Surface modification of silica-coated gadolinium oxide nanoparticles with zinc tetracarboxyphenoxy phthalocyanine for the photodegradation of Orange G

Journal of Molecular Catalysis A: Chemical 403 (2015) 64–76

Contents lists available at ScienceDirect

Journal of Molecular Catalysis A: Chemical journal homepage: www.elsevier.com/locate/molcata

Surface modification of silica-coated gadolinium oxide nanoparticles with zinc tetracarboxyphenoxy phthalocyanine for the photodegradation of Orange G Mpho Ledwaba, Nkosiphile Masilela, Tebello Nyokong, Edith Antunes ∗ Department of Chemistry, Rhodes University, Grahamstown 6140, South Africa

a r t i c l e

i n f o

Article history: Received 16 January 2015 Received in revised form 27 March 2015 Accepted 28 March 2015 Available online 31 March 2015 Keywords: Gd2 O3 nanoparticles Zinc tetracarboxyphenoxy phthalocyanine Photodegradation Azo dye

a b s t r a c t Zinc tetracarboxyphenoxy phthalocyanine was covalently linked to Gd2 O3 nanoparticles for the photocatalytic degradation of Orange G. Characterization of the composite was carried out using XRD, TEM, XPS, UV–vis spectroscopy and FT-IR spectroscopy. The composite showed improved photophysical properties over the phthalocyanine alone and the catalyst was found to be reusable. Analyses of the photodegradation rates of the azo dye indicated pseudo first-order kinetics. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Photocatalysis involves the use of photosensitizers which accelerate a reaction through the absorption of light and the photocatalyst is ideally capable of being recycled many times [1]. In environmental management, the photodegradation of organic pollutants, especially dye pollutants, is gaining enormous interest as a clean, efficient method to convert toxic pollutants to less noxious substrates. Phthalocyanines, particularly metallophthalocyanines, attract a great deal of attention in a wide variety of applications including photodynamic therapy of cancer [2], non-linear optics [3], as well as photosensitized catalytic applications [4–7]. Properties such as their high extinction coefficients, remarkable absorption of visible light and photoactivity, high chemical and thermal stability, and their ability to generate singlet oxygen makes these phthalocyanines ideally suited to applications such as photocatalysis [4]. Phthalocyanines have been anchored to support systems such as amberlite [8,9] and zeolite [10], and more recently in our group, to iron oxide nanoparticles [11,12] to successfully degrade Orange G and aid recovery of the photocatalytic system. This work reports for the first time the synthesis and use of a phthalocyanine–Gd2 O3 nanoparticle composite in the photodegradation of Orange G.

∗ Corresponding author. Tel.: +27 21 9594020; fax: +27 21 9593055. E-mail address: [email protected] (E. Antunes). http://dx.doi.org/10.1016/j.molcata.2015.03.023 1381-1169/© 2015 Elsevier B.V. All rights reserved.

Magnetic nanoparticles possess attractive properties in that they are capable of being manipulated by a magnetic field. The have found applications in biomedicine (drug delivery) [13], data storage [14], magnetic resonance imaging (MRI) [15], hyperthermia therapy [16], cell separation [17] and environmental remediation [18]. Furthermore, the surface of the nanoparticles may be easily modified to allow for attachment to other functional molecules or biomolecules [19]. In this way, bifunctional or multifunctional platforms are obtained. Owing to their excellent magnetic properties, the effortless recovery of the bifunctional magnetic nanoparticlephotocatalyst (e.g., phthalocyanine) from solution is anticipated. In this work, a zinc tetracarboxyphenoxy phthalocyanine was covalently linked to a Gd2 O3 NP as a support to facilitate an easily recoverable, heterogenous photocatalytic system. The composite was subsequently applied to the photodegradation of Orange G (Fig. 1), a common organic pollutant used in the dye or textile industry [20]. The photocatalytic degradation of Orange G has been reported previously by our group using phthalocyanines together with gold nanoparticles [21] or magnetite [11,12], however this is the first report of a Pc–Gd2 O3 NP photocatalytic system. Each component in the composite serves a purpose: i.e., the zinc tetracarboxyphenoxy phthalocyanine (ZnTCPPc) acts as the photosensitizer capable of generating singlet oxygen with ease [4] which would then accomplish the degradation of the azo dye [4,7,12]; while the magnetic nanoparticle serves as a support for the phthalocyanine (for heterogenous catalysis), enabling the easy recovery of the photocatalyst (by virtue of their magnetic

M. Ledwaba et al. / Journal of Molecular Catalysis A: Chemical 403 (2015) 64–76

HO N N SO3Na NaO 3S Fig. 1. Structure of Orange G.

properties). The magnetic nanoparticles are also expected to enhance the singlet oxygen generating ability of the Pc through enhanced intersystem crossing of the excited phthalocyanine to the triplet state. The photochemical properties of the phthalocyanines (and therefore their photocatalytic abilities) have consistently been enhanced in the presence of nanoparticles such as gold [21] and iron oxide [11,12]. 2. Materials and methods 2.1. Materials Dimethylsulphoxide (DMSO), ethanol (EtOH), dimethylformamide (DMF), methanol (MeOH) and toluene were obtained from SAARChem. Water collected from a Milli-Q water purification system (Millipore Corp., Bedford, MA, USA) was used for the preparation of all aqueous solutions. 1,3-diphenylisobenzofuran (DPBF), Anthracene-9, 10-bis-methylmalonate (ADMA), N,N dicyclohexylcarbodiimide(DCC), sodium hydroxide (NaOH) pellets, polyethylene glycol (PEG), (3-aminopropyl) triethoxysilane (APTES, 99%), Acetic acid [98%], formic acid [99.8%], Gadolinium nitrate hexahydrate [99.99 %], Orange G (OG) and Nhydroxysuccinimide (NHS) were purchased from Sigma–Aldrich. 2.2. Equipment UV–vis absorption spectra were measured at room temperature on a Shimadzu UV-2550 spectrophotometer using a 1 cm pathlength cuvette, whilst fluorescence emission and excitation spectra were obtained on a Varian Eclipse spectrofluorimeter using a 1 cm pathlength quartz cuvette. Fluorescence lifetimes were obtained using a time correlated single photon counting setup (TCSPC) (FluoTime 200, Picoquant GmbH). The fluorescence lifetime of the phthalocyanine and its composite was determined using a diode laser (LDH-P-670 with PDL 800-B, Picoquant GmbH, 670 nm, 20 MHz repetition rate, 44 ps pulse width). Fluorescence was detected under the magic angle with a Peltier cooled photomultiplier tube (PMT) (PMA-C 192-N-M, Picoquant) and integrated electronics (PicoHarp 300E, Picoquant GmbH). A monochromator with a spectral width of about 4 nm was used to select the required emission wavelength band. The response function of the system, which was measured with a scattering Ludox solution (DuPont), had a full width at half-maximum (FWHM) of 300 ps. All luminescence decay curves were measured at the maximum of the emission peak and lifetimes were obtained by deconvolution of the decay curves using the FLUOFIT software program (PicoQuant GmbH, Germany). The support plane approach was used to estimate the errors of the decay times. Laser flash photolysis experiments were performed with the light pulses produced by a Quanta-Ray Nd: YAG laser providing 400 mJ, 90 ns pulses of laser light at 10 Hz, pumping a Lambda -Physik FL3002 dye (Pyridin 1 dye in methanol). The energy of a single pulse ranged from 2 to 7 mJ. The analyzing beam source used was a ThermoOriel xenon arc lamp and a photomultiplier tube was used as a detector. Signals were recorded with a digital real-time oscilloscope (Tektronix TDS 360) and the

65

kinetic curves averaged over 256 laser pulses. Photo-irradiation for the photodegradation or singlet oxygen studies and determinations were performed using a General Electric Quartz line lamp (300 W). A 600 nm glass cut off filter (Schott) and water filters were used to filter off ultraviolet and infrared radiations, respectively. An interference filter (Intor 670 nm filter with a band width of 20 nm) was additionally placed in the light path before the sample. The light intensity was measured with a POWER MAX5100 (Molelectron Detector Incorporated) power meter and found to be 3.2 × 1020 photons cm−2 s−1 for the photodegradation studies and 1.3 × 1019 photons cm−2 s−1 for the singlet oxygen studies. Transmission electron microscopy (TEM) images were obtained using a ZEISS LIBRA® TEM and the sizes determined using ImageJ software. Infrared (FTIR) spectra were recorded using a Perkin-Elmer 100 FTIR Spectrometer equipped with a Universal Attenuated Total Reflectance (ATR) sampling accessory. X-ray powder diffraction (XRD) patterns were recorded on a Bruker D8 Discover equipped with a LynxEye detector, using Cu-K␣ radiation ( = 1.5405 Å, nickel filter). Data were collected in the range from 2 = 5◦ to 60◦ , scanning at 1◦ min−1 with a filter time-constant of 2.5 s per step together with a slit width of 6.0 mm. Samples were placed on a zero background silicon wafer slide. The X-ray diffraction data were treated using the Eva (evaluation curve fitting) software. Baseline correction was performed on each diffraction pattern. X-ray photoelectron spectroscopy (XPS) was carried out using a Kratos Axis UltraDLD spectrometer with an Al (monochromatic) anode. The system was equipped with a charge neutralizer, while the operating pressure was kept below 5 × 10−9 torr. For XPS survey scans, the following parameters were used: emission current was kept at 5 mA, the anode (HT) voltage at 15 kV and the resolution at 160 eV pass energy using a hybrid lens in the slot mode. The step size used was at 1 eV, while the dwell times were kept at 300 ms. High resolution spectra were obtained using a pass energy of 40 eV, also in slot mode, with the dwell times and step sizes at 500 ms and 0.1 eV, respectively. EPR measurements were accomplished using an X band EMXplus Bruker spectrometer, where the first derivative signal was obtained in solid state for the Gd2 O3 NPs, Si–Gd2 O3 NPs and their composite with ZnTCPPc.

2.3. Synthesis Zinc tetracarboxyphenoxy phthalocyanine (ZnTCPPc, complex 3) was synthesized as reported previously [22].

2.3.1. Synthesis of the Gd2 O3 NPs, Scheme 1 The method used in the synthesis of the ‘bare’ gadolinium oxide nanoparticles (1, Gd2 O3 NPs) was carried out as reported by Bazzi et al. [23] with slight modification as follows: Firstly, 1 g (2.7 mmol) of the Gd(NO3 )3 salt was added to 20 ml of PEG (as the capping agent) at 60 ◦ C under vigorous stirring overnight under an argon atmosphere. The temperature was kept at 140 ◦ C then decreased to 60 ◦ C. An aqueous solution of NaOH (2 M, 1 ml) was added dropwise to the reaction flask, after which the temperature was raised to 140 ◦ C for 1 h and subsequently increased to 180 ◦ C for a further 4 h. The resulting product was cooled and precipitated with ethanol, centrifuged and washed a further 3 times with ethanol. The Gd2 O3 NPs (1) were dispersed in ethanol (10 ml) until further use. The dried, uncoated Gd2 O3 NPs (1, 0.25 mg) were then washed 3 times with toluene, air-dried and suspended in a solvent mixture of DMF (12 ml) and toluene (8 ml). APTES (1 ml, 4.27 mmol) was added dropwise and the mixture stirred for 24 h at room temperature under an argon atmosphere. The resultant product was separated out by centrifugation, washed 4 times with toluene and thereafter stored in toluene (10 ml) until further use. The silica coated Gd2 O3 NPs are represented as SiGd2 O3 NPs (2). Yield: 3 g.

66

M. Ledwaba et al. / Journal of Molecular Catalysis A: Chemical 403 (2015) 64–76

1) PEG, 60 oC, 12 hrs 2) NaOH, 140 oC, 1 hr

DMF, Toluene, APTES, rt, Ar, 24 hrs

3) 180 oC, 4 hrs

H 3Si

NH 2

H 3Si NH 2

1

2

Scheme 1. Synthetic route for the preparation of the bare Gd2 O3 NPs (1) and the silica-coated Gd2 O3 NPs (Si–Gd2 O3 NPs, 2). PEG = polyethylene glycol; APTES = aminopropyltriethoxysilane.

2.3.2. Covalent linkage of the ZnTCPPc (3) to the amino-functionalized Gd2 O3 NPs (2), Scheme 2 ZnTCPPc (3) (0.1 g, 0.092 mmol) was added to DCC (1.03 g, 5 mmol) in DMSO (2 ml) and stirred at room temperature for 12 h to activate the carboxylic acid functional groups on the Pc [24]. Thereafter NHS (0.58 g, 5 mmol) and the silica-coated Gd2 O3 NPs (SiGd2 O3 , 2) (0.4 g) were sequentially added to the mixture, followed by stirring for 48 h. The product was purified using silica column chromatography with ethanol and toluene to separate the free Pc and the gadolinium oxide nanoparticles from the conjugate (4). Yield: 1.35 g. In addition to the conjugate (4) formed, a sample was prepared where the ZnTCPPc (3) (0.2 g, 1.84 mmol) was simply added to the silica-coated Gd2 O3 NPs (SiGd2 O3 , 2) (0.4 g) and stirred in DMSO (2 ml) for 48 h. The resultant mixture was centrifuged and the precipitate collected. This sample was intended to represent a sample that contains no covalent bond formation between the Pc and NP and is denoted as ZnTCPPc–SiGd2 O3 NPs (mixed, 5). Yield: 2.5 g.

2.4. Photophysical and photochemical parameters The fluorescence quantum yields (F ) were obtained using a comparative method [25] following Eq. (1):

F = F(std)

F × Astd × n2 Fstd × A × n2std

(1)

where F and Fstd are the areas under the fluorescence curves of the ZnTCPPc (or ZnTCPPc–SiGd2 O3 nanocomposite (4) or ZnTCPPc–SiGd2 O3 mix (5)) and the standard, respectively. A and Astd is the absorbance of the sample and reference at the excitation wavelength, while n and nstd are the refractive indices of the solvents used for the sample and standard, respectively. ZnPc in DMSO was used as a standard (where the F = 0.2 [26]. The triplet quantum yields were determined for the ZnTCPPc (or ZnTCPPc-nanocomposite (4) or ZnTCPPc-mix (5)) using a comparative method based on the triplet decay as outlined before [27] using ZnPc as the standard in DMSO [28,29]. The singlet oxygen (1 O2 ) quantum yields  were then determined using Eq. (2):

 = std 

std R × Iabs

Rstd × Iabs

(2)

where std is the singlet oxygen quantum yield for the standard 

ZnPc [26]. R and Rstd are the photodegradation rates of the singlet oxygen quencher DPBF in the presence of the ZnTCPPc and the stanstd are the rates of light absorption by dard, respectively. Iabs and Iabs the ZnTCPPc and standard, respectively [30].

3. Results and discussion 3.1. Characterization of the Gd2 O3 nanoparticles and ZnTCPPc–Gd2 O3 nanocomposite The synthetic procedure for synthesizing the magnetic gadolinium oxide nanoparticles (Gd2 O3 NPs, 1) as well as the procedure for coating the surface of the Gd2 O3 NPs with a silica shell to form Si–Gd2 O3 NPs (2) is shown in Scheme 1. The paramagnetic Gd2 O3 nanoparticles were prepared using a two-step polyol method in a similar manner to that reported by Bazzi et al. [23]. Minor adjustments in solvent, reaction time and temperature were found to improve the shape and dispersion of the nanoparticles. Since water soluble NPs were needed for the photocatalytic studies, this polyol method was found to be the most appropriate. The magnetic nanoparticles were capped with PEG, to give a white powder which exhibited a strong magnetic response. The nanoparticles were then functionalized with 3-aminopropyltriethoxy silane (APTES), also as reported in Bazzi et al. [23], to give amino functionalized nanoparticles which would facilitate conjugation to the phthalocyanine. APTES was found to produce highly stable nanoparticles which did not affect the dispersion of the nanoparticles by, for example, inducing aggregation. The zinc tetracarboxyphenoxy phthalocyanine (ZnTCPPc) was synthesized according to literature methods using 4nitrophthalonitrile and 4-(3,4-dicyanophenoxy) benzoic acid [22] and the structure confirmed as reported in the literature. Functionalizing the NP surface with a suitable moiety would enable the NP to serve as a platform for further conjugation/derivatization to achieve a desired application. In this case, ZnTCPPc is attached to the NP, allowing for the construction of a multi-functional nanocomposite with MRI (as given by the Gd2 O3 NPs) and PDT/photocatalytic properties (as imparted by the phthalocyanine). Chemical functionalization of the nanoparticle can be achieved either by covalent attachment or by non-covalent (adsorption) reactions and planar molecules are capable of adsorbing onto the nanoparticle surface via – interactions [31]. The as-synthesized Gd2 O3 NPs (1) were functionalised with 3(aminopropyl) triethoxysilane (APTES), which introduces amino groups to the surface of the nanoparticles and improves the dispersability of the NP in a variety of solvents. The presence of the amino group on the NP surface also facilitates conjugation of the carboxy substituted Pc to the amino functionalized NP to form an amide bond. N,N -dicyclohexylcarbodiimide (DCC) and N-hydroxy succinimide (NHS) were used to activate the carboxylic acid groups [24] on the ZnTCPPc (Scheme 2), while the covalent linkage was established through a variety of techniques particularly FT-IR and thermogravimetric analyses (TGA). The Gd2 O3 NPs and composites (i.e., 1, 2 and 4) synthesized were characterized using TEM, UV–vis and FT-IR spectroscopies, as well as thermogravimetric analyses (TGA), electron paramagnetic resonance (EPR), powder X-ray diffraction (XRD) and X-ray photoelectron spectroscopy (XPS).

M. Ledwaba et al. / Journal of Molecular Catalysis A: Chemical 403 (2015) 64–76

O

COOH

N H 3Si

Zn

N

H 3Si

+

NH 2

N

N

COO H

N

N

O

NH 2

67

N O N

2 HOO C

COO H

O

3 1) DCC, DMSO, 12 hrs 2) NHS, 48 hrs

O

COO H

N

Zn

N

N

N

H3Si

O

NOC H

N

N

O

COO H

N O N

COO H

H3Si

NH2

4

Scheme 2. Synthetic route showing the conjugation of the ZnTCPPc (3) to the surface of the silica-coated Gd2 O3 NPs (2) via an amide bond to form the conjugate (4). DCC = N,N -dicyclohexylcarbodiimide; NHS = N-hydroxysuccinimde.

3.1.1. Transmission electron microscopy (TEM) The TEM image obtained for the bare Gd2 O3 NPs (1) is shown in Fig. 2(i) and illustrates the size distribution and dispersion of the nanoparticles. The size of the uncapped Gd2 O3 NPs (1) were determined to range from approximately 5.9 nm to 7.9 nm. Formation of the silica shell to produce the Si–Gd2 O3 NPs (2) was found to increase the size of the Gd2 O3 NPs from ∼6 nm to 15 nm, Fig. 2(ii), as expected. Fig. 2(ii) shows that the silica coated NPs (2) are still well dispersed and uniform in size, although no clear boundary between the NP and the silica shell is observed. The TEM image obtained for ZnTCPPc–SiGd2 O3 nanocomposite (4) (Fig. 2(iii)) shows that the size of the Si–Gd2 O3 NPs (2) was further increased upon conjugation with the phthalocyanine to 17 nm, providing some evidence of successful nanocomposite (4) formation. These images show that there is a change in appearance in the sample and that the nanocomposite (4) is still well dispersed and somewhat uniform in size.

3.1.2. FT-IR spectroscopy Satisfactory FT-IR data were obtained for all the NPs. FT-IR data confirmed the presence of a silica-shell on the surface of the gadolinium oxide nanoparticles. Strong bands observed at 1150 cm−1 and 1560 cm−1 , corresponding to the Si O bend and primary amino functionalities [32], respectively, are observed after silica encapsulation and functionalization of the NPs (Fig. 3(iii)). The FT-IR spectrum of pure PEG is characterized by a strong C H stretching vibration at 2887 cm−1 ; this is also seen in the FT-IR spectra of the gadolinium oxide nanoparticles. The broad peak in the 3100–3600 cm−1 range is attributed to the PEG OH functionality [33]. The FT-IR spectra of ZnTCPPc alone (Fig. 4) showed peaks characteristic of the OH (3430 cm−1 ), C O (1640 cm−1 ), C C (1581 and 1480 cm−1 ), C O C (1251 and 1180 cm−1 ) and C H (840 cm−1 ) moieties. The covalent link expected to form between the Si–Gd2 O3 NPs (2) and ZnTCPPc (3) to form the nanocomposite (4) was con-

68

M. Ledwaba et al. / Journal of Molecular Catalysis A: Chemical 403 (2015) 64–76

Fig. 2. TEM images of (i) Gd2 O3 NPs (1), (ii) Si–Gd2 O3 NPs (2) (showing an increase in size for the nanoparticles), and (iii) ZnTCPPc–Gd2 O3 NP composite (4) (where a change in size and morphology is observed for the conjugates).

firmed using FT-IR spectroscopy (Fig. 4A and B). Peaks attributed to the primary amino ( NH2 ) moieties on the Si–Gd2 O3 NPs (2) appear at 1568 cm−1 and 1463 cm−1 (Fig. 4A(i)) [32], while the intense peak observed at ∼1060 cm−1 is assigned to the Si O Si bonding stretch for the NPs (2), as reported in the literature [32]. ZnTCPPc (3) alone (Fig. 4A(ii) shows the C O vibrational band at 1640 cm−1 , while weak bands in the 3100–3600 cm−1 region correspond to the COOH. The FT-IR spectra of the ZnTCPPc–Gd2 O3 NPs (4) composite (Fig. 4A(iii)) show that the peaks attributed to the COOH functional group have disappeared, whilst peaks due to the amide ( NHCO ) functionality at (1631 cm−1 and 1535 cm−1 ) were observed together with the appearance of an NH stretch at 3283 cm−1 (Fig. 4A(iii)). This suggests success-

Fig. 4. FT-IR spectra of the synthesized (A) (i) Si–Gd2 O3 NPs (2), (ii) ZnTCPPc (3), (iii) ZnTCPPc–SiGd2 O3 nanocomposite (4) and (iv) the simple mixture ZnTCPPc–SiGd2 O3 NP-mix (5). Graph (B) shows the zoomed in region from 2000–600 cm−1 to highlight the changes observed between (i) ZnTCPPc (3), (ii) ZnTCPPc–SiGd2 O3 nanocomposite (4) and ZnTCPPc–SiGd2 O3 NP-mix (5). Fig. 3. FT-IR spectra of the synthesized (A) (i) Si–Gd2 O3 NPs (2).

M. Ledwaba et al. / Journal of Molecular Catalysis A: Chemical 403 (2015) 64–76

69

ful formation of a covalent bond between the Si–Gd2 O3 NPs (2) and ZnTCPPc (3). In addition, these spectra were compared to ZnTCPPc–SiGd2 O3 NPs (mixed, 5). Fig. 4A(iv) shows that the C O vibrational peak (1710 cm−1 ) and the COOH (∼3300 cm−1 ) for the ZnTCPPc remains unchanged in sample 5. This spectrum served to show the differences between a nanocomposite (4) and that of a simple mixture (5). The most apparent differences between the mixture (5, Fig. 4B(iii)) and the nanocomposite (4, Fig. 4B(ii)) is the shift observed in the C O stretch at 1700 cm−1 together with the appearance of an amide stretch at 3300–3500 cm−1 upon conjugation. 3.1.3. Thermogravimetric analyses Thermogravimetric analyses (TGA) was used to confirm the compositional differences between the bare and silica coated gadolinium oxide nanoparticles (1 and 2), since the two materials are expected to give different thermal decomposition profiles. The thermal decomposition profiles were obtained under a N2 atmosphere at a heating rate of 10 ◦ C min−1 . The profiles for the starting materials (Gd(NO3 )3 salt (i) and polyethylene glycol (ii)), the bare Gd2 O3 NPs (iii, 1), and the silicacoated NPs (2, (iv)) are shown in Fig. 5. The Gd salt (Fig. 5(i)) shows two main weight loss steps with possibly a third taking place at approximately 500 ◦ C. The first step at ∼150 ◦ C i.e., an initial loss of 15% is attributed to the loss of water, while the slow decomposition step (50% weight loss) taking place at 420 ◦ C is thought to be due to the loss of nitro functional groups. The Gd metal content in the salt is calculated to be 35%, though the mass loss at the end of this TGA run is only at 53%, which suggests that removal of the functional groups is not complete at 500 ◦ C. The thermal decomposition profile of the capping agent alone, PEG (Fig. 5(ii)), shows a single, dramatic, complete (100% weight loss) decomposition step with an onset temperature of approximately 190 ◦ C. The decomposition profile for the bare Gd2 O3 NP (Fig. 5(iii)) revealed a single, featureless step (20% weight loss), which is attributed to the slow decomposition of the organic material (probably the PEG) on the surface of the nanoparticle. It is expected that the capping agent on the surface of the NP should be removed by 420 ◦ C based on the decomposition profile of PEG alone (Fig. 5(ii)). It is also possible to make the assumption that the PEG capping constituted approximately 20% of the NP sample. At first it was surprising to note that the thermal stability of the silica coated NP (Fig. 5(iv)) was reduced by 20% (upon comparison of the % weight loss at 500 ◦ C for the Gd2 O3 NP and Si–Gd2 O3 NP samples). However, the decrease in thermal stability for these nanoparticles can be accounted for as additional aminopropyl organic groups have now been added to the surface of the NP and will consequently

Fig. 5. TGA profiles of (i) Gd(NO3 )3 , (ii) PEG, (iii) Gd2 O3 NPs (1) and (iv) Si–Gd2 O3 NPs (2).

Fig. 6. TGA profiles of (i) Si–Gd2 O3 NP (2), (ii) ZnTCPPc (3) and (iii) ZnTCPPc–Gd2 O3 composite (4) at a heating rate of 10–500 ◦ C min−1 under a nitrogen atmosphere with a gas flow rate of 120 ml min−1 .

tally the increase in the overall weight loss for the silica coated NP at 500 ◦ C. The first, initial weight loss up to 150 ◦ C though, is due to the loss of adsorbed water or solvents. The Gd2 O3 NPs and the Si–Gd2 O3 NPs were shown to have greater thermal stability compared to the Gd(NO3 )3 salt and PEG as they retain over 50% of their weight at temperatures above 400 ◦ C. The thermal decomposition profiles obtained for ZnTCPPc (3) and the ZnTCPPc–Si–Gd2 O3 nanocomposite (4) is shown in Fig. 6. Two main decomposition steps are observed for ZnTCPPc (Fig. 6(i)), with the first, shallow step between 100 and 280 ◦ C, suggesting solvent loss [34]. The second, main decomposition step from ∼280–500 ◦ C is associated with functional group decomposition e.g., loss of COOH groups [35]. Overall, a 50 % weight loss occurred for the Pc at 500 ◦ C. This decomposition profile is typical for phthalocyanine complexes as they are known to be highly thermally stable, showing no evidence of melting only decomposition at higher temperatures [36]. The decomposition profile for the conjugate showed a marked increase in weight loss (45% at 250 ◦ C), followed by a further shallow decomposition step from 250–500 ◦ C, where an additional 25% was lost. This nanocomposite was found to be highly hygroscopic, thus the first weight loss may be attributed to loss of water. The 2nd decomposition step is due to the removal of the remainder of the organic groups on the surface of the nanoparticle. The conjugate was thus observed to be less thermally stable as overall a 70% weight loss was observed for (4) as compared to ZnTCPPc (3) alone and the Si–Gd2 O3 NP (2) which showed an overall weight loss of ∼50 % at 500 ◦ C. 3.1.4. X-ray powder diffraction XRD, in␪ addition to giving information about the crystallinity and purity of a sample, is often used to determine the size of the nanoparticle by making use of the Debye–Scherrer equation. The XRD patterns for the bare Gd2 O3 (1) and the Si–Gd2 O3 (2) nanoparticles are shown in Fig. 7(i) and (ii), respectively, and a broad peak is observed at ∼2␪ = 20◦ for both samples. A broad reflection is expected at this position due to the amorphous nature of carbon (in PEG) or silica which was used to cap the NPs (1) and encapsulate the NP (2), respectively [37]. For comparative purposes, the synthesized NPs (1 and 2) were compared to the starting material, the Gd(NO3 )3 salt (Fig. 7). However, the information derived from the X-ray diffractograms for the gadolinium oxide NPs (1 and 2) were not clear, with the patterns suggesting that either there is a substantial degree of amorphousness or that the nanoparticles are very small. Significant line broadening is observed when the Gd2 O3 NPs are smaller than 20 nm [38]. Indeed, the TEM data indicated that both sets of NPs were smaller than 20 nm (Section 3.1.1).

70

M. Ledwaba et al. / Journal of Molecular Catalysis A: Chemical 403 (2015) 64–76

Fig. 7. X-Ray diffraction patterns for (i) Gd(NO3 )3 , salt (ii) Gd2 O3 NPs (1) and (iii) Si–Gd2 O3 NPs (2).

From the XRD data, it was not possible to determine whether the NPs could be indexed to the cubic phase (JCPDS card No. 11-604) as has been obtained previously for larger sized Gd2 O3 NPs [38], but this is discussed later with the XPS data obtained in Section 3.1.5 below. The TEM images (Fig. 2(i and ii)) obtained and described above, however, clearly show the NP shape and crystal structure, although the XRD data does not. The XRD patterns for ZnTCPPc (3) and the nanocomposite (4) only showed a broad reflection at ∼2␪ = 28◦ which is typical for phthalocyanines [39], together with the reflections observed for the Si–Gd2 O3 NPs (2) alone. No real information about the composite structure was obtained from XRD. 3.1.5. X-ray photoelectron spectroscopy X-ray photoelectron spectroscopy (XPS) was used to analyse the surface chemistry and elemental composition of the NP surface and was carried out in this work on the synthesized materials. Wide scan and high resolution XPS spectra were acquired for the Gd2 O3 (1) and Si–Gd2 O3 (2) NPs (Fig. 7A(i) and (ii), respectively) and the experimental data analysed using deconvolution curve fitting. Fig. 7A (ii) clearly shows the presence of the characteristic Gd peaks, particularly the Gd 3d3/2 and 3d5/2 peaks at ∼1185 eV, together with C 1s, O 1s and Na 1s peaks for the bare Gd2 O3 NP (1). The O 1s (at ∼530 eV) was expected to be part of the Gd2 O3 NP core as well as the OH atoms of the capping agent PEG. The capping agent also accounted for the presence of carbon (at ∼284 eV) in the XPS spectra. Na peaks (Na 1s at ∼1100 eV) were also observed and are most likely due to the NaOH used in the synthesis of the NPs (1). It is difficult to estimate the experimental Gd/O ratio (which may be calculated from the total areas under the peaks of the Gd 3d and O 1s peaks in the survey XPS spectrum), since the organic capping agent is likely to suppress the signal from the NP core. However, this is dependent on the thickness of the capping agent layer. For this reason survey spectra of the sample annealed at 500 ◦ C were also analyzed, but the temperatures used for annealing were not high enough as some C was still observed in the spectra. The survey scan obtained for the silica coated Gd2 O3 (2), shows that the silica coating step has completely capped the NP (Fig. 8A(ii)), as none of the Gd peaks are now observed. This was to be expected as XPS is a surface sensitive technique (where typically the photoelectrons generated are from within the first 3 nm of the surface being analyzed) [40]. The survey scan obtained for NP 2 shows instead the characteristic peaks associated with Si (e.g., Si 2p at 100 eV) and an increase in intensity of the N 1s peak at 397 eV; the position of this N 1s peak also indicates successful amino group functionalization [42].

Fig. 8B shows the high resolution spectra obtained for the Gd 4d region of the Gd(NO3 )3 salt and Gd2 O3 NPs (1), while Fig. 8C shows the data for the high resolution O 1s regions for the Gd(NO3 )3 salt employed and both sets of synthesized Gd2 O3 NPs (1 and 2). The Gd2 O3 NP (1), the Gd 2d3/2 and Gd 2d5/2 peaks were found to reside at 142.5 and 148.8 eV (Fig. 8B(ii)) and this found to be in agreement with XPS values reported for cubic phase Gd2 O3 [40]. Thus, while the XRD data in Section 3.1.4 did not confirm the phase of the NPs, the XPS data indicates that the nanoparticles could be indexed to the cubic phase. Fig. 8C shows the high resolution spectra for the O 1s peak of the Gd(NO3 )3 salt (Fig. 7C(i)) employed and both sets of the Gd2 O3 NPs (1 and 2) obtained (Fig. 8C(ii) and (iii), respectively). Deconvolution of the high resolution O 1s spectra are shown in Fig. 8D–F. The differences are clear and the O 1s peaks are found at higher binding energies for the nitrate salt (Fig. 8C and D). The O 1s peak for the nitrate salt shows perhaps 3 peaks, one main peak (at 531.4 eV) with a second smaller peak (532.9 eV) at a higher binding energy and a third, almost negligible, peak (at 529.9 eV). Since oxygen is ubiquitous in materials, it is possible that at least one of these peaks is due to the normal adsorption of oxygen to the sample. The other two peaks are attributed to the N-O and Gd-O O 1s peaks for the gadolinium nitrate salt as shown in Fig. 8B(i) [40]. The O 1s high resolution spectra for the Gd2 O3 NP (1) sample, (Fig. 8C(ii)), shows two main peaks centered at 529.5 eV and 527.9 eV, together with two smaller peaks at ∼531.4 and 529.4 eV, clearly seen upon deconvolution of the O 1s spectra (Fig. 8E). The two main peaks are presumably due to Gd-O and Gd = O, while the third would be due to either the C-O or the O-H of the capping agent PEG. Finally, the high resolution O 1s spectra for Si–Gd2 O3 NP (2) (Fig. 8C(iii)) showed a complicated deconvolution pattern (6 peaks), however there is one main peak centered at 530.2 eV (Fig. 8F) which is attributed to the Si-O peak [41] (of the silica shell on the NP). Since this is a surface analytical technique, the core O 1s peaks (e.g., Gd-O) of the Gd2 O3 NP, while they can still be present, is not expected to be intense. The other peaks are therefore considered to be due to O adsorbed onto the surface of the NP, Gd-O (shell) and Gd-O (core). The sources of the other peaks are not yet known. These results, however, confirm that a silica shell has been assembled on the surface of the nanoparticle [42]. 3.1.6. Electron paramagnetic resonance Electron paramagnetic resonance (EPR) is considered to be an excellent technique for studying materials with unpaired electrons and detecting paramagnetic species such as free radicals [43]. In this study EPR was used to confirm the paramagnetic properties of the Gd2 O3 nanoparticles (1 and 2) and the nanocomposite. The solid state EPR spectra, expected to be broad, are shown in Fig. 9. Gd3+ is paramagnetic (due to seven unpaired electrons), but its EPR spectrum consists of only one broad line due to spin–spin interactions between unpaired electrons; hence the observed broad spectra. For comparison, the spectrum for the Gd(NO3 )3 salt is shown in Fig. 9 too, and it shows an intense, asymmetrical, broad curve. On the other hand, the EPR spectra acquired for Gd2 O3 (1) and SiGd2 O3 (2) NPs, shown in Fig. 9 (inset), show very weak assymetrical, broad signals. Gd(III) is a paramagnetic species, however since the relaxation time of Gd(III) is extremely short, detection of Gd complexes using EPR is difficult [44]; a stronger magnet is most likely needed and the data acquired at very low temperatures to detect paramagnetic species. It may be possible that the NPs 1 and 2, have even shorter relaxation times, making detection much more difficult [44]. Though the EPR spectra did no show conclusive results, Gd is present in the NP system – as it was also accounted for with the XPS measurements. Fig. 10 shows the EPR spectra acquired for the ZnTCPPc (3), the nanocomposite (4) and the mixed (5) complexes in the solid state,

M. Ledwaba et al. / Journal of Molecular Catalysis A: Chemical 403 (2015) 64–76

71

Fig. 8. Wide scan XPS spectra (A) of (i) Gd2 O3 NPs (1) and (ii) SiGd2 O3 NPs (2). High resolution spectra of: (B) Gd 4d region for (i) Gd(NO3 )3 salt and (ii) Gd2 O3 NP; and (C)O 1s region for (i) Gd(NO3 )3 salt, (ii) Gd2 O3 NPs, and (iii) SiGd2 O3 NPs. Deconvolution of the O 1s high resolution spectra is given in graphs D–F where: (D) is the Gd(NO3 )3 salt (i), (E) is the bare Gd2 O3 NP (ii), and (F) is the Si–Gd2 O3 NP (iii).

at room temperature. The nanocomposite, showing a small shift to a higher field from 2288 to 2345 G as compared to ZnTCPPc, appears to be rather similar to that of the Pc (3) alone. The mix (5), on the other hand, shows characteristics that are reminiscent of the EPR spectra acquired for the NPs themselves, i.e., weak, broad signals, suggesting that formation of the nanocomposite exerted some change in the NP’s physical properties, while a simple mix did not.

EPR spectra of unsubstituted zinc phthalocyanines have been studied and single, narrow, intense EPR signals centered at g = 2.0036 with a line width of 4.25 G both in chloroform and in the solid state (at 290 K and at 100 K) were obtained. In our case, in addition to the obvious broad signal obtained for the ZnTCPPc alone (Fig. 10(i)), a small, narrow, sharp signal centered at 3518 G was observed (Fig. 10, inset). The nanocomposite, on the other hand,

72

M. Ledwaba et al. / Journal of Molecular Catalysis A: Chemical 403 (2015) 64–76

Fig. 9. EPR spectra acquired in the solid state at room temperature for (i) Gd(NO3 )3 salt, (ii) Gd2 O3 NPs (1) and (iii) Si–Gd2 O3 NPs (2). Inset: zoomed in section to show the spectra obtained for the NPs.

Fig. 12. UV–vis spectra obtained for the samples in DMSO: (i) Si–Gd2 O3 NPs (2), (ii) ZnTCPPc (3), (iii) ZnTCPPc–SiGd2 O3 nanocomposite (4) and (iv) ZnTCPPc–SiGd2 O3 NP-mix (5).

The UV–vis absorbance spectra of the Si–Gd2 O3 NPs (2) and the ZnTCPPc (3) alone in DMSO is shown in Fig. 12(i) and (ii), respectively. The absorption for the NPs was minimal and no significant changes were observed in the UV–vis spectra of the NPs (2) and the Pc (3) upon formation of the composite (4) or the mix (5) (Fig. 12(iii) and (iv), respectively). No shifts were observed in the Q band of the Pc for either the nanocomposite (4) or the mixture (5). However, other studies examining the spectroscopic properties of unsubstituted and Zn substituted Pcs conjugated to Au NPs have also observed only slight or no changes in the absorption, emission and excitation peak positions [45].

Fig. 10. EPR spectra acquired at room temperature in the solid state for (i) ZnTCPPc (3), (ii) ZnTCPPc–SiGd2 O3 nanocomposite (4) and (iii) ZnTCPPc–SiGd2 O3 NP (mix) (5). Inset: zoomed in section (3400–3600 G) for the ZnTCPPc (i) and the composite (ii).

showed a broad, shallow step centered at 3544 G, suggesting some changes having taken place upon composite formation. 3.1.7. UV–vis absorption spectra The ZnTCPPc alone (3) showed a broad absorbance peak attributed to the B band at 349 nm in DMSO, Fig. 11, with the monomeric behavior of the ZnTCPPc evidenced by a single, narrow Q-band at 678 nm.

Fig. 11. UV–vis spectrum of ZnTCPPc in DMSO.

3.1.8. Fluorescence spectra, quantum yields and lifetimes The fluorescence spectra of ZnTCPPc (3) are shown in Fig. 13 and the excitation spectra were found to be similar to absorption spectra, with both being were mirror images of the emission spectra. The position of the Q band in the absorption and excitation spectra were observed at 678 and 679 nm, respectively, while the emission peak was observed at 689 nm; giving a Stokes shift of 10 nm. The fluorescence excitation and emission spectra obtained are typical for phthalocyanine complexes in DMSO, where Stokes shifts habitually range from 3 to 19 nm [46] Fig. 14. The fluorescence quantum yields (ФF ) values were determined by the comparative method and the values were found to decrease by 0.09 upon conjugation to the gadolinium oxide nanoparticles. Fluorescence quantum yields are often influenced by the heavy atom affect and by the aggregation tendencies experienced by phthalocyanines. These commonly increase upon conjugation and

Fig. 13. Normalized absorption, emission and excitation spectra of ZnTCPPc (3) in DMSO.

M. Ledwaba et al. / Journal of Molecular Catalysis A: Chemical 403 (2015) 64–76

73

Fig. 15. Fluorescence lifetime decay curves of (i) ZnTCPPc (3) alone, (ii) ZnTCPPc–Gd2 O3 nanocomposite (4) and (iii) the mixture (5) in DMSO. exc = 689 nm.

Fig. 14. Normalized absorption, emission and excitation spectra of (A) ZnTCPPc–SiGd2 O3 (nanocomposite (4) and (B) ZnTCPPc–SiGd2 O3 NP mix (5) in DMSO (exc = 610 nm).

thus a decrease in the fluorescence quantum yield [47] is usually observed. Low ФF values (as indicated in Table 1) are therefore expected for the nanocomposite (4) and mixed (5) samples as intersystem crossing (ISC) to the triplet state is expected to be enhanced due to the presence of Gd in the NP system (due to the heavy atom effect). ISC to the triplet state increases the triplet quantum yield but it inevitably shortens the fluorescence lifetimes and quantum yields of the singlet excited states, hence the observed low fluorescence quantum yields for the conjugates shown in Table 1. Fluorescence lifetimes for MPc complexes are short and strongly dependent on the solvent used, the nature of the central metal ion and the nature of substituents on the Pc molecule. Fluorescence lifetimes were obtained for the ZnTCPPc (3), ZnTCPPc–Gd2 O3 NPs nanocomposite (4) and ZnTCPPc–SiGd2 O3 NPs mix (5) and the decay curves are shown in Fig. 15. All the samples showed bi-exponential fluorescence decay curves. The decrease in fluorescence lifetimes for the phthalocyanines alone have been explained

in the literature to be due to the formation of aggregates which are non-fluorescent, which then can also quench the monomer [48] (though the aggregates are not easily apparent in the UV–vis spectra). The fluorescence lifetime of the Pc decreased upon conjugation (4) and also after forming a simple mix (5), which shows that the Pc’s fluorescence is quenched by the mere presence of the gadolinium oxide nanoparticles. The decrease observed was much greater, however, upon forming the nanocomposite compared to the mix (5). The two lifetimes obtained for the nanocomposites could be related to the different orientations of the phthalocyanine molecule on the nanoparticle surface. It has been reported that when a fluorophore is in close proximity to a metal, the fluorophore interacts with the free electrons on the surface of the metal modifying its fluorescence behavior [49]. This results in an increase or a decrease in the fluorescence lifetime depending on the orientation of the fluorophore to the metal and the distance between the fluorophore and the metal [50]. The quenching of the lifetimes may be explained using fluorescence radiative lifetimes ( 0 ) which were calculated using Eq. (3): 0 =

F F

(3)

The results, shown in Table 1, indicate that the radiative lifetimes increase markedly upon introducing the NP to the phthalocyanine (either covalently or by simple mix), from 9.37 to 19.58 ns, indicating that the NP interacts strongly with the Pc, thus quenching the Pc’s fluorescence. 3.1.9. Triplet quantum yields (ФT ) and lifetimes ( T ) Fig. 16 shows the triplet decay curves for the ZnTCPPc (3), the nanocomposite (4) and the mix (5) samples in DMSO. Triplet quantum yields (T ) give an indication of the fraction of absorbing

Table 1 Absorbance, photophysical and photochemical properties of ZnTCPPc (3), the composite (4) and the mix (5). Complex

Solvent

abs (nm)

em (nm)

F

 F (ns)

ZnTCPPc (3)

DMSO

679

689

0.27

ZnTCPPc–Si–Gd2 O3 (4)

DMSO

679

690

0.18

ZnTCPPc– Gd2 O3 mix (5)

DMSO

679

688

0.12

2.53 0.79 2.31 0.68 2.35 0.65

 0 (ns)

T

 T (␮s)

D

9.37

0.61

319.25

0.32

12.83

0.69

306.97

0.36

19.58

0.69

313.76

0.49

74

M. Ledwaba et al. / Journal of Molecular Catalysis A: Chemical 403 (2015) 64–76

Fig. 16. Triplet decay curves for (i) ZnTCPPc (3), (ii) ZnTCPPc–Gd2 O3 nanocomposite (4), (iii) the mixture (5) and (iv) Si–Gd2 O3 NPs (2) in DMSO as observed at 490 nm. Exc  = 682 nm.

molecules that undergo intersystem crossing to the meta-stable triplet excited state. The triplet decay curves of all the complexes displayed first order decay profiles and the triplet state parameters (T , ␶T ) were determined experimentally using laser flash photolysis. The values obtained for T and ␶T are given in Table 1. The triplet quantum yields were found to increase in the presence of the Gd2 O3 NPs (either as a composite or as a simple mix) from 0.61 for the ZnTCPPc alone to 0.69 for both the ZnTCPPc–Si–Gd2 O3 nanocomposite (4) and ZnTCPPc–Si–Gd2 O3 mix (5), respectively. While the triplet quantum yields increased for both the nanocomposite and the mixed samples as compared to the Pc alone, a concomitant decrease in triplet lifetimes (i.e., 319, 307 and 314 ␮s for 3–5, respectively) was observed as shown in Fig. 16. Fig. 16(iv) also shows the negligible triplet lifetime of the silica capped Gd2 O3 NP (2) at an excitation wavelength of 490 nm. Based on this increase in the triplet quantum yield, the nanocomposite and mix were expected to show an increased ability to produce singlet oxygen. 3.1.10. Singlet oxygen quantum yields In this work, the singlet oxygen quantum yields of the complexes were determined using a chemical method, using the known singlet oxygen quencher diphenylisobenzofuran (DPBF) in DMSO. As mentioned previously, the singlet oxygen quantum yield (Ф ) values were expected to increase since the Ф yields are dependent on the corresponding triplet quantum yields (ФT ) of the photosensitizer. Since it is the photosensitizer species in the triplet state which results in the production of singlet oxygen, the efficiency of the energy transfer give an indication of the singlet oxygen generation efficiency of the sensitizer [51]. That is, if the triplet state of a photosensitizer is highly populated, the excited sensitizer can then interact with ground state triplet molecular oxygen – exciting the molecular oxygen to its singlet excited state to a greater extent. Though there is no direct correlation, an increase in ФT signifies an increase in Ф and this is what was observed (as shown in Table 1). Fig. 17 shows the decay of DPBF upon irradiation with time for ZnTCPPc. No significant decrease was observed in the Pc Q band for the period of irradiation time, an indication that the phthalocyanine has not undergone any degradation and was stable over the study. DPBF degrades due to the production of singlet oxygen by the Pc complexes. The decrease in the DPBF absorbance was monitored at ∼417 nm, with the rate of decay of the DPBF directly related to the production of singlet oxygen. The nanocomposite and the mix showed stable Pc Q bands together with a steady decrease in the DPBF band. An improved ability to generate singlet oxygen, where  values of 0.36 and 0.49 were obtained for 4 and 5, respectively, compared to the Pc alone [0.32], was observed. Interestingly, the

Fig. 17. Photodegradation of DPBF in the presence of ZnTCPPc in DMSO. [ZnTCPPc] = 4.8 × 10−6 M and [DPBF] = 2.5 × 10−5 M.

Fig. 18. Absorption spectral changes of a 1.93 × 10−5 mol L−1 Orange G solution during visible light photocatalysis using ZnTCPPc–SiGd2 O3 nanocomposite (4). The experiments were carried out using unbuffered, distilled water and the spectra recorded at 5 min intervals.

increase was particularly significant for the mix (5) and thus the overall efficiency of the Pc to generate singlet oxygen is enhanced in the presence of the Gd2 O3 NPs, with or without the formation of a covalent bond. 3.2. Phototransformation of Orange G (OG) 3.2.1. Spectroscopic characterization and kinetic studies for the photodegradation of Orange G (OG) Fig. 18 shows the absorption spectral changes observed during the photolysis of OG at 5 min intervals using ZnTCPPc alone as a photocatalyst, as an example. The study was carried out in water, thus the Q band for the Pc is not observed since the Pc is not soluble in water itself. However, the photocatalytic degradation of the OG is still observed. All experiments were carried out in 5 ml deionised water in a glass vial and the photocatalytic setup used was the same as that used in Section 3.1.10 for singlet oxygen detection. The light reaching the reaction vessel was found to be 3.2 × 1020 photons cm−2 s−1 . The degradation was monitored by observing the change in absorbance of OG at 478 nm [6]. The rate of photodegradation of OG decreased with an increase in OG concentration for the samples with (4) and without (3) the nanoparticles. The photodegradation rate observed for the nancomposite is faster than the Pc alone at the same concentration of OG. This was expected since the nanocomposite is more efficient at producing singlet oxygen (i.e., it has a higher singlet oxygen quantum yield) than the Pc (3) alone. Similarly, a decrease in kobs was observed as the concentration of the Orange G increased and comparable trends were observed when comparing the kobs values of

M. Ledwaba et al. / Journal of Molecular Catalysis A: Chemical 403 (2015) 64–76

75

Table 2 The initial rate, rate constant (kobs ) and half-life (t1/2 ) of various initial concentrations of Orange G using ZnTCPPc and the nanocomposite (4) at 2 mmol. Catalyst

[OG] ( × 10−5 mol L−1 )

kobs (min−1 )

Initial rate (mol L−1 min−1 )

Half-life (t1/2 ) (min)

ZnTCPPc (3)

2.3 3.5 4.8

0.008 0.0058 0.001

1.8 × 10−7 2.0 × 10−7 4.8 × 10−8

87 119 693

ZnTCPPcnanocomposite (4)

2.3 3.5 4.8

0.03 0.01 0.002

6.9 × 10−7 3.5 × 10−7 9.6 × 10−8

25 69 347

the nanocomposite (4) and Pc (3). This observation may also be explained by the increased singlet oxygen generation ability for the nanocomposite. The improved efficiency induced by the NP in OG photodegradation is further evidenced by the reduced halflives obtained for the ZnTCPPc (3) and nanocomposite (4) at all OG concentrations. The OG photodegradation rate is slower for the ZnTCPPc alone than for the nanocomposite. Plots obtained for ln (Co /C) versus irradiation time were found to be linear, showing that the reaction follows first order kinetics. The kinetic data is listed in Table 2. In addition to the studies carried out above, the absorption spectra of OG during photolysis were also acquired without the phthalocyainine as the sensitizer, i.e., using the Gd2 O3 and SiGd2 O3 nanoparticles alone, in order to evaluate the actual contribution of the nanoparticles to the photodegradation of OG. A control experiment with a solution containing only OG was subjected to the same conditions. While there is some information on the photocatalytic ability of Gd2 O3 NPs, the material is usually in combination with well-known photosensitisers such as TiO2 [52]. Magnetic nanoparticles (as well as AuNPs) have been shown to enhance the photosensitizing ability of phthalocyainines [6,11,12,21]. The results (shown in the supplementary information) indicate that OG, without the presence of the Pc or NPs, is stable to photodegradation. Similarly, no degradation was observed with the bare Gd2 O3 nanoparticles. However, the SiGd2 O3 nanoparticles were found to contribute a small amount to the photodegradation of OG (i.e., less than 8%) after 30 min of photolysis. It therefore seems that the majority of the photodegradation of OG may be attributed to the presence of the phthalocyanine and the ability of the Pc to cause the photodegradation of OG is enhanced in the presence of the nanoparticles. These three samples were also found to be stable in the dark as expected.

4. Conclusions The phthalocyanine and the Pc nanocomposite was successfully synthesized and characterized. Introduction of the Gd2 O3 NPs to the Pc to form a nanocomposite or simple mixture was shown to improve the photophysical properties (T and D ) of the Pc. The phthalocyanine and phthalocyanine–gadolinium oxide nanocomposites were found to be photoactive and are promising photosensitizers for the conversion of the environmental pollutant, Orange G. The nanocomposite (4) formed between the NPs and ZnTCPPc showed an increase in the photodegradation rate of OG as a consequence of the increased singlet oxygen quantum yields observed upon incorporating the Gd2 O3 NPs into the photocatalytic system. The simple mix (5) also demonstrated increased singlet oxygen quantum yields as compared to the Pc (3) alone. Preliminary studies showed that it is the Pc that is primarily responsible for the degradation of Orange G, with a small contribution (less than 8%) effected by the SiGd2 O3 nanoparticle.

Acknowledgements This work was supported through a National Research Foundation CSUR/KFD grant (South Africa), Rhodes University and by the Department of Science and Technology (DST) South Africa through a DST/NRF South African Research Chairs Initiative for the Professor of Medicinal Chemistry and Nanotechnology. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.molcata. 2015.03.023. References [1] M. Fox, Photocatalytic oxidation of organic substances, in: Kluwer (Ed.), Photocatalysis and Environment: Trends and Applications, Academic Publishers, New York, 1988, p. 445. [2] B.D. Bonnett, J. Porphyrins Phthalocyanines 5 (2001) 652. [3] M. Dini, D. Hanack, in: K.M. Kadish, K.M. Smith, R. Guilard (Eds.), The Porphyrin Handbook: Physical Properties of Phthalocyanine-Based Materials, Vol. 17, Academic Press, USA, 2003, pp. 22–31. [4] T. Nyokong, E. Antunes, in: K.M. Kadish (Ed.), The Handbook of Porphyrin Science, Vol. 7, Academic Press, New York, 2010, pp. 247–349 (World Scientific, Singapore, Chapter 34). [5] V. Iliev, A. Mihaylova, L. Bilyarska, J. Mol. Catal. A: Chem. 184 (2002) 121. [6] V. Chauke, T. Nyokong, J. Mol. Catal. A: Chem. 289 (2008) 9. [7] D. Wöhrle, O. Suvorova, R. Gerdes, O. Bartels, Ł. Łapok, N. Baziakina, S. Makarov, A. Slodek, J. Porphyrins Phthalocyanines 8 (2004) 1020. [8] M. Hu, Y. Xu, J. Zhao, Langmuir 20 (2004) 6302. [9] E. Marais, R. Klein, E. Antunes, T. Nyokong, J. Mol. Catal. A: Chem. 261 (2007) 36. [10] M. Alvaro, E. Carbonell, M. Esplá, H. Garcia, Appl. Catal. B 57 (2005) 37–42. [11] P. Modisha, E. Antunes, T. Nyokong, J. Nanosci. Nanotechnol. 14 (2014) 1. [12] P. Modisha, T. Nyokong, E. Antunes, J. Mol. Catal. A: Chem. 380 (2013) 131. [13] J. Chomoucka, J. Drbohlavova, D. Huska, V. Adam, R. Kizek, J. Hubalek, Pharmacol. Res. 62 (2010) 144. [14] T. Hyeon, Chem. Commun. (2003) 927. [15] S. Mornet, S. Vasseur, F. Grasset, P. Verveka, G. Goglio, A. Demourgues, J. Portier, E. Pollert, E. Duguet, Prog. Solid State Chem. 34 (2006) 237. [16] H. Bin Na, I. Chan Song, T. Hyeon, Adv. Mater. 21 (2009) 2133. [17] E. Cheraghipour, S. Javadpour, A.R. Mehdizadeh, J. Biomed. Sci. Eng. 5 (2012) 715. [18] D.W. Elliott, W.-X. Zhang, Environ. Sci. Technol. 35 (2001) 4922. [19] Y.H. Deng, C.C. Wang, J.H. Hu, W.L. Yang, S.K. Fu, Colloids Surf. A: Physicochem. Eng. Aspects 62 (2005) 87. [20] T. Sauer, G.C. Neto, H.J. Jose, R.F.P.M. Moreira, J. Photochem. Photobiol. A: Chem. 149 (2002) 147. [21] S. Tombe, E. Antunes, T. Nyokong, J. Mol. Catal. A: Chem. 371 (2013) 125. [22] Y. Li, T.M. Pritchett, J. Huang, M. Ke, P. Shao, W. SunJ, Phys. Chem. A 112 (2008) 472. [23] R. Bazzi, M.A. Flores-Gonzalez, C. Louis, K. Lebbou, C. Dujardin, A. Brenier, W. Zhang, O. Tillement, E. Bernstein, P. Perriat, J. Lumin. 102 (2003) 445. [24] D.J. Lee, S.Y. Park, Y.T. Oh, N.M. Oh, K.T. Oh, Y.S. Youn, E.S. Lee, Macromol. Res. 19 (2011) 848. [25] S. Fery-Forgues, D.J. Lavabre, J. Chem Educ. 76 (1999) 1260. [26] A. Ogunsipe, J.Y. Chen, T. Nyokong, New J. Chem. 7 (2004) 822. [27] T.H. Tran-Thi, C. Desforge, C. Thiec, J. Phys. Chem. 93 (1989) 1226. [28] K. Ishii, N. Kobayashi, in: K.M. Kadish, K.M. Smith, R. Guilard (Eds.), The Porphyrin Handbook, Vol. 16, Elsevier Science, New York, 2003, Chapter 102. [29] J. Turro, Modern Molecular Photochemistry, The Benjamin/Cummings Publishing Co., Inc., New York, 1978. [30] N.A. Kusnetsova, N.S. Gretsova, E.A. Kalmykova, E.A. Makarova, S.N. Dashkevich, V.M. Negrimovsky, O.L. Kaliya, E.A. Lukyanets, Russ. J. Chem. 70 (2000) 133.

76

M. Ledwaba et al. / Journal of Molecular Catalysis A: Chemical 403 (2015) 64–76

[31] J.C. Love, L.A. Estroff, J.K. Kriebel, R.G. Nuzzo, G.M. Whitesides, Chem. Rev. 105 (2005) 1103. [32] J. Ho Chang, K. Ho Kang, J. Choi, Y.K. Jeong, Superlattices Microstruct. 44 (2008) 442. [33] D.W. Mayo, F.A. Miller, R.W. Hannah, Course Notes on the Interpretation of Infrared and Raman Spectra, John Wiley and Sons, Inc., New Jersey, 2004, pp. 5. [34] X. Wang, Y. Liu, W. Qiu, D. Zhou, J. Mater. Chem. 12 (2002) 1636. [35] B. Ballesteros, C. de la Torre, G.M.A. Rahman, F. Agullo-Rueda, D.M. Guldi, T. Torres, J. Am. Chem. Soc. 129 (2007) 5061. [36] R. Seoudi, G.S. El-Bahy, Z.A. El Sayed, J. Mol. Struct. 753 (2005) 119. [37] M.E. Park, J. Ho Chang, Mater. Sci. Eng. C 27 (2007) 1232. [38] L. Zhou, Z. Gu, X. Liu, W. Yin, G. Tian, L. Yan, S. Jin, W. Ren, G. Xing, W. Li, X. Chang, Z. Huc, Y. Zhao, J. Mater. Chem. 22 (2012) 966. [39] K.S. Achar, B.N. Lokesh, J. Organomet. Chem. 689 (2004) 2601. [40] D. Raiser, J.P. Deville, J. Electron Spectrosc. 57 (1991) 91. [41] M.R. Alexander, R.D. Short, F.R. Jones, W. Michaeli, C.J. Blomfield, Appl. Surf. Sci. 137 (1999) 179.

[42] S. Watcharinyanon, D. Nilsson, E. Moons, A. Shaporenko, M. Zharnikov, B. Albinsson, J. Martensson, L.S.O. Johansson, Phys. Chem. Chem. Phys. 10 (2008) 5264. [43] D.T. Edwards, Z. Ma, T.J. Meada, D. Goldfarb, S. Han, M.S. Sherwin, Phys. Chem. Chem. Phys. 15 (2013) 5707. [44] K. Takeshita, S. Kinoshita, S. Okazaki, Chem. Pharm. Bull. 60 (2012) 31. [45] N. Nombona, E. Antunes, C. Litwinski, T. Nyokong, Dalton Trans. 40 (2011) 11876. [46] M.O. Liu, C.H. Tai, A.T. Hu, T.H. Wei, J. Organomet. Chem. 689 (2004) 2138. [47] A. Gilbert, J.E. Baggott, Essentials of Molecular Photochemistry, Blackwell Scientific, Oxford, 1991. [48] C.D. Geddes, J.R. Lakowicz, J. Fluoresc. 12 (2002) 121. [49] J.A. Lacey, D. Philips, Photochem. Photobiol. Sci. 1 (2002) 378. [50] C.D. Geddes, Topics in Fluorescence Spectroscopy, in: J.R. Lakowicz (Ed.), Springer, New York, 2005. [51] L. Kaestner, M. Cesson, K. Kassab, T. Christensen, P.D. Edminson, M.J. Cook, T. Chambrier, G. Jori, Photochem. Photobiol. Sci. 2 (2003) 660. [52] G. Mamba, X.Y. Mbianda, A.K. Mishra, Environ. Sci. Pollut. Res. 21 (2014) 5597.