Physicochemical characterization of organosilylated halloysite clay nanotubes

Accepted Manuscript Physicochemical characterization of organosilylated halloysite clay nanotubes Andreia F. Peixoto, Ana C. Fernandes, Clara Pereira, João Pires, Cristina Freire PII:

S1387-1811(15)00432-1

DOI:

10.1016/j.micromeso.2015.08.002

Reference:

MICMAT 7245

To appear in:

Microporous and Mesoporous Materials

Received Date: 15 September 2014 Revised Date:

12 July 2015

Accepted Date: 4 August 2015

Please cite this article as: A.F. Peixoto, A.C. Fernandes, C. Pereira, J. Pires, C. Freire, Physicochemical characterization of organosilylated halloysite clay nanotubes, Microporous and Mesoporous Materials (2015), doi: 10.1016/j.micromeso.2015.08.002. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Graphical Abstract

ACCEPTED MANUSCRIPT

Physicochemical characterization of organosilylated halloysite clay nanotubes

1

RI PT

Andreia F. Peixoto,1* Ana C. Fernandes,2 Clara Pereira,1 João Pires,2 Cristina Freire1*

REQUIMTE, Departamento de Química e Bioquímica, Faculdade de Ciências, Universidade do Porto, Rua do Campo Alegre s/n, 4169-007 Porto, Portugal

Centro de Química e Bioquímica, Faculdade de Ciências, Universidade de Lisboa, 1749-

SC

2

*

TE D

M AN U

016 Lisboa, Portugal

Corresponding authors. Tel: +351 220402590(87), E-mail address: [email protected] (C.

AC C

EP

Freire), [email protected] (A. F. Peixoto)

1

ACCEPTED MANUSCRIPT

Abstract Halloysite nanotubes (HNTs) were functionalized with several organosilanes with different functional groups, by a post-grafting methodology, in aprotic and anhydrous conditions: 3-

(AEAPTMS),

(APTES),

N-2-aminoethyl-3-aminopropyltrimethoxysilane

(3-mercaptopropyl)trimethoxysilane

(MPTMS),

(3-bromopropyl)-

RI PT

aminopropyltriethoxysilane

trimethoxysilane (BrTMS), vinyltrimethoxysilane (VTMS) and phenyltriethoxysilane (PhTES).

SC

The pristine and silylated clay minerals were characterized by transmission and scanning electron microscopy, energy-dispersive X-ray spectroscopy, powder X-ray diffraction,

M AN U

nitrogen adsorption-desorption isotherms at -196 ºC, bulk elemental analysis, X-ray photoelectron spectroscopy, thermogravimetry, Fourier transform infrared spectroscopyattenuated total reflectance and 13C, 29Si and 27Al solid-state nuclear magnetic resonance. The techniques identified pristine HNTs as halloysite-7Å (dehydrated form) and proved their

TE D

successful silylation without the disruption of the nanotubes structure. The silylated HNTs showed bulk Si and C contents up to 7.30 and 1.92 mmol/g, respectively, with the APTES functionalized material containing the highest bulk and surface Si and C loadings, confirming

EP

its highest silylation efficiency. Some insights into the silylation reaction and mechanism 29

Si and

27

Al MAS NMR and

AC C

were also provided by the techniques used. Combination of

XRD data suggested that silylation reaction occurred at Al–OH groups from the inner lumen surface, as well as the Al-OH and Si-OH groups at the edges or external surface defects; no evidence was found for the existence of functionalization in the interlayer Al–OH groups. The silylation mechanism was found to proceed through reaction of the alkoxy moieties from the organosilane with the referred surface groups from the HNTs in a 3-fold (for VTMS and BrTMS), or 2-fold covalent grafting (for MPTMS) or a mixture of both approaches (for APTES, AEAPTMS and PhTES); in the case of APTES- and AEAPTMS-functionalized 2

ACCEPTED MANUSCRIPT HNTs, a polymerization side-reaction was also evidenced, as a parallel functionalization

RI PT

pathway.

AC C

EP

TE D

M AN U

SC

Keywords: halloysite, organosilane, silylation, clay minerals, post-grafting

3

ACCEPTED MANUSCRIPT

1.

Introduction Halloysite nanotubes (HNTs) with Al2Si2O5(OH)4•nH2O composition are a naturally

occurring two-layered (1:1) aluminosilicate, chemically similar to kaolinite, with a predominantly hollow tubular structure in the submicrometer range.1 Halloysite occurs mainly

RI PT

as two types of polymorphs: the anhydrous form, with an interlayer spacing of 7 Å, and the hydrated form, with expanded interlayer spacing of 10 Å, as a result of the incorporation of water in the interlamellar space. Halloysite is a cost-effective material that can be mined from

SC

deposits as a raw mineral and its nanosized dimensions and tubular arrangement depend on the region deposits: the tubes length typically varies from 500 to 1000 nm and the inner

M AN U

diameter is in the range of 15–100 nm.2,3,4 Hydroxyl groups are present as Al–OH groups at the inner surface, wall interlayers and edges, and as Si–OH groups at the edges and external surface defects of the material. Alumina innermost and silica outermost surfaces allow for HNTs inner/outer chemical modification, resulting in halloysite-based materials with

TE D

enhanced properties for applications as nanosized supports for chemical species5,6,7 as adsorbents,3,8 as drug-delivery systems,9,10 as nanofillers for polymers,11 as control release agents12 and as corrosion inhibitors.13,14 In fact, despite the considerable interest in a range of

EP

other synthetic nanotubular materials such as carbon nanotubes (CNTs) and boron nitride

AC C

nanotubes (BNNTs), recent works have highlighted the importance of the use of HNTs in catalysis, either as support of active phases or as an intrinsic catalyst. Organosilylation is a well-known method to selectively modify clay mineral surfaces efficiently and has already been used in the functionalization of pristine halloysite, but using a limited number of bifunctional organosilanes (typically containing amine and thiol groups).4,15 The covalent bonding between the organic functionalities of the organosilane and the hydroxyl groups of the clay mineral not only allows fine tuning the clay surface

4

ACCEPTED MANUSCRIPT chemistry, but also enables a robust immobilization of the organic moieties, preventing their leaching.16 Yuan et al. investigated the surface modification of different types of natural HNTs with 3-aminopropyltriethoxysilane (APTES) and concluded that the extension of amino-

RI PT

functionalization was strongly affected by the morphological parameters of the pristine halloysite (nanotubes length, outer and inner diameters and wall thickness), since they lead to differences on the amount of available hydroxyl groups; afterwards, the authors used the

SC

functionalized HNTs in the fabrication of nanocomposites, enzyme immobilization and controlled release of guest molecules.4,17 The same organosilane APTES was also used to

M AN U

chemically modify HNTs with N-2-pyridylsuccinamic acid (PSA) to produce a nanoadsorbent for the selective solid-phase extraction of Pb(II).18 An APTES-functionalized halloysite was also tested as host material for model dye (acid orange II) loading and controlled release: the study showed that the HNTs functionalization occurred at the internal

TE D

lumen surface and that the pH acted as external trigger for controlling the guest upload and subsequent controlled release.17 In a distinct work, the organosilane N-2-aminoethyl-3aminopropyltrimethoxysilane (AEAPTMS) was used in the functionalization of HNTs to

EP

allow immobilizing Au nanoparticles to be subsequently used as a surface-enhanced Raman scattering substrate, showing a remarkable enhancement effect due to the synergy between the

AC C

Au nanoparticles and the HNTs.19 In another application, HNTs functionalized with bis(triethoxysilylpropyl)-tetrasulphide (TESPT) were used to improve the dispersion and physical properties of a natural rubber matrix.1 Functionalized HNTs have also been used as supports for catalytic active species (metal complexes and metallic nanoparticles), although not in so great extent as other clay minerals, such as K10-montmorillonite.20 Calcined kaolinite and halloysite were used as supports for an anionic iron(III) porphyrin, which were subsequently used as heterogeneous catalysts in the oxidation of cyclooctene, cyclohexane 5

ACCEPTED MANUSCRIPT and n-heptane using iodosylbenzene as oxidant, giving expressive product yields for all the substrates.21 The complex CuBr-AEAPTMS supported in HNTs proved to mediate the living polymerization of methylmethacrylate showing increased conversion, relative low polydispersion and linear first-order rate plots.22 Ag nanoparticles were immobilized into

RI PT

AEAPTMS-functionalized HNTs and the resulting material showed high adsorption capability and photocatalytic activity towards methylene blue degradation.6 Pt nanoparticles were uniformly immobilized into APTES functionalized HNTs and were subsequently tested

SC

in the catalytic hydrogenation of styrene to ethylbenzene with 100% substrate conversion after 180 min; however, the catalytic activity slowly decreased with the increasing number of

M AN U

reaction cycles because of the low recovery of the catalyst.23 Finally, raw HNTs were used as intrinsic acid catalysts in methylic and ethylic esterifications of lauric acid with substrate conversions of 95.0 % and 87.1 % for the methylic and ethylic esterifications, respectively.24 In the present work, pristine HNTs were functionalized with several organosilanes with

TE D

different functional groups (-NH2, -NH-(CH2)2-NH2, -SH, -Br, -C6H6, -C=C) in aprotic and anhydrous media, Scheme 1, with the ultimate goal of being used as linkers or reactive groups for the immobilization of metal complexes/other active phases (metal and metal oxide

EP

nanoparticles) with catalytic activity. Furthermore, some of the introduced groups can be chemically modified in a subsequent step in order to impart intrinsic catalytic properties to

AC C

HNTs; the catalytic applications of the silylated HNTs will be the subject of a forthcoming paper. The influence of the type of organosilane on the silylation efficiency, amount of anchored organosilane, type of chemical bond and textural properties of the resulting functionalized HNTs were evaluated. Finally, with the selected group of organosilanes, we endeavor to get insights into the silylation reaction of HNTs and corresponding mechanism in order to establish correlations between HNTs structure/organosilane properties/organosilane loading/ functionalization mechanism. To the best of our knowledge, a complete study of the 6

ACCEPTED MANUSCRIPT silylation of this type of clay material with different types of organosilanes and the establishment of the aforementioned relations is lacking. In this context, this work provides relevant insights that will allow tuning the surface chemistry of HNTs and broadening the

RI PT

spectrum of applications through the selection of the most suitable organosilane.

Scheme 1

Experimental section

2.1

Materials

SC

2.

M AN U

All of the reagents and solvents used were used without further purification. HNTs were commercially obtained from Sigma-Aldrich and were used upon drying at 110 °C under vacuum for 12 h. Anhydrous toluene (99.8%, Sigma-Aldrich) was used as solvent and different types of organosilanes were used: (3-mercaptopropyl)trimethoxysilane (MPTMS, Fluka),

(3-bromopropyl)trimethoxysilane

vinyltrimethoxysilane

TE D

≥97%,

Dynasylan

VTMO

(VTMS,

(BrTMS,

≥97%,

Fluka),

Evonik

Degussa

GmbH),

phenyltriethoxysilane (PhTES, 98%, Sigma-Aldrich), APTES (99%, Sigma-Aldrich) and

2.2

AC C

EP

AEAPTMS Dynasylan DAMO ( Evonik Degussa GmbH).

HNTs functionalization with organosilanes The organosilylation of halloysite was performed using adapted published procedures25:

typically, 2 g of HNT (previous dried at 110° C under vacuum for 12 h) were dispersed in anhydrous toluene (100 cm3) and upon the addition of the organosilane (1.5 mmol), the suspension was kept in reflux under inert atmosphere (Ar) for 24 hours. After this period, the functionalized HNTs were filtered through 0.2 µm polyamide membrane filters (NL16 Whatman), rinsed with CH3CN, transferred to a round bottom flask with clean CH3CN (200 7

ACCEPTED MANUSCRIPT cm3) and washed by reflux for 1 h. Finally, the functionalized HNTs were filtered as previously described and were dried in an oven at 120 °C for 12 h. The materials will be denoted as HNTs_organosilane, where organosilane = MPTMS, BrTMS, VTMS, PhTES,

2.3

Physicochemical characterization

RI PT

APTES and AEAPTMS, Scheme 1.

Transmission electron microscopy (TEM) was performed at the Histology and Electron

SC

Microscopy Service, Instituto de Biologia Molecular e Celular (IBMC), Porto, Portugal, using a JEOL JEM 1400 electron microscope (Tokyo, Japan), equipped with an acquisition camera

M AN U

Gatan SC 1000 ORIUS CCD (Warrendale, PA, USA) operating at an accelerating voltage of 80 kV. The samples were dispersed in high-purity ethanol under sonication, after which a holey carbon film-coated 400 mesh copper grid was immersed in the suspension and then airdried.

TE D

Scanning electron microscopy (SEM) and energy-dispersive X-ray spectroscopy (EDS) studies were performed at “Centro de Materiais da Universidade do Porto” (CEMUP), Porto, Portugal, using a high-resolution environmental scanning electron microscope (FEI Quanta

EP

400 FEG ESEM) equipped with an energy-dispersive X-ray spectrometer (EDAX Genesis X4M). The samples were analyzed as powders, both uncoated and coated with a thin gold-

AC C

palladium film.

Powder X-ray diffraction (XRD) measurements were performed at room temperature with a Siemens D5000 diffractometer using Cu Kα radiation (λ = 1.5406 Å) and Bragg– Brentano θ/2θ configuration at “Unidade de Microscopia Eletrónica”, Universidade de Trásos-Montes e Alto Douro, Vila Real, Portugal. The measurements were performed over the 2θ range of 10–80º at a scan rate of 0.02° 2θ/s.

8

ACCEPTED MANUSCRIPT Nitrogen adsorption-desorption isotherms at -196 ºC were measured in an automatic apparatus Nova 2200e, Quantachrome. Before the adsorption experiments the samples were degassed at 120 ºC under vacuum during 2.5 h to avoid decomposition of organic groups and

Brunauer-Emmett-Teller (BET) model.

RI PT

at a pressure lower than 0.133 Pa. The specific surface areas were calculated from the

Elemental analyses (EA) – Si, C, N and S – were performed at “Laboratório de Análises do Instituto Superior Técnico”, Lisboa, Portugal.

SC

X-ray photoelectron spectroscopy (XPS) was performed at “Centro de Materiais da Universidade do Porto” (CEMUP), Porto, Portugal, using a Kratos AXIS Ultra HSA, with

M AN U

VISION software for data acquisition and CASAXPS software for data analysis. The analysis was carried out with a monochromatic Al Kα X-ray source (1486.7 eV), operating at 15 kV (90 W), in FAT mode (Fixed Analyzer Transmission), with a pass energy of 40 eV for regions ROI and 80 eV for survey. The powdered samples were pressed into pellets prior to the XPS

TE D

studies. Data acquisition was performed with a pressure lower than 1×10-6 Pa, and a charge neutralization system was used. The surface atomic percentages were calculated from the corresponding peak areas and using the sensitivity factors provided by the manufacturer. To

EP

correct possible deviations caused by electric charge of the samples, the C1s band at 285.0 eV

AC C

was taken as internal standard.

The thermogravimetric (TG) measurements were obtained in a STA 409 PC device from NETZSCH. Typically, 10–15 mg of the samples were heated under nitrogen flow (200 cm3/min) to 700 °C at a rate of 5 °C/min. The Fourier transform infrared spectroscopy-attenuated total reflectance (FTIR-ATR) spectra were obtained on a Perkin-Elmer Spectrum 100 Series spectrophotometer equipped with an ATR accessory (diamond/ZnSe) in the range of 650–4000 cm-1 at the Centre for Nanotechnology and Smart Materials (CENTI), V. N. de Famalicão, Portugal. 9

ACCEPTED MANUSCRIPT Solid-state nuclear magnetic resonance, spinning),

29

Si MAS (magic angle spinning),

13

C CPMAS (cross-polarization magic angle

29

Si CPMAS,

27

Al MAS, were recorded at

“Departamento de Química da Universidade de Aveiro”, Aveiro Portugal, on a 9.4 T Bruker

reference) and 14 KHz for aluminium (Al(NO3)3 reference).

Results and Discussion

3.1

Morphology and textural properties

SC

3.

RI PT

Avance 400 spectrometer at a spinning rate of 9 kHz for carbon, 5 kHz for silicon (TMS

TEM images presented in Figure 1, Panel A) reveal that the pristine halloysite is

M AN U

composed by cylindrical-shaped tubes with multilayer walls and open-ended lumen. The sample contains agglomerates of nanotubes with some irregularity in diameter, wall thickness and length. The tubes have an external diameter in the range of 60−100 nm and an inner diameter of 20−40 nm. Overall, the morphological parameters of the halloysite sample used in

TE D

this work are similar to those of previously reported samples.4 Upon HNTs silylation with the organosilanes, exemplified in Figure 1, Panel A) for HNT_APTES and HNT_MPTMS, the characteristic tubular morphology of the original clay mineral is retained. The SEM images,

EP

Figure 1, Panel B), confirm the tubular structure of the pristine halloysite, APTES- and MPTMS-functionalized HNTs and the presence of agglomerates. The corresponding EDX

AC C

spectra indicate that the pristine HNTs are mainly composed of Al, Si, and O elements, and the HNTs_APTES- and HNTs_MPTMS materials contain, besides Al, Si, and O elements from the HNTs intrinsic composition, the target elements from the organosilanes used in the functionalization: N and S for APTES- and MPTMS-functionalized HNTs, respectively.

Figure 1 – a)TEM & b) SEM/EDS

10

ACCEPTED MANUSCRIPT The X-ray diffractograms of the pristine and silylated-HNTs are shown in Figure S1, Supplementary Information, and the corresponding X-ray data are summarized in Table S1, Supplementary Information. The original HNTs XRD pattern exhibits a main diffraction peak at 2θ =11.97º assigned to the (001) reflection, corresponding to a basal spacing of 7.40 Å,

RI PT

determined using Bragg’s law, which identifies it as being halloysite-7Å (dehydrated form); the peak at 2θ = 8.5º, corresponding to halloysite-10Å was not observed.4 The diffractograms of all functionalized HNTs are very similar to that of the pristine HNTs, showing the

SC

diffraction peak around 2θ = 12.0º (basal spacing around 7.40 Å) typical of halloysite-7Å, which indicates that the HNTs structure was preserved upon the silylation reactions. This

M AN U

suggests that no significant organosilane intercalation into the interlayer of nanotube walls has occurred.

The textural characterization of the pristine and silylated HNTs was carried out by nitrogen adsorption-desorption at -196 ºC. Figure 2 shows the corresponding nitrogen

TE D

adsorption-desorption isotherms for all the materials and the specific surface areas (ABET) are given in Table 1. The isotherms are of type IIb,26 characteristic of natural clays, where the very minor hysteresis loops due to large mesopores result from disorder stacking of the clay 4,26

which is also noticed from the TEM and SEM results in Figure 1. Halloysite is

EP

tubes

AC C

polymorph of kaolinite 4 and, for the later, the ABET values are usually within 10 to 20 m2/g,26 depending on the type of tubes and the related stacking. The ABET value of the pristine Halloysite is 26 m2/g, Table 1, and is within the range of values reported in literature for this type of clay.4 In general, the ABET values decrease upon functionalization, with the exception of material HNTs_VTMS: this decrease is due to organosilane grafting and associated changes in nanotubes stacking. Curiously, the C constant of the BET equation, which can tentatively be related with the interaction between the adsorbed nitrogen molecules and the surface, presents its highest value for the pristine HNTs.26 This highest C value (highest 11

ACCEPTED MANUSCRIPT interaction energy) may be related with the highest number of free Al–OH/Si-OH groups in the pristine HNTs, which decrease after the silylation reactions.

Figure 2 - N2 adsorption-desorption isotherms

3.2

RI PT

Table 1 – N2 adsorption-desorption isotherms

Composition and spectroscopic data

SC

Table 2 shows the results of bulk elemental analysis for pristine and silylated HNTs: Si and C for all the materials and the relevant elements for specific functionalized materials: S

M AN U

for HNTs_MPTMS and N for HNTs_APTES and HNTs_AEAPTMS.

Table 2

EA (mmol/g)

TE D

All the materials contain Si element in the range of 6.75–7.30 mmol/g and C up to 1.92 mmol/g, with the pristine HNTs and HTNs_APTES exhibiting, respectively, the lowest and the highest Si and C loadings. The increase in Si and C loadings upon HNTs silylation

EP

confirms the success of organosilane grafting in the anhydrous conditions used. From the

AC C

group of organosilanes used in this work, APTES anchored with the highest efficiency, based in Si and C contents (7.30 and 1.92 mmol/g, respectively). This has already been observed for other works reporting HNTs functionalized with APTES4 and other materials with tubular morphology, c.a. multiwalled carbon nanotubes (MWCNTs). The authors used a similar group of organosilanes to graft to MWCNTs following a similar methodology:27 within the silylated MWCNTs the highest efficiency of APTES anchoring, compared with the other organosilanes was attributed to the existence of two grafting mechanisms: direct silylation reaction with the MWCNTs surface groups and APTES lateral polymerization via NH212

ACCEPTED MANUSCRIPT silicon oligomerization.27 Similarly, the highest APTES anchoring efficiency in HTNs can also be explained by the same mechanisms, see XPS data below. Although lower than for HTNs_APTES, the material HTNs_MPTMS also exhibits high C content (1.00 mmol/g), but XPS does not confirm this mechanism (see below). The materials HNTs_AEAPTMS

this implies a lower degree of silylation.

RI PT

(2N:1Si) and HNTs_APTES (1N:1Si) show similar N contents, but for HNTs_ AEAPTMS,

The surface atomic percentages were determined by XPS and are summarized in Table

SC

3. As the HNTs wall thickness is typically around 20 nm,4 which well exceeds the penetration depth of XPS (10 nm), it is difficult to disclose the chemical composition of the inner lumen

M AN U

surface (Al–OH groups). Nevertheless, since HNTs are randomly oriented on a substrate, the inner surface may be partially exposed to X-rays, as well as the Al–OH edge groups and Si– OH groups at the edges or external surface defects; consequently, the comparison with EA data cannot be made directly.

TE D

The pristine HNTs show oxygen, aluminium and silicon as main elements in Si/Al ratio of 1.1, which is in accordance with the typical structure of a two layered (1:1) aluminosilicate28 and with the EDS results. In this sample were also detected traces amounts

EP

of sodium, magnesium and chlorine, but there was no evidence of the presence of iron;29 a small amount of carbon was also observed, due to the presence of some adsorbed surface

AC C

organic contaminants. The silylated HNTs showed, besides the main elements (oxygen, aluminium and silicon), a similar or higher Si/Al ratio (from 1.0 to 1.4), the increase of the surface carbon atomic percentage and new target elements such as nitrogen, in the case of HNTs_APTES and AEAPTMS (2.6 and 4.1%, respectively), sulphur (0.6%) for HNTs_MPTMS and bromine (0.2%) for HNTs_BrTMS. The increase of the Si/Al ratio combined with the increase in C surface content and with the presence of the new elements for the referred organosilanes confirms the success of organosilane grafting onto the inner and 13

ACCEPTED MANUSCRIPT outer HNTs surfaces. Table 3 also shows that HNTs_APTES presents the highest carbon surface content (30.1%), followed by HNTs_BrTMS (23.4%), HNTs_MPTMS (22.5%) with almost similar carbon surface contents and HNTs_AEAPTMS (19.8%). In some aspects there is some similarity between the bulk EA analysis and XPS data, but the surface N% for

RI PT

HNTs_APTES is lower than that for HNTs_AEAPTMS. Taking into account that the bulk N content from EA for these two materials are similar (0.36 mmol/g), although AEAPTMS contains two nitrogen atoms, whereas APTES contains just one nitrogen atom, the highest

SC

surface N% for HNTs_AEAPTMS may suggest that AEAPTMS is anchored more externally

M AN U

than APTES, probably due the bigger size of the carbon chain.

Table 3: XPS atomic %

The core-level binding energies (BEs) of the different elements in all HNTs materials

TE D

are summarized in Table S2, Supplementary Information: the deconvolution of the bands in the N 1s, S 2p and Br 2p regions for the HNTs functionalized with APTES, AEAPTMS,

Information.

EP

MPTMS and BrTMS, respectively, are presented as examples in Figure S2, Supplementary

AC C

In the N 1s region, the HNT_APTES spectrum shows a broad band and a shoulder that can be deconvoluted into two components, at 399.6 and 401.6 eV, which can be tentatively assigned to N–H bonds of free amine groups and to a slight amount of protonated amine groups, respectively.27 Similarly, the N 1s high-resolution spectrum of HNTs_AEAPTMS shows a broad band and a weak shoulder that were deconvoluted into two peaks at 399.8 and 401.6 eV, respectively; the former peak can be tentatively attributed to the two types of N atoms (H2N–C and HN–C) and the peak at higher binding energy assigned to positively

14

ACCEPTED MANUSCRIPT charged quaternary nitrogen species.27 The protonated amine groups result from amine protonation by acidic sites at HNTs surface. In the S 2p high-resolution spectrum of HNTs_MPTMS, the observed band was deconvoluted into two peaks corresponding to the S 2p3/2 and S 2p1/2 components, which are

RI PT

split by 1.2 eV, at 163.5 eV and 164.7 eV, respectively: these peaks are ascribed to S–H and S–C bonds that have similar BEs.30 In the Br 3d high-resolution spectrum of HNTs_BrTMS, a broad band is detected at 70.0 eV, due to presence of C–Br bond from the BrTMS

SC

organosilane.31

In all materials the Si 2p region shows a broad band in the range of 102.6–103.7 eV

M AN U

which confirms the existence of Si–O bonds not only due to the intrinsic composition of HNTs, but also due to the grafted organosilanes, in the case of the silylated HNTs.32 This band comprises the contribution of the two components due to Si 2p3/2 and Si 2p1/2, separated by 0.6 eV, Table S2, Supplementary Information. An interesting result is that the BE of these

TE D

two components are similar for the pristine HNTs and all silylated materials (BEs ~ 103.0/103.1 and 103.6/103.7 eV) with exception of HNTs_APTES and HNTS_AEAPTMS, where the BEs are shifted to lower values relative to those of the parent HNTs: 102.6/102.7

EP

and 103.2/103.3 eV, suggesting the presence of a new Si environment (Si–N) at very similar BE as that of Si–O, that cannot be resolved but that leads to an overall shift of the Si 2p peaks. This result suggests the existence of lateral polymerization via NH2-silicon oligomerization

AC C

27

for both HNTs_APTES and HNTs_AEAPTMS silylated HNTs, as observed for similar functioanlziation in MWCNTs. 27 Finally, for all the materials, a band is also detected in the Al 2p region in the range of 74.5–75.0 eV, assigned to the Al–O bonds of the parent material.32 For both the parent and silylated HNTs the Al 2p band occurs at similar BEs (74.8–75.0 eV), except for

15

ACCEPTED MANUSCRIPT HNTs_APTES and HNTs_AEAPTMs where the band is shifted to lower BEs (74.5–74.6 eV), being in accordance with the results for the Si 2p region. Table 4 summarizes the mass loss values obtained by TG; the thermogravimetric curves and the corresponding differential profiles of the pristine and silylated-HTNs are presented in

RI PT

Figure S3, Supplementary Information. The mass loss profiles of all materials are very similar and show a very small mass loss in the range of 25–150 ºC and two main mass losses from 200–340 and 365–616 ºC. The first very small mass loss is due to the removal of physically

SC

adsorbed water from the HNTs surface and is not significant neither in the pristine HNTs (0.5%) nor in the silylated HNTs (0.2–0.3%);33 nevertheless, in the functionalized materials

M AN U

this mass loss decreases compared to that of the pristine material, which is due to the fact that some residual water present is consumed during the silylation reaction.4,22,33,34 For the pristine HNTs, the first substantial loss in the range of 220–310 ºC (Table 4) is attributed to the gradual loss of the residual interlayer water 35 and the second substantial loss at approximately

TE D

405–570 ºC is ascribed to the dehydroxylation of the structural aluminol groups (Al–OH).2 For the functionalized HNTs, the mass losses in both ranges 200–340 and 365–616 ºC include the decomposition of the organic component from the grafted silanes and the dehydroxylation

EP

of the structural aluminol groups. From the total mass loss values, presented in Table 4, it is possible to see that HNTs_APTES, HNTs_MPTMS and HNTs_AEAPTMS show the highest

AC C

mass losses, in nearly accordance with the results from bulk chemical analysis and XPS. However, these values cannot be used directly to conclude about the extent of silane grafting since they will also depend on the molecular weight of the organosilanes, and this justifies the higher weight loss observed for HNTs_AEAPTMS (18.0%). Some of the silylated materials – HNTs_VTMS, HNTs_PhTES and HNTs_BrTMS – show lower mass loss values than the pristine HNTs: a possible explanation for this is the eventual carbon condensation of the organosilanes inside the internal lumen promoted by the high temperature.33,34 16

ACCEPTED MANUSCRIPT

Table 4

TG

In Figure 3 are summarized the FTIR-ATR spectra of all HNTs materials. The spectrum

RI PT

of the pristine HNT shows intense vibration bands at 3692 and 3621 cm-1 that are attributed to the stretching vibrations of inner-surface Al–OH. The presence of some interlayer water is suggested by the observation of the O–H stretching and bending vibrations at 3553 and 1640

SC

cm-1, respectively, and the band at 1037 cm-1 is assigned to the Si–O vibrations of the silica network (Si–O–Si and O–Si–O stretching vibrations). The vibration band at 912 cm-1 is

M AN U

attributed to the bending vibrations of inner surface hydroxyl groups and finally the vibration band at 535 cm-1 is attributed to the bending vibration of Al–O–Si.4,6,36 The FTIR-ATR spectra of the silylated HNTs confirm the grafting of the organosilanes without the disruption of the HNTs original structure, as previously reported by electron

TE D

microscopy and XRD. They show, besides the characteristic vibrational bands from the original nanotubes, new bands corresponding to the different organosilanes. The spectra of HNTs_AEAPTMS and HNTs_APTES are very similar, showing vibrational bands at 1500-

EP

1551 cm-1 attributed to the N–H2 deformation and bands at respectively 2900 and 1400 cm-1

AC C

due to C–H2 stretching and bending vibrations.4,22 The FTIR spectra of HNTs_BrTMS and HNTs_MPTMS are very similar and exhibit vibration bands at 2899–2976 cm-1, corresponding to the CH2 stretching vibrations and at 1409 cm-1 related to the CH2 deformation vibration modes.37 The band due to the S–H stretching vibration (2600–2540 cm-1) has usually low intensity and consequently is not observed and the band due to the stretching vibrations of C–Br bonds, typically at 500–600 cm-1, may be overlapped with the vibrations from the HNTs structure (Al–O–Si bending).38 The FTIR-ATR spectrum of HNTs_VTMS presents new vibration bands at 1411 cm-1, due to 17

ACCEPTED MANUSCRIPT the =CH2 in-plane deformation modes, and at 1605 cm-1 related with the C=C stretching vibrations. Finally, the FTIR-ATR spectrum of HNTs_PhTES presents the characteristic bands of C–H stretching vibrations of the phenyl ring in the region of 3080–3010 cm−1 and of

Figure 3 - FTIR

The

13

C CPMAS,

29

29

Si and

Si MAS and

27

Al) of all HNTs materials are summarized in Table 5.

Al MAS NMR spectra of HNTs_APTES are shown as

M AN U

examples in Figure 4.

27

SC

The NMR data (13C,

RI PT

C–C stretching vibrations in the region of 1625–1430 cm−1.38,39

The 13C CPMAS NMR spectra of the silylated HNTs clearly confirmed the grafting of the organosilanes and the preservation of their structure upon immobilization. The

13

C

CPMAS NMR spectrum of HNTs_APTES (Figure 4(a)) shows three main signals at 42.7,

TE D

21.8 and 9.8 ppm assigned to carbons in the alkyl chain: CH2–NH2, –CH2– and CH2–Si– carbons, respectively. The spectrum of HNTs_AEAPTMS is very similar to that of HNTs_APTES, showing signals at 51.2, 40.8, 22.2 and 10.2 ppm, assigned to CH2–NH–,

EP

CH2–NH2, –CH2– and CH2–Si–, respectively.40 For HNTs_MPTMS the spectrum shows

AC C

similarly the three peaks due to the carbon atoms in the alkyl chain at 41.6, 27.6, 23.5 and 11.8 ppm, O–CH3, CH2–SH, –CH2– and CH2–Si–, respectively. The two main signals at 25.4 and 10.7 ppm are also observed in the spectrum of HNTs_BrTMS corresponding, respectively, to –CH2– of the alkyl chain and to CH2–Si. In the HNTs_VTMS spectrum the two signals at 134.6 and 130.5 ppm are assigned to =CH2 and =CH–Si, respectively, and finally, the spectrum of HNTs_PhTES presents two main peaks at 133.6 and 127.4 ppm, attributed to the aromatic carbons presented in the phenyl ring of the organosilane.38

18

ACCEPTED MANUSCRIPT The

29

Si NMR spectrum of the pristine HNTs, Table 5, shows a single band at -91.8

ppm in accordance with the typical chemical shift corresponding to Q3 silicon of halloysite, Si(OSi)3(OAl), which is assigned to the crystallographica equivalent Si atoms of the tetrahedral sheet.4,9,39 Moreover, it is also observed a low intensity peak at -86.5 / -87.2 ppm,

RI PT

that can be assigned to silicon in Q2 environment, c.a. Si(OH)(OSi)2(OAl), present at the edges or external surface defects. The 27Al spectrum of the pristine HNTs, Table 5 and Figure 4c, is composed by a strong resonance at 6 ppm, with side bands at high and low magnetic

SC

fields, corresponding to Al in octahedral coordination (Al[VI]); low-intense broad resonances at 69.6 and 60.0 ppm are also observed, that can be assigned to Al sites with tetrahedral

M AN U

coordination, that will be denoted respectively as, Al[IV]a and Al[IV]b signals.41,42,43 Upon HNTs silylation, there are some intensity changes in the original resonance signals in 29Si spectra, but Q3 occur at approximately the same resonance values, whereas for Q2 there are small peak shifts, Table 5 and Figure 4b, exemplified for HNTs_APTES.

TE D

Moreover, in the 29Si spectra, additional new low intensity peaks are observed that correspond to T3 (values within the range -60.3 / -67.5 ppm) and T2 signals (values within the range -55.9 /-59.7 ppm), due to silicon from grafted organosilanes.4,9,22,34 For

27

Al spectra, Table 5 and

EP

Figure 4c exemplified for HNTs_APTES, Al[VI] resonances occur at approximately the same values, but for Al[IV] signals there are some changes that are more evident for Al[IV]b: some

AC C

increase in intensity and peak shift are observed, suggesting the presence of Al sites that have reacted with the organosilanes. Signals T3 and T2 corresponds to, respectively, 3- and 2-fold covalent grafting of the organosilanes to HNTs, through reaction between three or two methoxy/ethoxy groups and the OH groups from the HNTs structure. For materials HNTs_APTES (Figure 4(b)), HNTs_AEAPTMS and HNTs_PhTES it was possible to identify silicon T3 and T2 signals, but for HNTs_VTMS and HNTs_BrTMS only T3 signals were observed, whereas for 19

ACCEPTED MANUSCRIPT HNTs_MPTMS only T2 was detected. Although the signals have low intensity, this result suggest different grafting mechanisms for the different organosilanes, but no apparent correlation between the organosilane structure and grafting mechanism could be found. The combination of 29Si and 27Al NMR spectra confirms silylation reactions at the Al–

defects.

SC

Table 5 : NMR data

RI PT

OH groups (inner lumen surface and edges) and Si-OH groups at the edges or external surface

4. Concluding remarks

M AN U

Figure 4: NMR spectra HNTs_APTES

Several organosilanes with the functional groups –NH2 (APTES, AEAPTMS), –SH (MPTMS), –Br (BrTMS), –C=C (VTMS) and phenyl (PhTES) were covalently grafted into

TE D

pristine HNTs using aprotic and anhydrous conditions. The extensive physicochemical characterization, which included morphological, structural, textural, chemical and

EP

spectroscopic techniques, identified pristine HNTs as halloysite-7Å and proved their successful silylation without the disruption of the nanotubes structure. The APTES-

AC C

functionalized HNTs showed the highest silylation efficiency, followed by both MPTMS and AEAPTMS. The silylation reaction occurred at Al–OH groups of the inner lumen surface, and at the edges as well as Si–OH groups at the edges or external surface defects. The silylation mechanism was found to take place through reaction of the alkoxy moieties from the organosilane with the referred HNTs surface groups in a 3-fold (for VTMS and BrTMS) or 2fold covalent grafting (for MPTMS) or a mixture of both approaches (APTES, AEAPTMS and PhTES). In the case of APTES- and AEAPTMS-functionalized HNTs, a polymerization 20

ACCEPTED MANUSCRIPT side-reaction via NH2-silicon oligomerization was also evidenced, as a parallel functionalization pathway. In this context, although APTES grafted with the highest content, some of the NH2 groups will not be available for subsequent chemical reactions. The variety of the functional groups of the grafted organosilanes allowed the selective

RI PT

modification of HNTs which is important in several applications. One of the applications that have not been extensively exploited is catalysis: i) as support for the covalent anchorage of metal complexes or metal/metal oxide nanoparticles with catalytic properties, or ii) as

SC

intrinsic catalyst upon subsequent chemical modification of the grafted groups. The nanotube morphology (nanometer scale) associated with the possibility of inner lumen functionalization

M AN U

can introduce in the HNTs-based catalysts synergistic effects, providing enhanced catalytic activity through the combination of confinement effect and shape selectivity, which are very important issues in heterogeneous catalysis.

TE D

Acknowledgements

The authors thank the Fundação para a Ciência e a Tecnologia (FCT, Portugal), the European Union, QREN, FEDER, COMPETE, for funding REQUIMTE through projects

EP

PEst-C/EQB/LA0006/2011, NORTE-07-0124-FEDER-000067-nanochemistry. The authors also thank Prof. P. Tavares and MSc. L. Fernandes from UTAD (Vila Real, Portugal) for the

AC C

XRD measurements and CeNTI - Centre for Nanotechnology and Smart Materials (Vila Nova de Famalicão, Portugal) for access to the FTIR-ATR equipment. AFP thanks FCT for a postdoctoral grant (SFRH/BPD/72126/2010).

21

ACCEPTED MANUSCRIPT

References S. Rooj, A. Das, V. Thakur, R. N. Mahaling, A. K. Bhowmick, G. Heinrich, Mater. Design

RI PT

1

31 (2010) 2151. 2

E. Joussein, S. Petit, J. Churchman, B. Theng, D. Righi, B. Delvaux, Clay Miner. 40 (2005)

SC

383.

D. Rawtani, Y. K. Agrawal, Rev. Adv. Mater. Sci. 30 (2012) 282.

4

P. Yuan, P. D. Southon, Z. Liu, M. E. R. Green, J. M. Hook, S. J. Antill, C. J. Kepert, J.

Phys. Chem. C 112 (2008) 15742. 5

M AN U

3

J. M. Pan, B. Wang, J. D. Dai, X. H. Dai, H. Hang, H. X. Ou, Y. S. Yan, J. Mater. Chem. 22

(2012) 3360.

M. Zou, M. Du, H. Zhu, C. Xu, Y. Fu, J. Phys. D: Appl. Phys. 45 (2012) 325302.

7

L. Wang, J. Chen, L. Ge, V. Rudolph, Z. Zhu, J. Phys. Chem. C 117 (2013) 4141.

8

H.-Y. Wu, C.-H. Liao, Y.-C. Pan, C.-L. Yeh, H.-M. Kao, Micropor. Mesopor. Mat. 119

(2009) 109.

D. Tan, P. Yuan, F. Annabi-Bergaya, H. Yu, D. Liu, H. Liu, H. He, Micropor. Mesopor.

AC C

9

EP

TE D

6

Mat. 179 (2013) 89. 10

V. Vergaro, Y. M. Lvov, S. Leporatti, Macromol. Biosci. 12 (2012) 1265.

11

K. Hedicke-Hochstotter, G. T. Lim, V. Altstadt, Compos. Sci. Technol. 69 (2009) 330.

12

Y. M. Lvov, D. G. Shchukin, H. Möhwald, Ronald R. Price, ACS Nano, 2 (2008) 814.

13

E. Abdullayev, R. Price, D. Shchukin, Y. Lvov, ACS Appl. Mater. Inter. 1 (2009) 1437.

14

E. Abdullayev, Y. Lvov, J. Mater. Chem. 20 (2010) 6681.

22

ACCEPTED MANUSCRIPT 15

L. R. Avila, E. H. de Faria, K. J. Ciuffi, E. J. Nassar, P. S. Calefi, M. A. Vicente, R.

Trujillano, J. Colloid Interface Sci. 341 (2010) 186. N. Takahashi, K. Kuroda, J. Mater. Chem. 21 (2011) 14336.

17

P. Yuan, P. D. Southon, Z. Liu, C. J. Kepert, Nanotechnology 23 (2012) 375705.

18

Q. He, D. Yang, X. Deng, Q. Wu, R. Li, Y. Zhai, L. Zhang, Water Res. 47 (2013) 3976.

19

H. Zhu, M. Du, M. Zou, C. Xu, Y. Fu, Dalton Trans. 41 (2012) 10465.

20

C. Pereira, A. R. Silva, A. P. Carvalho, J. Pires, C. Freire, J. Mol. Cat. A: Chem. 283 (2008)

SC

RI PT

16

5.

G. S. Machado, O. J. de Lima, K. J. Ciuffi, F. Wypych, S. Nakagaki, Catal. Sci. Technol. 3

M AN U

21

(2013) 1094. 22

S. Barrientos-Ramirez, G. Montes de Oca-Ramirez, E. V. Ramos-Fernandez, A. Sepulveda-

Escribano, M. M. Pastor-Blas, A. Gonzalez-Montiel, Appl. Cat. A: Gen. 406 (2011) 22. Y. Zhang, Y. Xie, A. Tang, Y. Zhou, J. Ouyang, H. Yang, Ind. Eng. Chem. Res. 53 (2014)

TE D

23

5507.

L. Zatta, J. E. F. C. Gardolinski, Fernando Wypych, App. Clay Sci. 51 (2011) 165.

25

C. Pereira, S. Patricio, A. R. Silva, A. L. Magalhaes, A. P. Carvalho, J. Pires, C. Freire, J.

EP

24

Colloid Interf. Sci. 316 (2007), 570. F. Rouquerol, J. Rouquerol, K. Sing, Adsorption by Powders and Porous Solids, Academic

AC C

26

Press, London, 1999, p 43-48. 27

H. Gaspar, C. pereira, S. L. H. Rebelo, M. F. R. Pereira, J. L. Figueiredo, C. Freire, Carbon

49 (2011) 3441. 28

L. Vast, G. Philippin, A. Destree, N. Moreau, A. Fonseca, J. B. Nagy, J. Delhalle, Z.

Mekhalif, Nanotechnology 15 (2004)781.

23

ACCEPTED MANUSCRIPT 29

K.-M. Ng, Y.-T. R. Lau, C.-M. Chan, L.-T. Weng, J. Wu, Surf. Interface Anal. 43 (2011)

795. 30

Y. Joseph, I. Besnard, M. Rosenberger, B. Guse, H-G. Nothofer, J. M. Wessels, J. Phys.

RI PT

Chem. B 107 (2003) 7406. M. E. Lipinska, S. L. H. Rebelo, C. Freire, J. Mater. Sci. 49 (2014) 1494.

32

Y. Zhang, X. He, J. Ouyang, H. Yang, Sci. Rep. 3 (2013) 2948.

33

M. Massaro, S. Riela, G. Cavallaro, M. Gruttadauria, S. Milioto, R. Noto, G. Lazzara, J.

SC

31

Organomet. Chem. 749 (2014) 410.

Y. Zhang, L. Fu, H. Yang, Colloids Surf., A 414 (2012) 115.

35

G.W. Brindley, K. Robinson, J. Goodyear, Am. Miner. 34 (1949) 337.

36

R. Li, Z. Hu, S. Zhang, Z. Li, X. Chang, Int. J. Environ. Anal. Chem. 2013, 93, 767.

37

M AN U

34

Y. Leng, Materials Characterization- Introduction to Microscopic and Spectroscopic

38

TE D

Methods, JohnWiley & Sons, Singapure, 2008.

E. Pretsch, P. Buhlmann, M. Badertscher, Structure Determination of Organic Compounds,

Springer-Verlag, Berlin, 2009, pg 69-156.

Y. S. Li, Y. Wang, S. Ceesay, Spectrochim. Acta A 71 (2009) 1819.

40

I. Warad, Molecules 15 (2010) 4652.

41

J. Rocha, J. Phys. Chem. B 103 (1999) 9801.

42

X. Pardal, F. Brunet, T. Charpentier, I. Pochard, A. Nonat, Inorg. Chem. 51 (2012) 1827.

43

P. Florian, E. Veron, T. F. G. Green, J. R. Yates, D. Massiot, Chem. Mat. 24 (2012) 4068.

AC C

EP

39

24

ACCEPTED MANUSCRIPT Table 1: Specific surface areas (ABET) and C values (BET model) from the nitrogen adsorption-desorption isotherms at -196 ºC for pristine and silylated HNTs

(m2/g)

C value

26 ± 0.5

177

HNTs_MPTMS

24 ± 0.5

112

HNTs_BrTMS

26 ± 0.8

56

HNTs_VTMS

32 ± 0.7

35

HNTs_PhTES

9 ± 0.1

139

HNTs_APTES

20 ± 0.7

77

HNTs_AEAPTMS

17 ± 0.2

SC

HNTs

RI PT

ABET

Material

AC C

EP

TE D

M AN U

44

1

ACCEPTED MANUSCRIPT Table 2: Silicon, carbon, nitrogen and sulphur bulk contents (mmol/g) for pristine and silylated HNTs determined by EA

Elemental Analysis Material

Si

C

N

S

RI PT

(mmol/g) HNTs

6.78

<0.58

-

-

HNTs_MPTMS

6.75

1.00

-

HNTs_BrTMS

6.98

0.83

-

HNTs_VTMS

7.01

0.58

-

HNTs_PhTES

6.91

0.75

-

-

HNTs_APTES

7.30

1.92

0.36

-

HNTs_AEAPTMS

6.91

0.67

0.36

-

0.62 -

AC C

EP

TE D

M AN U

SC

-

2

ACCEPTED MANUSCRIPT Table 3. Surface atomic percentages determined by XPS for pristine and silylated HNTs a) Surface atomic %

Material

C 1s

O 1s

Si 2p

Al 2p

N 1s

S 2p

Br 3d

Si/Al

HNTs

5.8

64.0

15.6

14.7

-

-

-

1.1

HNT_MPTMS

22.5

53.0

14.5

11.4

HNT_VTMS

12.2

60.6

14.6

12.6

-

HNT_PhTES

7.9

62.3

15.1

14.7

-

HNT_BrTMS

23.4

53.1

12.3

11.0

-

HNT_APTES

30.1

45.8

12.5

8.9

2.6

HNT_AEAPTMS

19.8

51.4

13.6

11.2

4.1

RI PT

1.3 -

1.2

-

-

1.0

-

0.2

1.1

-

-

1.4

-

-

1.2

SC

-

EP

TE D

M AN U

Determined by the areas of the respective bands in the high-resolution XPS spectra.

AC C

a)

0.6

3

ACCEPTED MANUSCRIPT

Table 4: Thermogravimetric data for pristine and silylated HNTs 2nd loss

3rd loss

Total mass loss

Material wt%

T (ºC)

wt%

(wt%)

HNTs

220-310

2.7

405-570

10.3

16.3

HNTs_MPTMS

210-340

2.7

425-540

9.5

17.3

HNTs_BrTMS

220-340

2.9

400-580

10.6

15.6

HNTs_VTMS

210-340

2.9

390-603

11.1

15.8

HNTs_PhTES

200-340

2.8

390-590

10.7

15.3

HNTs_APTES

240-310

3.0

400-590

HNTs_AEAPTMS

203-252

3.5

365-616

SC

RI PT

T (ºC)

16.7

12.6

18.0

AC C

EP

TE D

M AN U

11.0

4

ACCEPTED MANUSCRIPT

29

13

Si CPMAS (ppm)

HNTs

-86.2

-91.8

HNTs_MPTMS

-86.4

-91.8

HNTs_PhTES

-87.1

-91.8

-65.7

HNTs_VTMS

-86.4

-91.8

-62.8

HNTs_BrTMS

-87.0

-91.7

-60.3

HNTs_APTES

-86.6

-91.8

HNTs_AEAPTMS

-87.4

-91.9

T3

T2 -

-59.1

EP

TE D

-57.7

-67.5

-61.2

-59.7

-55.9

27

Al MAS (ppm)a)

SC

Q3

C CPMAS (ppm)

M AN U

Q2

AC C

Materials

RI PT

Table 5: Solid-state NMR data, 29Si CPMAS, 13C CPMAS and 27Al MAS for pristine and silylated HNTs

11.8 (CH2–Si), 23.5 (CH2), 27.6 (CH2–

SH), 41.6 (O–CH3)

6.0 ; ≈60, 69.6 Al; 86.5 6.0; 56.0; 69.9; 85.6

127.4, 133.6 (CHphenyl)

6.1; 56.2; 69.5; 85.5

130.5, 134.6 (CHvinyl)

6.0; 56.2; 70.1; 85.6

10.7 (CH2–Si), 25.4 (CH2)

6.0; 56.7; 69.7; 85.6

9.8 (CH2–Si), 21.8 (CH2), 42.7 (CH2–NH2)

5.9; 58.2; 70.5; 85.9

10.2 (CH2–Si), 22.2 (CH2), 40.8 (CH2– NH2), 51.2 (CH2–NH)

5.9 Al; 58.5; 71.2; 85.5

5

ACCEPTED MANUSCRIPT

Captions to Figures Scheme 1: Schematic representation of HNTs functionalization with organosilanes, and materials and organosilanes abbreviations.

RI PT

Figure 1: Electron microscopy images of pristine HNTs and functionalized HNT_APTES and HNT_MPTMS materials: Panel A) TEM, and Panel B) SEM and EDS spectra.

SC

Figure 2: Nitrogen adsorption-desorption isotherms at -196 ºC (closed symbols are desorption points) of pristine and functionalized HNTs.

M AN U

Figure 3: FTIR-ATR: a) full spectra of pristine in the range of 4000–500 cm-1 and b) magnification in the range of 3100–1400 cm-1 for pristine and functionalized HNTs.

Figure 4: Solid state NMR spectra of HNTs_APTES: a)

C CPMAS; b)

29

Si MAS

AC C

EP

TE D

and c) 27Al MAS.

13

1

ACCEPTED MANUSCRIPT

M AN U

SC

RI PT

Scheme 1

Organosilane names and abbreviations:

TE D

3-Mercaptopropyltrimethoxysilane: MPTMS (3-Bromopropyl)trimethoxysilane: BrTMS Vinyltrimethoxysilane: VTMS

EP

Phenyltriethoxysilane: PhTES

3-Aminopropyltriethoxysilane: APTES

AC C

N-2-aminoethyl-3-aminopropyltrimethoxysilane: AEAPTMS

2

ACCEPTED MANUSCRIPT Figure 1

AC C

EP

TE D

M AN U

SC

RI PT

Panel A)

3

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Panel B)

jj

4

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Figure 2

5

AC C EP TE D

Transmittance (a.u.)

b)

M AN U SC

Transmittance (a.u.)

Al-OH

RI PT

ACCEPTED MANUSCRIPT

Figure 3

a)

Al-O-Si

Si-O

6

ACCEPTED MANUSCRIPT Figure 4

RI PT

Q3

SC

Q2

M AN U

c) 27Al MAS NMR

side band

Al[VI]

Al[IV]a

AC C

EP

TE D

Al[IV]b

7

ACCEPTED MANUSCRIPT Highlights •

Halloysite nanotubes (HNTs) were silylated successfully with several organosilanes The silylation reaction occurred preferentially at the inner lumen surface Al-OH

RI PT

•

groups •

The silylation mechanism proceeded through reaction of alkoxy moieties from organosilane

Aminopropyltrimethoxysilane grafted into HNTs with the highest efficiency

•

Aminosilanes

polymerisation

side-reaction

M AN U

showed

SC

•

AC C

EP

TE D

oligomerization as a parallel functionalisation pathway

via

NH2-silicon

ACCEPTED MANUSCRIPT

Supplementary Information

Physicochemical characterization of organosilylated halloysite clay

RI PT

nanotubes

Andreia F. Peixoto,1 Ana C. Fernandes,2 Clara Pereira,1 João Pires,2 Cristina Freire1*

REQUIMTE, Departamento de Química e Bioquímica, Faculdade de Ciências,

SC

1

Centro de Química e Bioquímica, Faculdade de Ciências, Universidade de Lisboa,

EP

TE D

1749-016 Lisboa, Portugal

AC C

2

M AN U

Universidade do Porto, Rua do Campo Alegre s/n, 4169-007 Porto, Portugal

1

ACCEPTED MANUSCRIPT

RI PT

Figure S1: X-ray powder diffraction patterns of the original and functionalized HNTsbased materials.

HNTs_MPTMS

Intensity (a.u.)

HNTs_VTMS

HNTs_PhTES

HNTs_BrTMS

SC

HNTs_APTES

10

20

30

M AN U

HNTs_AEAPTMS

40

50

HNTs

60

AC C

EP

TE D

2θ2θ(º)

2

ACCEPTED MANUSCRIPT Figure S2: High-resolution XPS spectra in the N 1s, S 2p and Br 3d core-level regions for HNTs functionalized with a) MPTMS; b) BrTMS; c) APTES and d) AEAPTMS. a) S 2p (HNT_MPTMS)

166

164

162

73

160

72

71

c) N 1s (HNT_APTES)

402

400

69

68

67

398

SC

d) N 1s (HNT_AEAPTMS)

396

394

392

408

406

404

402

400

398

396

394

B.E.(eV)

AC C

EP

TE D

B.E. (eV)

M AN U

404

70

B.E. (eV)

B.E. (eV)

406

RI PT

168

b) Br 3d (HNT_BrTMS)

3

ACCEPTED MANUSCRIPT Figure S3: Thermogravimetric curves (lines) (TGA) and the corresponding derivatives

AC C

EP

TE D

M AN U

SC

RI PT

(DTG, dots) of pristine and silylated HNTs.

4

ACCEPTED MANUSCRIPT Table S1. XRD data of the original and functionalized HNTs

d(Å)

HNTs

11.97

7.40

HNT_MPTMS

11.98

7.38

HNT_VTMS

12.02

7.36

HNT_PhTES

12.07

7.33

HNT_BrTMS

12.05

7.34

HNT_APTES

12.15

7.28

HNT_AEAPTMS

11.97

7.40

AC C

EP

TE D

M AN U

SC

RI PT

2θ (º)

Material

5

ACCEPTED MANUSCRIPT Table S2. Core-level binding energies (BE) for pristine and functionalized HNTs obtained by curve fitting of XPS spectra. a)

HNT_PhTES

HNT_BrTMS

HNT_APTES

a)

Si 2p

103.1 (1.6)

Al2p

74.8 (1.5)

C 1s

285.0 (1.5)

O 1s

532.4 (1.7)

Si 2p

103.0 (1.4)

286.8 (1.7) 103.7 (1.6) 286.7 (1.5) 103.6 (1.4)

Al 2p

74.9 (1.4)

S 2p

163.5 (1.2)

164.7 (1.2)

C 1s

285.0 (1.6)

286.8 (1.6)

O 1s

532.5 (1.8)

Si 2p

103.1 (1.5)

Al 2p

75.0 (1.4)

C 1s

285.0 (1.8)

O 1s

532.4 (1.8)

Si 2p

103.1 (1.5)

291.3 (1.7)

RI PT

532.4 (1.9)

Al 2p

74.9 (1.5)

C 1s

285.0 (1.6)

O 1s

532.4 (2.0)

Si 2p

103.0 (1.6)

Al 2p

74.8 (1.5)

Br 3d

70.0 (2.5)

C 1s

285.0 (1.5)

O 1s

532.0 (1.8)

Si 2p

102.5 (1.4)

Al 2p

74.4 (1.4)

289.0 (1.5)

289.4 (1.6)

292.5 (1.6)

103.7 (1.5) 287.1 (1.8)

286.5 (1.6)

286.4 (1.5)

399.6 (2.0)

401.6 (2.0) 286.4 (1.6)

O 1s

532.3 (1.8)

Si 2p

102.7 (1.5) 74.6 (1.4) 399.8 (1.9)

288.7 (1.5)

103.2 (1.4)

285.0 (1.6)

N 1s

289.0 (1.6)

103.6 (1.6)

C 1s

Al 2p

289.8 (1.8)

103.7 (1.5)

N 1s

AC C

HNT_AEAPTMS

285.0 (1.7)

SC

HNT_VTMS

C 1s O 1s

M AN U

HNT_MPTMS

BE (eV)

TE D

HNTs

Element

EP

Material

288.3 (1.6)

103.3 (1.5) 401.6 (1.9)

The values between brackets refer to the full width at half-maximum of the bands.

6