Introduction

Solvent-free reaction conditions (SFRC) (Tanaka and Toda 2000; Metzger 1998) are alternative reaction conditions to the transformations in solution (Reichardt 2007) and attracted considerable attention due to important advantages. Shorter reaction times, lower consumption of organic solvents, simple isolation/purification procedures and less reaction waste are some of notable advantages. Nonconventional reaction media, such as water, supercritical CO2 or ionic liquids, are also a promising replacement for organic solvents (Adams et al. 2004). Another alternative is to perform reactions under highly concentrated reaction conditions (Walsh et al. 2007) or in the presence of a small amount of solvent vapor (Nakamatsu et al. 2005). A great number of different reactions under (SFRC) have been reviewed in both older and more recent literature (Cave et al. 2001; Martins et al. 2009; Singh and Chowdhury 2012; Gawande et al. 2014; Obst and König 2018; Zagande and Patil 2019).

N-bromosuccinimide (NBS) is one of the most common and versatile reagents in organic chemistry (Djerassi 1948; Paquette et al. 2009). It is easy to handle, inexpensive, readily available, non-toxic, and usually more selective than bromine. Traditional methods of bromination include molecular bromine or bromides under harsh reaction conditions, even though unconventional methods, for instance, in water (Shaw et al. 1997; Tanaka et al. 1999) or under SFRC (Toda and Schmeyers 2003; Pravst et al. 2009; Wang and Gao 2012), have been developed. In a more alternative approach, transformations with NBS are performed under microwave and ultrasound irradiation (Heropoulos et al. 2007), on water (Podgoršek et al. 2006), in ionic liquids (Togo and Hirai 2003; Ganguly et al. 2005), scCO2 (Tanko and Blackert 1994) and bio-waste extract (Appa et al. 2019). Under SFRC, aromatic (Prakash et al. 2004; Imanzadeh et al. 2006), Wohl–Ziegler (Winkler et al. 2014) and α-bromination of carbonyl compounds (Kumar et al. 2012; Pravst et al. 2008) have been reported, but these methods still require the use of acid catalysts or radical initiators. While more activated substrates react in the absence of catalysts or initiators (Rahman et al. 2005; Pravst et al. 2006a, b, 8; Bose and Mal 2014), other methods make the use of alternative modes of activation, such as microwave irradiation (Goswami et al. 2004) or concentrated solar radiation (Dinda et al. 2013; Deshpande et al. 2015). Photo-induced transformations with NBS have been extensively studied in solvents (Arbuj et al. 2007; Šterk et al. 2013; Cantillo et al. 2014); however, to the best of our knowledge there are rare reports on these reactions under SFRC (Jereb et al. 2009).

Molecular motion in solvent-free reactions is restricted, and the aggregate state of the reaction mixture plays an important role in SFRC transformations. Whereas molecules are more mobile in liquid/solid systems, between two solid reagents, the formation of a melt phase is crucial for increased molecular mobility (Rothenberg et al. 2001). More so than the aggregate state, the structure of the substrates largely affects the reactivity and selectivity of transformations. We have previously shown that a 40-W tungsten lamp promotes benzylic-, ring- or α-carbonyl bromination by NBS under SFRC (Jereb et al. 2009). One might wonder how selective bromination under visible-light conditions would be when a competitive aromatic-, radical- and α-carbonyl substitution was possible in a single substrate. Studies addressing this issue have been reported (Pravst et al. 2006a, b, 47; Podgoršek et al. 2009; Vražič et al. 2013), but in those cases regioselectivity of functionalization was modulated through careful tuning of the reaction conditions. To our knowledge, regioselectivity of visible-light-induced transformations of substrates containing several competitive reactive sites with NBS under SFRC remains unexplored.

In the present study, we investigated photochemical transformations of diverse aryl-substituted ketones with NBS under solvent-free reaction conditions: with an 8-W energy-saving lamp as an activation source, in the absence of any solvents, radical initiators and catalysts and in the presence of air. For testing the substrate scope, we selected a broad range of liquid and solid ketones that contained different functional groups, which could have an influence on the type of transformation.

Experimental

General information

Reagents and solvents were obtained from commercial sources. NBS was recrystallized from a mixture of CH2Cl2/petroleum ether, and solvents were used as received. All ketones were prepared by the method of the classical Friedel–Crafts acylation (Vogel et al. 1989). 14 was prepared according to the published procedure (Coan and Becker 1955).

Reactions were performed in closed flasks without inert atmosphere. The experimental procedure was the same for liquid and solid ketones: NBS was added to the substrate and the reaction mixture was stirred with a small stirring bar under solvent-free reaction conditions (SFRC). Reaction mixtures with solid ketones turned into a molten state and the reaction proceeded as efficiently as the reactions with liquid substrates. An 8-W energy-saving lamp was used throughout all the experiments, and the average reaction temperature was 27 °C. The experiments in the dark were carried out at 27 °C. The progress of the reactions was monitored by thin-layer chromatography (TLC). Light source: Philips Genie 8 W energy-saving light bulb. Column chromatography (CC): silica gel 60 (63–200 mm, 70–230 mesh ASTM; Fluka); CH2Cl2/petroleum ether (1:3). Preparative thin-layer chromatography (TLC): Merck-60-F254 plates; mixtures of light petroleum ether (b.p. 40–60 °C) and CH2Cl2. NMR spectra: Bruker DPX 300 (300 MHz) and Bruker AVANCE III (500 MHz) instruments; in CDCl3; δ in ppm relative to TMS as internal standard for 1H NMR and to the central line of CDCl3 (δ = 77.00 ppm) for 13C NMR, J in Hz. IR spectra: Bruker Alpha FT-IR spectrometer; in cm−1. HRMS: Agilent 6224 Accurate Mass TOF LC/MS system spectrometer; electrospray ionization; in m/z. Elemental analysis: Perkin Elmer CHNS/O Analyzer 2400 Series II elemental analyzer. Melting points: Büchi 535 melting point apparatus; uncorrected.

General procedure for photo-induced reactions with NBS under SFRC

To a ketone (0.3 mmol), NBS (0.33 or 0.66 mmol, 20 or 39 mg) was added and the resulted reaction mixture was stirred under illumination with an 8-W energy-saving light bulb for 19–29 h. At the end of the reaction, conversions and the product distribution were determined by 1H NMR spectroscopy of the crude reaction mixture. Pure products were obtained after separation by either CC or preparative TLC and were analyzed by 1H NMR, 13C NMR, IR, HRMS and melting point when solid. Spectroscopic data of the known compounds are available in Supporting Information and are in accordance with the literature data.

Characterization of the new products

2-Bromo-1-(4-(isopropyl)phenyl)ethanone 2c.

1c (0.049 g, 0.3 mmol), NBS (0.059 g, 0.33 mmol). Column chromatography on silica gel (Rf = 0.55, CH2Cl2/petroleum ether 1:3 → 2:1). Colorless oil (0.028 g, 39%). IR (neat): ν 2961, 1676, 1604, 1569, 1461, 1416, 1310, 1278, 1198, 1186, 1056, 1009, 990, 846, 822, 677, 661 \({\text{cm}}^{-1}\). 1H NMR (500 MHz, CDCl3): δ 1.28 (d, J = 6.9 Hz, 6H), 2.98 (sept, J = 6.9 Hz, 1H), 4.44 (s, 2H), 7.35 (d, J = 8.3 Hz, 2H), 7.93 (d, J = 8.3 Hz, 2H). 13C NMR (125 MHz, CDCl3): δ 23.6, 30.9, 34.3, 127.0, 129.2, 131.8, 155.7, 190.9. HRMS (ESI-TOF, m/z): found 241.0223 ([M + H]+, C11H14BrO+; calc. 241.0223).

2-Bromo-1-(3-methyl-4-methoxyphenyl)ethanone 2e.

1e (0.049 g, 0.3 mmol), NBS (0.059 g, 0.33 mmol). Column chromatography on silica gel (Rf = 0.47, CH2Cl2/petroleum ether 1:3 → 2:1). Colorless solid (0.013 g, 18%). M.p. 45–48 °C. IR (neat): ν 2941, 1683, 1596, 1497, 1438, 1410, 1250, 1212, 1126, 1020, 878, 818, 752, 682, 611 \({\text{cm}}^{-1}\). 1H NMR (500 MHz, CDCl3): δ 2.26 (s, 3H), 3.91 (s, 3H), 4.40 (s, 2H), 6.87 (d, J = 8.6 Hz, 1H), 7.80 (d, J = 1.7 Hz, 1H), 7.86 (dd, J = 1.7, 8.6 Hz, 1H). 13C NMR (125 MHz, CDCl3): δ 16.3, 30.8, 55.6, 109.4, 126.4, 127.3, 129.3, 131.5, 162.4, 190.2. HRMS (ESI-TOF, m/z): found 243.0013 ([M + H]+, C10H12BrO2+; calc. 243.0015).

2-Bromo-1-(3,5-dimethyl-4-methoxyphenyl)ethanone 2f.

1f (0.053 g, 0.3 mmol), NBS (0.059 g, 0.33 mmol). Column chromatography on silica gel (Rf = 0.38, CH2Cl2/petroleum ether 1:3 → 2:1). Pale yellowish solid (0.015 g, 19%). M.p. 43–45 °C. IR (neat): ν 2948, 1682, 1595, 1483, 1387, 1313, 1256, 1229, 1140, 1055, 1007, 887, 856, 769, 708, 641 \({\text{cm}}^{-1}\). 1H NMR (500 MHz, CDCl3): δ 2.34 (s, 6H), 3.77 (s, 3H), 4.41 (s, 2H), 7.67 (s, 2H). 13C NMR (125 MHz, CDCl3): δ 16.3, 30.9, 59.7, 129.5, 130.0, 131.6, 162.0, 190.7. HRMS (ESI-TOF, m/z): found 257.0172 ([M + H]+, C11H14BrO2+; calc. 257.0172).

2-Bromo-1-(4-(1-bromoethyl)phenyl)ethanone 3b.

1b (0.044 g, 0.3 mmol), NBS (0.059 g, 0.33 mmol). Preparative TLC (Rf = 0.45, CH2Cl2/petroleum ether 1:3). Brownish solid (0.002 g, 4%). M.p. 48–53 °C. IR (neat): ν 2948, 1694, 1603, 1571, 1412, 1388, 1280, 1197, 1162, 1069, 1042, 989, 822, 681 \({\text{cm}}^{-1}\). 1H NMR (300 MHz, CDCl3): δ 2.05 (d, J = 6.9 Hz, 3H), 4.42 (s, 2H), 5.20 (q, J = 6.9 Hz, 1H), 7.52–7.59 (m, 2H), 7.94–8.00 (m, 2H). 13C NMR (75 MHz, CDCl3): δ 26.4, 30.6, 47.5, 127.4, 129.4, 133.7, 149.1, 190.6. HRMS (ESI-TOF, m/z): found 304.9168 ([M + H]+, C10H11Br2O+; calc. 304.9171).

2,2-Dibromo-1-(4-(isopropyl)phenyl)ethanone 4c.

1c (0.049 g, 0.3 mmol), NBS (0.059 g, 0.33 mmol). Column chromatography on silica gel (Rf = 0.68, CH2Cl2/petroleum ether 1:3 → 2:1). Yellowish solid (0.011 g, 21%). M.p. 40–45 °C. IR (neat): ν 2961, 1684, 1599, 1567, 1460, 1414, 1311, 1270, 1223, 1198, 1055, 983, 850, 840, 770, 742, 709, 682 \({\text{cm}}^{-1}\). 1H NMR (300 MHz, CDCl3): δ 1.29 (d, J = 6.9 Hz, 6H), 2.99 (sept, J = 6.9 Hz, 1H), 6.69 (s, 1H), 7.36 (d, J = 8.4 Hz, 2H), 8.02 (d, J = 8.4 Hz, 2H). 13C NMR (75 MHz, CDCl3): δ 23.5, 34.4, 39.8, 127.1, 128.5, 130.0, 156.3, 185.6. HRMS (ESI-TOF, m/z): found 318.9330 ([M + H]+, C11H14Br2O+; calc. 318.9328).

2,2-Dibromo-1-(4-(cyclohexyl)phenyl)ethanone 4d.

1d (0.061 g, 0.3 mmol), NBS (0.059 g, 0.33 mmol). Column chromatography on silica gel (Rf = 0.75, CH2Cl2/petroleum ether 1:3 → 2:1). Colorless solid (0.003 g, 5%). M.p. 100–104 °C. IR (neat): ν 3027, 2920, 2848, 1683, 1600, 1566, 1445, 1416, 1315, 1273, 1212, 1191, 1145, 988, 841, 774, 707, 630 \({\text{cm}}^{-1}\). 1H NMR (500 MHz, CDCl3): δ 1.21–1.32 (m, 1H), 1.35–1.49 (m, 4H), 1.74–1.81 (m, 1H), 1.82–1.93 (m, 4H), 2.54–2.63 (m, 1H), 6.70 (s, 1H), 7.34 (d, J = 8.3 Hz, 2H), 8.01 (d, J = 8.3 Hz, 2H). 13C NMR (125 MHz, CDCl3): δ 25.9, 26.6, 33.9, 39.9, 44.8, 127.5, 128.4, 129.9, 155.5, 185.6. HRMS (ESI-TOF, m/z): found 358.9634 ([M + H]+, C14H17Br2O+; calc. 358.9641).

2,2-Dibromo-1-(3-methyl-4-methoxyphenyl)ethanone 4e.

1e (0.049 g, 0.3 mmol), NBS (0.059 g, 0.33 mmol). Column chromatography on silica gel (Rf = 0.68, CH2Cl2/petroleum ether 1:3 → 2:1). Colorless solid (0.015 g, 29%). M.p. 113–115 °C. IR (neat): ν 3033, 2923, 1665, 1591, 1503, 1455, 1438, 1416, 1259, 1217, 1127, 1022, 987, 908, 814, 761, 717, 680, 615 \({\text{cm}}^{-1}\). 1H NMR (500 MHz, CDCl3): δ 2.26 (s, 3H), 3.92 (s, 3H), 6.70 (s, 1H), 6.88 (d, J = 8.7 Hz, 1H), 7.88 (d, J = 2.1 Hz, 1H), 7.97 (dd, J = 2.1, 8.7 Hz, 1H). 13C NMR (125 MHz, CDCl3): δ 16.3, 40.0, 55.7, 109.5, 122.8, 127.6, 130.1, 132.2, 162.9, 184.9. HRMS (ESI-TOF, m/z): found 320.9117 ([M + H]+, C10H11Br2O2+; calc. 320.9120).

2,2-Dibromo-1-(3,5-dimethyl-4-methoxyphenyl)ethanone 4f.

1f (0.053 g, 0.3 mmol), NBS (0.059 g, 0.33 mmol). Column chromatography on silica gel (Rf = 0.48, CH2Cl2/petroleum ether 1:3 → 2:1). Colorless solid (0.014 g, 25%). M.p. 85–87 °C. IR (neat): ν 2940, 1688, 1593, 1479, 1310, 1221, 1142, 994, 903, 776, 749, 715, 652 \({\text{cm}}^{-1}\). 1H NMR (500 MHz, CDCl3): δ 2.34 (s, 6H), 3.79 (s, 3H), 6.71 (s, 1H), 7.75 (s, 2H). 13C NMR (125 MHz, CDCl3): δ 16.3, 39.8, 59.8, 126.2, 130.7, 131.8, 162.5, 185.4. HRMS (ESI-TOF, m/z): found 334.9276 ([M + H]+, C11H13Br2O2+; calc. 334.9277).

2-Bromo-1-(4-(methylthio)phenyl)propan-1-one 6c.

5c (0.054 g, 0.3 mmol), NBS (0.059 g, 0.33 mmol). Column chromatography on silica gel (Rf = 0.60, CH2Cl2/hexane 1:3 → 2:1). Pale brownish solid (0.053 g, 68%). M.p. 88–90 °C. IR (neat): ν 2988, 1666, 1583, 1405, 1346, 1247, 1192, 1164, 1093, 1058, 992, 944, 830, 768, 745, 689 \({\text{cm}}^{-1}\). 1H NMR (500 MHz, CDCl3): δ 1.89 (d, J = 6.6 Hz, 3H), 2.53 (s, 3H), 5.25 (q, J = 6.6 Hz, 1H), 7.28 (d, J = 8.5 Hz, 2H), 7.93 (d, J = 8.5 Hz, 2H). 13C NMR (125 MHz, CDCl3): δ 14.7, 20.1, 41.4, 125.0, 129.3, 130.1, 146.9, 192.3. HRMS (ESI-TOF, m/z): found 258.9785 ([M + H]+, C10H12BrOS+; calc. 258.9787).

2-Bromo-1-(5-methyl-2-methoxyphenyl)propan-1-one 6d.

5d (0.053 g, 0.3 mmol), NBS (0.059 g, 0.33 mmol). Column chromatography on silica gel (Rf = 0.59, CH2Cl2/hexane 1:2 → 2:1). Yellowish oil (0.056 g, 73%). IR (neat): ν 2925, 1671, 1608, 1495, 1463, 1441, 1404, 1332, 1283, 1248, 1178, 1160, 1136, 1020, 968, 809, 741, 607 \({\text{cm}}^{-1}\). 1H NMR (300 MHz, CDCl3): δ 1.85 (d, J = 6.7 Hz, 3H), 2.31 (s, 3H), 3.89 (s, 3H), 5.53 (q, J = 6.7 Hz, 1H), 6.86 (d, J = 8.5 Hz, 1H), 7.23–7.31 (m, 1H), 7.49–7.54 (m, 1H). 13C NMR (75 MHz, CDCl3): δ 20.2, 20.3, 48.0, 55.8, 111.5, 125.4, 130.4, 131.7, 134.5, 156.1, 196.6. HRMS (ESI-TOF, m/z): found 257.0164 ([M + H]+, C11H14BrO2+; calc. 257.0172).

2-Bromo-1-(3-bromo-4,5-dimethoxyphenyl)propan-1-one 6e.

5e (0.058 g, 0.3 mmol), NBS (0.059 g, 0.33 mmol). Column chromatography on silica gel (Rf = 0.42, CH2Cl2/petroleum ether 1:3 → 2:1). Yellowish oil (0.024 g, 42%). IR (neat): ν 2934, 1695, 1592, 1505, 1439, 1372, 1336, 1257, 1204, 1158, 1045, 1024, 972, 897, 855, 790, 760, 630 \({\text{cm}}^{-1}\). 1H NMR (500 MHz, CDCl3): δ 1.91 (d, J = 6.7 Hz, 3H), 3.90 (s, 3H), 3.92 (s, 3H), 5.40 (q, J = 6.7 Hz, 1H), 7.03 (s, 1H), 7.06 (s, 1H). 13C NMR (125 MHz, CDCl3): δ 19.9, 45.6, 56.2, 56.3, 110.6, 113.2, 115.9, 130.8, 148.4, 151.7, 196.6. HRMS (ESI-TOF, m/z): found 350.9225 ([M + H]+, C11H13Br2O3+; calc. 350.9226).

2-Bromo-1-(4-(bromomethyl)phenyl)propan-1-one 7a.

5a (0.044 g, 0.3 mmol), NBS (0.059 g, 0.33 mmol). Preparative TLC (Rf = 0.51, CH2Cl2/petroleum ether 1:3). Colorless solid (0.010 g, 20%). M.p. 80–83 °C. IR (neat): ν 2928, 1675, 1604, 1438, 1417, 1345, 1235, 1205, 1155, 1093, 993, 948, 857, 823, 767, 723, 702, 635 \({\text{cm}}^{-1}\). 1H NMR (500 MHz, CDCl3): δ 1.90 (d, J = 6.6 Hz, 3H), 4.50 (s, 2H), 5.26 (q, J = 6.6 Hz, 1H), 7.51 (d, J = 8.2 Hz, 2H), 8.00 (d, J = 8.2 Hz, 2H). 13C NMR (125 MHz, CDCl3): δ 20.0, 31.9, 41.4, 129.4, 129.4, 133.8, 143.4, 192.6. HRMS (ESI-TOF, m/z): found 304.9171 ([M + H]+, C10H11Br2O+; calc. 304.9171).

2-Bromo-2-methyl-1-(4-methylphenyl)propan-1-one 9a.

8a (0.049 g, 0.3 mmol), NBS (0.059 g, 0.33 mmol). Preparative TLC (Rf = 0.64, CH2Cl2/petroleum ether 1:3). Pale yellowish oil (0.017 g, 23%). IR (neat): ν 2927, 1672, 1606, 1459, 1370, 1267, 1163, 1104, 978, 886, 827, 748, 677 \({\text{cm}}^{-1}\). 1H NMR (500 MHz, CDCl3): δ 2.03 (s, 6H), 2.41 (s, 3H), 7.24 (d, J = 8.2 Hz, 2H), 8.07 (d, J = 8.2 Hz, 2H). 13C NMR (125 MHz, CDCl3): δ 21.6, 31.7, 60.5, 128.8, 130.4, 131.9, 143.2, 196.3. HRMS (ESI-TOF, m/z): found 241.0221 ([M + H]+, C11H14BrO+; calc. 241.0223).

2-Bromo-1-(2,4-dimethylphenyl)-2-methylpropan-1-one 9b.

8b (0.053 g, 0.3 mmol), NBS (0.059 g, 0.33 mmol). Column chromatography on silica gel (Rf = 0.63, CH2Cl2/petroleum ether 1:3 → 2:1). Pale yellowish oil (0.041 g, 53%). IR (neat): ν 2972, 2924, 1688, 1612, 1458, 1385, 1369, 1257, 1126, 1102, 973, 875, 817 \({\text{cm}}^{-1}\). 1H NMR (500 MHz, CDCl3): δ 1.97 (s, 6H), 2.26 (s, 3H), 2.34 (s, 3H), 6.98–7.03 (m, 1H), 7.06 (s, 1H), 7.60 (d, J = 7.9 Hz, 1H). 13C NMR (125 MHz, CDCl3): δ 20.1, 21.3, 30.7, 62.6, 125.5, 126.9, 131.8, 134.9, 135.9, 139.8, 202.2. HRMS (ESI-TOF, m/z): found 255.0379 ([M + H]+, C12H16BrO+; calc. 255.0379).

2-Bromo-1-(3-methyl-4-methoxyphenyl)-2-methylpropan-1-one 9d.

8d (0.058 g, 0.3 mmol), NBS (0.059 g, 0.33 mmol). Preparative TLC (Rf = 0.40, CH2Cl2/petroleum ether 1:5). Yellowish oil (0.013 g, 16%). IR (neat): ν 2928, 1664, 1598, 1501, 1458, 1386, 1317, 1298, 1256, 1127, 1104, 1021, 986, 916, 817, 760, 681, 617 \({\text{cm}}^{-1}\). 1H NMR (500 MHz, CDCl3): δ 2.04 (s, 6H), 2.25 (s, 3H), 3.90 (s, 3H), 6.83 (d, J = 8.7 Hz, 1H), 7.99 (d, J = 2.0 Hz, 1H), 8.15 (dd, J = 2.0, 8.7 Hz, 1H). 13C NMR (125 MHz, CDCl3): δ 16.3, 32.0, 55.5, 60.6, 108.7, 126.4, 126.4, 130.6, 133.1, 161.3, 195.2. HRMS (ESI-TOF, m/z): found 271.0329 ([M + H]+, C12H16BrO2+; calc. 271.0328).

2-Bromo-1-(4-(bromomethyl)phenyl)-2-methylpropan-1-one 10a.

8a (0.049 g, 0.3 mmol), NBS (0.059 g, 0.33 mmol). Preparative TLC (Rf = 0.48, CH2Cl2/petroleum ether 1:3). Yellowish solid (0.012 g, 23%). M.p. 36–38 °C. IR (neat): ν 2979, 1668, 1603, 1451, 1411, 1388, 1264, 1230, 1203, 1163, 1102, 979, 887, 849, 820, 764, 713, 679, 606 \({\text{cm}}^{-1}\). 1H NMR (500 MHz, CDCl3): δ 2.03 (s, 6H), 4.50 (s, 2H), 7.46 (d, J = 8.4 Hz, 2H), 8.13 (d, J = 8.4 Hz, 2H). 13C NMR (125 MHz, CDCl3): δ 31.5, 32.1, 60.2, 128.8, 130.7, 134.6, 142.0, 196.1. HRMS (ESI-TOF, m/z): found 318.9329 ([M + H]+, C11H13Br2O+; calc. 318.9328).

2-Bromo-1-(3-(bromomethyl)-4-methoxyphenyl)-2-methylpropan-1-one 10d.

8d (0.058 g, 0.3 mmol), NBS (0.059 g, 0.33 mmol). Preparative TLC (Rf = 0.21, CH2Cl2/petroleum ether 1:5). Colorless solid (0.011 g, 18%). M.p. 79–81 °C. IR (neat): ν 2929, 1655, 1595, 1503, 1458, 1444, 1265, 1223, 1142, 1105, 1034, 1018, 1001, 937, 916, 842, 819, 763, 617 \({\text{cm}}^{-1}\). 1H NMR (500 MHz, CDCl3): δ 2.04 (s, 6H), 3.97 (s, 3H), 4.57 (s, 2H), 6.91 (d, J = 8.8 Hz, 1H), 8.22 (d, J = 2.3 Hz, 1H), 8.26 (dd, J = 2.3, 8.8 Hz, 1H). 13C NMR (125 MHz, CDCl3): δ 28.1, 31.8, 55.9, 60.4, 110.0, 125.9, 126.8, 133.4, 133.6, 160.7, 194.5. HRMS (ESI-TOF, m/z): found 348.9432 ([M + H]+, C12H15Br2O2+; calc. 348.9433).

2-Bromo-1-(4-methoxyphenyl)-2-phenylethan-1-one 12b.

11b (0.068 g, 0.3 mmol), NBS (0.059 g, 0.33 mmol). Column chromatography on silica gel (Rf = 0.47, CH2Cl2/petroleum ether 1:3 → 2:1). Pale yellowish solid (0.055 g, 60%). M.p. 67–70 °C. IR (neat): ν 2928, 1671, 1601, 1571, 1508, 1452, 1422, 1357, 1312, 1264, 1217, 1173, 1027, 846, 824, 774, 734, 688 \({\text{cm}}^{-1}\). 1H NMR (500 MHz, CDCl3): δ 3.85 (s, 3H), 6.36 (s, 1H), 6.89–6.93 (m, 2H), 7.29–7.38 (m, 3H), 7.51–7.55 (m, 2H), 7.95–8.00 (m, 2H). 13C NMR (125 MHz, CDCl3): δ 51.1, 55.5, 114.0, 126.9, 128.9, 129.0, 129.0, 131.5, 136.3, 163.9, 189.6. HRMS (ESI-TOF, m/z): found 305.0165 ([M + H]+, C15H14BrO2+; calc. 305.0172).

2-Bromo-1-(5-methyl-2-methoxyphenyl)-2-phenylethan-1-one 12c.

11c (0.120 g, 0.5 mmol), NBS (0.107 g, 0.6 mmol). Column chromatography on silica gel (Rf = 0.45, CH2Cl2/hexane 1:2 → 2:1). Yellowish solid (0.100 g, 63%). M.p. 44–46 °C. IR (neat): ν 2943, 1676, 1607, 1579, 1495, 1453, 1401, 1285, 1246, 1157, 1126, 1020, 810, 695, 626 \({\text{cm}}^{-1}\). 1H NMR (300 MHz, CDCl3): δ 2.27 (s, 3H), 3.85 (s, 3H), 6.59 (s, 1H), 6.81 (d, J = 8.5 Hz, 1H), 7.21–7.36 (m, 4H), 7.44–7.53 (m, 3H). 13C NMR (75 MHz, CDCl3): δ 20.2, 55.1, 55.7, 111.5, 125.5, 128.6, 128.6, 129.4, 130.5, 131.9, 134.8, 136.5, 155.9, 193.9. HRMS (ESI-TOF, m/z): found 319.0326 ([M + H]+, C16H16BrO2+; calc. 319.0328).

2-Bromo-2-phenyl-1-(thiophen-2-yl)ethan-1-one 12d.

11d (0.303 g, 1.5 mmol), NBS (0.320 g, 1.8 mmol). Column chromatography on silica gel (Rf = 0.53, CH2Cl2/hexane 1:2 → 2:1). Yellow solid (0.304 g, 72%). M.p. 36–39 °C. IR (neat): ν 3089, 1660, 1514, 1494, 1454, 1408, 1353, 1285, 1237, 1222, 1178, 1062, 948, 859, 719, 691, 634 \({\text{cm}}^{-1}\). 1H NMR (500 MHz, CDCl3): δ 6.20 (s, 1H), 7.11 (dd, J = 3.9, 4.8 Hz, 1H), 7.30–7.40 (m, 3H), 7.53–7.59 (m, 2H), 7.68 (dd, J = 0.9, 4.8 Hz, 1H), 7.76 (dd, J = 0.9, 3.9 Hz, 1H). 13C NMR (125 MHz, CDCl3): δ 51.4, 128.3, 129.0, 129.0, 129.2, 133.4, 135.2, 135.8, 140.6, 184.1. HRMS (ESI-TOF, m/z): found 280.9628 ([M + H]+, C12H10BrOS+; calc. 280.9630).

2,2-Dibromo-2-phenyl-1-(thiophen-2-yl)ethan-1-one 13d.

11d (0.303 g, 1.5 mmol), NBS (0.320 g, 1.8 mmol). Column chromatography on silica gel (Rf = 0.67, CH2Cl2/hexane 1:2 → 2:1). Brownish solid (0.052 g, 16%). M.p. 65–69 °C. IR (neat): ν 3109, 2921, 1667, 1491, 1444, 1406, 1350, 1238, 1222, 1190, 1084, 1064, 961, 845, 801, 749, 722, 704, 689, 648 \({\text{cm}}^{-1}\). 1H NMR (500 MHz, CDCl3): δ 6.92 (dd, J = 4.0, 4.9 Hz, 1H), 7.30 (dd, J = 1.1, 4.0 Hz, 1H), 7.33–7.42 (m, 3H), 7.56 (dd, J = 1.1, 4.9 Hz, 1H), 7.66–7.71 (m, 2H). 13C NMR (125 MHz, CDCl3): δ 68.4, 127.3, 127.8, 128.8, 129.8, 135.1, 136.1, 136.3, 140.7, 180.1. HRMS (ESI-TOF, m/z): found 358.8727 ([M + H]+, C12H9Br2OS+; calc. 358.8735).

2-Bromo-2,4-diphenyl-3-oxobutanenitrile 15.

14 (0.071 g, 0.3 mmol), NBS (0.059 g, 0.33 mmol). Column chromatography on silica gel (Rf = 0.70, CH2Cl2/petroleum ether 1:3 → 2:1). Yellowish oil (0.070 g, 75%). IR (neat): ν 2923, 1731, 1495, 1449, 1320, 1031, 1002, 705, 691, 646 \({\text{cm}}^{-1}\). 1H NMR (500 MHz, CDCl3): δ 3.99 (d, J = 16.7 Hz, 1H), 4.03 (d, J = 16.7 Hz, 1H), 7.01–7.06 (m, 2H), 7.22–7.29 (m, 3H), 7.43–7.48 (m, 3H), 7.62–7.68 (m, 2H). 13C NMR (125 MHz, CDCl3): δ 43.1, 54.1, 116.0, 127.5, 127.6, 128.6, 129.4, 129.5, 130.6, 131.3, 132.1, 190.9. HRMS (ESI-TOF, m/z): found 314.0173 ([M + H]+, C16H13BrNO+; calc. 314.0175).

2-Bromo-2-phenylacetonitrile 18.

16 (0.048 g, 0.3 mmol), NBS (0.059 g, 0.33 mmol). Preparative TLC (Rf = 0.35, CH2Cl2/petroleum ether 1:3) Yellowish oil (0.031 g, 53%). IR (neat): 2971, 1496, 1455, 1188, 1138, 971, 817, 763, 690, 653, 621 \({\text{cm}}^{-1}\). 1H NMR (300 MHz, CDCl3): δ 5.49 (s, 1H), 7.39–7.48 (m, 3H), 7.52–7.60 (m, 2H). 13C NMR (75 MHz, CDCl3): δ 27.4, 116.2, 127.8, 129.5, 130.3, 133.5.

1-(3-bromo-4-methoxynaphthalen-1-yl)ethanone 21a.

19a (0.200 g, 1.0 mmol), NBS (0.356 g, 2.0 mmol). Column chromatography on silica gel (Rf = 0.24, CH2Cl2/hexane 1:2 → 2:1). Brownish solid (0.008 g, 3%). M.p. 36–39 °C. IR (neat): ν 2940, 2848, 1663, 1568, 1501, 1401, 1352, 1275, 1229, 1134, 1073, 1008, 974, 835, 759, 712, 675, 627 \({\text{cm}}^{-1}\). 1H NMR (300 MHz, CDCl3): δ 2.72 (s, 3H), 4.05 (s, 3H), 7.55–7.68 (m, 2H), 8.07 (s, 1H), 8.16–8.21 (m, 1H), 8.74–8.80 (m, 1H). 13C NMR (125 MHz, CDCl3): δ 29.7, 61.7, 110.8, 122.3, 126.6, 127.3, 128.6, 129.6, 131.1, 132.8, 133.6, 156.9, 199.6. HRMS (ESI-TOF, m/z): found 279.0013 ([M + H]+, C13H12BrO2+; calc. 279.0015).

(3-Bromo-4-methoxynaphthalen-1-yl)(phenyl)methanone 21c.

19c (0.079 g, 0.3 mmol), NBS (0.117 g, 0.66 mmol). Column chromatography on silica gel (Rf = 0.67, CH2Cl2/petroleum ether 1:3 → 2:1). Yellowish solid (0.101 g, 99%). M.p. 81–84 °C. IR (neat): ν 2922, 1649, 1567, 1501, 1448, 1372, 1270, 1243, 1152, 1070, 1054, 1022, 963, 863, 832, 763, 710, 693, 670, 656, 613 \({\text{cm}}^{-1}\). 1H NMR (500 MHz, CDCl3): δ 4.08 (s, 3H), 7.45–7.51 (m, 2H), 7.51–7.56 (m, 1H), 7.57–7.65 (m, 2H), 7.71 (s, 1H), 7.84–7.89 (m, 2H), 8.08 (d, J = 8.5 Hz, 1H), 8.22 (d, J = 8.4 Hz, 1H). 13C NMR (125 MHz, CDCl3): δ 61.7, 111.1, 122.4, 126.2, 127.4, 127.8, 128.6, 129.4, 130.4, 131.8, 132.1, 133.5, 133.9, 137.9, 155.6, 196.1. HRMS (ESI-TOF, m/z): found 341.0170 ([M + H]+, C18H14BrO2+; calc. 341.0172).

2-Bromo-1-(3-bromo-4-methoxynaphthalen-1-yl)ethanone 22a.

19a (0.200 g, 1.0 mmol), NBS (0.356 g, 2.0 mmol). Column chromatography on silica gel (Rf = 0.33, CH2Cl2/hexane 1:2 → 2:1). Brownish solid (0.007 g, 2%). M.p. 64–69 °C. IR (neat): ν 2934, 1686, 1566, 1503, 1373, 1265, 1213, 1165, 1124, 1089, 1074, 973, 908, 860, 836, 789, 761, 718, 701, 680, 639 \({\text{cm}}^{-1}\). 1H NMR (500 MHz, CDCl3): δ 4.07 (s, 3H), 4.53 (s, 2H), 7.60–7.70 (m, 2H), 8.07 (s, 1H), 8.19–8.23 (m, 1H), 8.62–8.67 (m, 1H). 13C NMR (125 MHz, CDCl3): δ 33.3, 61.8, 110.6, 122.5, 126.3, 127.7, 129.0, 129.7, 129.8, 131.6, 133.4, 157.7, 192.4. HRMS (ESI-TOF, m/z): found 356.9117 ([M + H]+, C13H11Br2O2+; calc. 356.9120).

Results and discussion

The ketones under investigation were arranged into the groups of distinct alkyl substituents. First, we investigated the regioselectivity of functionalization at differently phenyl-functionalized model acetophenones 1af. Neat ketones were reacted with NBS under stirring and irradiated using an 8-W energy-saving light bulb at 27 °C. The results are summarized in Table 1.

Table 1 Transformation of neat substituted acetophenones 1a–f with NBS under visible-light irradiation

All substrates 1a–f responded to the above reaction conditions in the same way, mainly affording α-bromoacetophenones 2a–f, which were accompanied by α,α-dibrominated products 4a–f. The selectivity was poorer in the cases of 4’-methylacetophenone 1a and 4’-ethylacetophenone 1b, where α-bromination was accompanied with benzylic bromination of the corresponding alkyl (Me and Et) group, resulting in a small amount of 3a and 3b, respectively (entries 1 and 2).

Surprisingly, selective benzylic bromination of 4’-methylacetophenone 1a into 4’-bromomethylacetophenone took place with NBS in solvent-free reaction under concentrated solar radiation (Deshpande et al. 2015). A comparison of the photochemical functionalization of 4’-methylacetophenone 1a with NBS under different reaction conditions revealed interesting features. On water, 4’-bromomethylacetophenone was the only product formed (Podgoršek et al. 2009), whereas the treatment of 1a with NBS/SiCl4 in MeCN furnished the compound 3a as the sole product (Salama and Novák 2011). Conversely, compound 1a was selectively α-brominated using NBS/PTSA under ultrasound irradiation in methanol (Adhikari and Samant, 2002). In contrast to the straight-chain functionalized derivatives, at the branched isopropyl 1c and cyclohexyl 1d acetophenones only α-bromination took place with no benzylic bromination observed (entries 4 and 5). Increasing the complexity of the substrate with additional reactive sites, e.g., in 3’-methyl-4’-methoxyacetophenone 1e and 3’,5’-dimethyl-4’-methoxyacetophenone 1f, did not diminish the reaction selectivity, furnishing solely α-brominated ketones 2e/4e and 2f/4f, respectively (entries 6 and 8). Conducting the reactions with 1b and 1e in the presence of a radical scavenger TEMPO (2,2,6,6-tetramethylpiperidin-1-yloxy) strongly influenced the selectivity outcome. As indicated in entries 3 and 7, benzylic bromination and introduction of the second bromine atom to the α-position were completely suppressed; hence, the only products formed were 2b and 2e, respectively. Transformation of 1f did not take place in the absence of light (entry 9).

Next, the selectivity of photo-transformations of model aryl-substituted propiophenones 5 with NBS under SFRC was examined. As with acetophenones 1, α-bromination was the main reaction pathway, but in this case no α,α-dibrominated products could be detected. Transformations of most of the propiophenones exhibited a somewhat different selectivity compared to acetophenones. The results are summarized in Table 2.

Table 2 Regioselectivity of functionalization of neat substituted propiophenones with NBS under visible light

Functionalization of 4’-methylpropiophenone 5a yielded α-brominated ketone 6a as the main product, accompanied with a product of mixed α- and benzylic bromination 7a (entry 1). Benzylic bromination of 5a in the presence of TEMPO was completely suppressed, and only α-bromination could be detected with low conversion into 6a (entry 2). This might indicate that the main reaction channels are of a radical nature. The reaction of 5a did not occur in the dark (entry 3). Transformation of 4’-methoxypropiophenone 5b gave the corresponding α-bromoketone 6b as the main product. To some extent, this was further brominated at the aromatic ring also yielding a minor product 7b (entry 4). The sulfur analogue, 4’-methylthiopropiophenone 5c, exhibited the highest selectivity with only the α-carbon atom being brominated, producing 6c (entry 5). Unexpectedly, the methylthio group did not affect the type of functionalization as no aromatic substitution took place. The sulfur atom, being itself sensitive to oxidation (Ali et al. 2006), remained unchanged. The reaction of 2’-methoxy-5’-methylpropiophenone 5d, which contained at least three potential reactive sites, gave the related α-bromoketone 6d as the main product as well as an additional side-chain brominated minor product 7d (entry 6). Functionalization of 3’,4’-dimethoxypropiophenone 5e yielded α- and ring brominated 6e as the main product, accompanied by 10% of α-bromoketone 7e (entry 7). Interestingly, no reaction took place in the presence of TEMPO to 5e (entry 8). It appears that aromatic substitution of 5e takes place via a radical pathway. Reactions of α-bromination of acetophenones and propiophenones in the presence of TEMPO suggest involvement of electrophilic and radical reaction pathways without clear boundary.

Subsequently, we selected model isobutyrophenones (aryl isopropyl ketones) (Table 3).

Table 3 Regioselectivity of functionalization of neat substituted isobutyrophenones with NBS under visible light

In the group of acetophenones 1, the isopropyl and cyclohexyl moieties in 1c and 1d did not undergo benzylic bromination (Table 1, entries 4 and 5). In this step, the reactivity of model isopropyl ketones was examined. The general feature of ketones 8 was that, regardless of the nature of the substituents under the studied reaction conditions, α-bromoketones 9 were formed as the main products. Functionalization of 4’-methylbutyrophenone 8a in addition to the α-brominated ketone 9a also leads to the formation of benzyl bromide 10a (entry 1). The reaction of 8a was completely suppressed in the presence of TEMPO (entry 2) as well as in the dark (entry 3). Reaction of 2’,4’-dimethylisobutyrophenone 8b, although being similar to 8a, did not result in benzylic bromination, but the only product formed was α-bromoketone 9b (entry 4). Transformation of 4’-methoxyisobutyrophenone 8c exclusively produced α-bromoketone 9c with no ring substitution noted (entry 5). The reaction of 3’-methyl-4’-methoxybutyrophenone 8d furnished a mixture of α-bromoketone 9d and benzyl bromide 10d (entry 6). The addition of TEMPO to the reaction mixture resulted in the recovery of the starting material 8d (entry 7). In contrast to acetophenones and propiophenones, the results of functionalization of isobutyrophenone suggest that α-bromination proceeds exclusively by a radical mechanism. This suggests that functionalization under solvent-free conditions might depend on substrate structure, reminiscent of radical α-bromination of highly substituted ketones with bromine in the presence of 1,2-epoxycyclohexane under photochemical conditions (Caló et al. 1977). Photochemical α-bromination of ketones with NBS under UV–Vis irradiation in diethyl ether is believed to follow a radical pathway (Arbuj et al. 2007).

In the next step, we explored the reactivity of the selected model benzyl ketones 11 (Table 4).

Table 4 Regioselectivity of functionalization of neat substituted benzyl ketones with NBS under visible light

The highest regioselectivity was observed for this group, with all substrates being brominated at the carbonyl carbon atom. This outcome was anticipated, because the site of functionalization was an α-carbonyl and a benzylic position at the same time. The presence of either methyl group 11a, methoxy group 11b or both 11c in the substrate did not influence the selectivity (entries 1, 3 and 4). Also, for the substrate with a sensitive thiophene ring like in 11d, only α-bromination into 12d was noticed, even though some α,α-dibrominated product 13d was formed as well (entry 6). TEMPO completely stopped any transformation for 11a and 11d (entries 2 and 7), and no reaction took place for 11c in the dark (entry 5). Functionalization of benzyl ketones in the presence of TEMPO indicates that α-bromination might proceed via radical pathway.

To examine the effect of the structure at the carbonyl group on the reactivity of transformations under the applied reaction conditions, two substrates, 14 and 16, with an additional electron-withdrawing nitrile group bonded to one of the α-carbon atoms, were selected (Table 5).

Table 5 Regioselectivity of functionalization of neat substituted α-cyanoketones with NBS under visible light

Clearly, this triple activation in 14 and 16 had a strong influence on regioselectivity and the reaction afforded exclusively products 15 and 17, respectively (entries 1 and 3). Both reactions of 14 and 16 occurred even in the dark (entries 2 and 4). When purifying the product 17 by means of column chromatography, we observed an interesting phenomenon, in which the acetyl group was partly cleaved on the column and a mixture of 17 and 18 was eluted. Attempts to separate these two compounds by a preparative thin-layer chromatography resulted in complete deacetylation of 17 into product 18.

Finally, the reactivity of naphthyl ketones was examined (Table 6). The most interesting and unexpected result was observed by irradiating (4-methoxynaphthyl) methyl ketone 19a with 2.2 equiv. of NBS under SFRC.

Table 6 Regioselectivity of functionalization of neat substituted naphthyl ketones with NBS under visible light

Instead of the anticipated bromination at the alkyl side chain, the main reaction pathway was ipso-substitution of the acetyl group by a bromine atom alongside with aromatic and α-carbonyl bromination as minor pathways (entry 1). This observation stimulated us to examine the influence of the acyl moiety structure of (4-methoxynaphthyl) alkyl ketones on potential bromodeacylation. The fact that a similar distribution of the products was found for (4-methoxynaphthyl) ethyl ketone 19b (entry 4) indicates that the ipso-substitution under these conditions could be more general. Although the illumination of (4-methoxynaphthyl) benzyl ketone resulted in a complex mixture of products, a significant amount of 20 was found in the crude reaction mixture, indicating that bromodeacylation occurs with NBS under SFRC with this substrate as well. To test the hypothesis that an enolizable carbonyl group must be present for the ipso-substitution, the non-enolizable (4-methoxynaphthalen-1-yl)(phenyl)methanone 19c was subjected to the same reaction conditions. The reaction exclusively furnished the ring-brominated product 21c with no deacylation observed (entry 5). Interestingly, ipso-substitution was also noticed in the case of the considerably electron-rich aryl ketone 2’-methyl-4’,5’-dimethoxyacetophenone 23, in which the main bromodeacylated product 24 was accompanied by a small amount of α-bromoketone 25 (entry 8). Bromodeacylation is rather a rare process, known to take place with molecular bromine in a solution (Beirne et al. 1970). The addition of TEMPO to the reaction mixture resulted in different effects for different substrates. While inhibiting α-carbonyl bromination, it had no effect on ipso- and ring substitution at 19a and 19c (entries 2 and 6). In contrast, TEMPO completely suppressed all transformations of the compound 23 (entry 9). Selectivity of the transformation of 19a in the dark was decreased considerably resulting in a complex mixture of products (entry 3), while no reaction took place with 19c in the absence of light (entry 7).

Some further experiments were carried out in order to shed light on mechanism of bromodeacylation. 4-Methoxy-1-acetonaphthone 19a was reacted as a model substrate under different reaction conditions with NBS and bromine. Transformation of 19a with NBS/NH4OAc in diethyl ether (Tanemura et al. 2004) led to 2-bromo-1-(4-methoxynaphthalen-1-yl)ethan-1-one with 35% conversion. Reaction of 19a with NBS/NH4NO3 in acetonitrile (Tanemura et al. 2003) furnished a complex mixture of ring-brominated products and a small proportion of bromodeacylated product 20. Functionalization of 19a with bromine in acetic acid furnished 2-bromo-1-(4-methoxynaphthalen-1-yl)ethan-1-one with 65% conversion. This reaction is assumed to take place in an electrophilic fashion via enol form of 19a. The system NBS/pTSA in DCM (Pravst et al. 2008; Vyas et al. 2016) expectedly furnished 2-bromo-1-(4-methoxynaphthalen-1-yl)ethan-1-one with 80% conversion. It can be concluded that electrophilic functionalization of α-carbonyl position takes place in the present of an acid (Br2/AcOH and NBS/pTSA in DCM). The system NBS/NH4NO3 in acetonitrile led to the expected electrophilic ring functionalization; in addition, a minor portion of ipso-product 20 was observed. Reactions in the presence of TEMPO suggest that the ring functionalization of 19a might proceed via electrophilic and radical reaction pathways, while formation of 20 is supposed to be attributed to a different reaction channel—potentially a radical ion reaction pathway (Fasani et al. 1994).

Conclusions

In this study, we disclosed regioselectivity of visible-light-induced transformations of model aryl methyl-, aryl ethyl-, aryl isopropyl-, aryl benzyl-, α-cyano-substituted- and naphthyl alkyl ketones, containing several competitive reactive sites with NBS in the absence of solvents. The entire study was carried out under heterogeneous reaction conditions. In most cases, α-brominated ketones were the main products, accompanied with α,α-dibrominated ketones. Wohl–Ziegler bromination and aromatic ring bromination were noted as minor reaction pathways in the case of several alkyl- and alkoxy-substituted ketones. The α-bromination of ketones under solvent-free reaction conditions appears to depend on the structure of the substrates. While in the case of acetophenones and propiophenones it is proposed to proceed by the electrophilic and radical pathways, in the case of isobutyrophenones and aryl benzyl ketones it is likely to take place by the radical pathway. α-Cyano-substituted ketones were converted into their α-bromo-α-cyano-substituted derivatives even in the dark. An unexpected exception was the reaction of enolizable (4-methoxynaphthyl) alkyl ketones, in which the main product was formed because of an ipso-substitution of the acyl group with a bromine atom. Bromodeacylation is suggested to proceed via a radical cation intermediates. The radical scavenger TEMPO slowed or suppressed α- and ring bromination, whereas benzylic bromination and α,α-dibromination with NBS were entirely suppressed. It was demonstrated in several cases that reactions might proceed in electrophilic and radical fashion without a distinct boundary. In summary, we contributed to understanding reaction pathways of these sensitive compounds in reactions with NBS under solvent-free reaction conditions.