Research ArticleORGANIC CHEMISTRY

Near-IR squaraine dye–loaded gated periodic mesoporous organosilica for photo-oxidation of phenol in a continuous-flow device

See allHide authors and affiliations

Science Advances  11 Sep 2015:
Vol. 1, no. 8, e1500390
DOI: 10.1126/sciadv.1500390

Abstract

Periodic mesoporous organosilica (PMO) has been widely used for the fabrication of a variety of catalytically active materials. We report the preparation of novel photo-responsive PMO with azobenzene-gated pores. Upon activation, the azobenzene gate undergoes trans-cis isomerization, which allows an unsymmetrical near-infrared squaraine dye (Sq) to enter into the pores. The gate closure by cis-trans isomerization of the azobenzene unit leads to the safe loading of the monomeric dye inside the pores. The dye-loaded and azobenzene-gated PMO (Sq-azo@PMO) exhibits excellent generation of reactive oxygen species upon excitation at 664 nm, which can be effectively used for the oxidation of phenol into benzoquinone in aqueous solution. Furthermore, Sq-azo@PMO as the catalyst was placed inside a custom-built, continuous-flow device to carry out the photo-oxidation of phenol to benzoquinone in the presence of 664-nm light. By using the device, about 23% production of benzoquinone with 100% selectivity was achieved. The current research presents a prototype of transforming heterogeneous catalysts toward practical use.

Keywords
  • microfluidic device
  • periodic mesoporous organosilica
  • photo-oxidation
  • singlet oxygen
  • squaraine dye

INTRODUCTION

Periodic mesoporous organosilica (PMO) prepared by the polycondensation of small hydrocarbon-bridged silane molecules has been proven to be a promising material for various applications including catalysis (14). Relatively hydrophobic nanoconfinement inside the mesopores inherited from the hydrocarbon bridges makes PMO a suitable porous domain for hosting a variety of organic molecules in solution (1). During the past decades, several mechanized, mesoporous silica materials were fabricated and successfully used to entrap different cargos inside the mesopores for controlled release. In addition, various stimuli-responsive, gated mesoporous silica materials were achieved by adopting different synthetic pathways of surface modifications (58). Among these gated materials, photothermal triggered systems can be considered clean-energy powered systems on account of the reversible mode of operation without by-products. For instance, Mal et al. reported coumarin-functionalized, MCM-41–type, mesoporous silica material and demonstrated its uptake, storage, and release of organic molecules through photo-controlled reversible dimerization of coumarin derivatives (9). Apart from coumarin, azobenzene and its derivatives were also proven as photo-switchable gates for mesoporous materials, and some azobenzene-assisted photothermal-powered mesoporous systems were successfully used in the areas of controlled cargo loading and release (10).

The chemical modification of the organic bridges in PMO has been considered a very effective approach to achieve highly functionalized pores with uniform mesopore space. Covalently bonded organic bridges such as phenylene and ethenylene units in the siloxane network of PMO can be functionalized chemically via direct sulfonation (11), bromination (12), amination (13, 14), hydroxylation (15), and Diels-Alder reaction (16). Among these transformations, direct amination is advantageous, because the amino functional group offers the versatility for further modifications via amide, imine, urea, and diazo bond formations. Ohashi et al. successfully demonstrated the direct amination of phenylene bridges in the mesoporous benzene-silica network of PMO via a two-step chemical transformation (14). This methodology becomes more significant for the preparation of switchable gate–functionalized PMO nanomaterials. For the design of switchable organic gates, azobenzene and its derivatives are commonly used because of their unique cis-trans isomerization properties upon alternate treatment with ultraviolet/visible light (UV/Vis) irradiation or heat.

Inspired by the unique features of PMO and azobenzene, we developed an approach to functionalize azobenzene moieties directly onto the phenylene bridges of a benzene-PMO by using the post-synthetic amination methodology. Here, the azobenzene-appended benzene-PMO (denoted as trans-azo@PMO) obtained serves as a nanocontainer with hydrophobic confinement, in which the azobenzene units in the pores perform the functions of (i) photo-responsive gates to tune the internal pore size of PMO and (ii) spacers to prevent guest molecules from undesired aggregation.

We selected a squaraine derivative as the guest molecule. Squaraine (Sq) dyes are a class of zwitterionic dyes that exhibit excellent photophysical properties in the near-infrared (NIR) region and serve as efficient photosensitizers to produce singlet oxygen (1O2) (1721). This fascinating class of dyes has also been proven to be potential candidates in various applications such as metal ion sensing (2224), NIR fluorescent labeling (25), two-photon absorption (26, 27), and the detection of amino thiols in blood plasma (2830). Recently, we and others have demonstrated different strategies to protect Sq dyes from nucleophilic attack for effective utilization (3133).

In azo@PMO, the opening and closure of the gate were reversibly controlled by light-induced cis-trans isomerization of azobenzene units. An unsymmetrical NIR Sq dye was loaded inside open pores as the guest molecule when the azobenzene gate was at its cis conformation. The azobenzene gate was then closed through its cis-trans isomerization under visible light irradiation, allowing the loaded Sq to be intercalated by the trans-azobenzene spacers in the pores without the formation of Sq aggregates. The dye-loaded hybrid (Sq-trans-azo@PMO) could act as a photosensitizer, exhibiting the capability to generate reactive oxygen species upon light excitation at 664 nm (Fig. 1). Finally, Sq-trans-azo@PMO was used as a catalytic bed in a microfluidic free-flow device, where the photo-oxidation of phenol to benzoquinone (BQ) with a constant conversion yield was carried out.

Fig. 1 Schematic representation of the preparation and 1O2 generation of Sq-azo@PMO.

Overall process for the preparation of Sq-azo@PMO and its capability for the photo-oxidation of phenol in water.

RESULTS

First, we prepared a phenylene-bridged PMO (denoted as Ph-PMO) by adopting an established synthesis route (34). During the preparation, we selected a pluronic P123 surfactant as the template to achieve a pore size of more than 5 nm under acidic conditions. Then, the prepared Ph-PMO was subjected to a nitration process to generate nitro Ph-PMO (denoted as Ph-PMO-NO2), followed by a reduction to generate amino functionality in PMO (denoted as Ph-PMO-NH2, fig. S1). After the successful preparation of Ph-PMO-NH2, an azobenzene compound (compound 4 in fig. S2) was grafted through amide coupling, leading to trans-azo@PMO for further uses. Ph-PMO has a relatively higher hydrophobic environment inside the pore channels according to previous reports (3537). The functionalization of the azobenzene group with a terminal ethoxy unit presumably reinforces the internal hydrophobicity of the pore channels. In addition, direct amination onto the phenylene bridges could facilitate further grafting of the azobenzene unit.

Successful functionalization was characterized by Fourier transform infrared spectroscopy (FT-IR), 13C cross-polarization magic angle spinning nuclear magnetic resonance (CP-MAS NMR), and 29Si CP-MAS NMR techniques (see the Supplementary Materials for more details). The analytical results obtained indicate the successful reduction of –NO2 to –NH2 functionality to form Ph-PMO-NH2, followed by the introduction of the azobenzene derivative through amide linkage.

UV/Vis absorption studies (Fig. 2) were carried out to obtain an insight into the photothermal response of azo@PMO in aqueous solution. The UV/Vis absorption spectrum of Ph-PMO-NH2 exhibits broad scattering in water with a maximum of 320 nm corresponding to the phenyl unit in the PMO skeleton (Fig. 2A). Similarly, trans-azo@PMO in water presents a signature absorption peak at 380 nm corresponding to the π-π* transition of the azobenzene unit (Fig. 2A). The irradiation of trans-azo@PMO under 380-nm light for 20 min shows a decrease in the absorption intensity at 380 nm, with concomitant formation of a new band at 460 nm corresponding to the cis isomer, that is, cis-azo@PMO (Fig. 2A). Furthermore, we demonstrated the reversible process of the trans-cis isomerization between trans-azo@PMO and cis-azo@PMO up to six cycles (Fig. 2B), proving the efficient trans-cis isomerization capability of azo@PMO in aqueous solution.

Fig. 2 Photoisomerization of azo@PMO.

(A) Schematic representation of reversible isomerization between trans-azo@PMO and cis-azo@PMO under light irradiation and heat. a.u., absorbance units. (B) UV/Vis absorption spectra of Ph-PMO-NH2, cis-azo@PMO, and trans-azo@PMO (1.5 mg ml−1 in aqueous solution). The changes in the UV/Vis absorption spectra due to the isomerization of trans-azo@PMO to cis-azo@PMO (1.5 mg ml−1) were recorded under the irradiation of 383-nm UV light for 20 min. (C) Absorbance changes of the UV/Vis spectra of azo@PMO at 383 nm as a function of cycles upon alternating UV light irradiation and heating at 55°C.

We then synthesized an unsymmetrical Sq dye by coupling half-squaraine with dimethoxy styrylpyrrole under azeotropic refluxing conditions in benzene/butanol (fig. S3). The dye was purified by column chromatography over silica gel and characterized by 1H NMR, 13C NMR, and high-resolution mass spectrometry. Sq exhibits a characteristic absorption maximum at 667 nm (ε = 165,620 M−1 cm−1, Fig. 3A) and a corresponding emission maximum at 685 nm (Fig. 3B) in dichloromethane.

Fig. 3 Photophysical studies of Sq-azo@PMO.

(A) UV/Vis absorption spectra of Sq in dichloromethane (DCM) (red curve), trans-azo@PMO in water (black curve), and Sq-azo@PMO in water (blue curve). (B) Emission spectra of Sq in DCM (red curve) and Sq-azo@PMO in water (blue curve).

The synthesized Sq was subsequently loaded inside azo@PMO to generate the final gated hybrid (Sq-trans-azo@PMO or Sq-azo@PMO). The loading procedure is described in Materials and Methods. Furthermore, the amount of Sq loaded in Sq-azo@PMO was determined by thoroughly extracting all dye molecules from the pores in dichloromethane under UV light irradiation to open the gates (fig. S8). All the extracted Sq dye in dichloromethane solution was quantified by its UV/Vis absorbance intensity and was calculated to be about 0.2 wt %.

Transmission electron microscope (TEM) images of Ph-PMO, Ph-PMO-NH2, trans-azo@PMO, and Sq-azo@PMO are shown in Fig. 4. In the case of Ph-PMO, mesopores having a diameter of ~6 nm are uniformly arranged in honeycomb-like hexagonal arrays throughout the sample (Fig. 4A). The image also indicates a wall thickness of ~3 nm (fig. S9A). Upon subsequent modifications, the regular mesoporous structure of pristine Ph-PMO was basically preserved in Ph-PMO-NH2 and trans-azo@PMO (Fig. 4, B and C). However, after the loading of Sq dye, the contrast between the pore channels and the pore walls markedly decreased because of the intercalation of the dye molecules inside the pores (Fig. 4D and fig. S10, A and B). The TEM images of trans-azo@PMO obtained after the removal of Sq from Sq-azo@PMO also show a regular mesoporous structure, indicating the retention of mesostructure in Sq-azo@PMO (fig. S10, C and D) with a constant wall thickness (fig. S9B). These textural observations for the four phases of PMOs are in agreement with the results obtained from the N2 adsorption/desorption measurements shown below.

Fig. 4 TEM images of PMOs.

(A to D) TEM images of (A) Ph-PMO, (B) Ph-PMO-NH2, (C) trans-azo@PMO, and (D) Sq-azo@PMO.

We performed isothermal N2 adsorption/desorption measurements for Ph-PMO, Ph-PMO-NH2, trans-azo@PMO, and Sq-azo@PMO to determine their surface areas and pore size distributions (Table 1). From the N2 adsorption isotherms under the lowest pressure, the Brunauer-Emmett-Teller (BET) method was used to calculate the specific surface areas of the materials, whereas we adopted the nonlocal density functional theory (NLDFT) equilibrium model for cylindrical pores, considering N2 as an adsorbate and silica as an adsorbent at 77 K to achieve the pore size distributions of the PMOs. We found that Ph-PMO, Ph-PMO-NH2, trans-azo@PMO, and Sq-azo@PMO showed typical type IV isotherms with different kinds of adsorption hysteresis. Ph-PMO and Ph-PMO-NH2 exhibited typical H1 hysteresis, indicating the presence of a mesoporous matrix with uniform pore distribution (fig. S11, A and B) (38). However, the surface modification in the case of Ph-PMO-NH2 reduced its surface area and the pore size (Table 1). On the other hand, the functionalization of the azobenzene unit led to a marked reduction in the pore diameter and surface area for trans-azo@PMO, offering an isotherm with H3 type hysteresis (fig. S11C) (38). Upon the loading of Sq dye into the pores of trans-azo@PMO, the resulting Sq-azo@PMO underwent further reduction of both surface area and pore size. The N2 isotherm of Sq-azo@PMO was also assigned to type IV isotherm without any distinguishable adsorption hysteresis (fig. S11D), indicating that the loaded dye occupies the mesoporous channels.

Table 1 Textural and porosity data of different PMOs.
View this table:

To understand the photophysical properties of the prepared hybrids, UV/Vis absorption and fluorescence spectroscopy studies were carried out. In the UV/Vis absorption spectra, we observed that the aqueous suspension of Sq-azo@PMO exhibits a characteristic peak at ~380 nm corresponding to the trans-azobenzene moiety and a peak around 664 nm assigned to the dye intercalated inside the pores (Fig. 3A). As compared to the absorption of Sq in dichloromethane (Fig. 3A), Sq-azo@PMO shows no obvious change of the absorption maximum at 664 nm in aqueous solution. Sq-azo@PMO reveals an ~5-nm blue shift of the emission maximum at 680 nm under excitation at 650 nm when compared to that of Sq in dichloromethane (Fig. 3B). The observed blue shift in the emission maximum could be attributed to the internal absorption by the hydrophobic interior of the azobenzene-intercalated PMO pores. Hence, it was proven that the azobenzene unit grafted in the pore wall of PMO serves as not only a closed gate in its trans-form but also a spacer between Sq molecules to effectively prevent Sq aggregation inside the pores. Furthermore, the UV/Vis spectrum of Sq-azo@PMO showed a decrease in the absorption intensity at 380 nm in aqueous suspension after 380-mn UV irradiation for 30 min on account of the formation of Sq-cis-azo@PMO (fig. S12). However, the absorption maximum at 664 nm corresponding to the Sq dye disappeared in the case of Sq-cis-azo@PMO (fig. S12). The open pores of Sq-cis-azo@PMO were not able to protect the trapped Sq from aggregation under the hostile aqueous environment due to the formation of cis-azobenzene isomer. Thus, aggregated Sq lost its photophysical properties in aqueous solution. These observations provide solid evidence in support of the role of an azobenzene unit as the gate in azo@PMO.

Furthermore, we investigated the photo-induced 1O2 generation capability of the Sq-azo@PMO hybrid through an indirect chemical method using 9,10-anthracenediyl-bis-(methylene)dimalonic acid (ADMA) as the 1O2 trap (39). ADMA reacts with 1O2 to produce corresponding endoperoxide that can be monitored by the absorption changes of ADMA. An aqueous solution of ADMA (8.2 × 10−5 M) was mixed with an aqueous suspension of Sq-azo@PMO (0.3 mg ml−1) under stirring. A typical absorption spectrum of the mixture shows two characteristic regions (fig. S13), that is, the sharp absorption corresponding to ADMA at 350 to 415 nm and the absorption of Sq from 630 to 700 nm. Upon the irradiation of 664-nm light for 50 min, the absorption of ADMA decreased continuously (fig. S14A). The distinct photobleaching of ADMA indicates the photo-induced generation of 1O2 by monomeric Sq protected inside Sq-azo@PMO in aqueous medium. In addition, the 1O2 quantum yield of Sq-azo@PMO was calculated to be 9.9% (see the Experimental Section in the Supplementary Materials for more details). In another control experiment, the same concentration of ADMA aqueous solution containing trans-azo@PMO was irradiated by 664-nm light to check the photo-induced 1O2 generation capability of trans-azo@PMO without Sq. No photo-induced 1O2 generation was observed after the light irradiation for 60 min (fig. S14B). Therefore, effective 1O2 generation of Sq-azo@PMO in aqueous condition provides a proof of concept for protecting Sq dye in its monomeric state inside mesoporous channels of PMO to preserve its photophysical properties for further applications.

Among various applications of 1O2, using 1O2 as a reactive reagent for the synthesis of fine chemicals and the treatment of wastewater has attracted a lot of attention. Phenol and its derivatives are toxic compounds commonly found in the wastewater produced from paper and dye manufacturing industries as well as oil refineries because of their resistance to biodegradation. One common approach of dealing with such water contaminants is to oxidize them into corresponding oxygenated compounds that are easily biodegradable (9, 36, 37). Many studies have demonstrated the oxidation of phenol and its derivatives in organic solvents through in situ generation of 1O2 in the presence of photosensitizers such as eosin, rose bengal, methylene blue, riboflavin, and Zn(II) tetraphenylporphyrin (40, 41). Similar reactions have also been carried out in aqueous solution using various dyes and metal complexes as the photosensitizers, and it was evident that the presence of monomeric photosensitizers in solution is critical to achieve the highest efficacy (41). However, most of the organic dyes undergo severe aggregation in the aqueous medium, which substantially decreases their ability to generate 1O2 for practical applications.

In this context, we carried out the oxidation of phenol to BQ using 1O2 generated from Sq-azo@PMO under 664-nm light irradiation in aqueous medium (Fig. 1). The main challenges associated with this method that we encountered were (i) how to reduce the reaction time to achieve 100% conversion of phenol into BQ because a slow rate of conversion as well as a rapid decline in conversion rate were observed as the reaction proceeded (fig. S15) and (ii) how to address the limitation over the area for light irradiation. It is also important to mention that 1O2 exhibits a short half-life (0.6 μs) and a short diffusion distance (0.1 μm) in aqueous medium (41), for which efficient utilization of this reactive oxygen species is a great challenge, especially in macroscale. As an approach to circumventing the abovementioned issues, we used a microfluidic device that has a short diffusion time for the reactants, which contributes to the improved mass transportation in the reactor (42). For this purpose, we designed and fabricated a two-inlet–single-outlet “Y”-type microfluidic device (fig. S16) to carry out the free-flow photo-oxidation of phenol using Sq-azo@PMO. Figure 5A shows the photograph of the ready-to-use microfluidic device fitted with two inlets that allow the inputs of aqueous phenol solution (0.1 M) and oxygen. A central chamber was designed to load the catalyst and was later filled with Sq-azo@PMO (3 mg). To investigate the photo-oxidation of phenol to BQ, an optimized condition was used after several trial experiments. Basically, aqueous solution of phenol (0.1 M) was allowed to flow through inlet 1 using a syringe pump at a rate of 2 ml hour−1. O2 was simultaneously flowed through inlet 2 under 1-atm pressure. The central catalytic bed was then irradiated by 664-nm light (2.5 mW cm−2) to generate 1O2 in situ. Keeping a constant flow of both reactants, we carried out the oxidation reaction at room temperature upon the irradiation of a laser light with a wavelength of 664 nm (Fig. 5B). The products were collected from the outlet every 30 min.

Fig. 5 Microfluidic free-flow device loaded with Sq-azo@PMO for photo-oxidation of phenol.

(A) Photograph of the microfluidic device under normal light. (B) Operational mode of Sq-azo@PMO loaded microreactor under the irradiation of 664-nm light. Scale bar, 1 cm. (C and D) Photo-oxidation of phenol inside the microfluidic channel as a function of cycles upon (C) alternating irradiation of light and (D) continuous irradiation of light.

The flame ionization detector gas chromatography (FID-GC) analysis indicates that the photo-oxidation could lead to a selective conversion of phenol to BQ. The microfluidic device–based microreactor showed 22 to 23% yield of the BQ formation during every cycle of the reaction under the irradiation of 664-nm light for 30 min each cycle (Fig. 5D). The yield is about 1.5 times higher than that obtained from a cuvette-based aqueous phase reaction, for a reaction time of 30 min. Furthermore, we investigated the effect of light by carrying out a periodic switch “on” and “off” process, proving that an effective oxidation process takes place only in the presence of light (Fig. 5C). The reaction was further investigated under the same microfluidic conditions with continuous irradiation of light for longer reaction time, and the products were collected and analyzed periodically. The results showed that, for every 30 min of reaction under the given conditions up to a total of 210 min, a consistent production of BQ with 21 to 23% yield was obtained (Fig. 5B), whereas a rapid declination of the BQ production was observed in the cuvette-based aqueous phase reaction (fig. S15). After the oxidation reaction, the used Sq-azo@PMO was recovered, and its photophysical properties were analyzed by UV/Vis spectroscopy. After comparing the spectra of Sq-azo@PMO before and after the photo-oxidation, it could be concluded that the loaded Sq dye remains stable in its monomeric state for a long time in an aqueous environment, after the generation of 1O2 (fig. S17). In another control experiment, we performed the photo-oxidation reaction in a microfluidic device using Sq-loaded benzene-PMO (Sq@PhPMO) as the catalytic bed under the same conditions [see the Supplementary Materials (fig. S18) for the details on preparation and characterization]. However, no production of BQ was observed. This observation validates the necessity of monomeric Sq for achieving the catalytic activity. It also proves that only gated PMO, that is, trans-azo@PMO, can prevent the unwanted aggregation of Sq in aqueous medium.

DISCUSSION

Here, we have reported the development of azobenzene gate–appended phenylene-bridged PMO (azo@PMO). The azobenzene moieties inside the pores serve as the gates of the mesoporous channels, which can be reversibly closed and opened by virtue of trans-cis isomerization under light irradiation. The photo-responsive gate functionalized inside the pore channels of PMO has been less investigated in the literature because of the unavailability of suitable synthetic approaches. We have adopted a post-synthetic grafting method to introduce an azobenzene unit inside the pore channels. The obtained azo@PMO has unique properties capable of loading and releasing cargos in a light-controlled manner. Thus, an unsymmetrical Sq dye has been loaded inside the gated pores in its monomeric state because of the intercalation of trans-azobenzene spacers, leading to the formation of the Sq-azo@PMO hybrid. The monomeric Sq inside the mesoporous channels of azo@PMO shows high efficiency to generate 1O2 in an aqueous environment upon light irradiation. To efficiently use the generated 1O2 in situ, a microfluidic reactor loaded with Sq-azo@PMO has been fabricated, where the photo-oxidation of phenol to BQ has been carried out. By using the microfluidic reactor, up to 23% production of BQ with 100% selectivity has been achieved. Thus, the current research presents the first gated heterogeneous catalyst and paves the way for applying such gated catalysts within microfluidic reactors toward practical uses.

MATERIALS AND METHODS

Preparation of Sq-azo@PMO

In a typical process, trans-azo@PMO (20 mg) was suspended in acetonitrile (10 ml), and the suspension was stirred under UV light with a wavelength of 380 nm for 1 hour to convert surface-functionalized azobenzene units from the trans form to the cis form. The photostationary state of the cis-azobenzene unit was confirmed by UV/Vis absorption. An Sq solution in acetonitrile (0.4 mM, 5 ml) was added to this suspension under UV light irradiation for 15 min. Subsequently, the mixture was stirred at 55°C under visible light for 2 hours. Solid material was isolated by centrifugation and thoroughly washed by acetonitrile followed by dichloromethane under sonication. The greenish blue material was dried in vacuum to generate Sq-azo@PMO.

Microfluidic chip fabrication

Soft lithography was used for the fabrication of polydimethylsiloxane (PDMS) chips. It involves two steps: the creation of a mold, followed by the creation of PDMS chips (43). The mold was created by photolithography, whereby a negative photoresist (SU-8 25, MicroChem Inc.) was coated on a silicon substrate and exposed to UV radiation at 365 nm through a photomask. Once it was exposed, the substrate was baked and developed to obtain the final mold with a channel thickness of about 40 μm. PDMS (Sylgard 184, Dow Corning) was prepared by mixing an elastomer and a curing agent in a weight ratio of 10:1. Once PDMS was prepared, it was poured over the mold, and the curing was performed at 70°C for 2 hours, at the end of which PDMS was polymerized and the chip could be cut out from the mold.

Photo-oxidation of phenol in the microfluidic chip

Sq-azo@PMO (3 mg) was loaded into the cavity in the prepared microfluidic chip (fig. S16). A thin cotton plug was used in front of the outlet channel to prevent unwanted leakage of Sq-azo@PMO from the cavity. A microscopic cover glass was used to close the cavity using vacuum grease and transparent cello tape. The openings of the two inlets in the microfluidic chip were respectively connected to a syringe pump containing 0.1 M aqueous solution of phenol and to a balloon containing O2 by using a plastic capillary tube with a 1-mm diameter. Likewise, the opening of the outlet was connected to a sample vial through a similar capillary tube to collect the reaction mixture after the reaction. The 0.1 M phenol solution was allowed to pass through the reactor cavity along with O2 via a microchannel, at a continuous flow rate of 2 ml hour−1 with the help of the syringe pump. Under such reaction conditions, a laser light with a 664-nm wavelength was used to irradiate by focusing at the reactor cavity. The reaction mixture collected from the outlet was analyzed after an interval of 30 min. After determining the response factors for the reactant and product individually, the product and unreacted reactant were identified and quantified by using an Agilent 6890 GC system equipped with an FID.

SUPPLEMENTARY MATERIALS

Supplementary material for this article is available at http://advances.sciencemag.org/cgi/content/full/1/8/e1500390/DC1

Experimental section

Fig. S1. Synthetic route for two-step amination of Ph-PMO.

Fig. S2. Synthetic route for the preparation of azobenzene compounds.

Fig. S3. Synthetic route for Sq dye.

Fig. S4. Difference between consecutive absorbance at each time intervals was plotted against time for (a) the absorption decrease of ADMA upon the irradiation of 664 nm light by mixing with aqueous suspension of Sq-azo@PMO under stirring, and (b) the absorption decrease of ADMA upon the irradiation of 660 nm light by mixing with aqueous suspension of MB under stirring.

Table S1. EA results of various PMO materials.

FT-IR analyses of PMOs

Fig. S5. FT-IR spectra of (a) Ph-PMO, Ph-PMO-NO2, and Ph-PMO-NH2 as well as (b) trans-azo@PMO.

13C CP-MAS NMR analyses of PMOs.

Fig. S6. 13C CP-MAS solid-state NMR spectra of Ph-PMO, Ph-PMO-NH2, and trans-azo@PMO.

29Si CP-MAS NMR analyses of PMOs

Fig. S7. 29Si CP-MAS solid-state NMR spectra of (a) Ph-PMO-NH2 and (b) trans-azo@PMO.

UV/Vis analyses of Sq-azo@PMO and azo@PMO

Fig. S8. UV/Vis spectra of (a) Sq-trans-azo@PMO measured after 48 hours of stirring in DCM and (b) azo@PMO obtained after the removal of Sq in DCM.

TEM images of Sq-azo@PMO and azo@PMO

Fig. S9. Wall thickness from TEM images of (a) Ph-PMO and (b) azo@PMO obtained after the removal of Sq from Sq-trans-azo@PMO.

Fig. S10. TEM images of (a and b) Sq-azo@PMO and (c and d) azo@PMO obtained after the removal of Sq from Sq-trans-azo@PMO.

N2 adsorption/desorption measurements of PMOs

Fig. S11. Plots for N2 adsorption/desorption isotherms with NLDFT pore size distribution (in the insets) of (a) Ph-PMO, (b) Ph-PMO-NH2, (c) trans-azo@PMO, and (d) Sq-azo@PMO.

UV/Vis analysis of Sq-trans-azo@PMO and Sq-cis-azo@PMO

Fig. S12. UV/Vis absorption spectra of (a) Sq-trans-azo@PMO in water and (b) Sq-cis-azo@PMO obtained after irradiation of 383 nm UV light on Sq-trans-azo@PMO in water for 60 min.

Generation of 1O2 by Sq-azo@PMO in aqueous medium

Fig. S13. UV/Vis absorption spectra of (a) ADMA in aqueous solution, (b) Sq-azo@PMO in aqueous suspension, and (c) ADMA along with Sq-azo@PMO in aqueous solution.

Fig. S14. UV/Vis absorption changes of ADMA aqueous solution in the presence of (a) Sq-azo@PMO and (b) trans-azo@PMO upon the irradiation of 664 nm light.

Photo-oxidation of phenol in aqueous medium

Fig. S15. Production of BQ as a function of time in cuvette-based aqueous phase.

Fig. S16. (a) Photograph of microfluidic channels, (b) loading of Sq-azo@PMO in the cavity, and (c) schematic representation of the cross-sectional view of the prepared microreactor.

UV/Vis analysis of Sq-azo@PMO before and after photo-oxidation reaction

Fig. S17. UV/Vis absorption spectra of Sq-azo@PMO in water (a) before the photo-oxidation reaction and (b) after recovered from the photo-oxidation reaction in a microfluidic reactor.

Preparation of Sq@PhPMO for control study

Fig. S18. UV/Vis absorption spectrum of Sq@PhPMO in water.

References (4446)

This is an open-access article distributed under the terms of the Creative Commons Attribution-NonCommercial license, which permits use, distribution, and reproduction in any medium, so long as the resultant use is not for commercial advantage and provided the original work is properly cited.

REFERENCES AND NOTES

Funding: This research was supported by the National Research Foundation (NRF), Prime Minister’s Office, Singapore under its NRF Fellowship (NRF2009NRF-RF001-015) and Campus for Research Excellence and Technological Enterprise Programme–Singapore-Peking University Research Centre for A Sustainable Low-Carbon Future, as well as the Nanyang Technological University–A*STAR (Agency for Science, Technology and Research) Silicon Technologies, Centre of Excellence under grant no. 11235100003. Author contributions: P.B., S.S., P.A., A.A., and Y.Z. designed and planned this project. P.B., S.S., and P.A. carried out the synthesis and characterizations of the materials and performed the photo-oxidation of phenol. N.V.M. and Y.K. were involved in the design and fabrication of the microfluidic device. P.B., S.S., P.A., A.A., and Y.Z. cowrote this paper, and all authors discussed the results and commented on the manuscript. Competing interests: The authors declare that they have no competing interests. Data and materials availability: The authors will make data available upon request.
View Abstract

Navigate This Article