Dendrimer-functionalized halloysite nanotubes for effective drug delivery

Dendrimer-functionalized halloysite nanotubes for effective drug delivery

Applied Clay Science 153 (2018) 134–143 Contents lists available at ScienceDirect Applied Clay Science journal homepage: www.elsevier.com/locate/cla...

1MB Sizes 0 Downloads 65 Views

Applied Clay Science 153 (2018) 134–143

Contents lists available at ScienceDirect

Applied Clay Science journal homepage: www.elsevier.com/locate/clay

Research paper

Dendrimer-functionalized halloysite nanotubes for effective drug delivery a,⁎

a

b

Joanna Kurczewska , Michał Cegłowski , Beata Messyasz , Grzegorz Schroeder a b

T

a

Faculty of Chemistry, Adam Mickiewicz University in Poznań, Umultowska 89b, 61-614 Poznań, Poland Department of Hydrobiology, Faculty of Biology, Adam Mickiewicz University in Poznań, Umultowska 89, 61-614 Poznań, Poland

A R T I C L E I N F O

A B S T R A C T

Keywords: Halloysite nanotubes Polyamidoamine dendrimers Drug delivery Adsorption Release

Halloysite nanotubes functionalized with polyamidoamine dendrimer were prepared and characterized. The material studied was applied as a carrier of three model therapeutic compounds - chlorogenic acid, ibuprofen and salicylic acid. It showed higher adsorption capacity for the drugs studied (123.16 mg/g for chlorogenic acid; 182.72 mg/g for ibuprofen; 39.52 mg/g for salicylic acid) compared to raw halloysite and 3-aminopropyltrimethoxysilane functionalized – halloysite nanotubes. The experimental adsorption data fits the Langmuir model. As a result of surface functionalization of halloysite with the dendrimer, the release rate of chlorogenic and salicylic acid decreased, while the release profile of ibuprofen was similar to that of 3-aminopropyltrimethoxysilane functionalized nanotubes. The release kinetics of chlorogenic acid and salicylic acid followed Higuchi model and the release exponents indicated a Fickian diffusion mechanism. The release mode of ibuprofen followed the first order kinetics and the mechanism was described as non-Fickian (anomalous) transport. The in vivo toxicity studies showed that the dendrimer –functionalized halloysite had no effect on the living organisms used in the bioassays.

1. Introduction Halloysite (Hal), a clay mineral of the kaolin group, is of great interest due to a variety of its potential applications. In contrast to kaolinite (Al2Si2O5(OH)4), halloysite (Al2Si2O5(OH)4·nH2O) contains additional water molecules, and it is composed of aluminosilicate layers. Each layer contains octahedral alumina and tetrahedral silica sheets (1:1 stoichiometric ratio). Tubular morphology, with alumina sheet inside and silica sheet outside, is most commonly found in this clay mineral. The advantages of halloysite nanotubes compared to other materials with the same morphology, are their natural origin, low price, environmental friendliness and biocompatibility (Churchman et al., 2016; Du et al., 2010; Joussein et al., 2005; Rawtani and Agrawal, 2012; Yuan et al., 2015). Hollow tubular structure and the properties of halloysite influenced development of scientific research on biomedical applications of these nanotubes (Abdullayev and Lvov, 2013; Aguzzi et al., 2007; Fakhrullin and Lvov, 2016; Hanif et al., 2016; Lazzara et al., 2017; Lvov et al., 2008, 2016a, 2016b; Naumenko et al., 2016; Khodzhaeva et al., 2017; Yendluri et al., 2017a, 2017b). There are a number of literature reports on the use of halloysite as a carrier of bioactive compounds. Different treatments are used to improve adsorption and controlled release of active compounds from halloysite material. One of the commonly used methods is acid- and/or heat-treatment of the material (Wang et al.,



2014). Under favorable conditions, acid treatment can lead to increased halloysite nanotube lumens and consequently to improved loading efficiency of a drug (Abdullayev et al., 2012). Surface modification of Hal with dopamine can be helpful in more effective immobilization of a desired compound (Chao et al., 2013). The addition of halloysite to other drug carriers, nanocomposite hydrogels, can enhance the mechanical properties of such materials (Tu et al., 2013). Furthermore, hybrid systems based on halloysite nanotubes can also contribute to improvement of loading and release of drugs. A glycoclaster composed of the clay nanotubes and carbohydrate functionalized cyclodextrin can be used for drug transport into living cells (Massaro et al., 2016b). In addition to the advanced research on the use of raw halloysite as a nanocontainer of drugs, there are numerous works on a modification of Hal surfaces in order to improve the properties of the material (Tan et al., 2016). Chemical modification of inner lumen and outer surface depends on adsorbed compounds and medical application of a designed hybrid system (Massaro et al., 2014; Massaro et al., 2016a; Massaro et al., 2017). 3-Aminopropyl triethoxysilane/trimethoxysilane (APTES/ APTS) very often appears as an agent for surface functionalization of different inorganic materials (Yuan et al., 2008; Kurczewska et al., 2009; Narkiewicz et al., 2010). In case of halloysite, it is used to facilitate loading process (Shi et al., 2011), as well as for slowing down of drug release from the material (Tan et al., 2013, 2014). In addition to a simple one-step functionalization, a very interesting concept is to

Corresponding author. E-mail address: [email protected] (J. Kurczewska).

https://doi.org/10.1016/j.clay.2017.12.019 Received 19 October 2017; Received in revised form 11 December 2017; Accepted 11 December 2017 0169-1317/ © 2017 Elsevier B.V. All rights reserved.

Applied Clay Science 153 (2018) 134–143

J. Kurczewska et al.

(PAMAM dendrimer, Fig. 1b). 1 H NMR (CDCl3, 400 MHz), δ: 2.21 (12H, m, CH2), 2.38–2.73 (48H, m: overlapping CH2, NH2), 3.05 (12H, m, CH2), 8.07 (6H, bs, CONH); MS (ESI, positive) m/z 831.72.

combine by chemical bonding of different drug carriers. Dendrimers, including polyamidoamine (PAMAM) dendrimers, form host-guest complexes with small molecules through electrostatic and hydrophobic interactions. This class of compounds is also widely studied for their biomedical and biotechnological applications (Sato and Anzai, 2013). Therefore, chemical functionalization of halloysite surface with a dendrimer should significantly affect both the adsorption of drug molecules and their release from such a carrier. At the moment, there are reports in a literature concerning the use of similar hybrid systems only for environmental applications (Shahamati Fard et al., 2016). However, there are few works showing the use of other nanostructured materials which, in combination with PAMAM dendrimers, can slow down the release rate of small molecules compared to unmodified materials (Torres et al., 2016). The objective of this study was to obtain and characterize a drug delivery system, PAMAM-dendrimer functionalized halloysite nanotubes. The material studied was investigated as a carrier of three model drug compounds – chlorogenic acid, ibuprofen and salicylic acid. All the small drug molecules have carboxyl groups. We suspect that it should promote a formation of stronger ionic bonds between the drug molecules and the organic unit in the carrier studied compared with systems based only on hydrogen bonding. We present the evaluation of drug adsorption isotherms and drug release kinetics, as well as the toxic effect of the material studied on living organisms. Based on our knowledge, this is the first report presenting an application of PAMAM dendrimer-functionalized halloysite nanotubes for loading and release of drug molecules.

2.3. Synthesis of PAMAM-functionalized halloysite nanotubes The synthesis of polyamidoamine dendrimer-functionalized halloysite nanotubes (Hal_PAMAM) consisted of several steps (Fig. 2). The nanotubes were initially functionalized with 3-aminopropyltrimethoxysilane (APTS) following the procedure previously reported (Kurczewska et al., 2017). In the next process, Hal_APTS (10.0 g) was dispersed in chloroform (100 mL) and reacted with suberic acid bis (Nhydroxysuccinimide ester) (DSS; 1.3 mmol per 1.0 mmol of APTS). Then the mixture was stirred and heated for 72 h. The solid was filtered, washed and dried. The intermediate product, labeled as Hal_DSS, was reacted with polyamidoamine dendrimer (PAMAM; 1.1 mmol per 1.0 mmol of DSS) in methylene chloride. The mixture was stirred for 72 h at room temperature. The obtained solid – Hal_PAMAM - was centrifuged, washed several times with methanol and demineralized water and dried for 24 h at 40 °C. 2.4. Characterization The infrared spectra were taken on an IFS 66v/s Fourier transform infrared (FTIR) spectrophotometer from Bruker, equipped with an MCT detector (125 scans, resolution 2 cm− 1). The spectra were recorded in the 400–4000 cm− 1 range for KBr pellets. The thermogravimetric studies were carried out in a Setsys 1200 apparatus (Setaram) at a heating rate of 5 °C/min under helium atmosphere. X-ray diffraction (XRD) was measured using Brucker AXS D8 Advance powder diffractometer equipped with Johansson monochromator (λCu Kα1 = 1,5406 Å). The surface morphology was studied in a Carl Zeiss EVO-40 scanning electron microscope, SEM (resolution 2 nm, operating voltage 80 kV). Transmission electron microscope (TEM) images were recorded on a Hitachi HT7700 microscope, operating at accelerating voltage of 100 kV. NMR spectra were recorded on a Bruker (Billerica, MA, USA) NanoBay 400 MHz spectrometer. Elemental analysis of Hal_APTS was carried out on a Vario ELIII (Elementar, USA) analyzer. Nitrogen adsorption-desorption isotherms were measured on a sorptometer Quantachrome Autosorb iQ (Boynton Beach, Florida, USA). The samples were degassed at 150 °C (Hal) and 40 °C (Hal_PAMAM). The specific surface area (SBET) and the pore size distribution were calculated by Brunauer–Emmett–Teller (BET) and Barrett–Joyner–Halenda (BJH) methods, respectively. The ESI mass spectra were obtained on an amaZon SLBruker mass spectrometer equipped with an electrospray ion (ESI) source in infusion mode. The sample solution was introduced into the ionization source at a flow rate of 5 μL min− 1 using a syringe pump. The apparatus was operated using the so-called “enhanced resolution mode” (mass range: 50–2200 m/z, scanning rate: 8100 m/z per second). The capillary voltage was set at − 4.5 kV and the endplate offset at − 500 V. The source temperature was 80 °C and the desolvation temperature was 250 °C. Helium was used as the cone gas and desolvating gas (nitrogen) at flow rates of 50 L h− 1 and 800 L h− 1, respectively. The mass spectrometer was operated in the ESI positive and negative ionization mode. For all the experiments, 0.1 mM water/methanol solutions of the dendrimer and the complexes of the dendrimer with the drugs studied were used.

2. Materials and methods 2.1. Materials The clay mineral - halloysite (Hal) with tubular structure was purchased from Sigma-Aldrich (product of Applied Minerals, Inc., Dragon Mine, USA). According to the supplier, the diameters of the halloysite nanotubes were 30–70 nm with a length of 1–3 μm, pore volume – 1.261.34 mL/g, specific surface area – 64 m2 g− 1, cation exchange capacity – 8.0 meq/g and relative density – 2.53 g/cm3. The chemical composition of the halloysite from Dragon Mine was 43.50% SiO2, 0.02% TiO2, 38.88% Al2O3, 0.33% Fe2O3, 0.12% MgO, 0.26% CaO, 0.07% Na2O, 0.07% K2O, 0.83% P2O5, 0.26% SO3 and 15.70% others (Pasbakhsh et al., 2013). Chlorogenic acid (CHLG), ibuprofen (IBU), salicylic acid (SAL) and all other chemicals were obtained from SigmaAldrich and used as received without further purification. All solvents were of the p.a. grade, purchased from POCH (Poland). Demineralized water was used for aqueous solutions preparation. 2.2. Synthesis of polyamidoamine (PAMAM) dendrimer To a stirred solution of methyl acrylate (55 g, 638.87 mmol) in methanol (50 mL) cooled in an ice-water bath a solution of tris(aminoethyl)amine (10 g, 68.38 mmol) in methanol (50 mL) was added dropwise. After addition of tris(aminoethyl)amine solution the resulting mixture was allowed to warm to room temperature and was stirred for further six days. The solvent and excess methyl acrylate were removed under reduced pressure using rotary evaporator, yielding intermediate dendrimer (Fig. 1a). 1 H NMR (CDCl3, 400 MHz), δ: 2.45 (12H, t, CH2), 2.51 (12H, bs, CH2), 2.78 (12H, t, CH2), 3.67 (18H, s, CH3); MS (ESI, positive) m/z 663.1. Then, to a stirred solution of ethylenediamine (13.6 g, 226 mmol) in methanol (50 mL) cooled in an ice-water bath a solution of intermediate dendrimer (5 g, 7.5 mmol) in methanol (50 mL) was added dropwise. The resulting solution was allowed to warm to room temperature and was stirred for further seven days. Methanol and excessive ethylenediamine were removed in vacuum, yielding a yellow oil

2.5. Drug adsorption experiments Adsorption isotherm studies were performed for three model drugs following the general procedure: 5 mg of halloysite samples (Hal, Hal_APTS and Hal_PAMAM) were introduced into 5 mL of drug solution (0.2–1.0 mM in ethanol for CHLG and IBU; methanol for SAL). The mixtures were stirred for 24 h at constant temperature (298 ± 1 K) to 135

Applied Clay Science 153 (2018) 134–143

J. Kurczewska et al.

Fig. 1. Structures of intermediate dendrimer (a) and PAMAM dendrimer (b).

Fig. 2. Scheme of synthesis of PAMAM-functionalized halloysite nanotubes.

reach the drug encapsulation equilibrium. The starting (C0; mg L− 1) and equilibrium (Ce; mg L− 1) concentration of the drugs in the supernatant were determined using UV–Vis Spectrophotometer (Agilent 8453) at 331 nm (CHLG), 264 nm (IBU) and 305 nm (SAL). The amount of drug adsorbed at equilibrium (qe; mg/g) was calculated by using the following equation:

the Langmuir constant. The values of qm and KL can be calculated from the slope and intercept of the plot of Ce/qe versus Ce. The linear form of Freundlich adsorption isotherm is given as:

log qe =

1 log Ce + log KF n

(3) − 1 1/n

qe =

(C0 − Ce ) V m

where n and KF (mg/g(L mg ) ) represent the Freundlich constants related to adsorption intensity and capacity. KF and 1/n can be calculated by plotting logqeversus logCe.

(1)

where V (L) is the volume of drug solution and m is the weight of halloysite samples (g). Langmuir and Freundlich models were used for fitting of the experimental data. The linear form of Langmuir equation is expressed as:

Ce C 1 = e + qe qm KL qm

2.6. In vitro drug release studies and kinetic modeling For in vitro release studies, the weighted samples of the drug carriers under study were immersed in phosphate buffered saline (PBS, pH 7.4), and stirred at 37 °C. At a specified time intervals,- aliquots (5 mL) were withdrawn from the release medium and replaced with fresh PBS in order to keep constant the volume of the release medium. The samples

(2)

where qm (mg/g) is the maximum adsorption of drug and KL(L mg− 1) is 136

Applied Clay Science 153 (2018) 134–143

J. Kurczewska et al.

Fig. 3. (a) FTIR spectra of halloysite materials, (b) XRD patterns of Hal_PAMAM without and with the drugs loaded, (c) Nitrogen adsorption-desorption isotherms for Hal and Hal_PAMAM.

137

Applied Clay Science 153 (2018) 134–143

J. Kurczewska et al.

3. Results and discussion

were analyzed on a UV–Vis spectrophotometer to determine the amount of released drugs. Total amounts of the drug released (Ft) were calculated as follows:

3.1. Characterization of the materials studied

t−1

Ft = Vm Ct +

∑ Va Ci i=0

Dendrimer-functionalized halloysite nanotubes (Hal_PAMAM) were prepared according to the procedure presented in Fig. 2. Raw halloysite (Hal) was initially functionalized with 3-aminopropyltrimethoxysilane (Hal_APTS). According to the elemental analysis, for Hal_APTS, the molar ration of C/N was 3, suggesting that complete hydrolysis of APTS occurred. Then, polyamidoamine dendrimer was covalently attached to the support by using a linker, suberic acid bis(N-hydroxysuccinimide ester) (DSS). The nanomaterials studied (Hal, Hal_APTS, Hal_PAMAM) were characterized using several different physicochemical methods. Organic functionalization of halloysite surfaces was confirmed by FTIR spectroscopy (Fig. 3a). The bands derived from raw halloysite (Hal) are as follow: 3700 and 3625 cm− 1 (stretching vibrations of AlOH), 3527, 3456 and 1650 cm− 1 (stretching and bending vibrations of water molecules), 1052, 776, 758 and 692 cm− 1 (stretching vibrations of SieO), 913 cm− 1 (deformation of OeH in Al-OH), 541 and 472 cm− 1 (bending vibrations of Al-O-Si and Si-O-Si) and 436 cm− 1 (deformation of SieO). The signals appearing at 2943 and 2877 cm− 1 in the spectra of functionalized halloysite are related to the CeH stretching vibrations in alkyl chain. The band at 1509 cm− 1 in the spectrum of Hal_APTS corresponds to deformation of NH2. Two bands at 1655 and 1551 cm− 1 in the spectrum of Hal_PAMAM originate from C]O vibrations of primary amide groups. Additionally, the signals at 3527 and 3456 cm− 1 in functionalized halloysite increase due to stretching vibrations of both - hydroxyl and amine groups. The organic functionalization as well as thermal stability of the materials studied was also confirmed by thermogravimetric measurements. The mass loss in the Hal curve was observed at 100–500 °C as a consequence of elimination of water molecules from halloysite interlayer and dehydroxylation of aluminol groups. The Hal_APTS and Hal_PAMAM materials showed additional mass loss at 225–425 °C caused by a decomposition of organic unit at the halloysite surface. The TGA curves of Hal_PAMAM with noncovalently bound drugs were characterized by another mass loss in the temperature region below 250 °C due to decomposition of immobilized drug molecules (see Supporting Information). Thermogravimetry was also used to estimate the amount of PAMAM dendrimer on the halloysite surfaces and a loading content of the drugs studied in Hal_PAMAM. The percentage of APTS grafted was 6 wt.% that corresponds to 0.34 mmol g− 1. After multiple step synthesis, the amount of the dendrimer in Hal_PAMAM drug carrier was 0.04 mmol g− 1 (3.4 wt.%). The loading values of the drugs studied were as follows: 0.45 mmol g− 1 (16.1 wt.%; CHLG), 1.13 mmol g− 1 (23.3 wt.%; IBU) and 0.39 mmol g− 1 (5.4 wt.%; SAL). The amount of ibuprofen loaded was higher (Tan et al., 2013) or comparable (Li et al., 2015) to the content of the drug in other hybrid carriers based on halloysite. On the other hand, the ability of Hal_PAMAM for encapsulation of salicylic acid was lower than that of a raw halloysite of a different origin (Makaremi et al., 2017). The XRD patterns of Hal and Hal_PAMAM (Fig. 3b) are very similar. It indicates that organic functionalization of halloysite does not affect the structure of the inorganic material. The diffraction reflections at 2 Theta (11.8, 19.9, 24.8, 35.0, 54.5, 62.4) are consistent with the standard structure of halloysite (Joint Committee for Powder Diffraction Studies, JCPDS No. 29-1487). Additionally, the samples contain some associated minerals (kaolinite, gibbsite and quartz; Pasbakhsh et al., 2013). The presence of the drugs loaded on Hal_PAMAM surface was manifested by small reflections (marked with black triangles). The low intensity of these reflections could be explained by lower structural order. The nitrogen adsorption - desorption isotherms of Hal and Hal_PAMAM are shown in Fig. 3c. These isotherms are classified as a type IV with a characteristic hysteresis loop of H3 type, characteristic of mesoporous structure. The specific surface area, SBET, and the total pore

(4)

where Vm and Ct are volume and concentration of the drug at time t, Va is the volume of the sample withdrawn and Ci is the drug concentration at time i (i < t). The drug release data were fitted to several kinetic models. The data were analyzed using zero order (5), first order (6), Higuchi (7) and Hixon-Crowell (8) mathematical models (Costa and Lobo, 2001; Dash et al., 2010; Reddy et al., 2014). The data were also fitted to KorsmeyerPeppas model (9) to understand the drug release mechanism.

Ft = k 0 t

(5)

Ft = 1 − e−k1 t

(6)

Ft = kH t

(7)

3

F0 −

3

Ft = kHC t

Ft = kKP t n

(8) (9)

where Ft is the fraction of the drug released at time “t”, F0 is the initial amount of the drug in the material, k0, k1, kH, kHC and kKP – the release constants of the respective equations, n – the diffusion exponent characteristic of the release mechanism. 2.7. Toxicity investigation For bioassays pure cultures of Acutodesmus acuminatus (Lag.) Tsarenko (syn. Scenedesmus acuminatus (Lag.) Chodat) from the algal bank of Department of Hydrobiology, Faculty of Biology, Adam Mickiewicz University in Poznan were used. The culture of Acutodesmus acuminatus was grown in sterile conditions at 23–25 °C under artificial lighting with white light (4300 lx). The bioassays were carried out in flasks after inserting the culture medium (100 mL) with Acutodesmus acuminatus cells (stabilized phase of growth). Moreover, the toxicity bioassays were conducted with cladocerans Daphnia magna Straus, which were cultivated in accordance with International Standard (ISO) recommendations. Daphnia bioassays were carried out in Petri dishes in water (100 mL) at temperature 20–21 °C under low exposure of light. The bioassay samples contained 50 individuals of Daphnia magna. The amount of Acutodesmus acuminatus cells and Daphnia magna individuals was counted after 0, 8, 24, 72, 148 and 296 h. A number of Acutodesmus acuminatus cells was counted under a light microscope (chamber of 0.2 mm × 0.0625 mm2), while a number of Daphnia magna individuals – under a stereoscopic microscope. The cultures of Acutodesmus acuminatus and Daphnia magna were treated with Hal_PAMAM without any drug, and loaded with chlorogenic acid, ibuprofen and salicylic acid. Three different weights of Hal_PAMAM were used for the bioassays: 0.1790 ± 0.0017 g, 0.3581 ± 0.0033 g and 0.5371 ± 0.0050 g. The samples were dispersed in 100 mL of the culture medium. Additionally, the reference samples without a drug carrier were also studied. 2.8. Statistical analysis The quantitative results were performed in triplicate and expressed as mean ± standard deviation (SD). Statistical significance was tested by One-Way ANOVA with post-hoc Tukey HSD test. The p value < 0.05 was considered as statistical significance. 138

Applied Clay Science 153 (2018) 134–143

J. Kurczewska et al.

the dendrimer is loaded not only at the external surface but also in the lumen. Another source of information is given from the pore size distribution calculated by BJH method. Hal demonstrates three pore populations at about 3.5, 11 and 22 nm, which is consistent with the literature data (Tan et al., 2013). The dendrimer-functionalized halloysite has only two pore populations – 11 and 22 nm – with lower intensity than Hal. It indicates that the organic unit covers internal surface (halloysite lumen) of the material studied. On the other hand, absence of 3.5 nm mesopores is probably due to their complete filling by the organic unit, which in consequence excludes loading of drugs in these pores. According to the Transmission Electron Microscope imaging, raw halloysite is characterized by cylindrical shape (Fig. 4a). A single tube length varies from 200 to 1000 nm, and it has external diameter of 25–50 nm and internal diameter of 9–20 nm. These dimensions were determined statistically based on several TEM images. A multistep modification of halloysite surface with organic unit did not change the tubular structure of the material (Figs. 4b-c). However, the average dimensions of the nanotubes changed. The difference of the average outer diameter is particularly significant, because it increased about 8 nm compared with the raw material. On the other hand, the inner diameter decreased by only about 2 nm. It suggests that the organic unit is grafted on inner and outer surfaces of halloysite. However, it seems that the availability of lumen surface for the dendrimer molecules is limited and probably a larger amount of PAMAM dendrimer is located on the outer surface. 3.2. Adsorption experiments of the model drugs Adsorption isotherm models were used to describe interactive behaviour between halloysite carries and the model drugs. The study of the adsorption capacity of the functionalized halloysite material is a preliminary assessment to determine whether a given drug carrier has a potential for a practical application. The adsorbed compounds are small molecules and they all have carboxyl groups. This should promote the adsorption process on a carrier, Hal_PAMAM, having numerous of amino groups. However, the model drugs are also characterized by different physicochemical properties, which should affect their different adsorption on the materials studied. The experimental data for Hal, Hal_APTS and Hal_PAMAM samples were analyzed using Langmuir and Freundlich models. Langmuir model assumes a monolayer adsorption on a structurally homogeneous support. The sorption sites are energetically equivalent. On the other hand, Freundlich model assumes heterogenous adsorptive energies on a surface of an adsorbent. All values calculated for the isotherm models applied are given in Table 1. The correlation coefficients R2 of Langmuir fitting are higher than those of the Freundlich model. The R2 values are about 0.99 for chlorogenic acid and ibuprofen indicating that Langmuir model is appropriate for describing of adsorption process on all the carriers studied. In case of salicylic acid adsorption, the R2 value for Hal_PAMAM is slightly lower (R2 > 0.96), but still this adsorption

Fig. 4. Micrographs of Hal (TEM - a) and Hal_PAMAM (TEM – b; SEM – c).

volume, Vpore, for Hal are 71.8 m2 g− 1 and 0.31 cm3 g− 1, respectively. Organic functionalization of halloysite surfaces caused a decrease of porosity manifested by reduction in SBET and Vpore values for Hal_PAMAM (29.5 m2 g− 1 and 0.16 cm3 g− 1, respectively). It confirms that Table 1 Comparison of isotherm models for drug adsorption onto Hal, Hal_APTS and Hal_PAMAM. Drug

CHLG

IBU

SAL

Drug carriers

Hal Hal_APTS Hal_PAMAM Hal Hal_APTS Hal_PAMAM Hal Hal_APTS Hal_PAMAM

Langmuir model

Freundlich model

qm (mg/g)

KL (L mg− 1)

R2

1/n

34.64 ± 2.95 54.95 ± 5.12 123.16 ± 4.95 95.56 ± 4.13 136.02 ± 3.67 182.72 ± 9.18 12.17 ± 0.81 16.75 ± 1.06 39.52 ± 0.93

0.014 0.018 0.041 0.018 0.037 0.046 0.008 0.012 0.016

0.993 0.993 0.998 0.989 0.989 0.986 0.999 0.997 0.967

0.64 0.51 0.29 0.25 0.44 0.47 0.20 0.19 0.44

± ± ± ± ± ± ± ± ±

0.001 0.001 0.002 0.002 0.004 0.002 0.001 0.001 0.001

139

± ± ± ± ± ± ± ± ±

0.03 0.07 0.03 0.01 0.03 0.01 0.03 0.02 0.04

KF (mg/g(L mg− 1)1/n)

R2

0.48 2.05 8.03 2.46 3.23 4.26 0.79 1.58 2.08

0.983 0.972 0.964 0.894 0.981 0.929 0.897 0.953 0.893

± ± ± ± ± ± ± ± ±

0.06 0.08 0.12 0.23 0.06 0.33 0.03 0.09 0.06

Applied Clay Science 153 (2018) 134–143

J. Kurczewska et al.

loading into porous structure of Hal carriers. The maximum adsorption capacity on Hal_PAMAM is 123.16 mg/g for CHLG, 182.72 mg/g for IBU and 39.52 mg/g for SAL. Despite this is the first report for drug adsorption on this type of carrier – Hal_PAMAM, the results obtained can be compared with the data present in the literature. The beneficial influence of halloysite functionalization with APTES on loading of ibuprofen has been already reported (Tan et al., 2014). It was proven that IBU content increased from about 11.7 mass% for halloysite to even 14.8 mass% for APTES functionalized halloysite. A stronger affinity of IBU loading was caused by the presence of stronger interactions, i.e., electrostatic attraction between the drug and the amino groups on the carrier surfaces. In the present study, a similar relationship was observed, which was manifested by an increase in qm value from 95.56 mg/g for Hal to 136.02 mg/g for Hal_APTS. In the new material, Hal_PAMAM, the increase in adsorption of all the drugs corresponds to the presence of dendrimer on the Hal surface. The affinity between the drugs studied and PAMAM dendrimer was confirmed by ESI mass spectrometry (see Supplementary Information). The mass spectrum of PAMAM dendrimer in the positive ion mode shows peaks at m/z 831.72, 416.39, 771.63, 717.65 and 657.57 that correspond to [PAMAM + H]+, [PAMAM + 2H]2 + and fragmentation ions ([PAMAM - (NH2CH2CH2NH2) + H]+, [PAMAM – 2 (NH2CH2CH2NH2) + H]+, [PAMAM – 3 (NH2CH2CH2NH2) + H]+,) respectively. The presence of complexes between the drugs studied and the PAMAM dendrimer was observed in the negative ion mode. As a result of PAMAM dendrimer protonation, molecules of the dendrimer and the drugs studied are electrostaticlly attracted to form more stable binding on the surfaces of the carrier compared with drug adsorption on the raw halloysite surfaces. The presence of ionically bonded chlorogenic acid is confirmed by peaks at m/z 1183.92 [PAMAM + CHLGH]−, 591.47 [PAMAM + CHLG-2H]2 − and 1123.83 (fragmentation ion). The ESI MS spectrum of the dendrimer with ibuprofen shows peak at m/z 1035.97 [PAMAM + IBU-H]− and several peaks from fragmentation ions (975.86, 921.85, 861.77). Finally, the formation of complex between PAMAM dendrimer and salicylic acid is demonstrated by peak at m/z 967.84 [PAMAM + SAL-H]− as well as peaks at 907.73, 853.73 and 793.68 (fragmentation ions). In addition to peaks assigned to PAMAM dendrimer-drug complexes, the ESI MS spactra contain also peaks at 353.20 [CHLG-H]−, 205.07 [IBU-H]− and 137.05 [SAL-H]− that confirm further adsorption of drug molecules through intermolecular hydrogen bonding. Therefore, the actual content of the drugs studied in Hal_PAMAM carrier is higher than maximum adsorption capacities calculated from the Langmuir model. Despite well fitness of this mathematical model, the actual drug loading is probably not a monolayer adsorption. The drug molecules anchored to PAMAM dendrimer can further adsorb another molecules through intermolecular hydrogen bonding. The beneficial effect of a dendrimer presence on an inorganic surfaces has been already observed for TiO2 nanotubes functionalized with PAMAM dendrimer of third generation (Torres et al., 2016). The amounts of drug adsorbed were higher not only in relation to the pristine TiO2, but also native PAMAM dendrimers. It indicates a beneficial cooperation between nanotubes and dendrimers on the ability of the hybrid system studied to drug adsorption. Additional information is given from the Freundlich isotherm. The Freundlich constants, KF and n, are related to the adsorption capacity and intensity. The constant KF increased in the following order Hal < Hal_APTS < Hal_PAMAM indicating higher affinity of dendrimer functionalized halloysite for drug adsorption compared to the other carriers studied. The values of 1/n are all smaller than 1, which indicates a normal Langmuir isotherm. The adsorption process was favorable on all the carriers studied.

Fig. 5. Release profiles of CHLG (a), IBU (b) and SAL (c) from halloysite materials (● Hal, ■ Hal_APTS, ♦ Hal_PAMAM).

model better describes the process compared to the Freundlich one. The maximum adsorption capacities, qm, calculated for the Langmuir model are the lowest for Hal and the highest for Hal_PAMAM. It indicates that the presence of PAMAM dendrimer influences the increase of drug adsorption compared to raw halloysite and APTS-functionalized material. The adsorption capacity values differ for each of the model drugs and they increase in the following order: qm(SAL) < qm(CHLG) < qm(IBU). The different values of qm for the drugs studied are caused by differences in the affinity between the drug and the halloysite carrier. It is related to the nature of a drug and a porous structure of the carrier. Salicylic acid molecules are readily to form intramolecular hydrogen bonding, which hinders the further loading on halloysite material. Moreover, chlorogenic acid shows a larger molecule dimension as compared to the other drugs studied, which limits its

3.3. Drug release from the carriers studied The in vitro release studies of the model drugs from halloysite 140

Applied Clay Science 153 (2018) 134–143

J. Kurczewska et al.

release mechanism was analyzed based on the n values in the Korsmeyer-Peppas model. Diffusion exponent lower than 0.45 was obtained for chlorogenic and salicylic acids indicating that the release mechanism was Fickian diffusion. On the other hand, the release of ibuprofen was non-Fickian (anomalous) in nature (0.45 < n < 0.89).

carriers were carried out in PBS solution, pH 7.4. The release profiles demonstrated different behaviour depending on the drug-carrier system involved. The largest difference was observed for chlorogenic acid (Fig. 5a), for which the release rate decreased in the following order Hal > Hal_APTS > Hal_PAMAM. Initially, CHLG was released very fast (in 12 h) and then release rate slowed down significantly. The release profile of the drug from functionalized halloysite carriers was linear at the beginning. A cumulative CHLG release decreased from 90% for Hal to about 55% for Hal_PAMAM. Statistical analysis showed significant difference between cumulative CHLG release from each halloysite carrier (p < 0.05). Presumably, the presence of amino groups in the Hal_APTS carrier and increased number of these groups in Hal_PAMAM affected a stronger affinity between the drug and the carriers. The process of ibuprofen releasing is somewhat different (Fig. 5b). As a consequence of organic functionalization of halloysite surfaces, the IBU release rates slowed down. A similar trend has been observed for other halloysite modifications intended to reduce the IBU release rate (Tan et al., 2014; Li et al., 2015). Unfortunately, surface functionalization with PAMAM dendrimer had no effect on the release profile of ibuprofen. Some differences between Hal_APTS and Hal_PAMAM were noticeable, but they were not statistically significant (p > 0.05). Probably, in this case the IBU release rate is determined not only by a type of interactions (hydrogen bonding or electrostatic interactions) but also a surface on which the drug is loaded (inner lumen or outer surface). Presumably in the Hal_APTS carrier, the drug is mainly inside the lumen, while in Hal_PAMAM it may be mostly located on the outer surface. The profiles of salicylic acid release from the three halloysite carriers are also different (Fig. 5c). The amount of drug released in time from Hal and Hal_APTS in the first 12 h is comparable (p > 0.05). In the following hours, salicylic acid release is slower from Hal_APTS as compared to raw halloysite. In case of Hal_PAMAM, the release rate of SAL decreases with respect to the other materials. Thus, the affinity of the drug to the carrier increases significantly only after functionalization of inorganic surfaces with the dendrimer. The adsorbed compounds are acidic with the following pKa values: 2.66 for CHLG; 5.20 for IBU and 2.97 for SAL. In the PBS solution all the model drugs are ionized (pH > pKa), thus the differences in migration of the compounds to the dissolution medium could be explained by differences in a structure, including various organic units in the molecules of adsorbed drugs, affecting the strength of interactions in each carrier – drug pairs. The experimental release data for Hal_PAMAM material were analyzed using several mathematical models, which are often applied to describe a release of guest molecules from porous materials. The release profiles of CHLG, IBU and SAL from PAMAM - dendrimer functionalized halloysite fitted to zero order (a), first order (b), Higuchi (c), HixsonCrowell (d) and Korsmeyer-Peppas (e) theoretical models are shown in Supporting Information. The values of the respective release constants k, correlation coefficients R2 and diffusion exponents n are collected in Table 2. The release of chlorogenic and salicylic acids was well fitted with Higuchi model (R2 > 0.98), while the release of ibuprofen – with the first order kinetics (R2 > 0.99). In case of other mathematical models, regression coefficients were significantly different from unity. Therefore, the release process of CHLG and SAL is mainly controlled by diffusion, while the IBU release depends on its concentration. The drug

3.4. Toxicity of Hal_PAMAM The in vivo toxicity studies were carried out with model organisms. In addition to in vitro studies using human cell cultures, in vivo toxicity of nanoclays is often investigated by adding the nanomaterials to living organisms (Fakhrullina et al., 2015; Kryuchkova et al., 2016). Here, bioassays with Acutodesmus acuminatus and Daphnia magna were applied. The cultures of A. acuminatus and D. magna were treated with Hal_PAMAM with and without the drugs studied using three different weights of the solid carrier. The bioassays were carried out from 0 to 296 h and compared with the reference sample without halloysite. Regardless of the drug used in the bioassays, the limiting influence of the compounds studied on the living organisms was not observed. A number of A. acuminatus cells (Table 3) and D. magna individuals (Table 4) were similar to those observed for the reference sample. It indicates no influence of the material studied – Hal_PAMAM – on the organisms applied in the bioassays. Some reduction in the number of the organisms were noticed for the samples with higher concentration of the support studied. However, the calculated numbers have no statistically significant difference (p > 0.05) in comparison to the other samples with Hal_PAMAM and the reference sample. Additionally, mobility of D. magma individuals decreased during the experiment. It was especially observed for the samples with ibuprofen. Nevertheless, despite slower movements of these organisms, the condition of their life remained unchanged. 4. Conclusions Halloysite nanotubes functionalized with 3-aminopropyltrimethoxysilane (Hal_APTS) and, subsequently with polyamidoamine dendrimer (Hal_PAMAM) have been prepared for adsorption and release of model drug compounds. The presence of the dendrimer had a beneficial effect on adsorption and/or release of all the drugs studied characterized by acidic nature. The amount of chlorogenic acid adsorbed on Hal_PAMAM increased, while its release rate slowed down as compared to Hal and Hal_APTS. The adsorption capacity of Hal_PAMAM was higher for ibuprofen with respect to the other carriers studied. However, the release rate of IBU from both functionalized materials was comparable. On the other hand, release of salicylic acid decreased only after organic functionalization of halloysite surface with PAMAM dendrimer. The Langmuir isotherm was found to be the best in describing the adsorption process of all the model drugs. The release kinetic studies showed that Higuchi model was appropriate to describe a release process of CHLG and SAL, while first order equation – a process of IBU release from Hal_PAMAM material in PBS buffer solution (pH 7.4). There were also differences in the release mechanism. Based on the diffusion exponent n from the Korsmeyer-Peppas model, it was found that CHLG and SAL was released according to the Fickian diffusion mechanism, while IBU - non-Fickian diffusion. Additionally, the material studied –

Table 2 Release kinetic data of chlorogenic acid, ibuprofen and salicylic acid from Hal_PAMAM. Drug

CHLG IBU SAL a

Zero order

First order

Higuchi

Korsmeyer-Peppasa

Hixson-Crowell

k0 (h− 1)

R2

k1 (h− 1)

R2

kH (h− 1/2)

R2

kHC (h− 1/3)

R2

kKP (h− n)

n

R2

0.866 ± 0.006 1.317 ± 0.110 0.642 ± 0.074

0.906 0.839 0.911

0.017 ± 0.002 0.045 ± 0.001 0.022 ± 0.001

0.966 0.991 0.967

8.262 ± 0.805 12.840 ± 0.878 6.116 ± 0.095

0.982 0.950 0.986

−0.021 ± 0.003 0.043 ± 0.007 −0.022 ± 0.002

0.950 0.956 0.963

14.605 ± 1.205 9.806 ± 1.311 4.167 ± 0.356

0.30 ± 0.05 0.59 ± 0.09 0.13 ± 0.03

0.929 0.869 0.965

Data for 60% of drug released.

141

Applied Clay Science 153 (2018) 134–143

J. Kurczewska et al.

Table 3 A number of Acutodesmus acuminatus cells (106 ×/mL) in 100 mL (average value ± 0.01) after addition of different weights of Hal_PAMAM with/without the drugs studied. Cultivation time [h]

8 24 48 72 148 296

Reference sample

4.97 4.98 4.98 5.02 5.46 5.61

0.1790 ± 0.0017 g

0.3581 ± 0.0033 g

0.5371 ± 0.0050 g

No drug

CHLG

IBU

SAL

No drug

CHLG

IBU

SAL

No drug

CHLG

IBU

SAL

4.96 4.97 4.97 5.01 5.43 5.62

4.96 4.96 4.97 4.99 5.23 5.47

4.97 4.97 4.97 4.98 5.35 5.43

4.97 4.97 4.97 4.99 5.33 5.41

4.97 4.96 4.96 4.98 4.99 5.12

4.96 4.96 4.96 4.98 5.11 5.20

4.97 4.97 4.97 4.98 5.08 5.27

4.96 4.96 4.96 4.98 5.28 5.34

4.97 4.95 4.96 4.99 4.99 5.23

4.96 4.96 4.95 4.97 5.09 5.16

4.95 4.95 4.95 4.96 5.01 5.12

4.96 4.95 4.95 4.96 5.09 5.29

Table 4 A number of Daphnia magna individuals in 100 mL (average value ± 1) after addition of different weights of Hal_PAMAM with/without the drugs studied. Cultivation time [h]

8 24 48 72 148 296

Reference sample

50 50 50 50 48 47

0.1790 ± 0.0017 g

0.3581 ± 0.0033 g

0.5371 ± 0.0050 g

No drug

CHLG

IBU

SAL

No drug

CHLG

IBU

SAL

No drug

CHLG

IBU

SAL

50 50 50 50 49 48

50 50 50 50 48 48

50 50 50 50 48 47

50 50 50 50 48 47

50 50 50 50 49 47

50 50 50 50 47 46

50 50 50 50 46 45

50 50 50 50 47 45

50 50 50 50 48 46

50 50 50 50 46 45

50 50 50 49 44 44

50 50 50 50 45 44

Hal_PAMAM – without and with the model drugs, demonstrated a neutral character to the living organisms applied in the bioassays. The results obtained indicate that functionalization of halloysite surfaces with polyamidoamine dendrimer can improve the properties of a drug carrier based on halloysite. This type of a hybrid system could be used as an effective and versatile carrier of various drugs characterized by a low molecular size and the presence of a carboxyl groups. Therefore, Hal_PAMAM is a promising material for biomedical applications.

Pharm. Sci. 13, 123–133. Dash, S., Murthy, P.N., Nath, L., Chowdhury, P., 2010. Kinetic modeling drug delivery release from controlled drug delivery systems. Acta Pol. Pharm. 67, 217–223. Du, M., Guo, B., Jia, D., 2010. Newly emerging applications of halloysite nanotubes: a review. Polym. Inter. 59, 574–582. Fakhrullin, R.F., Lvov, Y.M., 2016. Halloysite clay nanotubes for tissue engineering. Nanomed. 11, 2243–2246. Fakhrullina, G.I., Akhatova, F.S., Lvov, Y.M., Fakhrullin, R.F., 2015. Toxicity of halloysite clay nanotubes in vivo: a Caenorhabditis elegans study. Environ. Sci. Nano 2, 54–59. Hanif, M., Jabbar, F., Sharif, S., Abbas, G., Farooq, A., Aziz, M., Churchman, J., 2016. Clay Miner. 51, 469–477. Joussein, E., Petit, S., Churchman, J., Theng, B., Righi, D., Delvaux, B., 2005. Halloysite clay minerals - a review. Clay Miner. 40, 383–426. Khodzhaeva, V., Makeeva, A., Ulyanova, V., Zelenikhin, P., Evtugyn, V., Hardt, M., Rozhina, E., Lvov, Y., Fakhrullin, R., Ilinskaya, O., 2017. Binase immobilized on halloysite nanotubes exerts enhanced cytotoxicity toward human colon adenocarcinoma cells. Front. Pharmacol. 8, 631. Kryuchkova, M., Danilushkina, A., Lvov, Y., Fakhrullin, R.F., 2016. Evaluation of toxicity of nanoclays anf grapheme oxide in vivo: a Paramecium caudatum study. Environ. Sci. Nano 3, 442–452. Kurczewska, J., Gierczyk, B., Cegłowski, M., Schroeder, G., 2009. Inorganic magnetic support for sodium cation scavenging. Thin Solid Films 517, 6076–6080. Kurczewska, J., Pecyna, P., Ratajczak, M., Gajęcka, M., Schroeder, G., 2017. Halloysite nanotubes as carriers of vancomycin in alginate-based wound dressing. Saudi. Pharm. J. 25, 911–920. Lazzara, G., Riela, S., Fakhrullin, R.F., 2017. Clay-based drug-delivery systems: what does the future hold? Therap. Deliv. 8, 633–646. Li, H., Zhu, X., Zhou, H., Zhong, S., 2015. Functionalization of halloysite nanotubes by enlargement and hydrophobicity for sustained release of analgesic. Colloids Surf. A Physicochem. Eng. Aspects 487, 154–161. Lvov, Y.M., Shchukin, D.G., Möhwald, H., Price, R.R., 2008. Halloysite clay nanotubes for controlled release of protective agents. ACS Nano 2, 814–820. Lvov, Y.M., DeVilliers, M.M., Fakhrullin, R.F., 2016a. The application of halloysite tubule nanoclay in drug delivery. Expert Opin. Drug Deliv. 13, 977–986. Lvov, Y., Wang, W., Zhang, L., Fakhrullin, R., 2016b. Halloysite clay nanotubes for loading and sustained release of functional compounds. Adv. Mater. 28, 1227–1250. Makaremi, M., Pasbakhsh, P., Cavallaro, G., Lazzara, G., Aw, Y.K., Lee, S.M., Milioto, S., 2017. Effect of morphology and size of halloysite nanotubes on functional pectin bionanocomposites for food packaging applications. ACS Appl. Mater. Interfaces 9, 17476–17488. Massaro, M., Riela, S., Lo Meo, P., Noto, R., Cavallaro, G., Milioto, S., Lazzara, G., 2014. Functionalized halloysite multivalent glycocluster as a new drug delivery system. J. Mater. Chem. B 2, 7732–7738. Massaro, M., Amorati, R., Cavallaro, G., Guernelli, S., Lazzara, G., Milioto, S., Noto, R., Poma, P., Riela, S., 2016a. Direct chemical grafted curcumin on halloysite nanotubes as dual-responsive prodrug for pharmacological applications. Colloids Surf. B Biointerfaces 140, 505–513. Massaro, M., Riela, S., Baiamonte, C., Blanco, J.L.J., Giordano, C., Lo Meo, P., Milioto, S., Noto, R., Parisi, F., Pizzolanti, G., Lazzara, G., 2016b. Dual drug-loaded halloysite hybrid-based glycocluster for sustained release of hydrophopic molecules. RSC Adv. 6, 87935–87944. Massaro, M., Lazzara, G., Milioto, S., Noto, R., Riela, S., 2017. Covalently modified halloysite clay nanotubes: synthesis, properties, biological and medical applications. J.

Conflicts of interest The authors declare that they have no personal financial or nonfinancial conflicts of interest. Acknowledgments This work was financially supported by the National Science Centre, Poland, under grant number 2016/21/B/ST4/02082. Appendix A. Supplementary data Supplementary material associated with this article: Structures of chlorogenic acid, ibuprofen and salicylic acid; TGA curves of the samples studied; ESI-MS spectra of PAMAM dendrimer and the complexes between the dendrimer and the drugs studied; Drug release data for Hal_PAMAM fitted to various mathematical models. Supplementary data associated with this article can be found in the online version, at https://doi.org/10.1016/j.clay.2017.12.019. References Abdullayev, E., Lvov, Y., 2013. Halloysite clay as a ceramic “skeleton” for functional biopolymer composites with sustained drug release. J. Mater. Chem. B 1, 2894–2903. Abdullayev, E., Joshi, A., Wei, W., Zhao, Y., Lvov, Y., 2012. Enlargment of halloysite clay nanotube lumen by selective etching of aluminum oxide. ACS Nano 6, 7216–7226. Aguzzi, C., Cerezo, P., Viseras, C., Caramella, C., 2007. Use of clays as drug delivery systems: possibilities and limitations. Appl. Clay Sci. 36, 22–36. Chao, C., Liu, J., Wang, J., Zhang, Y., Zhang, B., Zhang, Y., Xiang, X., Chen, R., 2013. Surface modification of halloysite nanotubes with dopamine for enzyme immobilization. ACS Appl. Mater. Inter. 5, 10559–10564. Churchman, G.J., Pasbakhsh, P., Hillier, S., 2016. The rise and rise of halloysite. Clay Miner. 51, 303–308. Costa, P., Lobo, J.M.S., 2001. Modeling and comparison of dissolution profiles. Eur. J.

142

Applied Clay Science 153 (2018) 134–143

J. Kurczewska et al.

Tan, D., Yuan, P., Annabi-Bergaya, F., Liu, D., Wang, L., Liu, H., He, H., 2014. Loading and in vitro release of ibuprofen in tubular halloysite. Appl. Clay Sci. 96, 50–55. Tan, D., Yuan, P., Liu, D., Du, P., 2016. Surface modifications of halloysite. In: Yuan, P., Thill, A., Bergaya, F. (Eds.), Developments in Clay Science. vol. 7. Elsevier, pp. 167–201 (Chapter 8). Torres, C.C., Campos, C.H., Diáz, C., Jiménez, V.A., Vidal, F., Guzmán, L., Alderete, J.B., 2016. PAMAM-grafted TiO2 nanotubes as novel versatile materials for drug delivery applications. Mater. Sci. Eng. C 65, 164–171. Tu, J., Cao, Z., Jing, Y., Fan, Ch., Zhang, Ch., Liao, L., Liu, L., 2013. Halloysite nanotube nanocomposite hydrogels with tunable mechanical properties and drug release behavior. Composites. Sci. Technol. 85, 126–130. Wang, Q., Zhang, J., Zheng, Y., Wang, A., 2014. Adsorption and release of ofloxacin from acid- and heat-treated halloysite. Colloids Surf. B Biointerfaces 113, 51–58. Yendluri, R., Lvov, Y., De Villiers, M.M., Vinokurov, V., Naumenko, E., Tarasova, E., Fakhrullin, R., 2017a. Paclitaxel encapsulated in Halloysite clay nanotubes for intestinal and intracellular delivery. J. Pharm. Sci. 106, 3131–3139. Yendluri, R., Otto, D.P., De Villiers, M.M., Vinokurov, V., Lvov, Y.M., 2017b. Application of Halloysite clay nanotubes as a pharmaceutical excipient. Int. J. Pharm. 521, 267–273. Yuan, P., Southon, P.D., Liu, Z., Green, M.E.R., Hook, J.M., Antill, S.J., Kepert, C.J., 2008. Functionalization of halloysite clay nanotubes by grafting with γ-aminopropyltriethoxysilane. J. Phys. Chem. C 112, 15742–15751. Yuan, P., Tan, D., Annabi-Bergaya, F., 2015. Properties and applications of halloysite nanotubes: recent research advances and future prospects. Appl. Clay Sci. 112-113, 75–93.

Mater. Chem. B 5, 2867–2882. Narkiewicz, U., Pełech, I., Podsiadły, M., Cegłowski, M., Schroeder, G., Kurczewska, J., 2010. Preparation and characterization of magnetic carbon nanomaterials bearing APTS–silica on their surface. J. Mater. Sci. 45, 1100–1106. Naumenko, E.A., Guryanov, I.D., Yendluri, R., Lvov, Y.M., Fakhrullin, R.F., 2016. Clay nanotube-biopolymer composite scaffolds for tissue engineering. Nano 8, 7257–7271. Pasbakhsh, P., Churchman, G.J., Keeling, J.L., 2013. Characterisation of properties of various halloysites relevant to their use as nanotubes and microfibre fillers. Appl. Clay Sci. 74, 47–57. Rawtani, D., Agrawal, Y.K., 2012. Multifarious applications of halloysite nanotubes: a review. Rev. Adv. Mater. Sci. 30, 282–295. Reddy, N.S., Sowmya, S., Bumgardner, J.D., Chennazhi, K.P., Biswas, R., Jayakumar, R., 2014. Tetracycline nanoparticles loaded calcium sulphate composite beads for periodontal management. Biochim. Biophys. Acta 1840, 2080–2090. Sato, K., Anzai, J.I., 2013. Dendrimers in layer-by-layer assemblies: synthesis and applications. Molecules 18, 8440–8460. Shahamati Fard, F., Akbari, S., Pajootan, E., Arami, M., 2016. Enhanced acidic dye adsorption onto the dendrimer-based modified halloysite nanotubes. Desal. Water Treat. 57, 26222–26239. Shi, Y.F., Tian, Z., Zhang, Y., Shen, H.B., Jia, N.Q., 2011. Functionalized halloysite nanotube-based carrier for intracellular delivery of antisense oligonucleotides. Nano Express 6, 608. Tan, D., Yuan, P., Annabi-Bergaya, F., Yu, H., Liu, D., Liu, H., He, H., 2013. Natural halloysite nanotubes as mesoporous carriers for the loading of ibuprofen. Micropor. Mesopor. Mater. 179, 89–98.

143