Introduction

Morphine and ketamine are controlled drugs and routinely dispensed at pharmacies and hospitals. However, their uses and particularly that of methamphetamine are not limited to medical needs but often abused [1–3]. In recent years, the occurrence and distribution of pharmaceutical and personal care products (PPCPs) including controlled drugs have gained much attention due to their potential risk to aquatic lives particularly through bioaccumulation and biomagnification [4, 5]. Controlled substances enter the natural waterways with human metabolic wastes. From 2008, many developed countries began examining controlled drugs in surface waters and wastewater treatment plants. Van Nuijs et al. [6] investigated the occurrence of cocaine (COC) and its metabolite benzoylecgonine (BE) in municipal wastewaters and rivers of Belgium. They found as high as 115 ng/L of COC and 520 ng/L of BE in rivers, and 680 ng/L of COC and 550 ng/L of BE in the wastewater of treatment plants. In 2008, the daily loading rate of cocaine was 40–128 g/d to Brussel–Noord wastewater treatment plants, serving 850,000 inhabitants at 2–3 m3/s [7]. In South Wales, average daily loads of amphetamine and cocaine were 2.5 and 0.9 g/day/1,000 people, respectively, primarily of abuse uses [8]. Lin et al. [9] sampled waters of three rivers and effluents from 5 regional hospitals and rwo wastewater treatment plants in Taiwan; they found concentrations of morphine, codeine, methamphetamine, and ketamine in rivers to be as high as 108, 57, 405, and 341 ng/L, in order, and 1240, 378, 260, and 206 ng/L in hospital effluents. In addition, methamphetamine and ketamine were found in wastewater treatment plants at 296 and 147 ng/L, respectively, in the influents and at 61 and 183 ng/L, respectively, in the effluents. Boleda et al. [10] reported normorphine, morphine, codeine, EDDP, and methadone up to 107, 81, 397, 1150, and 732 ng/L, in order, in the effluents of 15 wastewater treatment plants in Catalonia, Spain. Traditional domestic wastewater treatment plants are designed to remove suspended solids and biodegradable organics and not PPCPs. Based on influent and effluent concentrations, removals were estimated at 70–100 % for cocaine and amphetamine by conventional activated sludge process [11, 12]. Postigo et al. [13] reported >90 % removal of cocaine and amphetamine but no removal of 3,4-methylenedioxymethamphetamine (“ecstasy”) and methamphetamine through wastewater treatment plants. The removal of illicit drugs via biodegradation or biosorption has not been adequately addressed [11, 13, 14]. Lin et al. [9] suggested from their study that domestic wastewater treatment plants were incapable of removing controlled drugs adequately to prevent them from entering the aqueous environment.

Photocatalytic degradation of organic pollutants can be an effective alternative to biological methods for removal of organic contaminants. At present, UV disinfection is widely deployed in wastewater treatment plants. Among many semiconductor materials, titanium dioxide (TiO2) has shown promise as a photocatalyst because of its high chemical and photocatalytic reactivity and stability [15, 16]. At contact with an illuminated photocatalyst, organic chemicals are oxidized and eventually mineralized to carbon dioxide. As widely accepted for the degradation of organics by UV/TiO2 [17], valence holes are created in TiO2 after absorption of sufficiently energetic photons (λ < 360 nm) that result in the oxidation of surface bound OH− or H2O to potent hydroxyl radicals at the surface, which in turn oxidize and break down contaminant compounds. The separation of valence holes from conduction band electrons are enhanced by adsorbed oxygen that acts as an electron trap for the conduction band electrons, preventing the recombination of photogenerated electrons and holes thus increasing photocatalytic efficiency. A main focus of this manuscript is to examine the possible incorporation of semiconductor material such as TiO2 for plants with existing UV disinfection process in removal of controlled substances as a means to arrest their release to the environment.

In this study, we examined the photocatalytic degradation of morphine, methamphetamine, and ketamine using illuminated TiO2 and ZnO. The degradation kinetics and optimal operating parameters were determined. The results have provided an assessment in the utility of photocatalysis for abating the release of controlled substances into the aqueous environments.

Materials and methods

Materials

TiO2 Degussa P 25 (Degussa, Germany) with 20 % rutile and 80 % anatase was used; ZnO was of 20 nm nano grade from a local supplier (Yu-Ho, Taiwan). Both photocatalysts were used as received. Ketamine hydrochloride and morphine sulfate were purchased from Sigma-Aldrich (USA), and methamphetamine hydrochloride was obtained from Pharmacopeial Convention (USA) with permission of the Department of Health, Taiwan. The chemical structures and properties of morphine, methamphetamine, and ketamine are summarized in Table 1. Other chemicals were of reagent grade. Deionized Milli-Q water with specific resistance of 18.2 MΩ cm was used in solution preparations. Stock solutions of the studied drugs were prepared at 1,000 mg/L and stored in amber glass bottles at 4 °C. Standard solutions of different concentrations were prepared by appropriate dilution of stock solutions before each experiment.

Table 1 Chemical structures and properties of ketamine, methamphetamine, and morphine
Full size table

Photocatalytic experiments

A hollow, cylindrical photoreactor with an effective volume of 2 L was used. Two kinds of UV light sources were used: a 9 W UV lamp (Philips) that generated 3.61 mW/cm2 of light at 254 nm and a UVLED array of 10 LED bulbs (S-bend, SB1100UV-365) that generated 1.97 mW/cm2 of light at 365 nm. Control experiments were conducted with 100 μg/L solutions of ketamine, methamphetamine, and morphine to evaluate the extent of evaporation (without light and without catalyst), adsorption (0.4 g/L of catalyst without light and without aeration), and photolysis (without catalyst but with light at 254 or 365 nm). Photocatalytic experiments were performed with the test drugs under varied conditions of photocatalysts and UV systems. All experiments were performed at 25 ± 1 °C at 1 atm. The pH of the solution was controlled at 5.5 ± 0.2 during reaction by manually adding HNO3 or NaOH. Studied TiO2 doses were 0.001, 0.005, 0.01, 0.04, 0.05, 0.1, 0.4, 0.7, and 1.0 g/L and ZnO doses were 0.01, 0.04, 0.05, 0.1, 0.4, 0.7, and 1.0 g/L. Experiments were conducted with an initial drug concentration of 100 μg/L and continuous stirring (500 rpm), with samples periodically withdrawn at prescribed time intervals. The samples were filtered through a 0.45 μm filter and the filtrate stored at 4 °C until analysis by LC/MS–MS.

Chemical analysis

The analysis of ketamine, methamphetamine, and morphine was carried out with HPLC–MS/MS as previously described [9]. The HPLC–MS/MS system consisted of a liquid chromatograph (Agilent 1200 module; Agilent, USA) equipped with a ZORBAX Eclipse XDB-C18 column and the tandem mass spectrometer (Sciex API 4000; Applied Biosystem, USA) with an electrospray ionization interface. Ions were acquired in multiple reaction monitoring (MRM) mode with a dwell time of 50 ms. The mass spectrometer conditions were: ion spray voltage of 5.5 kV, curtain gas at 10 L/h, nebulizer gas at 50 L/h, turbo gas at 60 L/h, heated capillary temperature of 500 °C, interface heater switched on, and collision activated dissociation at 5. Analyses were performed in positive mode. The HPLC–ESI–MS/MS conditions by MRM in positive ion modes for the drugs are listed in Table 2. Calibration curves for ketamine and methamphetamine were established at concentrations of 0.1, 0.25, 0.5, 1, 2.5, 5, 10, 25, 50, and 100 μg/L and for morphine at 0.5, 1, 2.5, 5, 10, 25, 50, and 100 μg/L.

Table 2 Conditions of HPLC–ESI–MS/MS by MRM in positive ion modes for drug analyses
Full size table

Results and discussion

The results of control experiments to delineate the effects of evaporation, adsorption, and direct photolysis are shown in Fig. 1. Negligible losses were due to evaporation (Fig. 1a) or adsorption to TiO2 or ZnO solids (Fig. 1b) after 30 min, a typical duration of experiments. Fig. 1c shows that, after 30 min, no degradation of the drugs by UVLED light at 365 nm but significant degradation, i.e. ketamine by 93.7 %, methamphetamine by 45.1 %, and morphine by 95.8 %, by UV light at 254 nm. The ability of UV light at 254 nm to destruct compounds and the inability of the UVLED light at 365 nm to do the same was consistent with the ability to break bonds with the more energy-intense UV light of the shorter wavelength, as observed in Fig. 1c. It should be noted that the destruction of drugs with UV light at 254 nm was much more than twice that of with UVLED light at 365 nm, while the light intensity of the former (3.6 mW/cm2) was about twice that of the latter (2.0 mW/cm2). Fig. 2 shows various extents of degradation of ketamine according to different UV and catalyst conditions, specifically UV-TiO2, UV-ZnO, UVLED-TiO2, and UVLED-ZnO. The destruction of ketamine was most rapid with UV-TiO2, showing complete removal in 20 min with all tested TiO2 concentrations (0.001–0.4 g/L) and within 5 min with > 0.01 g/L of TiO2 (Fig. 2a). Similarly to UV-ZnO, destruction of ketamine was complete in 30 min, albeit with a higher required dose of ZnO (0.01–1 g/L) (Fig. 2b). The destruction of ketamine with UVLED-TiO2 was also significant, capable of achieving complete destruction in 20 min when a higher dose of TiO2 (0.04–1.0 g/L) was used (Fig. 2c). The destruction of ketamine (Fig. 2d) was modest with UVLED-ZnO, achieving 18–99 % removal in 30 min with increased removal at increased ZnO concentration (0.01–0.7 g/L). The results showed greater destruction effectiveness with UV (254 nm) than with UVLED (365 nm), as well as with TiO2 than with ZnO.

Fig. 1
figure 1

Control experiments to identify effects of a evaporation, b adsorption, and c direct photolysis

Full size image
Fig. 2
figure 2

Concentration vs. time profiles of ketamine during photocatalytic degradation under different conditions: a ketamine (UV lamp–TiO2), b ketamine (UV lamp–ZnO), c ketamine (UVLED–TiO2), and ketamine (UVLED–ZnO)

Full size image

The increased contaminant degradation with UV light of shorter wavelength was previously reported by Lo et al. [18] and Lin et al. [19], both showing increased 4-chlorophenol degradation with decreased wavelength of the light source. This study confirmed it. We also found increasing destruction of ketamine with increasing photocatalyst loading, which was explained by increased adsorption of the photonic output of light by increased catalyst loading that turned a larger portion of the photonic outputs into valence holes leading to increased extent of photocatalytic oxidation of the compound. Thus, the increased destruction was attributed to increased utilization of the light’s photonic output.

Fig. 3 shows the degradation results of methamphetamine with varied combinations of UV lights and the photocatalysts. The UV-TiO2 system (Fig. 3a) facilitated the most rapid degradation, with complete removal of methamphetamine in less than 5 min in some cases. The UVLED-ZnO system offered the least degradation of methamphetamine, with the target compound remaining after 30 min (Fig. 3d). By comparison, TiO2 played a more important role in the rapid degradation than ZnO did, even when UVLED of a longer wavelength was used (Fig. 3b, c). Wu [20] found the rate of organic degradation using UV-ZnO at pH 4 markedly slower than at pH 7 or 10. However, the rate obtained using UV-TiO2 contradicted it with the order of pH 4 > pH 7 > pH 10 [21, 22]. Gouvea et al. [23] and Sakthivel et al. [24] found ZnO to be a more powerful photocatalyst than TiO2. Conversely, Kanmoni et al. [25] indicated that TiO2 exhibited a better photocatalytic activity than ZnO. However, ZnO was less effective for oxidative degradation under acidic conditions because it corroded readily under such conditions [26]. As the reaction pH was controlled at 5.5 in the present study, the higher photocatalytic activity observed of TiO2 relative to ZnO was reasonable. Overall, the photocatalytic degradation of methamphetamine (Fig. 3) exhibited the same patterns as of ketamine (Fig. 2) in terms of the catalyst kind and dose.

Fig. 3
figure 3

Concentration vs. time profiles of methamphetamine during photocatalytic degradation under different conditions a methamphetamine (UV lamp–TiO2), b methamphetamine (UV lamp–ZnO), c methamphetamine (UVLED–TiO2), and d methamphetamine (UVLED–ZnO)

Full size image

The photocatalytic degradation results of morphine are presented in Fig. 4. Again, the UV-TiO2 (Fig. 4a) provided clearly superior efficiency; morphine was completely destroyed within 5 min. The UVLED-TiO2 system also completely removed morphine in 5 min under the same conditions (Fig. 4c), asserting TiO2 being a more effective photocatalyst than ZnO as contrasted in Fig. 4b, d. In general, TiO2 provided stronger photocatalytic power than did ZnO even under illumination by UVLED. When TiO2 was used, all three target compounds were removed in 30 min regardless of the light source. However, ZnO had to be coupled with the stronger UV source (254 nm) in order to be effective.

Fig. 4
figure 4

Concentration vs. time profiles of morphine during photocatalytic degradation under different conditions a morphine (UV lamp–TiO2), b morphine (UV lamp–ZnO), c morphine (UVLED–TiO2), and d morphine (UVLED–ZnO)

Full size image

The degradation results under different conditions can be approximated by the pseudo-first order kinetics with fitted pseudo-first order rate constants k (min−1) [21, 22, 27, 28]. The fitted rate constants k under different conditions are summarized in Table 3. The k values first increased with an increasing catalyst dose and then decreased with further increase in dose (Table 3). The optimal dose of TiO2 with UV was 0.04 g/L, while the optimal dose in UVLED-TiO2, UV-ZnO, and UVLED-ZnO systems was 0.4 g/L for all test compounds. The amount of valence holes was increased with increased catalyst loading that captured more photons resulting in enhanced contaminant degradation. However, adding an excess amount beyond the level necessary to capture the light output reduced light penetration and produced a screening effect on the UV light, resulting in reduced degradation [29–32]. This explained the optimal dose observed for catalyst loading.

Table 3 Fitted pseudo-first order rate constants for contaminant degradation under different conditions
Full size table

To compare the degradation rates of different contaminants under different conditions of light and catalyst loading, the fitted pseudo-first order rate constants are plotted in Fig. 5 along with half-lives calculated for each target compound under different conditions. The plots reveal that morphine was most rapidly removed, and methamphetamine and ketamine were removed at roughly equal rates. The order of catalyst activity follows the decreasing order of UV-TiO2 > UVLED-TiO2 > UV-ZnO > UVLED-ZnO, confirming the catalyst kind being the most important factor (i.e. TiO2 being more active than ZnO). The direct photolytic results also showed the highest removal rate for morphine, which suggested that morphine more readily absorbed UV light resulting in more extensive degradation than ketamine and methamphetamine. As a result, morphine showed the highest rate of removal by photocatalysis as well as by photolysis. TiO2 exhibited a stronger capability for degradation of methamphetamine than for ketamine (Fig. 5a, b), while ZnO was a better catalyst for degradation of ketamine than for methamphetamine (Fig. 5c, d). Calculated times to reach 3-log destruction efficiency for each drug are shown in Table 4, which show clearly that the removal of morphine is fastest among these three drugs.

Fig. 5
figure 5

Fitted first order rate constants under different conditions a UV lamp + TiO2, b UVLED + TiO2, c UV lamp + ZnO, and d UVLED + ZnO

Full size image
Table 4 Calculated times required to reach 3-log destruction efficiency by photocatalysis
Full size table

Conclusions

Controlled amounts of ketamine, methamphetamine, and morphine can be removed from wastewater by direct photolysis using UV light of 254 nm even without the presence of TiO2 or ZnO catalyst. However, UVLED light of 365 nm must be coupled with TiO2 or ZnO in order to be effective. The UV-TiO2 (3.6 mW/cm2, 0.04 g/L) system was most effective, achieving 99.9 % removal of all compounds in 1–6 min, with pseudo-first order rate constants of 1.7 min−1 for katamine, 2.7 min−1 for methamphetamine, and 6.2 min−1 for morphine. The results suggest a possible technique of removing controlled substances from municipal wastewater by photocatalysis.