Prakash, Babu, & Hayakawa, 2015). Then, this signal
must be adjusted with an ammonia solution since the addition of a slightly increased its binding energy to 288.2 eV, which could be at-
sodium hydroxide solution could favour the reduction of the -NCS tributed to the additional presence of thioamide bonds when the DITC
terminal group to an -NH2 group, which prevents thioamide bond for- and βCDNH2 monolayers were incorporated (Zhang, Zhou, Ding, Fu, &
mation (Lee, Guivarch, Van Dau, Tessier, & Krstulovic, 2003). Wang, 2017).
Table 3 shows a decrease in the contact angle values for βCDM as The curve fittings for the O1s peak show a peak of greater intensity
the surface becomes hydrophilic since the secondary hydroxyl groups of at 532.4 eV, which can be attributed mainly to silicon oxide. Finally,
the βCD molecules are exposed to the surface. The thickness determined the incorporation of βCDNH2M was confirmed by the presence of a
was 0.93 nm, which coincides with the length reported for the βCD peak at 533.4 eV, which was attributable to cyclic CeOeC bonds
unit, evidencing monolayer formation (Del Valle, 2004; Onclin et al., (Beanson & Briggs, 1992).
2004). The characteristic N1s peak at approximately 400.0 eV consisted of
To characterize the chemical composition and electronic state in three distinct peaks at 398 eV, 400 eV and 402 eV, which can be

Table 2
1
H-NMR chemical shifts of βCDNH2 in its free state and in the respective inclusion complexes.
H assignments Pure βCDNH2 βCDNH2@4CPA complex βCDNH2@4ABA complex βCDNH2@PhEA complex

δ βCDNH2 δ complex (ppm) Δδ δ complex(ppm) Δδ δ complex(ppm) Δδ

(ppm) (ppm) (ppm) (ppm)

H1 5.138 5.128 −0.010 5.143 −0.005 5.112 −0.026

H2 3.648 3.634 −0.014 3.652 0.004 3.626 −0.022
H3 3.955 3.905 −0.050 3.923 −0.032 3.901 −0.054
H4 3.556 3.550 −0.006 3.560 0.004 3.526 −0.030
H5 4.167 4.081 0.086 4.142 −0.025 4.116 −0.051
H6a 3.416 3.410 −0.006 3.424 0.008 3.378 −0.038
H6b 3.240 3.232 −0.008 3.240 0.000 3.198 −0.042

4
B.A. Herrera, et al. Carbohydrate Polymers 233 (2020) 115865

Fig. 2. Cross-peaks in the ROESY spectra for the aromatic protons of the guest molecules with the H3 and H5 protons of βCDNH2 and the proposed inclusion
geometry for the a) βCDNH2@4CPA, b) βCDNH2@4ABA and c) βCDNH2@PhEA complexes.

5
B.A. Herrera, et al. Carbohydrate Polymers 233 (2020) 115865

Fig. 3. Schematic representation of the βCDNH2M preparation on silicon surfaces.

Table 3 The sorption kinetics were investigated using two kinetic models,
Ellipsometric and contact angle measurements for each monolayer. comprising pseudo-first order (2) and pseudo-second order (3) equa-
Monolayer θa (°) θr (°) Experimental ellipsometric Expected
tions, expressed as:
thickness (nm) thickness (nm)
ln(qe − qt ) = ln (qe ) − k1 t (2)
-NH2 61 ± 1 28 ± 2 0.80 ± 0.05 0.80
-NCS 64 ± 2 38 ± 3 0.60 ± 0.10 1.10 t 1 t
= +
-βCD 49 ± 1 27 ± 2 0.93 ± 0.02 0.79 qt k2 qe2 qe (3)

where k1 and k2 are the pseudo-first and pseudo-second order velocity

assigned to the -NH2 group, NeC bonds and NH3+ group, respectively. constants, respectively, and qe and qt are the sorption capacities (mmol/
The signal broadening and increase in the binding energy of the peak at substrate) at equilibrium and at time t , respectively. Applying the
400 eV when the monolayers of DITC and βCDNH2 were incorporated pseudo-first order kinetic model, the values of k1 and qe can be calcu-
can be attributed to the formation of thioamide bonds (Graf et al., 2010; lated from the slope and the intercept of the line ln(qe − qt ) plotted
Tischer et al., 2012). against t . Likewise, k2 and qe can be determined from the intercept and
The doublet signal characteristic of S2p, which appeared at 168 eV, slope of the line t / qt plotted against t for the pseudo-second order
confirms the presence of sulfur in the thioamide bond formed after the model (Aly, Graulet, Scales, & Hanley, 2014; Ho & McKay, 1999).
incorporation of the molecular monolayer of DITC (Tischer et al., 2012; The three pollutants show similar sorption behaviours with two
Zhang et al., 2017). The signals found at 162 eV and 170 eV corre- sections of different slopes, as displayed in Fig. 6. The contact times in
sponded to the presence of a S species with sp2 hybridization and to the first section vary between 0 and 6 h for 4ABA and 0−10 h for PhEA
interactions between nearby sulfur atoms, respectively (Beanson & and 4CPA. The graph shows that the slope of the first section, de-
Briggs, 1992; Graf et al., 2010). pending on the sorbate, has values between 14 and 21 times larger than
Fig. 5 presents topographic AFM images representative of surfaces the second section before reaching equilibrium conditions (see Sup-
with SiO2, APTES, DITC and βCDNH2 monolayers. These monolayers plementary Material, S5). Finally, a plateau was observed after 144 h
have homogeneous topography with regular morphologies and root- for 4ABA and 120 h for PhEA and 4CPA, which suggests that the
mean-square (RMS) roughness ranging between 0.2 and 0.5 nm for all sorption capacity of the pollutants in equilibrium has been reached
stages (see Supplementary Material, S4). These micrographs comple- (Fig. 6).
ment the results obtained by ellipsometry, corroborating the prepara- Pseudo-first order kinetics are associated with an adsorption (phy-
tion of ordered and densely packed monolayers in each of the stages. sisorption) process, and pseudo-second order corresponds to an ab-
sorption process (chemisorption) (Zou et al., 2016). The results ob-
3.3. Uptake assays of the pollutants and regeneration tests of the modified tained by applying the two kinetic models over time until reaching the
substrates sorption capacity at equilibrium show that a linear fit for the pseudo-
second order kinetic behaviour is adequate (see Supplementary Mate-
The sorption capacity of the substrates in each assay was calculated rial, S5). The adsorption and absorption processes could occur si-
according to Eq. (1): multaneously. Physisorption is favoured by surface hydration due to the
hydrophilic nature of the substrates functionalized with βCDNH2M,
(C0 − Ct ) × V
qt = improving the interaction with pollutant molecules dissolved in water.
n (1)
The substrate can continue to capture pollutant molecules through in-
where C0 and Ct are the concentrations at time = 0 and t , respectively; clusion in the cavities of modified βCD after its accumulation on the
V is the volume of the respective pollutant solution; and n is the number surface.
of substrates used in each experiment. The sorption capacity at equili- Table 4 shows the results obtained by applying the kinetic models
brium, qe , was determined using an analogous equation, considering the for 4CPA sorption; the results for 4ABA and PhEA sorption are shown in
equilibrium concentration, Ce . Fig. 6 shows the increasing sorption the Supplementary Material, S5. The highest values of R2 and the best
capacity of each pollutant over time until equilibrium conditions were coincidence between the experimental and determined qe values con-
obtained. firm pseudo-second-order kinetic behaviour. The second-order model

6
B.A. Herrera, et al. Carbohydrate Polymers 233 (2020) 115865

Fig. 4. (a) Full XPS spectra acquired during all modification stages. (b) Curve fittings of each characteristic peak in the XPS spectra.

suggests that the rate-limiting step may be a supramolecular inclusion Table 4 shows a 4CPA experimental sorption capacity at equilibrium
process, demonstrating that the pollutant sorption behaviour on the of 4413 × 10−5 mmol/substrate; this value is 50.76 % higher than the
substrates modified with βCDNH2M is dominated by matrix-guest in- capture capacity of the unmodified substrate (see Supplementary Ma-
teractions (Triki, Tanazefti, & Kochkar, 2017; Yang et al., 2013; Zou terial, S5.1).
et al., 2016). Additionally, the presence of included 4CPA was corroborated by

7
B.A. Herrera, et al. Carbohydrate Polymers 233 (2020) 115865

Fig. 5. AFM micrographs of (a) an SiO2 surface, and monolayers terminated in (b) -NH2 (APTES), (c) -NCS (DITC), and (d) βCDNH2.

Table 4
Summary of results obtained from the kinetic models applied to 4CPA capture.
Kinetic model qe (exp.) qe (calc.) [mmol/ k1 [h−1] R2
[mmol/substrate] substrate]

Pseudo-first 4.413 × 10−5 3.513 × 10−5 2,662 × 10−2 0.892

order

Kinetic model qe (exp.) qe (calc.) [mmol/ k2 [substrate/ R2

[mmol/substrate] substrate] (mmol*h)]

Pseudo-second 4.413 × 10−5 4.238 × 10−5 3,359 × 103 0.980

order

XPS measurements (see Supplementary Material, S6).

Fig. 7 shows the reusability results for the capture of 4CPA after four
consecutive regeneration cycles measured every 24 h. The graph shows
that the substrates can operate with more than 90 % efficiency, with
100 % sorption capacity considered the maximum amount adsorbed
until 24 h. This is similar to other reported results on βCD-based sorbent
materials. This substrate does not leave a residue because it is a com-
pact material that can be easily removed from a water matrix (Alzate-
Sánchez, Smith, Alsbaiee, Hinestroza, & Dichtel, 2016; Flaherty et al.,
2013; Schofield et al., 2012).
Fig. 6. Sorption capacity over time in contact with (a) 4CPA, (b) 4ABA and (c)
PhEA in aqueous solution.
4. Conclusions

Silicon wafers functionalized with βCDNH2M were prepared suc-

cessfully, yielding stable, robust and homogeneous material for the

8
B.A. Herrera, et al. Carbohydrate Polymers 233 (2020) 115865

Fig. 7. (a) Graphical representation of the reusability test for 4CPA capture. (b) Schematic representation of 4CPA release after each cycle.

sorption of pollutants. The sorption capacity of this solid was evaluated Journal of Environmental Science and Health, Part C, 28, 231–271.
by performing assays with three potential pollutants: 4-chlorophenox- Aly, Z., Graulet, A., Scales, N., & Hanley, T. (2014). Removal of aluminium from aqueous
solutions using PAN-based adsorbents: Characterisation, kinetics, equilibrium and
yacetic acid, 4-aminobenzoic acid and phenylethylamine. This substrate thermodynamic studies. Environmental Science and Pollution Research, 21, 3972–3986.
showed a high sorption capacity at equilibrium between 2.5 × 10−5 Alzate-Sánchez, D. M., Smith, B. J., Alsbaiee, A., Hinestroza, J. P., & Dichtel, W. R.
and 6.0 × 10−5 mmol/substrate for the studied pollutants. This surface (2016). Cotton fabric functionalized with a β-cyclodextrin polymer captures organic
pollutants from contaminated air and water. Chemistry of Materials, 28, 8340–8346.
can be reused four times with an efficiency equal to the initial effi- Araya, M., Díaz-Droguett, D. E., Ribeiro, M., Albertin, K. F., Avila, J., Fuenzalida, V. M.,
ciency, being a versatile platform for the capture of other organic et al. (2012). Photoluminescence in silicon/silicon oxide films produced by the
pollutants from water and leaving no residue. Our studies show that the pulsed electron beam ablation technique. Journal of Non-Crystalline Solids, 358,
880–884.
kinetic behaviour in the sorption of pollutants conforms to a pseudo- Asela, I., Noyong, M., Simon, U., Andrades-Lagos, J., Campanini-Salinas, J., Vásquez-
second order model, coinciding with the results obtained for other Velásquez, D., et al. (2017). Gold nanoparticles stabilized with βcyclodextrin-2-
sorbent βCD-based materials. amino-4-(4-chlorophenyl)thiazole complex: A novel system for drug transport. PloS
One, 12, e0185652.
Ashton, P. R., Königer, R., Stoddart, J. F., Alker, D., & Harding, V. D. (1996). Amino acid
Author contributions derivatives of β-cyclodextrin. The Journal of Organic Chemistry, 61, 903–908.
Awad, A. M., Shaikh, S. M. R., Jalab, R., Gulied, M. H., Nasser, M. S., Benamor, A., et al.
Nicolás I Yutronic: Group leader, research design, Contribute re- (2019). Adsorption of organic pollutants by natural and modified clays: A compre-
hensive review. Separation and Purification Technology, 228, 115719.
search funds, analysis and discussion of results. Beanson, G., & Briggs, D. (1992). High resolution XPS of organic polymers: The scienta ESCA
Bárbara A. Herrera: Postdoctoral researcher, design and carrying 300 database. Chichester: Wiley Interscience.
out of experiments, bibliographic review, manuscript and figure Connors, K. A. (1997). The stability of cyclodextrin complexes in solution. Chemical
Reviews, 97, 1325–1358.
writing, analysis and discussion of results. Crini, G. (2014). Review: A history of cyclodextrins. Chemical Reviews, 114,
Marcela del P. Urzúa: Funding, ellipsometry measurements and 10940–10975.
contact angles. Del Valle, E. M. M. (2004). Cyclodextrins and their uses: A review. Process Biochemistry,
39, 1033–1046.
Erika P. Lang: NMR measurement and analysis. Derylo-Marczewska, A., Blachnio, M., Marczewski, A. W., Swiatkowski, A., & Buczek, B.
Marcos I. Flores: Measurement and analysis of XPS and AFM. (2017). Adsorption of chlorophenoxy pesticides on activated carbon with gradually
Rodrigo A. Sierpe: interpretation and analysis of results, design removed external particle layers. Chemical Engineering Journal, 308, 408–418.
Flaherty, R. J., Nshime, B., DeLaMarre, M., DeJong, S., Scott, P., & Lantz, A. W. (2013).
and construction of figures. Cyclodextrins as complexation and extraction agents for pesticides from con-
Tamara C. Bruna: Synthesis of systems and development of ex- taminated soil. Chemosphere, 91, 912–920.
periments, literature review. Flores, M., Fuenzalida, V., & Häberle, P. (2005). Thermal effects in the size distribution of
SiC nanodots on Si(111). Physica Status Solidi, 202, 1959–1966.
Paul S. Jara: research design; contribute research funds, analysis
Geissen, V., Mol, H., Klumpp, E., Umlauf, G., Nadal, M., van der Ploeg, M., et al. (2015).
and discussion of results and manuscript writing. Emerging pollutants in the environment: A challenge for water resource manage-
ment. International Soil and Water Conservation Research, 3, 57–65.
Acknowledgements Graf, N., Lippitz, A., Gross, T., Pippig, F., Holländer, A., & Unger, W. E. S. (2010).
Determination of accessible amino groups on surfaces by chemical derivatization
with 3,5-bis(trifluoromethyl)phenyl isothiocyanate and XPS/NEXAFS analysis.
The authors thank the funding granted by FONDECYT for projects Analytical and Bioanalytical Chemistry, 396, 725–738.
No 3140485, 3180706, 1160114, 1171611 and 1151221, Projects PIA Halberstadt, A. L., van der Zee, J. V. F., Chatha, M., Geyer, M. A., & Powell, S. B. (2019).
Chronic treatment with a metabotropic mGlu2/3 receptor agonist diminishes beha-
Anillo ACT-1412 and Millennium Nucleus Multifunctional Material for vioral response to a phenethylamine hallucinogen. Psychopharmacology, 236,
Surface Science Applications. T. Bruna thanks CONICYT for doctoral 821–830.
fellowship No 21120856. Higuchi, T., & Connors, K. A. (1965). Phase solubility techniques. In C. N. Reilly (Ed.).
Advances in analytical chemistry and instrumentation (pp. 117–212). New York, NY:
Wiley-Interscience.
Appendix A. Supplementary data Ho, Y. S., & McKay, G. (1999). Pseudo-second order model for sorption processes. Process
Biochemistry, 34, 451–465.
Ivshina, I., Tyumina, E., & Vikhareva, E. (2018). Biodegradation of emerging pollutants:
Supplementary material related to this article can be found, in the Focus on pharmaceuticals. Microbiology Australia, 39, 117–122.
online version, at doi:https://doi.org/10.1016/j.carbpol.2020.115865. Lagrost, C., Alcaraz, G., Bergamini, J.-F., Fabre, B., & Serbanescu, I. (2007).
Functionalization of silicon surfaces with Si–C linked β-cyclodextrin monolayers.
Chemical Communications, 10, 1050–1052.
References
Lee, C. R., Guivarch, F., Van Dau, C. N., Tessier, D., & Krstulovic, A. M. (2003).
Determination of polar alkylating agents as thiocyanate/isothiocyanate derivatives
Ahmad, T., Rafatullah, M., Ghazali, A., Sulaiman, O., Hashim, R., & Ahmad, A. (2010). by reaction headspace gas chromatography. The Analyst, 128, 857–863.
Removal of pesticides from water and wastewater by different adsorbents: A review. Leopold, A. C., van Schaik, P., & Neal, M. (1960). Molecular structure and herbicide

9
B.A. Herrera, et al. Carbohydrate Polymers 233 (2020) 115865

adsorption. Weeds, 8, 48–54. and cyclodextrin complexes. Chemical Reviews, 98, 1755–1786.
Ling, X. Y., Reinhoudt, D. N., & Huskens, J. (2008). Reversible attachment of nanos- Schofield, W. C. E., Bain, C. D., & Badyal, J. P. S. (2012). Cyclodextrin-functionalized
tructures at molecular printboards through supramolecular glue. Chemistry of hierarchical porous architectures for high-throughput capture and release of organic
Materials, 20, 3574–3578. pollutants from wastewater. Chemistry of Materials, 24, 1645–1653.
Lu, P., Cheng, J., Li, Y., Li, L., Wang, Q., & He, C. (2019). Novel porous β-cyclodextrin/ Sierpe, R., Lang, E., Jara, P., Guerrero, A. R., Chornik, B., Kogan, M. J., et al. (2015). Gold
pillar[5]arene copolymer for rapid removal of organic pollutants from water. nanoparticles interacting with β-cyclodextrin–phenylethylamine inclusion complex:
Carbohydrate Polymers, 216, 149–156. A ternary system for photothermal drug release. ACS Applied Materials & Interfaces, 7,
Manning, B., Leigh, S. J., Ramos, R., Preece, J. A., & Eritja, R. (2010). Fabrication of 15177–15188.
patterned surfaces by photolithographic exposure of DNA hairpins carrying a novel Sierpe, R., Noyong, M., Simon, U., Aguayo, D., Huerta, J., Kogan, M. J., et al. (2017).
photolabile group. Journal of Experimental Nanoscience, 5, 26–39. Construction of 6-thioguanine and 6-mercaptopurine carriers based on
Méndez-Torres, A. M., Sandoval-Altamirano, C., Sánchez-Arenillas, M., Marco, J. F., & βcyclodextrins and gold nanoparticles. Carbohydrate Polymers, 177, 22–31.
Yáñez, C. (2018). Amino β-cyclodextrins immobilized on gold surfaces: Effect of Silva, R. A., Urzúa, M. D., Petri, D. F. S., & Dubin, P. L. (2010). Protein adsorption onto
substituents on host-guest interactions. Electrochimica Acta, 282, 860–869. polyelectrolyte layers: Effects of protein hydrophobicity and charge anisotropy.
Morin-Crini, N., & Crini, G. (2013). Environmental applications of water-insoluble β-cy- Langmuir, 26, 14032–14038.
clodextrin–epichlorohydrin polymers. Progress in Polymer Science, 38, 344–368. Taka, A. L., Pillay, K., & Mbianda, X. Y. (2017). Nanosponge cyclodextrin polyurethanes
Naik, P., Abdellah, I. M., Abdel‐Shakour, M., Acharaya, M., Pilicode, N., El‐Shafei, A., and their modification with nanomaterials for the removal of pollutants from waste
et al. (2018). An efficient aniline‐based co‐sensitizer for high performance water: A review. Carbohydrate Polymers, 159, 94–107.
N3‐sensitized solar cells. ChemistrySelect, 3, 12297–12302. Thangaraju, D., Karthikeyan, R., Prakash, N., Babu, S. M., & Hayakawa, Y. (2015).
Niu, Y., Huang, T., Zhou, Z., Xu, G., Zhang, L., & Wei, T. (2015). Formation of cyclo- Growth and optical properties of Cu2ZnSnS4 decorated reduced graphene oxide na-
dextrin monolayer through a host–guest interaction with tailor-made phenyl- nocomposites. Dalton Transactions, 44, 15031–15041.
triethoxysilane self-assembled monolayer. Colloids and Surfaces A: Physicochemical Tischer, T., Goldmann, A. S., Linkert, K., Trouillet, V., Börner, H. G., & Barner-Kowollik,
and Engineering Aspects, 470, 224–229. C. (2012). Modular ligation of thioamide functional peptides onto solid cellulose
Onclin, S., Mulder, A., Huskens, J., Ravoo, B. J., & Reinhoudt, D. N. (2004). Molecular substrates. Advanced Functional Materials, 22, 3853–3864.
printboards: Monolayers of β-cyclodextrins on silicon oxide surfaces. Langmuir, 20, Triki, M., Tanazefti, H., & Kochkar, H. (2017). Design of β-cyclodextrin modified TiO2
5460–5466. nanotubes for the adsorption of Cu(II): Isotherms and kinetics study. Journal of Colloid
Palik, E. D. (1985). Handbook of optical constants of solids. London, UK: Academic Press. and Interface Science, 493, 77–84.
Pan, Y., Mu, N., Shao, S., Yang, L., Wang, W., Xie, X., et al. (2015). Selective surface Wang, Y.-Y., Zhao, X.-L., Qian, H.-F., & Huang, W. (2018). Mono/double 3-methox-
acoustic wave-based organophosphorus sensor employing a host-guest self-assembly ypropan-1-amine substituted pyridone azo dyes having isomeric ortho/para-amino-
monolayer of β-cyclodextrin derivative. Sensors, 15, 17916–17925. benzoic acids and corresponding methyl esters components. Dyes and Pigments, 153,
Pasternack, R. M., Amy, S. R., & Chabal, Y. J. (2008). Attachment of 3-(Aminopropyl) 44–52.
triethoxysilane on silicon oxide surfaces: Dependence on solution temperature. Yang, S., Zong, P., Hu, J., Sheng, G., Wang, Q., & Wang, X. (2013). Fabrication of β-
Langmuir, 24, 12963–12971. cyclodextrin conjugated magnetic HNT/iron oxide composite for high-efficient de-
Rao, V. M., & Stella, V. J. (2003). When can cyclodextrins be considered for solubilization contamination of U(VI). Chemical Engineering Journal, 214, 376–385.
purposes? Journal of Pharmaceutical Sciences, 92, 927–932. Zhang, G., Zhou, Y., Ding, Z., Fu, L., & Wang, S. (2017). Nanosilica-supported thiose-
Sabelli, H., Fink, P., Fawcett, J., & Tom, C. (1996). Sustained antidepressant effect of PEA micarbazide–glutaraldehyde polymer for selective Au(iii) removal from aqueous so-
replacement. The Journal of Neuropsychiatry and Clinical Neurosciences, 8, 168–171. lution. RSC Advances, 7, 55215–55223.
Salazar, S., Guerra, D., Yutronic, N., & Jara, P. (2018). Removal of aromatic chlorinated Zou, Y., Wang, X., Ai, Y., Liu, Y., Ji, Y., Wang, H., et al. (2016). β-Cyclodextrin modified
pesticides from aqueous solution using β-cyclodextrin polymers decorated with Fe3O4 graphitic carbon nitride for the removal of pollutants from aqueous solution:
nanoparticles. Polymers, 10, 1038–1057. Experimental and theoretical calculation study. Journal of Materials Chemistry A, 4,
Schneider, H.-J., Hacket, F., Rüdiger, V., & Ikeda, H. (1998). NMR studies of cyclodextrins 14170–14179.

10