Concentration dependent effect of CuCl₂ on the photocatalytic degradation of phenol over anatase, rutile and brookite TiO₂

Abstract

Effect of Cu²⁺ ions on the TiO₂-photocatalyzed degradation of organic pollutants in water has been reported in the literature, but the results are controversial. In this work, the effects of Cu²⁺ concentration, TiO₂ structures, and light intensity have been re-investigated, by using phenol degradation in an acidic aqueous suspension as a model reaction. As the concentration of CuCl₂ increased, the rates of phenol degradation over anatase, rutile, and brookite increased, and then decreased, but the reaction rate measured with P25 TiO₂ always decreased. Such unusual behavior of P25 was also observed when CuCl₂ was replaced with NaCl and MgCl₂, ascribed to the salt-induced particle aggregation. In all cases, Cu⁺ species were formed, followed by reoxidation to Cu²⁺ by O₂. However, at a high CuCl₂ loading, metallic Cu was also detected, due to the competitive disproportionation of Cu⁺ ions. An open circuit potential measurement with anatase and P25 film electrodes revealed that the interfacial electron transfer from the irradiated TiO₂ to Cu²⁺ was faster than that to O₂. Furthermore, the reaction rate was proportional to light intensity. The relevant quantum yield of phenol degradation over rutile upon the addition of CuCl₂ was increased by approximately 4 times.

---

1. Introduction

Semiconductor photocatalysis for environmental remediation has been widely studied.^1–3 Although many semiconductors have been examined, TiO₂ is still the best in terms of the cost, activity, and stability. More importantly, over the irradiated TiO₂, the reduction of O₂ and the oxidation of water can occur at the same time, with the formation of various reactive species including H₂O₂ and ˙OH. Because of that, many recalcitrant organic pollutants can be dagraded to CO₂ and small fragments at room temperature.^4–6 However, the quantum efficiency of organic degradation over TiO₂ is usually very low.⁷ In general, TiO₂ after excited with UV light will generate electrons (e_cb⁻) and holes (h_vb⁺) in its conduction and valence bands, respectively. These charge carriers may recombine to heat, and/or migrate to surface reacting with target sorbates. Therefore, to improve the photocatalytic activity of TiO₂, the interfacial charge transfer to O₂ and organic substrates need to be accelerated.

Transition metal ions are commonly present in aquatic environments. Their effect on the direct photolysis and the TiO₂-photocatalyzed degradation of organic pollutants in aqueous solution have been studied, and reviewed.^8,9 For example, Cu²⁺ ions can promote organic degradation, ascribed to the reduction of Cu²⁺ by e_cb⁻ of TiO₂, followed by regeneration through O₂ in aqueous solution.¹⁰ However, both the positive and negative effects of Cu²⁺ have been reported in the literature. For the photocatalytic degradation of dioxane,¹¹ phenol,¹² and resorcinol,¹³ the rates are increased upon addition of 0.1, 0.4, and 0.01–5.21 mM Cu²⁺, respectively, whereas for the photocatalytic degradation of 4-nitrophenol,¹⁴ and phenol,^10,15 the rates are decreased upon addition of 1.0, and 0.06–1.2 mM Cu²⁺, respectively. In these studies, the photocatalyst is P25, which composes of 80% anatase and 20% rutile. Strangely, when a synthetic mixture of 80% anatase and 20% rutile is used as photocatalyst, the rate of methamidophos degradation increases, and then decreases with the increase of Cu²⁺ concentration.¹⁶ A maximum degradation rate of methamidophos appears at 6 μM Cu²⁺. Such concentration effect of Cu²⁺ is also observed from toluene degradation over anatase,¹⁷ and from phenol degradation on a reduced anatase.¹⁸ But their maximum rates are located at 1 μM and 1 mM Cu²⁺, respectively. When rutile is used as photocatalyst, the rate of phenol degradation is increased upon addition of 0.90 mM Cu²⁺.¹⁰ In nature, TiO₂ exists in three crystal forms of anatase, rutile, and brookite. But no study of the Cu²⁺ effect on the brookite photocatalyzed reaction has been found. Furthermore, the mechanism proposed for the Cu²⁺ effect is also controversial. On one hand, the positive effect of Cu²⁺ is due to the reduction of Cu²⁺ improving the efficiency of charge separation,^12,18 and due to the Cu(II)-assisted toluene degradation not related with TiO₂,¹⁷ and due to the Cu(II)-assisted adsorption and degradation of organic intermediates on TiO₂,^11,13 and due to the Cu²⁺-promoted desorption of O₂⁻˙ from TiO₂ helping HCHO oxidation in solution.¹⁹ On the other hand, the negative effect of Cu²⁺ is due to the solution filter of excess Cu²⁺,¹⁷ and/or the formation of Cu or Cu₂O that reduces the number of photons reaching TiO₂, and consequently deactivates the TiO₂-photocatalyzed reaction.^10,15 The negative effect of Cu²⁺ is also ascribed to the hole oxidation of Cu^+/0 making Cu²⁺ in null recycles.^12,14,18 According to the above literature survey, the observed effect of Cu²⁺ remains incompletely understood.

In this work, we report a great increase in the photocatalytic activities of anatase, rutile, and brookite TiO₂, but not of P25 TiO₂, upon the addition of CuCl₂ (1 μM to 2 mM). In all of the studies, phenol degradation in an aerated aqueous suspension at initial pH 3.0 was used as a model reaction, as so to avoid the possible effect of organic photolysis and adsorption on the evaluation of the real photocatalytic activity of TiO₂. Experiments were carried out under UV light at wavelengths longer than 320 nm. To clarify the discrepant literature result, the effect of CuCl₂ concentration and light intensity were investigated. The interfacial electron transfer from the irradiated TiO₂ to Cu²⁺ or to O₂ was examined through an open circuit potential technique. The possible formation of Cu^+/0 species from the Cu²⁺ reduction was analyzed with a colorimetric method, X-ray powder diffraction (XRD), and X-ray photoelectron spectroscopy (XPS). Furthermore, a plausible mechanism for the observed effect of Cu²⁺ is discussed.

2. Experimental section

2.1 Materials

Bathocuproine disulfonic acid disodium salt (Cu reagent), and titanium bis(ammonium lactate) dihydroxide (TALH) were purchased from Sigma-Aldrich. Other chemicals were purchased from Shanghai Chemicals, Inc., including CuCl₂, NaCl, MgCl₂, K₂Cr₂O₇, and phenol. Anatase (cAT) and rutile (cRT) TiO₂ were purchased from Sigma-Aldrich, and P25 TiO₂ from Degussa. Brookite (sBT) was homemade by the hydrothermal reaction of TALH in the aqueous solution of urea at 160 °C for 24 h.²⁰ The precipitate was collected by centrifugation, washed with ethanol and distilled water, dried at 60 °C, and finally annealed in air at 500 °C for 3 h. Solid was characterized by XRD, N₂ adsorption, Raman, and UV-vis diffuse reflectance spectroscopes (Fig. S1 of the ESI†). Table 1 shows the relevant physical parameters for those samples.

Table 1 Physical parameters of TiO₂ photocatalysts^a

Samples	Phase (%)	dXRD (nm)	Asp (m^2 g^−1)	Eg (eV)
a A, B and R represent anatase, brookite, and rutile, respectively; dXRD, average crystallite size; Asp, BET surface area; Eg, band gap energy.
cAT	A100	A13.4	144.0	3.29
P25	A80, R20	A22.0, R32.0	50.0	3.10
sBT	B100	B22.4	33.6	3.28
cRT	R99, A1	R70.4	0.9	3.00

---

2.2 Photocatalysis and analysis

Reactions were carried out at 25 °C in a Pyrex-glass reactor. The aqueous suspension (50 mL) containing necessary components (0.43 mM phenol, 0.27 mM chromate, 1.00 g L⁻¹ TiO₂, pH 3.0) was stirred in the dark for 1 h, and then irradiated with a high pressure mercury lamp (300 W, Shanghai Yamin). The light intensity reaching the external surface of the reactor was 4.50 mW cm⁻². At given intervals, 2.0 mL of the suspension was withdrawn, filtered, and analyzed by HPLC (high performance liquid chromatography) on a Dionex P680 (Apollo C18 reverse column, and 50% CH₃OH eluent). Cu(I) species was analyzed by using a Cu reagent,²¹ and the resulting complex spectrum was recorded on an Agilent 8453 UV-visible spectrophotometer. Chromate was measured at 540 nm through its complex with 1,5-diphenylcarbazide.²² Chloride ions were analyzed by IC (ionic chromatography) on a Dionex ICS90, equipped with an AS14 column (mobile phase, 3.5 mM Na₂CO₃ and 1.0 mM NaHCO₃; flow rate, 1.2 mL min⁻¹).

2.3 Electrode fabrication and measurement

A TiO₂ film electrode was prepared by the doctor blade method. A gel containing 2 wt% TiO₂ and 7.5 wt% PVA (polyvinylalcohol) was used to coat an indium-doped tin oxide (ITO) glass (Nippon Sheet Glass), followed by annealing in air at 500 °C for 3 h. The resulting ITO glass was cut into several pieces. Each of them had an exposed area of 1 cm × 1 cm, whereas other part was sealed by epoxy resin. The (photo)electrochemical measurements were carried out on a CHI660E Electrochemical Station(Chenghua, Shanghai), with a saturated calomel electrode (SCE) as the reference electrode, and a platinum gauze as the counter electrode. The TiO₂ film was used as working electrode, and illuminated with a 500 W Xe lamp from the electrode/electrolyte side through a quartz window. The supporting electrolyte was 0.5 M NaClO₄ at pH 3.0.

3. Results and discussion

3.1 Effect of crystal structure

Fig. 1 shows the result of phenol degradation in aqueous solution at initial pH 3.0. As the irradiation time increased, the concentration of phenol in aqueous phase exponentially decreased (Fig. S2†). This time profile of phenol degradation well fit the pseudo-first-order rate equation. Such pseudo-first-order kinetics is often observed in TiO₂ photocatalysis for organic degradation at low concentration.^23–25 Under a fixed condition, the concentration of the reactive species photogenerated from TiO₂ is constant. Therefore, the rate equation of phenol degradation is first order in phenol. Since the adsorption of phenol on TiO₂ in aqueous solution is negligible, a higher rate of phenol degradation would represent a higher photocatalytic activity of TiO₂ for the generation of the reactive species. In the absence of CuCl₂, the photocatalytic activity of TiO₂ increased in the order of cRT > cAT > sBT. In the presence of 0.50 mM CuCl₂, the photocatalytic activity of TiO₂ also increased in the order of cRT > cAT > sBT. Note that the above activity order will change into cRT > sBT > cAT, when the apparent rate of phenol degradation is normalized with the BET surface area for those oxides (Table 1). However, in all cases, the photocatalytic activity of TiO₂ in the presence of CuCl₂ was much higher than that in the absence of CuCl₂. Control experiments in a homogeneous aqueous solution showed that there was negligible reaction between CuCl₂ and phenol either in the dark or under UV light. These observations indicate that CuCl₂ is beneficial to all of the photocatalytic reactions over anatase (cAT), rutile (cRT), and brookite (sBT). To eliminate the possible effect of Cl⁻ ion, a separate experiment with 0.50 mM NaCl was performed. The rate constants of phenol degradation over cAT, cRT and sBT upon the addition of NaCl only changed a little (±2%). Therefore, the observed positive effect of CuCl₂ is due to Cu²⁺, not due to Cl⁻. However, P25 was special. Upon addition of 0.50 mM CuCl₂ and NaCl, the photocatalytic activity of P25 for phenol degradation was decreased by 0.42 and 0.19 times, respectively. This negative effect of CuCl₂ on phenol degradation over P25 has been reported,^10,14,15 which is not due to the so-called mixed phase effect of P25. A control experiment with a mechanical mixture of 80% cAT and 20% cRT showed that the rate of phenol degradation upon the addition of 0.50 mM CuCl₂ was increased by 0.74 times, whereas the effect of NaCl on the rate of phenol degradation was negligible (2.2%). Furthermore, the rate of phenol degradation obtained from the mixture of cRT and cAT was larger than the rate of phenol degradation measured from cRT or cAT. Since the TiO₂ loading was fixed at 1.00 g L⁻¹, this is ascribed to the molecular oxygen transfer from anatase to rutile, that exploits the intrinsic photocatalytic activity of rutile, and consequently improves the apparent activity of rutile for phenol degradation in an aerated aqueous susspenion.²⁹ Therefore, the negative effect of CuCl₂ and NaCl is only special to P25, which will be further discussed below.

Fig. 1 Apparent rate constants (k_obs) of phenol degradation in the aqueous suspensions of TiO₂ at initial pH 3.0, measured in the absence and presence of 0.50 mM CuCl₂ or NaCl, respectively. Herein, mixed TiO₂ stands for a mechanical mixture containing 80% cAT and 20% cRT.

3.2 Effect of CuCl₂ concentration

Fig. 2 shows the results of phenol degradation in aqueous suspension at initial pH 3.0. As the concentration of CuCl₂ added in the suspension increased, the rate of phenol degradation increased, and then decreased. A maximal rate of phenol degradation was observed at 0.25, 0.50, and 0.50 mM CuCl₂ for sBT, cAT, and cRT, respectively. At this CuCl₂ loading, the photocatalytic activities of sBT, cAT, and cRT were increased by 1.93, 1.57, and 1.77 times, respectively. However, in such concentration region of CuCl₂ (1 μM to 4 mM), the photocatalytic activity of P25 always decreased with the increase of CuCl₂ concentration. In comparison with sBT, cAT, and cRT, P25 was highly dispersed in aqueous solution with a long time stability. Then, it is highly possible that there is an inorganic salt effect facilitating the particle aggregation of P25 in aqueous solution. As a result, the number of photons adsorbed by P25 is reduced, and the rate of the P25-photocatalyzed reaction is decreased. To practice this hypothesis, CuCl₂ was replaced with MgCl₂. Surprisingly, the rate of phenol degradation over P25 also always decreased with the increase of MgCl₂ concentration. Such salt effect has been examined in the literature. Yamazaki and coworkers report that in the N₂-purged suspension of P25 and formate, the photocatalytic reduction of Cu²⁺ is independent of the ionic strength (0.1–0.5 μM), adjusted with NaClO₄ or NaNO₃.²⁶ Minero and coworkers report that the aqueous suspension of P25 (0.100 g L⁻¹) at pH 3.5 becomes unstable upon addition of 10 mM NaF. This is ascribed to the fluoride adsorption reducing the density of the surface positive charge, and consequently promoting the particle aggregation of P25.²⁷ In the present case, a high ionic strength (0.75–12.0 mM), a high loading of P25 (1.00 g L⁻¹), and a low solution pH (3.0) were used, all of which would favor the particle aggregation of P25. Through an ionic chromatography, the amount of Cl⁻ adsorbed on TiO₂ in aqueous solution at pH 3.0 was measured, which was 61, 45, 38, and 26 μmol g⁻¹ for P25, sBT, cAT, and cRT, respectively. Such anion adsorption on TiO₂ would be more critical to P25, as compared with other TiO₂ samples. The possible salt-induced particle aggregation of P25 was further analyzed by SEM (scanning electron microscopy), and the result was shown in Fig. S3.† To ensure the detection of particle aggregation, a high concentration of the relevant components was used (2.00 g L⁻¹ TiO₂, 0.43 mM phenol, pH 3.0, and 50 mM CuCl₂ or 50 mM MgCl₂). After the suspension was irradiated with UV light for 2 h, the particles were collected by filtration, followed by washing with distilled water, and drying at 45 °C overnight. In the absence of CuCl₂ or MgCl₂, the particles of P25 were well dispersed. However, after salt treatment, some of the primary particles of P25 stacked together, with the formation of a plate-like particle. Therefore, the salt present in aqueous suspension of P25 has a positive effect on the particle aggregation. However, the negative effect of MgCl₂ was about half that of CuCl₂ (curves d′ versus d, Fig. 2). Other factors are also possible, such as the formation of metallic Cu, which will be shown in Section 3.5.

Fig. 2 Effect of CuCl₂ concentration on the photocatalytic degradation of phenol in the aerated aqueous suspensions of (a) sBT, (b) cAT, (c) cRT, and (d) P25. Curve (d′) represents the effect of MgCl₂ concentration, measured with P25.

The observed positive effect of CuCl₂ implies that there is a lot of e_cb⁻ and h_vb⁺ available to reactants on the surface of TiO₂. However, upon the addition of CuCl₂, sBT, cAT, and cRT samples showed different activity enhancement. On one hand, these samples may have different sorption capacities toward Cu²⁺ in an acidic aqueous suspension. On the other hand, these samples may have different intrinsic photocatalytic activities. The amount of Cu²⁺ adsorbed on TiO₂ in aqueous solution was too low to be quantitatively determined. The intrinsic activity of TiO₂ exponentially increases with its synthesis temperature, regardless of the solid structures in the forms of anatase, rutile, brookite, and their mixture.^28–31 A high intrinsic photocatalytic activity of TiO₂ means a large number of the photogenerated e_cb⁻ and h_vb⁺ that have migrated and reached the oxide surface. To evaluate the intrinsic photocatalytic activity of TiO₂, the reduction of Cr(VI) to Cr(III), with the concurrent oxidation of H₂O to O₂, was used as a model reaction,^31,32 and the result is shown in Table 2. The initial rate of Cr(VI) reduction (R₀) increased in the order of cAT > cRT > sBT, but the initial amount of Cr(VI) adsorption (q₀) increased in another order of cAT > sBT > cRT. Since R₀ is proportional to q₀, the specific rate (R₀/q₀) was then calculated. The value of R₀/q₀ would represent the intrinsic activity of TiO₂, which increased in the order of cRT > cAT > sBT. This activity trend among the samples is in agreement with that observed from the maximum rate of phenol degradation in the presence of CuCl₂ (Fig. 2). These observations indicate that the intrinsic photocatalytic activities of anatase, rutile, and brookite are all exploited upon the addition of CuCl₂. As a result, the apparent photocatalytic acitivity of TiO₂ for phenol degradation in an aerated aqueous suspension is greatly improved. Under a fixed condition the number of e_cb⁻ and h_vb⁺ on the surface of TiO₂ would be limited. Therefore, with each of TiO₂ samples, an optimal loading of CuCl₂ was observed, at which the maximum rate of phenol degradation was reached (Fig. 2). In this respect, a higher intrinsic photocatalytic activity of TiO₂ would correspond to a higher optimal loading of CuCl₂. In practice, however, it is not the case. For example, the intrinsic activity of cAT is much lower than that of cRT (Table 2), but cAT shows the same optimal CuCl₂ loading as does cRT (Fig. 2). It is highly possible that Cu²⁺ ions are recyclable during the TiO₂-photocatalyzed of phenol in an aerated aqueous suspension.

Table 2 Photocatalytic reduction of Cr(VI) over TiO₂^a

a q0, the initial amount of Cr(VI) adsorption, measured in the dark; R0, the initial rate of Cr(VI) reduction, measured at the first 10 min under UV light. Experiment was carried out in an aerated aqueous suspension at pH 3.0.
Catalysts	sBT	cAT	cRT	P25
q0 (μmol g^−1)	51.88	100.10	1.72	47.34
R0 (μM min^−1)	2.21	5.41	3.85	10.37
R0/q0 [g (min L)^−1]	0.04	0.05	2.24	0.22

---

3.3 Effect of light intensity

Fig. 3A shows the apparent rate constants (k_obs) of phenol degradation, measured in the aqueous suspension of cRT. In this case, a Xenon lamp was used, and its intensity reaching the reactor (I_hν) was measured by an irradiance meter. At given I_hν, the rate of phenol degradation was a function of CuCl₂, as observed in Fig. 2. At given concentration of CuCl₂, the rate of phenol degradation increased with the light intensity. In general, the rate of organic degradation over TiO₂ is proportional to I_hν^θ, where θ = 1 at low I_hν, and θ = 0.5 at high I_hν.^23–25 In the present case, the plots of ln k_obs vs. ln I_hν were all satisfactorily linear (Fig. 3B), the slopes of which were 0.72 ± 0.07, 1.05 ± 0.06, 1.00 ± 0.05, 1.00 ± 0.03, and 1.11 ± 0.08 for the reactions at 0, 0.5, 1.0, 2.0 and 4.0 mM CuCl₂, respectively. Interestingly, the slopes obtained in the presence of CuCl₂ were all larger than that in the absence of CuCl₂. A small value of θ implies a serious recombination of e_cb⁻ and h_vb⁺. In other words, the charge separation of TiO₂ is improved upon addition of CuCl₂. To estimate the quantum yield of phenol degradation (Q) at given wavelength, the exact number of photons adsorbed by TiO₂ in aqueous suspension is needed, but it is difficultly measured. Nevertheless, the value of Q may be estimated from the plot of k_obs (10⁻³ min⁻¹) vs. I_hν (mW cm⁻²), which was also linear (not shown here). This plot gave a slope of 0.059 ± 0.008, 0.307 ± 0.012, 0.321 ± 0.013, 0.321 ± 0.007, and 0.337 ± 0.017 for the reactions at 0, 0.5, 1.0, 2.0 and 4.0 mM CuCl₂, respectively. The absolute value of Q is meaningless, but the number of photons adsorbed by TiO₂ would be influenced a little by CuCl₂. Then, according to the relative value of Q, one can say that the quantum yield of phenol degradation over rutile upon the addition of CuCl₂ was increased by approximately 4 times.

Fig. 3 (A) Apparent rate constants (k_obs) of phenol degradation over cRT, obtained under different light intensity (I_hν) of (a) 22.0, (b) 25.2, (c) 36.1, and (d) 51.4 mW cm⁻². (B) The corresponding plots of ln k_obs versus ln I_hν, where CuCl₂ concentration is indicated by the legend.

Furthermore, under different light intensity, the maximum rates of phenol degradation were all located around 1.0 mM CuCl₂ (Fig. 3A), which is not expected. The increase of light intensity would result into increase in the number of e_cb⁻ and h_vb⁺ photogenerated, and thus in the amount of Cu²⁺ consumed. These observations suggest again that Cu²⁺ ions once reduced are regenerated in situ. This recycling behavior of Cu²⁺ will be further discussed below.

3.4 (Photo)electrochemical measurement

To examine the electron transfer from the irradiated TiO₂ to Cu²⁺, an open circuit potential (OCP) technique was applied. In this case, cAT and P25 were used as the representative TiO₂. Fig. 4 shows the time profiles of OCP, measured under N₂ with the TiO₂ film electrodes in 0.50 M NaClO₄ at pH 3.0. When the electrode was illuminated with a Xenon lamp, the electrode potential was negatively shifted, due to the photogenerated electrons of TiO₂. After a certain time, the potential became stable, due to the balance between the electron generation, recombination, and interfacial transfer. After the light was blocked off, the electrode potential dropped toward a more positive value. This decay of OCP with time is due to recombination of the electrons with trapped holes, and/or due to the electron transfers to oxidants in the solution. When the electrode was illuminated again, the OCP turned to shift negatively, and then reached a stable state. After the light was switched off, an aqueous solution of CuCl₂ (80 μL, 0.50 M) was added immediately to the electrolyte. At this moment, the potential of the electrode dropped again to a more positive value. However, the OCP decay with time in the presence of Cu²⁺ was faster than that in the absence of Cu²⁺. These phenomena were observed with both cAT and P25. In general, the electron density on the electrode exponentially increases with the potential. A faster decay rate of OCP implies a faster electron transfer. Since no other electron acceptors were introduced, there would be an interfacial electron transfer from the irradiated TiO₂ to Cu²⁺. Similar observations with the cAT and P25 film electrodes were also obtained under air (Fig. S4†). In this case, both of Cu²⁺ and O₂ were present in the electrolyte. However, the OCP decay with time was notably faster in the presence of CuCl₂ and air, as compared to that in the presence of air alone. Moreover, the OCP decay with time in the presence of Cu²⁺ under N₂ (Fig. 4) is obviously faster than that observed under air in the absence of Cu²⁺ (Fig. S4†). These observations clearly indicate that Cu²⁺ ions are more efficient than O₂ to capture the photogenerated electrons of TiO₂ in aqueous suspension.

Fig. 4 Open circuit potentials of (A) cAT, and (B) P25 film electrodes, measured under N₂ in 0.5 M NaClO₄ at pH 3.0. The amount of CuCl₂ added was 40 μmol.

3.5 The reduced Cu(II) species

The photocatalytic reduction of Cu(II) over TiO₂ may result in the formation of Cu(I), Cu(0), and/or both. At the first, the Cu(I) species produced in aqueous solution was detected by a Cu reagent.²¹ To accelerate the reduction of Cu²⁺, and to avoid the re-oxidation of Cu^+/0, the reactions were carried out under N₂ by using excess phenol as the hole scavenger of TiO₂. First, an aqueous suspension containing 1.00 g L⁻¹ TiO₂, 0.43 mM phenol, and 4.0 mM CuCl₂ was irradiated under N₂ for 1.5 h. Then, to the irradiated suspension, a certain amount of copper reagent was added under N₂, followed by stirring for 5 min. After the particles were removed, the filtrate was collected. The Cu(I) complex was orange (Fig. S5†), and showed a broad absorption band centered at 485 nm (Fig. 5). Comparatively, the un-irradiated suspension upon the addition of the Cu reagent did not show color change. By using the dark sample as reference, the net absorbance at 485 nm was calculated, which was 0.13, 0.31, and 0.76 for sBT, cAT, and cRT, respectively. That is, for the reduction of Cu(II) to Cu(I), the photocatalytic activity of TiO₂ increases in the order of cRT > cAT > sBT, which is the same as that observed from the photocatalytic degradation of phenol in the presence of 4.0 mM CuCl₂ (Fig. 2). When the above irradiated suspensions were exposed to air for 10 min, followed by addition of the copper reagent, no orange color was observed. The resulting spectra nearly overlapped with those of the un-irradiated sample (Fig. S6†). These observations indicate that the photocatalytic reduction of Cu(II) to Cu(I) can occur with all of anatase, rutie, and brookite, followed by quick reoxidation of Cu(I) to Cu(II) through O₂ reduction.

Fig. 5 UV-vis absorption spectra of a Cu(I) complex with bathocuproine disulfonic acid, obtained from the filtrates of (a) sBT, (b) cAT, (c) cRT, and (d) P25. The suspension containing TiO₂ (1.00 g L⁻¹), CuCl₂ (4.0 mM), and phenol (0.43 mM) was degassed by N₂, and then irradiated with UV light for 1.5 h. The dotted spectra correspond to those recorded before light irradiation.

The formation of metallic Cu was analyzed by XRD and XPS. To ensure the detection of solid Cu, experiment was carried out at a high concentration (2.00 g L⁻¹ TiO₂, 50 mM CuCl₂, 3.20 mM phenol, and pH 3.0). After the suspension was irradiated under air for 22 h, the particles were collected by filtration, followed by washing with distilled water, and drying at 50 °C in a vacuum oven. Interestingly, on the filter paper, there were many brown spots (Fig. S7†), probably indicative of the formation of Cu₂O or Cu⁰ on TiO₂.¹⁵ However, in the dried samples, neither Cu nor Cu₂O was detected by XRD (Fig. S8†), mostly due to their low content. Then, the sample was analyzed by XPS. There were two peaks at 932.9 and 952.8 eV (Fig. 6). These binding energies of Cu 2p can be assigned either to metallic Cu or to Cu₂O.³³ Recall that Cu⁺ is not stable against oxidation by O₂ (Fig. S6†). Since these samples were prepared under air, we speculate that the observed XPS signals result only from metallic Cu. Such XPS signals were observed from all of the irradiated aqueous suspensions of sBT, cAT, cRT, and P25 under air and at a high CuCl₂ concentration.

Fig. 6 XPS spectra of Cu 2p recorded with (a) sBT, (b) cAT, (c) cRT, and (d) P25. The suspension (2.00 g L⁻¹ TiO₂, 3.20 mM phenol, and 50 mM CuCl₂, pH 3.0) was irradiated for 22 h under air, followed by collection, washing and drying.

3.6 Possible mechanism

In thermodynamics, the electron transfers from the irradiated anatase, rutile, and brookite to Cu²⁺ are all possible. In an aqueous solution at pH 0, the redox potential for the Cu²⁺/Cu⁺ couple is 0.16 V versus normal hydrogen electrode (NHE). This potential is more positive than any of the conduction band edge potentials for anatase, rutile, and brookite, which are −0.12, 0.10, and −0.24 V vs. NHE, respectively.^2,4,20,31 In fact, the electron transfer from the irradiated TiO₂ to Cu²⁺, and the formation of Cu(I) are observed by an OCP measurement (Fig. 4) and absorption spectroscopy (Fig. 5), respectively. Then the resulting Cu(I) species are quickly re-oxidized back to Cu²⁺ by O₂ (Fig. S6†), mostly with the formation of H₂O₂.³⁴ In aqueous solution at pH 0, the redox potentials for the O₂/HO₂⁻˙ and O₂/H₂O₂ couples are −0.05, and 0.68 V vs. NHE, respectively. The former and latter are less and more positive than that for the Cu²⁺/Cu⁺ couple, respectively. Therefore, during phenol degradation in the aerated aqueous solution of TiO₂, Cu²⁺ ions are recyclable. On the other hand, the reduction of Cu²⁺ to Cu⁺ is faster than the one-reduction of O₂ over anatase and brookite,^4,20,31 as well as the two-electron reduction of O₂ over rutile.³⁵ This is predicted in the views of both the driving force and orbital overlapping between the conduction band edge of TiO₂ and the unfilled orbital of oxidant. In fact, a faster reduction of Cu²⁺ than that of O₂ is observed from a study of the light intensity effect (Fig. 3), and the time-dependent decay of OCP as well (Fig. 4). Therefore, all of the photocatalytic degradation of phenol over anatase, rutile, and brookite in the presence of CuCl₂ are faster than those in the absence of CuCl₂.

However, as the Cu²⁺ concentration is further increased, the rate of phenol degradation turns to decrease after reaching a maximum (Fig. 2). Such negative effect of Cu²⁺ can result from several factors. First, the aqueous solution of CuCl₂ is blue in color (Fig. 5). Second, there is formed a metallic Cu in the aerated suspension (Fig. 6). All of these would reduce the number of photons absorbed by TiO₂, consequently decelerating the TiO₂-photocatalyzed reactions. In an acidic aqueous solution, Cu⁺ ions can quickly disproportionate into Cu²⁺ and Cu⁰. At a high concentration, such disproportionation of Cu⁺ may easily occur, as observed from metallic Cu present in the aerated aqueous suspension of TiO₂ (Fig. 6). In other words, there is a competition between the reoxidation and disproportionation of Cu⁺ during the TiO₂-photocatalyzed reactions. Therefore, the rate of phenol degradation after reaching a maximum turns to decrease with the increase of Cu²⁺ concentration. As P25 is concerned, not only its particles in aqueous suspension are easily aggregated upon addition of salt, but also its photocatalytic activity is high, as compared with those of cAT, cRT and sBT (Fig. 1). This may result in a large rate in the formation and growth of metallic Cu on P25. As a result, the rate of phenol degradation over P25 always decreases with the increase of CuCl₂ concentration.

4. Conclusions

In this work, we have shown that Cu²⁺ ions are positive to all of the photocatalytic degradation of phenol over anatase, rutile, and brookite in an aerated aqueous solution. The observed negative effect of Cu²⁺ at a high concentration is due to the combination between salt effect, solution filter effect, and metallic Cu effect, all of which would reduce the number of photons absorbed by TiO₂. Such combined negative effect is very serious to the P25-photocatalyzed reaction, because the particles of P25 are fine and highly dispersed in aqueous solution. Through the measurements about the fate of e_cb⁻, the formation of Cu(I), and the quantum yield of phenol degradation, it become clear that there is an interfacial electron transfer from the irradiated TiO₂ to Cu²⁺, followed by the re-oxidation of Cu⁺ to Cu²⁺ by O₂. This recycle of copper species would improve the efficiency of the charge separation of TiO₂, and consequently increase the rate of phenol degradation. However, at a high concentration of Cu²⁺, the disproportionation of Cu⁺ into Cu²⁺ and metallic Cu becomes fast, which is harmful to the TiO₂-photocatalyzed reaction. Therefore, for the practical application of TiO₂ photocatalysis, a low concentration of Cu²⁺ should be used. On the other hand, when TiO₂ is modified with a solid copper compound such as CuO, the dissolved Cu²⁺ ions in aqueous solution should be taken into account.^36,37 The present work is important to the further development of a photocatalyst for water treatment.

Acknowledgements

This work was supported by NSFC (No. 21377110).

References

1. M. R. Hoffmann, S. T. Martin, W. Choi and D. W. Bahnemann, Chem. Rev., 1995, 95, 69.
2. O. Carp, C. L. Huisman and A. Reller, Prog. Solid State Chem., 2004, 32, 33.
3. S. G. Kumar and L. G. Devi, J. Phys. Chem. A, 2011, 115, 13211.
4. T. Tachikawa, M. Fujitsuka and T. Majima, J. Phys. Chem. C, 2007, 111, 5259.
5. P. Salvador, J. Phys. Chem. C, 2007, 111, 17038.
6. J. Ryu and W. Choi, Environ. Sci. Technol., 2008, 42, 294.
7. A. V. Emeline, X. Zhang, M. Jin, T. Murakami and A. Fujishima, J. Phys. Chem. B, 2006, 110, 7409.
8. P. Cieśla, P. Kocot, P. Mytych and Z. Stasicka, J. Mol. Catal. A: Chem., 2004, 224, 17.
9. M. I. Litter, Appl. Catal., B, 1999, 23, 89.
10. L. Wan, J. Sheng, H. Chen and Y. Xu, J. Hazard. Mater., 2013, 262, 114.
11. S. W. Lam, M. Hermawan, H. M. Coleman, K. Fisher and R. J. Amal, J. Mol. Catal. A: Chem., 2007, 278, 152.
12. N. Wang, Z. Chen, L. Zhu, X. Jiang, B. Lv and H. Tang, J. Photochem. Photobiol., A, 2007, 191, 193.
13. S. W. Lam, K. Chiang, T. M. Lim, R. Amal and G. K.-C. Low, Appl. Catal., B, 2005, 55, 123.
14. N. San, A. Hatipoğlu, G. Koçtürk and Z. Çinar, J. Photochem. Photobiol., A, 2002, 146, 189.
15. V. Brezová, A. Blažková, E. Borošová, M. Čeppan and R. J. Fiala, J. Mol. Catal. A: Chem., 1995, 98, 109.
16. W. Liu, S. Chen, W. Zhao and S. J. Zhang, J. Hazard. Mater., 2009, 164, 154.
17. E. C. Butler and A. P. Davis, J. Photochem. Photobiol., A, 1993, 70, 273.
18. K. Okamoto, Y. Yamamoto, H. Tanaka, M. Tanaka and A. Itaya, Bull. Chem. Soc. Jpn., 1985, 58, 2015.
19. Y. Du, S. Goldstein and J. Rabani, J. Photochem. Photobiol., A, 2011, 225, 1.
20. T. A. Kandiel, A. Feldhoff, L. Robben, R. Dillert and D. W. Bahnemann, Chem. Mater., 2010, 22, 2050.
21. H. Noma, T. Koga, C. Hirakawa, K. Nonaka, K. Shobu, T. Kaibuki and S. Moriyama, ECS Trans., 2014, 58, 77.
22. G. Zhang and Y. Xu, Chem. Commun., 2005, 8, 520.
23. D. F. Ollis, J. Phys. Chem. B, 2005, 109, 2439.
24. A. V. Emeline, V. K. Ryabchuk and N. Serpone, J. Phys. Chem. B, 2005, 109, 18515.
25. A. Mills, C. O'Rourke and K. Moore, J. Photochem. Photobiol., A, 2015, 310, 66.
26. S. Yamazaki, S. Iwai, J. Yano and H. Taniguchi, J. Phys. Chem. A, 2001, 105, 11285.
27. C. Minero, G. Mariella, V. Maurino and E. Pelizzetti, Langmuir, 2000, 16, 2632.
28. Q. Sun and Y. Xu, J. Phys. Chem. C, 2010, 114, 18911.
29. S. Cong and Y. Xu, J. Phys. Chem. C, 2011, 115, 21161.
30. Z. Li, R. Liu and Y. Xu, J. Phys. Chem. C, 2013, 117, 24360.
31. Z. Li, S. Cong and Y. Xu, ACS Catal., 2014, 4, 3273.
32. C. R. Chenthamarakshan and K. Rajeshwar, Langmuir, 2000, 16, 2715.
33. M. C. Biesinger, L. W. Lau, M. A. R. Gerson and R. S. C. Smart, Appl. Surf. Sci., 2010, 257, 887.
34. R. Cai, Y. Kubota and A. Fujishima, J. Catal., 2003, 219, 214.
35. Z. Li, J. Sheng, Y. Zhang, X. Li and Y. Xu, Appl. Catal., B, 2015, 166, 313.
36. H. Chen and Y. Xu, J. Phys. Chem. C, 2012, 116, 24582.
37. X. Xiong, H. Chen and Y. Xu, J. Phys. Chem. C, 2015, 119, 5946.

---

Footnote

† Electronic supplementary information (ESI) available: XRD patterns, N₂ adsorption isotherm, UV-vis absorption spectra, Tauc plots, Raman spectrum and SEM images, kinetics for phenol degradation, OCP under air, photographs and XRD spectra for the irradiated samples. See DOI: 10.1039/c6ra04471a

---