Selective desorption characteristics of halloysite nanotubes for anionic azo dyes

Abstract

Halloysite nanotubes show selective desorption characteristics for anionic azo dyes. The dye desorption behaviors from halloysite using deionized water, strong acid and strong base were studied. Methyl orange was easily and completely eluted by deionized water, whereas Congo red is hardly desorbed even by strong acid or base.

---

One-dimensional tubular nanomaterials have received much attention in recent years.¹ Due to suitable nanopores (usually with mesopores), they have been widely used in many fields, such as nano-reactors,² catalysts,³ adsorbents⁴ and hydrogen storage materials.⁵ Among these materials, halloysite is a natural layered silicate with the same crystal structure as kaolinite, i.e. the ratio of silicon–oxygen tetrahedral layers to aluminum–oxygen octahedral layers is 1 : 1. Due to different forces in the process of formation, the sheets of halloysite curl in the direction of the aluminum–oxygen octahedral layer, resulting in the unique tubular configuration. The inner and outer surface of the tube are composed of Al–OH and Si–O–Si, respectively.⁶ The halloysite nanotube is a few hundred nanometers to two micrometers in length, 30–100 nm in outer diameter and 10–40 nm in inner diameter. The ideal formula of halloysite nanotube is Al₂Si₂O₅(OH)₄·nH₂O, where n = 0 and n = 2 represent halloysite nanotubes with the layer spacing of 0.7 nm and 1 nm, respectively.⁷ Because of its tubular structure, abundant mesopores, and excellent chemical and thermal stability, has been currently applied in the fields of catalysis,⁸ drug delivery,⁹ wastewater treatment¹⁰ and nanocomposites.¹¹

Recently, halloysite has been used as sorbent for azo-dyes in the literature,¹² but desorption of the dye/recovery of the halloysite has not been studied yet. Here we show a significant difference in desorption behavior of two common anionic azo dyes, methyl orange (MO) and Congo red (CR) from halloysite. The difference is mainly determined by the functional group of the two dye molecules.

The images of pristine halloysite, dye-loaded samples before and after thoroughly washed with deionized water (DI water) are shown in Fig. 1. In comparison with halloysite, the dye-loaded samples show their characteristic colors, implying good adsorption of halloysite for the dyes. After sufficient washing with DI water, the MO-loaded sample almost completely faded, whereas the color of the Congo red-loaded sample did not change significantly. This implied a difference in binding mechanisms of halloysite to the two dyes and a stronger interaction between Congo red and halloysite.

Fig. 1 Pristine halloysite (a), methyl orange-loaded samples before (b) and after washing (c), Congo red-loaded samples before (d) and after washing (e).

Based on the above finding about the selective desorption properties of halloysite, we firstly investigated the desorption properties and reusability of halloysite for methyl orange (Fig. 2). The adsorbed methyl orange can be removed almost completely by deionized water. With the increase in the cycles of adsorption–desorption, no significant change in the adsorption capacity and desorption efficiency of halloysite for methyl orange was observed. These results indicate that halloysite is a new recyclable adsorbent for methyl orange.

Fig. 2 Repeated adsorption/desorption of halloysite for methyl orange.

We also studied desorption of Congo red using concentrated HCl (36%), HNO₃ (65%), H₂SO₄ (98%), and 4 M NaOH. The CR-loaded sample after being washed with DI water (hereinafter referred to hybrid) was immersed in the various solutions for 24 h at room temperature according to solid-to-liquid ratio of 1 : 100 and then washed by centrifugation until supernatant was colorless and its pH value was close to 7. The obtained solids were dried and milled to be measured by solid UV-vis spectrophotometer (Fig. 3). For the dye-loaded sample four significant absorption bands were observed at 245, 350, 500 and 650 nm in UV-vis spectra. The first three bands belong to characteristic absorption bands of Congo red also found for samples in aqueous solutions.¹³ The band at 650 nm implies that a small amount of Congo red molecules existed in halloysite in azonium form.¹⁴ The UV-vis spectra change significantly after treated by strong acids. The characteristic bands at 500 and 650 nm disappear, the intensity of the band at 245 increases, and the band at 350 nm shifts to ∼360 nm. Congo red, a pH indicator, is blue at pH < 3.0 and red when at pH > 5.2. After acid treatment, Congo red molecules in the halloysite are likely to change from a non-protonated form to a protonated form,¹⁵ which could be demonstrated by the colors of the samples immersed in strong acid solutions for 24 h (a shift from red to blue was observed except for samples submerged in HNO₃ due to its strong oxidizing property, Fig. S1†). After washing thoroughly by DI water, the colors of the hybrids become blue-gray (Fig. S2†), which implies that the original acid form partially or fully changes to other forms. In order to better distinguish the color characteristics of the samples, CIE L*a*b* values of the samples were measured. As shown in Table S1,† after the acid and alkali treatment, the values of L* (brightness) of the samples all increased but the relative changes were limited, indicating that these treatments do not result in significant removal of Congo red molecules in hybrid. This should be consistent with the result of UV-visible spectra (Fig. 3). Compared to the hybrid, the position of the a* and b* values of the acid-treated samples slid closer to the median value between red and green/yellow and blue, indicated by a loss of intensity in the yellow and red hues.¹⁶ The L*a*b* coordinates of the untreated and alkali treated samples are very close, indicating that strong base did not have a substantial effect on the color of the hybrid, which can be verified by an entirely similar spectra for the two samples (Fig. 3).

Fig. 3 UV-vis spectra of halloysite, Congo red/halloysite hybrid and strong acid- and base-treated samples.

It is noteworthy that even if the sample was sufficiently washed to neutral, colors of the hybrids were still significantly different compared with that of the original hybrid. This indicated that Congo red in the halloysite could be chemically and irreversibly changed by acid treatment, which was reflected by the change of the UV-vis spectra mentioned above. According to the intensity of the bands, it can be concluded that the hybrid material has a good chemical resistance to the three acids with the order of H₂SO₄ > HCl > HNO₃. For the sample immersed in strong base, the position and intensity of its characteristic bands showed no significant change. Only the band at 650 nm disappeared, implying that only a small amount of Congo red were desorbed and most of the dye molecules in the halloysite were still in the basic form after alkali soak. This means that the hybrid is also very resistant to strong alkaline solution. In order to further identify the impact of strong acid and base on halloysite, we checked the UV-vis spectra and XRD patterns of the acid and base-treated halloysites under otherwise identical conditions in the absence of dyes. It can be noticed that for natural halloysite, the spectrum shape and the peak position in UV-visible region both changed little after acid or base treatment, only the peak intensity increased by different degrees (Fig. S3†). However, this change had almost no effect on the UV-vis spectra of the acid- or base-treated hybrids due to very weak peak intensity of the blank samples as compared with the hybrid. Obvious change for the positions of characteristic peaks of halloysite in XRD pattern was not found (Fig. S4†), indicating that the acid and base treatment did not affect the crystal structure of halloysite. These results definitively confirmed that under the experimental conditions, the introduction of strong acids and base into the Congo red/halloysite hybrid can not have an substantial impact on halloysite. Therefore, the above explanation about the UV-visible spectra of the samples based on the molecular structure of Congo red is reliable.

The XRD results (Fig. S5†) showed that halloysite has (001) diffraction reflection corresponded to a multilayer wall spacing of 0.74 nm, which identified it as halloysite-7 Å.¹⁷ After adsorption, the characteristic reflections of halloysite have almost no change, suggesting that the dye molecules do not affect crystal structure of halloysite, that is, dye molecules were not inserted into the inter-layer region. So they could only exist in the external or inner surface of halloysite nanotubes.

Transmission electron microscopy (TEM) images revealed that halloysite has a hollow tubular configuration. The nanotubes have an external diameter of 40–60 nm, an inner diameter of 15–40 nm and a wall thickness of about 15–30 nm (Fig. 4a and c), which was identical to those of previously reported samples.^6,7,9 After adsorption for Congo red, the outer diameter and inner diameter of the nanotube became larger and smaller, respectively (Fig. 4b and d). Note that the variation degree of the latter was more obvious than that of the former, indicating Congo red molecules have been mainly loaded into the lumen. The TEM results demonstrated unambiguously that Congo red was mainly in the interior of nanotube. We selected halloysite, physically grinded sample and the hybrid to determine TG-DTG (Fig. S6†). For the physically grinded sample and the hybrid, the similar weight loss variation occurred. However, there were significant differences in the weight loss temperature. The former showed an exothermic peak at 386 °C corresponding to the decomposition temperature of methyl orange, but at 423 °C, which is likely to be related to the following fact for the latter. Methyl orange molecules were mainly in the lumen of halloysite and the nanotubes act as a buffer to thermal conductivity, resulted in a significant enhancement of dye thermal decomposition temperature. The TG-DTG results well proved methyl orange molecules are mainly present in the lumen of halloysite nanotube. The above results provided an important basis for clarifying the adsorption mechanisms of the two dyes on the halloysite.

Fig. 4 TEM images of (a and c) halloysite and (b and d) Congo red/halloysite hybrid.

In this study, the neutral initial dye solutions (pH = 7) were selected. The pHs of the supernatants became 7.13 and 8.05 for methyl orange and Congo red, respectively, at adsorption equilibrium. This means the outer surface of the nanotubes is always negatively charged during the adsorption, while the inner surface contains a number of positively charged sites (Al–OH₂⁺).⁸ Based on the previous conclusion about dye's place on the halloysite (mainly in the lumen of the nanotube), methyl orange and Congo red are both anionic dyes and contain –SO₃⁻ group, which can interact with Al–OH₂⁺ situated in halloysite lumen by electrostatic attraction. However, this force (physical adsorption) is so weak that the adsorbed dye can be easily desorbed by washing with DI water. If both the two dye molecules are adsorbed using this approach, the selective desorption would be impossible. We noted that Congo red contains –NH₂ group, which did not contain in methyl orange molecule. To verify if a hydrogen bond between the O in AlOAl and the H of NH₂ existed, ¹H NMR of Congo red and hybrid were tested and the results shown in Fig. S7.† It can be observed that compared with Congo red, the positions of various hydrogen contained in Congo red molecule located on halloysite all shifted to high field, but the extent of this shift was very limited. As we all know the formation of intermolecular hydrogen bond results in a shift to a lower field for hydrogen of molecule in ¹H NMR spectrum. So ¹H NMR result confirmed that the hydrogen bonding interaction did not exist in the hybrid. Even the presence of a small amount of such hydrogen bonding in the hybrid material, they would be positively removed in a subsequent immersion in strong acids. However, Congo red/halloysite hybrid exhibited excellent chemical resistance, which implied that there must be another more strong interaction. After strong acid treatment, ¹H NMR spectra of three acid-treated samples were similar, but significant changes in peak positions occurred as compared with that of the hybrid (Fig. S7†). This demonstrated the hybrids changed by strong acid treatment. To determine whether Congo red was degraded by these acids, we performed a control experiment, i.e., solid Congo red was immersed into the three acids for 24 h, and then dried and milled to give a blue powder. When the powders encountered the alkaline solution, it rapidly turned red. This suggested that the acid treatment did not cause degradation of the dye molecules in the hybrid under the experimental conditions. This is a strong evidence for Congo red/halloysite hybrid with excellent acid erosion endurance. Furthermore, taking into account that Congo red molecule changed to the protonated form under acidic conditions, namely the H⁺ ions were added to amino groups, this change was unlikely to have such a huge change in the hydrogen spectrum. In addition, we noted that there were still some different peaks for the three acid-treated samples in ¹H NMR spectra. According to the above results, it can be speculated that unique interaction between Congo red molecules and halloysite must exist in the acid-treated samples, and the interaction mode were also different for each sample.

To further clarify the interaction mechanism, halloysite and the hybrid were measured by X-ray photoelectron spectroscopy (XPS) and the results are shown in Fig. S8†. Halloysite is mainly composed of oxygen, silicon and aluminum, of which the silica-to-alumina ratio was about 1.18, higher than the theoretical value (1 : 1). This mainly resulted from the presence of an outer surface riched in silicon layer.⁹ A small amount of carbon was also detected in halloysite, which was likely to be introduced during sample preparation by using carbon tape to fix sample in place. The intensity of C characteristic peak (C_1s) significantly enhanced and a weak N_1s peak appeared after sufficient washing, which proved the presence of Congo red. For the hybrid, the N_1s peak at a binding energy of 399.9 eV attributed to N=N group in Congo red¹⁸ was observed in the hybrid. It meaned that the nitrogen atom in N=N group seemed not to be adsorbed onto halloysite surface by chemical interaction. It can be clearly observed from the Si_2p and Al_2p detail spectra of the samples (Fig. S9† and 5) that after adsorption, the peak bond energy of an Si_2p of a silicon–oxygen bond and an Al_2p of aluminum–oxygen bond shifted respectively 0.28 eV and 0.37 eV to higher energy area, while the energy resolution of the X-ray photoelectron spectroscopy used in this experiment is 0.48 eV, so the above change is probably caused by the accuracy of the instrument itself. We noted that Si_2p of the halloysite before and after adsorption both presented a single peak, which can not be split by the XPS peak 4.1 program, indicating that the Congo red did not interact strongly with silicon contained in the halloysite. The Al_2p peak located in 74.17 eV in XPS spectra was unimodal pattern, which can be splitted into the two new peaks (74.96 eV and 74.23 eV) for the hybrid. Based on the peak area ratio of 1 : 2, they should belong to 2p_1/2 and 2p_3/2, respectively. The XPS results demonstrated that strong interactions between aluminum of halloysite and certain elements in Congo red molecules undoubtedly occurred, that is, the aluminum combined with the elements having electron-withdrawing effect (such as the nitrogen contained in amino group) in addition to with the oxygen on the surface of the nanotubes, to reduce the outer electron cloud density, thus resulting in the change of electron binding energy of Al_2p.

Fig. 5 Al_2p detail spectra of halloysite and Congo red/halloysite hybrid.

We know that the probing depth of XPS was less than 10 nm, and most of the wall thickness of halloysite nanotubes exceeded this size. However, it must be mentioned that halloysite nanotubes inevitably showed random and disordered arrangement during the XPS measurement, part of the HNTs formed a certain angle with the probe, which ensured the probe detecting inner surface of nanotubes. So, although the results obtained by XPS were not very accurate, it still can be used as an important basis for judging interaction mode of Congo red dye and the inner wall, i.e. aluminum oxide layer of halloysite nanotube. These results implied the formation of new chemical bonds between Congo red and halloysite. It was speculated that a strong coordination, a kind of chemical bond, between N of the amino and Al located on the inner wall of the nanotubes may occur. The proposed adsorption mechanisms of the two dye molecules onto halloysite are shown in Fig. 6.

Fig. 6 Schematic illustration of the interactions between halloysite nanotube and the two anionic azo dyes.

It should be noted that the hybrids quickly changed from red to blue when they were immersed in strong acids (except nitric acid, Fig. S1†), i.e., Congo red molecules in the hybrids were rapidly converted from a non-protonated form to a protonated form. In the subsequent washing process the materials were always blue even when the pH of the supernatant was closed to neutral. Considering the fact that Congo red should be red under such pH condition, so this indicated that Congo red molecules in the hybrid were irreversibly changed during acid treatment, which may be related to a more stable complexation between the protonated form of Congo red and halloysite as compared with the non-protonated form. In addition, after drying, the colors of the above samples changed from blue to blue-gray (Fig. S2†). It is a reminder that the drying process (80 °C, the removal of adsorbed water) is likely to further affect the existence forms of Congo red in halloysite. When the acid-treated hybrids were submerged again in concentrated hydrochloric acid or base, colors of the samples remained unchanged. This implied that interaction mode of Congo red and halloysite in the acid-treated samples obtained by drying should be the most stable. Such result was quite interesting, and it provided a new idea for preparing pigments with stable colors.

In conclusion, halloysite has a distinct desorption behavior for the two anionic dyes. A high desorption rate and excellent reusability was obtained for methyl orange, in contrast, Congo red desorption is extremely difficult even in strong acids or strong alkaline environment. The two dye molecules both mainly existed in halloysite lumen. For methyl orange, electrostatic attraction is the dominant mechanism, while for Congo red, there was a strong coordination between nitrogen of the amino groups in molecules and aluminum located on the inner surface tubes. These findings showed that halloysite clay has an attractive prospect in many fields such as the adsorbents in dye wastewater treatment and new hybrid pigments. Our ongoing work is to select more anionic dyes with different functional groups to further clarify the reasons for this selective desorption characteristics.

This work was supported by the National Natural Science Foundation of China (no. 20903070). The authors are grateful to the anonymous referees for their constructive comments and suggestions, which materially improved this work.

References

1. V. N. Popov, Mater. Sci. Eng., R, 2004, 43, 61; J. Jang and H. Yoon, Chem. Commun., 2003, 720.
2. M. Remškar, A. Mrzel, M. Viršek and A. Jesih, Adv. Mater., 2007, 19, 4276; A. N. Khlobystov, ACS Nano, 2011, 5, 9306.
3. J. M. Planeix, N. Coustel, B. Coq, V. Brotons, P. S. Kumbhar, R. Dutartre, P. Geneste, P. Bernier and P. M. Ajayan, J. Am. Chem. Soc., 1994, 116, 7935.
4. Y. Cai, G. Jiang, J. Liu and Q. Zhou, Anal. Chem., 2003, 75, 2517.
5. F. L. Darkrim, P. Malbrunot and G. P. Tartaglia, Int. J. Hydrogen Energy, 2002, 27, 193; L. Schlapbach and A. Züttel, Nature, 2001, 414, 353.
6. Y. Joo, J. Sim, Y. Jeun, S. U. Lee and D. Sohn, Chem. Commun., 2013, 49, 4519.
7. P. Yuan, P. D. Southon, Z. Liu, M. E. Green, J. M. Hook, S. J. Antill and C. J. Kepert, J. Phys. Chem. C, 2008, 112, 15742.
8. R. Zhai, B. Zhang, L. Liu, Y. Xie, H. Zhang and J. Liu, Catal. Commun., 2010, 12, 259.
9. W. O. Yah, A. Takahara and Y. M. Lvov, J. Am. Chem. Soc., 2012, 134, 1853.
10. P. Luo, Y. Zhao, B. Zhang, J. Liu, Y. Yang and J. Liu, Water Res., 2010, 44, 1489.
11. W. O. Yah, H. Xu, H. Soejima, W. Ma, Y. Lvov and A. Takahara, J. Am. Chem. Soc., 2012, 134, 12134; U. A. Handge, K. Hedicke-Höchstötter and V. Altstädt, Polymer, 2010, 51, 2690.
12. R. Liu, K. Fu, B. Zhang, D. Mei, H. Zhang and J. Liu, J. Dispersion Sci. Technol., 2012, 33, 711; P. Sharma, H. Kaur, M. Sharma and V. Sahore, Environ. Monit. Assess., 2011, 183, 151; Y. Zhao, E. Abdullayev, A. Vasiliev and Y. Lvov, J. Colloid Interface Sci., 2013, 406, 121; P. Yuan, P. D. Southon, Z. Liu and C. J. Kepert, Nanotechnology, 2012, 23, 375705.
13. J. Wang, R. H. Li, Z. H. Zhang, W. Sun, R. Xu, Y. P. Xie, Z. Q. Xing and X. D. Zhang, Appl. Catal., A, 2008, 334, 227.
14. E. Pigorsch, A. Elhaddaoui and S. Turrell, Spectrochim. Acta, Part A, 1994, 50, 2145.
15. Z. Yermiyahu, I. Lapides and S. Yariv, Clay Miner., 2003, 38, 483.
16. Q. Gao, X. Wu, Y. Fan and X. Zhou, Dyes Pigm., 2012, 95, 534.
17. G. Tarì, I. Bobos, C. S. F. Gomes and J. M. F. Ferreira, J. Colloid Interface Sci., 1999, 210, 360.
18. C. D. Wagner, W. M. Riggs, L. E. Davis, J. F. Moulder and G. E. Muilenberg, Handbook of X-ray photoelectron spectroscopy, Physical Electronics Inc., Eden Prairie, 1979, p. 40.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c3ra47561a

---