Kleber T. de Oliveira*ab,
L. Zane Millera and
D. Tyler McQuade*a
aFlorida State University, Department of Chemistry and Biochemistry, Tallahassee, FL 32306-4390, USA. E-mail: tylermcquade@gmail.com
bUniversidade Federal de São Carlos, Departamento de Química, 13565-905, São Carlos, SP, Brazil. E-mail: kleber.oliveira@ufscar.br; Web: http://www.lqbo.ufscar.br
First published on 26th January 2016
Photooxygenation reactions are a powerful synthetic tool to produce oxidized organic compounds; however, these reactions often exhibit experimental limitations including the production of complex mixtures that hinder desired product isolation and scale-up. Herein, we present a photocatalysed protocol under continuous flow conditions using a simple home built photoreactor and porphyrinoids as photocatalysts. Reaction conditions, long-term experiments, and scope demonstrate a protocol that is cost-effective, safe, reproducible and robust, thus allowing the production of relevant substituted naphthoquinones with interest in natural product synthesis and biological activity.
The generation of singlet oxygen using photosensitizers is well-known.7 While many photosensitizers have been described,8 porphyrin derivatives are one of the most efficient classes of compounds for this purpose as supported by recent and relevant applications in organic synthesis.5,9
Since their discovery, photooxygenation processes have presented restrictions for their use in large scale reactors due to limitations imposed by the potential to form explosive intermediates, thus requiring the use of high dilutions and small scales. However, this limitation has recently been addressed by utilization of flow-based approaches to carry out photochemical transformations.10 Current advances in chemical synthesis under continuous flow conditions have in fact changed the way in which reactions are developed and scaled up in the pharmaceutical industry and research laboratories.11 Numerous advantages have been provided such as controlled mixing, fast heat transfer, control of residence time and process automation.12 Particularly, photochemical and photocatalysed transformations accelerated under continuous flow are one of the most prominent processes since significant improvements have been accomplished at the micro- and meso-scale, and successful reaction classes improved using these techniques due to the high efficiency of light irradiation and enhanced safety.10
Regarding the photooxygenations of naphthols, previous work performed in batch conditions has been published,13 as well as preliminary versions of photochemical flow devices. The pioneering work of Oelgemöller14 and co-workers showed the possibility to use both plug-flow and parallel falling films for photooxygenations under different continuous flow conditions. However, both methodologies and devices of these previous work displayed limitations and required many reaction cycles, high concentrations of photosensitizers (up to 5 mol%) with low conversions in some cases,14c inspiring us to explore improved and cost-competitive conditions for this continuous flow photooxygenation.
Herein, we have built and applied a simple, yet effective in-house engineered photoreactor to perform a comprehensive study of the photooxygenation of naphthols (Scheme 1). Different porphyrinoid and phthalocyanine derivatives were evaluated as photosensitizers (0.1–0.5 mol%), achieving high efficiency. Scope, robustness, and most importantly, scalability with a 24 h extended experiment are presented under continuous flow conditions, which allowed the production of relevant substituted naphthoquinones of interest in natural product synthesis and with well-recognized biological activities.15
We began the optimization by pumping of 100 mL of 3 mM solution of 1 with 0.1 mol% of the photocatalyst tetraphenylporphyrin (3) under three different plug-flow conditions (Table 1) in order to select the best flow rate and light source for the transformations (Scheme 1 and Fig. 1).
| Entry | Visible light source | Solution flow rate (mL min−1) | Juglone (2) yield (%) | Residence time (min) | STY (2)c (g per day) | ||
|---|---|---|---|---|---|---|---|
| 1st | 2nd | Avg | |||||
a Reactions performed by using 100 mL of a solution of the substrate 1 at 3.0 mmol L−1, photocatalyst 3 at 0.1 mol% in CH3CN : CH2Cl2 (95 : 5), 1 cm plug-flow (regular oxygen and solution) in a 25 mL PFA tubing photoreactor (0.125 in (OD) × 0.065 in (ID)).b Isolated yield by column chromatography.c Considering the average yield (Avg). |
|||||||
| 1 | FLC, 45 W | 0.50 | 52 | 55 | 53 | 25.0 | 0.20 |
| 2 | FLC, 45 W | 0.75 | 58 | 60 | 59 | 16.7 | 0.33 |
| 3 | FLC, 45 W | 1.00 | 48 | 49 | 48 | 12.5 | 0.36 |
| 4 | LED, 24 W | 0.50 | 75 | 72 | 74 | 25.0 | 0.28 |
| 5 | LED, 24 W | 0.75 | 58 | 56 | 57 | 16.7 | 0.32 |
| 6 | LED, 24 W | 1.00 | 43 | 43 | 43 | 12.5 | 0.32 |
Using the FLC lamp (45 W) the best result was obtained with a flow rate of 0.75 mL min−1 (entry 2, Table 1) considering the yield, residence time, and space-time yield (STY). Similarly, the use of white LED lamp (24 W) and a flow rate of 0.75 mL min−1 (entry 5, Table 1) produced the best result in terms of yield, residence time and STY, and requiring approximately half of the energy compared to the FLC source (24 W vs. 45 W). The superiority of the white LED lamp can be explained by the comparison of the emission and absorption spectra (Fig. 2) for each light source and the porphyrin 3. Specifically, the broad emission band of white LED source (480–700 nm) encompasses the entire visible region of the absorption spectrum of 3 (500–670 nm).
Thus, from the first six experiments it was possible to suggest the flow rate of 0.75 mL min−1 as the most adequate due to the productivity, and also because the most significant increase of STY was found in the range of 0.5 to 0.75 mL min−1. It is important to highlight that the reproducibility was also evaluated by performing each reaction twice (Table 1) with only minor variations of 1–3% yield, as observed in Table 1.
After establishing the best light source, different photocatalyst concentrations were examined from 0.1 to 0.5 mol% at three different flow rates (Table 2). From the results we conclude that 0.3 and 0.5 mol% are the best photosensitizer concentrations, but it was not clear if the use of 0.5 mol% would furnish the best cost-benefit.
| Entry | Solution flow rate (mL min−1) | Juglone (2) yield (%) using TPP (3) at 0.1 mol% | Juglone (2) yield (%) using TPP (3) at 0.2 mol% | Juglone (2) yield (%) using TPP (3) at 0.3 mol% | Juglone (2) yield (%) using TPP (3) at 0.5 mol% |
|---|---|---|---|---|---|
a Reactions performed by using 100 mL of a solution of the substrate 1 at 3.0 mmL L−1, photocatalyst 3 in different concentrations in CH3CN : CH2Cl2 (95 : 5), 1 cm plug-flow (regular oxygen and solution) in a 25 mL PFA tubing photoreactor (0.125 in (OD) x 0.065 in (ID)). |
|||||
| 1 | 0.50 | 73 | 81 | 85 | 85 |
| 2 | 0.75 | 57 | 71 | 80 | 83 |
| 3 | 1.00 | 43 | 60 | 75 | 81 |
Subsequently, different concentrations of substrate 1 (Table 3) were assessed from 3.0 to 12.0 mM while maintaining the flow rate at 0.75 mL min−1 and employing two different photocatalyst concentrations (0.3 and 0.5 mol%). Comparing the experiments from entries 1–5 and 6–9, very similar yields and STY were observed between all comparable entries, however, the use of lower amounts of photocatalyst (0.3 mol%, entries 1–5, Table 3) were decisive for the choice of 0.3 mol% as the most ideal catalyst loading.
| Entry | Substrate 1 mmol L−1 | TPP (3) as photocatalyst (mol%) | Juglone (2) yield (%) | STY (2) (g per day) | Productivity mmol product per mmol catalyst per h |
|---|---|---|---|---|---|
a Reactions performed by using 100 mL of a solution of the substrate 1, photocatalyst 3 at 0.3 mol% or 0.5 mol% in CH3CN : CH2Cl2 (95 : 5), solution flow rate (0.75 mL min−1), 1 cm plug-flow (regular oxygen and solution) in a 25 mL PFA tubing photoreactor (0.125 in (OD) × 0.065 in (ID)).b In these conditions it was observed a small amount of TPP (3) as a precipitate after the experiment with no serious blockage of the pump system. |
|||||
| 1 | 3.0 | 0.3 | 80 | 0.45 | 120 |
| 2 | 6.0 | 0.3 | 85 | 0.96 | 128 |
| 3 | 9.0 | 0.3 | 85 | 1.44 | 128 |
| 4 | 10.0 | 0.3 | 82 | 1.54 | 123 |
| 5b | 12.0 | 0.3 | 78 | 1.76 | 117 |
| 6 | 3.0 | 0.5 | 83 | 0.47 | 75 |
| 7 | 6.0 | 0.5 | 83 | 0.93 | 75 |
| 8b | 9.0 | 0.5 | 83 | 1.41 | 75 |
| 9b | 12.0 | 0.5 | 78 | 1.76 | 70 |
Another parameter which highlights this preference is the productivity of the photocatalyst which is almost two times greater for 0.3 mol% photocatalyst loading. The selected conditions to proceed with the methodology study are shown in entry 4, Table 3, since greater concentrations of substrate (up to 10 mM) required amounts of photocatalyst 3 (mol%) which caused precipitation to occur (entries 5, 8 and 9, Table 3). It is important to highlight that this optimized amount of TPP (3) (0.3 mol%, US dollars 33.1 mmol−1)16 is advantageous compared to similar photooxygenation procedures found in the literature14 using bengal rose (≥5 mol%, US dollars 40.4 mmol−1).16
After screening to find the best flow rate and substrate/photocatalyst concentrations we decided to evaluate different photocatalysts including additional porphyrinoids 4–6 and the phthalocyanine 7 (Table 4). Different photoactivities were observed most likely due to the different ability of these photosensitizers (3–7) to produce singlet oxygen since nearly the entire absorption spectra of 3–7 (Q-bands, 500–700 nm) were covered by the emission spectra of the LED lamp (430–700 nm). Clearly, the simplest and most cost-competitive photocatalyst 3 proved to be the most efficient giving us all of the optimized parameters required to advance the scope of this photooxygenation protocol. Before continuing with the scope of this protocol, we decided to showcase the robustness of this method and performed two 24 h experiments using the optimized conditions (substrate 1 at 10 mM, TPP 3 at 0.3 mol%, flow rate at 0.75 mL min−1 in CH3CN
:
CH2Cl2 95
:
5, and 1 cm plug-flow of oxygen and solution). Juglone (2) was obtained in both experiments in 72% (1.35 g) and 74% (1.40 g) yield, respectively, proving this protocol as reproducible and in agreement with STY (1.54 g) for these conditions (entry 4, Table 3).
| Entry | Photocatalysts at 0.3 mol% | Juglone (2) yield (%) | STY juglone (2) (g per day) | Productivity mmol product per mmol catalyst per h |
|---|---|---|---|---|
a Reactions performed by using 100 mL of a solution of the substrate 1 at 10.0 mmol L−1, different photocatalysts at 0.3 mol% in CH3CN : CH2Cl2 (95 : 5), solution flow rate (0.75 mL min−1), 1 cm plug-flow (regular oxygen and solution) in a 25 mL PFA tubing photoreactor (0.125 in (OD) × 0.065 in (ID)). |
||||
| 1 | 3 | 82 | 1.54 | 123 |
| 2 | 4 | 15 | 0.28 | 22 |
| 3 | 5 | 72 | 1.35 | 108 |
| 4 | 6 | 6 | 0.11 | 9 |
| 5 | 7 | 18 | 0.34 | 27 |
In order to evaluate the scope, we selected different naphthol derivatives 8–17 containing various substituent groups in different positions (Table 5). The first general result is that naphthol derivatives containing electron-donating groups are invariably more reactive than the ones with electron-withdrawing groups, which is consistent with the accepted [4 + 2] pericyclic mechanism for this photooxygenation.1,13
| Entry | Substrate | Product and yield (%) under continuous flow conditionsb | Product and yield (%) in batch conditionsb,c |
|---|---|---|---|
a All the reactions under continuous flow conditions were performed using the optimized conditions: LED lamp 24 W, 10 mM solutions of the substrates in CH3CN : CH2Cl2 (95 : 5) (100 mL), TPP (3) at 0.3 mol%, solution flow rate = 0.75 mL min−1, 1 cm plug-flow (oxygen-solution), residence time = 16.7 min, 25 mL photoreactor.b Isolated yields by using column chromatography.c Reactions performed using the residence time of the comparable continuous flow conditions (16.7 min).d An attempt to run this reaction under continuous flow conditions was performed, but due to the low solubility of the starting material it was not possible to finish it because the blockage of the system.e An attempt to perform this reaction in batch conditions was performed but no products were observed after 1 h.f Starting material totally recovered. |
|||
| 1 | ![]() |
![]() |
![]() |
| 2 | ![]() |
![]() |
![]() |
| 3 | ![]() |
![]() |
![]() |
| 4 | ![]() |
![]() |
![]() |
| 5 | ![]() |
No reactiond | No reactione |
| 6 | ![]() |
![]() |
![]() |
| 7 | ![]() |
No reactionf | No reactionf |
| 8 | ![]() |
![]() |
![]() |
| 9 | ![]() |
No reactionf | No reactionf |
| 10 | ![]() |
![]() |
![]() |
| 11 | ![]() |
![]() |
![]() |
Comparing entries 1 and 3 (Table 5) the starting materials 1 and 8 yielded the expected corresponding naphthoquinones 2 and 19, respectively, with the same yield (82%) under continuous flow conditions, compared to 59% yield obtained for 18 (entry 2, Table 5). Also, the yields for batch conditions (using the same residence time, 16.7 min) were substantially lower than under continuous flow, providing a real advantage with the use of this photoreactor.
The activated 6-hydroxynaphthol (10) (entry 4, Table 5) yielded 20 in 46% yield under continuous flow conditions together with minor by-products which were difficult to separate. In an identical reaction performed using batch conditions only traces of 20 could be identified.
Compound 11 was submitted to photooxygenation under both flow and batch conditions, but unsuccessfully, most likely due to the very low solubility of this compound (entry 5, Table 5). On the other hand, the corresponding acetylated compound 12 yielded the corresponding naphthoquinone 21 in both flow (72% yield) and batch (8% yield) conditions (entry 6, Table 5). Similarly, the deactivated naphthol 13 (entry 7, Table 5) was unreactive in both flow and batch reactions, but the equivalent reduced compound 14 furnished the corresponding product 22 under both flow (75% yield) and batch (7% yield) conditions (entry 8, Table 5). The deactivated compound 15 also did not react (entry 9, Table 5), however, both compounds 16 and 17 yielded the corresponding naphthoquinone 18 (8 and 26% yield, respectively) under continuous flow conditions together with many by-products (entries 10 and 11, Table 5). It is important to highlight in entry 11 (Table 4) that we were able to isolate the brominated derivative 23 in 22% yield under continuous flow conditions, and we conclude that this was most likely possible via reaction of 18 and HOBr produced during the process.
1H and 13C NMR spectra were recorded with a Bruker Avance 600 spectrometer at 600 and 150 MHz, respectively. CDCl3 or DMSO-d6 was used as solvent and TMS (tetramethylsilane) as the internal reference. The chemical shifts are expressed in δ (ppm) and coupling constants (J) are given in Hertz (Hz). The UV-vis spectra were recorded with a Perkin-Elmer Lambda 25 spectrophotometer using 1 cm optical length quartz cuvettes at 25 °C and dichloromethane as solvent. Emission spectra of FLC and LED lamps were recorded an Ocean Optics Spectrometer HR2000CG-UV-NIR. MS analysis were performed using HP model 5973 mass selective detector with HP model 6890 + gas chromatograph; scanned from 50 to 550 amu; T = 70 °C for 3 min and then to 300 °C at 30 °C min−1.
All continuous flow experiments were carried out using a micro HPLC pump from ThalesNano and an in-house engineered photoreactor as specified in the ESI.† Analytical thin-layer chromatography was performed on glass plates (3 × 6 cm, 1 mm thick), Merck TLC silica-gel 60 F254.
:
hexanes 6
:
4 as eluent, Rf = 0.38, 59% yield (94.0 mg, 0.594 mmol). 1H NMR (600 MHz, CDCl3) δ (ppm): 6.99 (s, 2H); 7.75–7.79 (m, 2H); 8.08–8.12 (m, 2H). 13C NMR (150 MHz, CDCl3) δ (ppm): 126.4; 131.9; 133.9; 138.7; 185.0.
:
MeOH 9.5
:
0.5, Rf = 0.50, 46% yield (79.3 mg, 0.455 mmol). 1H NMR (600 MHz, DMSO-d6) δ (ppm): 6.97 (d, J = 10.2 Hz, 1H); 7.00 (d, J = 10.2 Hz, 1H); 7.18 (dd, J = 8.7 Hz, J = 2.5 Hz, 1H); 7.28 (d, J = 2.5 Hz, 1H); 7.86 (d, J = 8.7 Hz; 1H); 11.0 (br s, 1H). 13C NMR (150 MHz, DMOS-d6) δ (ppm): 111.6; 120.8; 123.7; 128.9; 133.7; 138.2; 139.0; 162.8; 183.6; 185.0.
:
ethyl acetate 6
:
4, Rf = 0.60, 72% yield (154.9 mg, 0.720 mmol). 1H NMR (600 MHz, CDCl3) δ (ppm): 2.31 (s, 3H); 6.93 (d, J = 10.2 Hz); 6.96 (d, J = 10.2 Hz, 1H); 7.74 (dd, J = 8.5 Hz, J = 1.1 Hz); 7.83 (dd, J = 7.5 Hz, J = 1.1 Hz, 1H); 9.09 (dd, J = 8.5 Hz, J = 1.1 Hz); 11.87 (br s, 1H). 13C NMR (150 MHz, CDCl3) δ (ppm): 25.6; 116.0; 121.9; 126.0; 132.2; 135.7; 138.0; 139.9; 141.3; 169.9; 184.4; 189.1.
:
ethyl acetate 9.5
:
0.5, Rf = 0.30, 75% yield (152.1 mg, 0.752 mmol). 1H NMR (600 MHz, CDCl3) δ (ppm): 1.52 (d, J = 6.4 Hz, 3H); 2.49 (d, J = 4.9 Hz, 1H); 5.02 (ddq, J = 6.4 Hz, J = 4.9 Hz, J = 1.5 Hz, 1H); 7.01 (d, J = 1.5 Hz, 1H); 7.74–7.78 (m, 2H); 8.06–8.12 (m, 2H). 13C NMR (150 MHz, CDCl3) δ (ppm): 22.6; 65.3; 126.2; 126.5; 131.9; 132.2; 132.9; 133.8; 134.0; 152.7; 185.3; 185.6.Footnote |
| † Electronic supplementary information (ESI) available: Copy of 1H and 13C NMR spectra. See DOI: 10.1039/c6ra00285d |
| This journal is © The Royal Society of Chemistry 2016 |