Reversible Triplet Excitation Transfer in a ... - ACS Publications

Article pubs.acs.org/joc

Cite This: J. Org. Chem. 2018, 83, 8995−9007

Reversible Triplet Excitation Transfer in a Trimethylene-Linked Thioxanthone and Benzothiophene-2-Carboxanilide that Photochemically Expels Leaving Group Anions Gilbert N. Ndzeidze, Lingzi Li, and Mark G. Steinmetz* Department of Chemistry, Marquette University, P.O. Box 1881, Milwaukee, Wisconsin 53201-1881, United States

Downloaded via KAOHSIUNG MEDICAL UNIV on August 17, 2018 at 10:09:30 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

S Supporting Information *

ABSTRACT: The triplet excited state of thioxanthone produced by photolysis undergoes reversible triplet energy transfer with a trimethylene-linked benzothiophene-2-carboxanilide ring system. The ensuing electrocyclic ring closure of the anilide moiety produces a putative zwitterionic intermediate that is capable of expelling leaving groups (LG−) from the C-3 position of the benzothiophene ring. Stern− Volmer quenching studies with cyclohexadiene as quencher furnish the rate constants for the triplet excitation transfer in the forward and reverse directions, which can be expressed as an equilibrium constant K = 0.058. Overall, the rate of the triplet excited state reaction becomes K × kr = 5.7 × 104 s−1 for LG− = Cl−, where kr is the triplet decay rate of the C-3 chloro-substituted benzothiophene-2-carboxanilide, found through Stern−Volmer quenching. The high quantum efficiencies found for the trimethylene-linked systems are due to K × kr being competitive with the triplet excited state decay of the thioxanthone of kd = 7.7 × 104 s−1. On the basis of Φisc = 0.68, the overall expected quantum yield for direct photolysis should be 0.50 for LG− = Cl− as compared to 0.41 at 25 °C experimentally. Φ decreases with increasing basicity of the leaving group (LG− = Cl−, (EtO)2PO2−, PhCH2CO2−, PhS−, and PhCH2S−).

■

INTRODUCTION Photoremovable protecting groups are used to facilitate research in biology, biochemistry, and physiology; practical applications have also emerged in materials science and synthesis.1−6 Our research in the area has focused on photochemical rearrangements that produce putative zwitterionic intermediates, which are capable of expelling leaving group anions having a wide range of basicities.7−13 Recent examples of such photochemistry are the anilides of benzothiophene-2-carboxamides, where various leaving groups are incorporated at the C-3 position of the thiophene ring (Scheme 1).14 Chemical yields of the released leaving groups are typically quantitatitve, and quantum yields range from 0.1 to 0.2, depending on the leaving group basicity. An attempt to extend the photochemistry to longer wavelengths by replacing the N-phenyl group of the benzothiophene carboxanilide with a thioxanthone chromophore15 gave mixed results. Quantitative expulsion of chloride and thiolate leaving groups by photolysis at 386 nm was achieved (eq 1), but the quantum yields were rather low (Φ = 0.01−06). Computational study of the triplet excited state potential surface involved in the reaction showed that triplet excitation transfer from thioxanthone to benzothiophene was a prerequisite for the reaction to occur. The observed inefficiency was thus thought to be due to the endothermic nature of the energy transfer, considering that the thioxanthone chromophore has ET = 65 kcal mol−1 16−19 while the ET estimated for benzothiophene is 69 kcal mol−1.20 This © 2018 American Chemical Society

Scheme 1

explanation was recently called into question when we observed that thioxanthone can bimolecularly sensitize leaving group release from the benzothiophene carboxanilide 7 with quantum yields as high as Φ = 0.41, depending on the concentration of the anilide (eq 2). Received: May 8, 2018 Published: June 27, 2018 8995

DOI: 10.1021/acs.joc.8b01173 J. Org. Chem. 2018, 83, 8995−9007

Article

The Journal of Organic Chemistry

The above studies were prompted by the aforementioned thioxanthone sensitized photolyses of anilide 7 in eq 2, at which time it was found that the sensitized quantum yield strongly increased with the increasing concentration of anilide 7, the triplet energy acceptor. Analysis of the sensitized quantum yields as a function of anilide 7 concentration was performed to afford an estimate for the bimolecular rate constant for the endothermic energy transfer step. Given the high efficiencies for LG− expulsion observed for the trimethylene-linked system 9a, synthetic efforts were made to incorporate a variety of LG− at C-3 of the benzothiophene2-carboxanilide ring system for photochemical studies of derivatives 9a,c−e and 10f−h. We note that the C-3 hydroxyl group in 10c was the precursor to the phosphate ester, and acetate derivatives 10f−h. Unfortunately, we were unable to selectively demethylate esters 10f−h to obtain the corresponding acids.

■

The efficient thioxanthone sensitized photolysis of anilide 7 led us to suspect that the inefficient reaction of 5 may be due to a substituent effect. Alternatively, the direct attachment of the thioxanthone chromophore to nitrogen of the carboxamide in 5 might provide an unfavorable geometry for triplet excitation transfer to the benzothiophene ring system via either through-space or through-bond mechanisms.21−25 In this paper, we report the results obtained when the thioxanthone triplet energy donor is covalently attached to the benzothiophene-2-carboxanilide nitrogen via a trimethylene linkage (Scheme 2). This structure allows the electro-

RESULTS Synthesis. Most of the photochemical studies of trimethylene-linked benzothiophenes used compound 9a, which was synthesized according to Scheme 3. Its synthesis involved the coupling of the known acid chloride 1515 with the trimethylene-tethered aniline 14 to give 10a, followed by demethylation of the methyl ester with trimethyltin hydroxide to obtain 9a. The C-3 hydroxyl-substituted compound 9c was obtained via an analogous route involving the coupling of acid chloride 16 to alkylaniline 14, followed by demethylation of the aromatic methoxide 10b using BBr3. Further demethylation of ester 10c using trimethyltin hydroxide then furnished 9c. Various leaving groups could be introduced at the C-3 position using 10a,c as the reactants. The synthesis of benzothiophene 16 is described in Scheme 4. The first step in the synthesis of 16 (Scheme 4) involved oxidation26 of the known15 carboxylic acid 17 to give crude sulfoxide 18, which was esterified to obtain the crystalline tertbutyl ester 19. The oxidation was problematic, in that sulfoxide 18 always contained some unoxidized 17, which could not be removed by crystallization or chromatography until two steps later in the synthesis with 20. This problem has been discussed previously.26 The overall yield of 20 starting from 17 was ca. 20%. The tert-butyl ester of sulfoxide 19 was needed to successfully perform the nucleophilic aromatic substitution of the C-3 chloro group with the methoxy group. The reduction of sulfoxide 20 to give 21 also converted the tert-butyl ester into the corresponding carboxylic acid. While the sulfoxide reduction is normally performed with methyltrichlorosilane,27 trimethylchlorosilane is just as effective. The acid chloride 16 is obtained by reaction of 21 with thionyl chloride. With compounds 10a and 10c, various leaving groups could be introduced at the C-3 position of the benzothiophene ring system. Nucleophilic aromatic substitution of 10a by benzylmercaptan was effected at 60 °C with DBU in DMF and gave 9d in 67% yield. Demethylation of the ester occurred during the course of the reaction. The analogous reaction of 10a with thiophenol also gave 9e in 70% yield. Upon the treatment of 10c with sodium hydride in THF, reaction with diethyl chlorophosphate produced 10f in 45% yield, while a similar reaction with phenylacetyl chloride gave 10g in 47% yield. The C-3 acetate derivative 10h was obtained in 76% yield upon the reaction of 10c with acetic anhydride followed by aqueous workup (Scheme 5).

Scheme 2

cyclization of the benzothiophene ring to occur with the Nphenyl group, which is known to be efficient. Moreover, when linked together in this fashion, the thioxanthone undergoes triplet energy transfer to the benzothiophene ring system reversibly in an equilibrium process. The reversible triplet energy transfer and ensuing electrocyclization can take place within the lifetime of the long-lived thioxanthone triplet excited state, and the photochemistry can therefore be efficient. The combined effects of these changes result in 10-fold higher quantum yields for leaving group expulsion from 9a, i.e., Φ = 0.41, as compared to Φ = 0.03415 for 5 for LG− = Cl−. Because of the endothermic nature of the triplet excitation transfer, quantum yields increase with temperature, consistent with the energy transfer being a thermally activated process with an apparent Ea = 3.98 kcal mol−1 (experimentally). The triplet excitation transfer is likely to be reversible in the trimethylene-linked donor−acceptor system; therefore, the Stern−Volmer plot obtained upon quenching of the photoreaction of 9a with cyclohexadiene as the triplet excited state quencher was successfully analyzed to obtain the rate constants for the forward and reverse energy transfer steps. 8996


Article

The Journal of Organic Chemistry Scheme 3. Synthesis Linking Thioxanthone and Modification of Benzothiophene-2-carboxanilides

Scheme 4. Substitution of Chloride with Methoxide in Benzothiophene-2-carboxylic Acid

Scheme 5. Synthetic Elaboration To Vary Leaving Groups at C-3 of Thiophene Ring

Absorption Spectra. The absorption spectrum of reactant 9a exhibits λmax = 385 nm (ε 6020 M−1 cm−1) (Figure 1), while photoproduct 11 has a structured band centered at λmax = 365 nm (ε 9070 M−1 cm−‑1). In contrast, a long wavelength absorption was observed in the 430−440 nm region of the previously reported12 methyl ester of photoproduct 6, which was obtained from the methyl ester of 5 (LG− = Cl−). Like 11, the longest wavelength band for benzothiophene-2-carboxanilide photoproduct 8 was in the 365 nm region (Figure S2). This is also true for the methyl ester photoproduct 8 obtained upon photolysis of the methyl ester of 7 (Figure S3). Therefore, the long wavelength absorption band exhibited by

the methyl ester of 6 is evidently due to the extended conjugation of the ring-fused thioxanthone that is inherent in photoproduct 6. Minor differences were observed between the absorption spectra of the linked benzothiophene/thioxanthone system 9a and that of an equimolar mixture of thioxanthone and benzothiophene-2-carboxanilide 7 (Figure S4). A possible weak interaction of the linked thioxanthone and benzothiophene-2-carboxamide moieties was manifested by a 6 nm red shift in wavelength to 385 nm and a shift from 255 to 259 nm in the linked system. Otherwise, the linked system mostly resembled the sum of the absorptions of the isolated 8997


Article


yields also are found to decrease with increasing basicity of the released leaving groups LG− (Table 1). Note that for 9e the effect of pH was briefly assessed. However, the quantum yield was unchanged (Φ = 0.19) when the pH was increased from 7.1 to 9.2. During the photolyses the thioxanthone is expected to selectively absorb all of the light. Subsequent intersystem crossing (Φisc = ca. 0.618,19) would generate its triplet excited state (ET = 65 kcal mol−1 16−19). Endothermic triplet excitation transfer then must occur to effect the observed photorearrangement of the linked benzothiophene2-carboxamide (ET = ca. 69 kcal mol−1 20) to give photoproduct 10. Direct photolysis of benzothiophene-2-carboxamide anilide 7 in 20% water and 80% dioxane containing 100 mM phosphate buffer at pH 7 gave the cyclized photoproduct 8 as a precipitate, which was isolated by filtration and fully characterized in 94% yield. The quantum yield was 0.14. The thioxanthone triplet sensitized photolysis (eq 2) gave the same photoproduct, compound 8, according to NMR analyses, and the quantum yield was determined to be Φ = 0.22 for 0.02 M thioxanthone and 0.003 M anilide 7. However, quantum yields were found to increase with increases in the concentration of the anilide. This concentration dependence were studied in detail (vide infra). Temperature Dependence of Quantum Yields for Photolysis of Trimethylene-Linked Thioxanthone Energy Donor and Benzothiophene-2-carboxamide Energy Acceptor Compound 9a. As noted above, the photocyclization of 9a is thought to involve a endothermic triplet energy transfer step that is ca. 4 kcal mol−1. Consistent with this, it was found that quantum yields increased with temperature. An Arrhenius plot of log Φ vs T−1 (Figure 2) gave a slope of 870 K−1 and an intercept of 2.53 (R2 = 0.9777). The slope corresponds to an Ea of 3.98 kcal mol−1.

Figure 1. Absorption spectra of reactant 9a () and photoproduct 11 (---). Concentrations were 3.37 × 10−5 and 3.43 × 10−5 M, respectively, in 20% water and 80% dioxane containing 100 mM phosphate buffer. The corresponding methyl esters are given in Figure S1.

molecules. Thioxanthone absorbs light in the 380−400 nm region, whereas N-methyl benzothiophene-2-carboxanilide 7 absorbs light at much shorter wavelengths below 360 nm (Figure S5). In the trimethylene-linked system, 390 nm light will be entirely absorbed by the thioxanthone chromophore and not by the benzothiophene moiety. Photolyses of Trimethylene-Linked and Unlinked Thioxanthone and Benzothiophene-2-carboxanilide. Upon direct photolysis of 9a,c−e and 10f−h at 395 nm in 20% water and 80% dioxane containing 100 mM phosphate buffer at pH 7.1, the benzothiophene-2-carboxamide moiety is observed to undergo photocyclization with leaving group release to form photoproduct 11 (Scheme 2) in yields summarized in Table 1. Quantum yields are about 10-fold higher (Table 1) than previously studied examples with thioxanthone and benzothiophene closely coupled via the carboxamide group in 5. As previously reported for benzothiophene-2-carboxanilides 1 (Scheme 1), quantum Table 1. Quantum Yields for Product 11 and 12 Formation and Chemical Yields for Photolyses of 9 and 10 Bearing Various Leaving Groups (LG−) at 395 nm LG−

reactant 9a 9c 9d 9e 10f 10g 10h

−

Cl HO− PhCH2S− PhS− (EtO)2PO2− PhCH2CO2− CH3CO2−

LG-H, % b

nd ndb 80 ndb 93 93 ca. 91

unreacted start. mater, % nd 2 2 3 3 2 3

b

11 or 12, %a

Φ

100 98 98 97 97 98 97

0.41c 0.011c 0.14c 0.18c 0.35d 0.28d 0.23d

Figure 2. Temperature dependence of quantum yields for photolysis of 9a.

Quenching of the Photoreaction of 9a and 7 with Cyclohexadiene. The photoreaction of 9a was readily quenched by the known triplet quencher, cyclohexadiene. Since the triplet excitation transfer from thioxanthone to benzothiophene-2-carboxamide is ca. 4 kcal mol−1 endothermic, one would expect that the triplet energy transfer from the initial triplet excited state of thioxanthone, T1, to generate the benzothiophene triplet excited state, T2, should also occur in the energetically favorable reverse direction. In other words, the energy transfer should be reversible with the equilibration of the T1 and T2 excited states, when the triplet donor and

a

Chemical yields obtained in DMSO-d6 by NMR integration for carboxylic acids 9c−e; CD3CN was used as solvent for 10f,g. bNot determined. cThe solvent was 20% H2O containing 100 mM phosphate buffer and 80% dioxane; photolysates were acidified and then lyophilized followed by product yield determination by 1H NMR spectroscopy with PhCH2OCH3 as the integration standard. dSame procedure as acids 9 except that the solvent system was 10% H2O and 90% dioxane with phosphate buffer. 8998


Article


is defined as τe = [X1kd1 + X2(kd2 + kr)]−1. From b = kq1kq2τe(k12 + k21), one obtains (k12 + k21) = 1.6 × 108 s−1. From c = kq1/k12, k12 = 8.7 × 106 s−1 and thus k21 = 1.5 × 108 s−1. If states T1 and T2 are not in equilibrium, yet both states are quenched by cyclohexadiene, then the kinetic expression for Φ°/Φ becomes quadratic.28 The data in Figure 3 could not be fit by such a quadratic expression unless the coefficient for the [Q]2 term is negative, which in turn requires a negative rate constant. A linear plot for Figure 3 would result if one of the excited states, i.e., state T2 (Scheme 6), is not quenched, which would be possible if that state is too short-lived to be quenched. However, upon direct photolysis of anilide 7 as a model system for T2 of 9a, the triplet excited state generated upon intersystem crossing is readily quenched to give a linear Stern−Volmer plot (Figure S6). Note that for direct photolysis of 7, Φ° = 0.14. From the Stern−Volmer plot for quenching with cyclohexadiene, kqτ = 4.51 × 103 M−1 s−1. For kq = 4.4 × 109 M−1 s−1 in aq dioxane the triplet lifetime τ = 1.02 μs. This lifetime for the triplet excited state of anilide 7 with LG− = Cl− is very similar to the lifetime of 932 ns previously reported for the triplet excited state of a similar anilide that lacks the C-6 carboxylic acid group.14 In that prior study it was also found that the lifetime did not change significantly when the carboxylate leaving group was swapped for a more basic phenolate leaving group.30 Sensitized Photolysis of Anilide 7 with Thioxanthone. The quantum yield of the thioxanthone sensitized photochemical electrocyclic ring closure of 7 with release of chloride (eq 2) increases with increasing concentration of anilide 7, and at a high concentration of 0.0120 M, Φ = 0.45. The usual double reciprocal plot of the sensitized quantum of anilide 7 is linear with R2 = 0.9994. (Figure 4).

Scheme 6. Kinetic Scheme for Excited State Reactivity and Quenching of Thioxanthone-Linked Benzothiophene-2carboxanilides

Figure 3. Stern−Volmer plot for quenching of the photoreaction of 9a by cyclohexadiene.

acceptor are linked together, as shown in Scheme 6. The corresponding rate constants k12 and k21 can be obtained by analysis of a plot of Φ°/Φ vs [Q] for cyclohexadiene as quencher Q (Figure 3). Φ°/Φ is expressed in terms of rate constants by eq 3, which conforms to the kinetics shown in Φ° = Φ 1 + (X1kq1 + X 2kq2)τe[Q] + kq1kq2τe[Q]2 (k12 + k 21)−1 1 + kq2[Q] /k12 (3)

Scheme 6. This equation is similar to that derived by Wagner28 for a case whereby the initial triplet excited state is the one that undergoes reaction, whereas in the present example it is T2, which reacts instead. The only difference here is that the denominator contains a k12 term, whereas the original Wagner equation would instead have k21. Equation 3 has the general form of eq 4 where a−c are adjustable parameters. The y=

1 + ax + bx 2 1 + cx

(4)

Figure 4. Φ−1 vs anilide 7 with constant concentration of 2.00 × 10−2 M thioxanthone.

nonlinear least-squares fit to the quenching data in Figure 3 gave a = 1220, b = 34000, and c = 504 (R2 = 0.9699). Accordingly, parameter a = (X1kq1 + X2kq2)τe. The kq1 and kq2 values of 4 × 109 M−1 s−1 can be estimated from the Debye equation using the viscosity of 20% H2O and 80% dioxane reported in the literature.29 Assuming the mole fraction of triplets T1 is X1 = k21/(k21 + k12) > 0.9 such that X2 = k12/(k21 + k12) can be neglected, the equilibrium kinetic lifetime is estimated as τe = 2.8 × 10−7 s. The equilibrium kinetic lifetime

In this case, the unlinked triplet energy donor and acceptor, the 3TX and 37 excited states, are not necessarily equilibrated. Consequently, the sensitization kinetics are assumed to conform to Scheme 7, from which eq 5 is derived. The observed slope of the plot in Figure 4 is 0.00886 M, and the intercept is 1.47. From the slope, k12 = ca. 1.15 × 107 M−1 s−1, taking kd1 = 7.69 × 104 s−1 for thioxanthone,31 and kr = ca. τ−1 = 9.80 × 105 s−1, which was found for Stern−Volmer quenching of anilide 7 by cyclohexadiene (vide supra). Here 8999


Article


to generate the benzothiophene triplet excited state, which ultimately leads to reaction. Thus, the overall rate for energy transfer and reaction of 9a is estimated to be 0.058kr = ca. 5.7 × 104 s−1, where kr = 9.8 × 105 s−1 is obtained from τ−1 (7) from Stern−Volmer quenching of the anilide with cyclohexadiene, which assumes that the rate constant for radiationless decay of the anilide triplet, kd2, can be neglected relative to kr. The overall rate constant of 5.7 × 104 s−1 is quite competitive with decay of the initial thioxanthone triplet excited state (kd1 = 7.7 × 104 s−1) such that the trimethylenelinked system would be ca. 70% efficient once intersystem crossing of the thioxanthone takes place. Taking into account Φisc = 0.68, the overall quantum yield for the linked system could be ca. 0.5, as compared to Φ = 0.4 that is actually observed. For the sensitized reaction of 7 the rate of triplet energy transfer from thioxanthone is [0.012 M] × k12 = 6.24 × 104 s−1, which can be compared to the decay rate kd1 for the thioxanthone sensitizer. The overall efficiency after intersystem crossing has occurred would be ca. 0.8. Assuming the efficiency for electrocyclization and LG− release in the triplet excited state of 7 is essentially unity, taking into account Φisc, the reaction is estimated to be 50−60% efficient, as estimated above for the trimethylene-linked system. For both the trimethylene-linked system and the intermolecular sensitized case the estimated quantum yields are somewhat higher than those actually observed. This could be due to the actual electrocyclization step and the leaving group expulsion being somewhat less than completely efficient. The rapid electrocyclization of triplet excited benzothiophene-2-carboxanilide was expected, since the computational study of the triplet excited state cyclization of 5 (LG− = Cl−) showed the barrier for this process to be virtually nonexistent, 295 °C. HRMS (ESI-TOF), m/z: [M − H]− calcd for C17H10NO3S, 308.0387; found, 308.0364.1H NMR (400 MHz, DMSO-d6), δ: 13.24 (br s, 1H), 8.68 (d, J = 8.7 Hz, 1H), 8.58 (s, 1H), 8.53 (d, J = 8.7 Hz, 1H), 7.97 (d, J = 7.9 Hz, 1H), 7.64−7.50 (br m, 2H), 7.39−7.29 (br m, 1H), 3.67 (s, 3H).13C NMR (75 MHz, DMSO-d6), δ: 164.2, 154.4, 138.6, 135.5, 131.7, 131.2, 126.6, 126.5, 123.3, 122.9, 122.7, 121.2, 120.2, 115.4, 113.4, 27.2. Photochemical Synthesis of Methyl 5-Methyl-6-oxo-5,6dihydrobenzo[4,5]thieno[2,3-c]quinoline-9-carboxylate (Methyl Ester of Acid 8). The same procedure was used as in the photochemical synthesis of photoproduct 8 above. We used 120 mg (0.33 mmol) of the ester of acid 7 to give 109 g (99% yield) of photoproduct for the ester of acid 7, mp 250−251 °C. HRMS (ESITOF), m/z: [M + H]+ calcd for C18H14NO3S, 324.0689; found, 324.0660. Anal. Calcd for C18H13NO3S: C, 66.86; H, 4.05; N, 4.33. Found: C, 66.90; H, 4.18; N, 4.30. 1H NMR (300 MHz, CDCl3), δ: 8.67−8.41 (m, 3H), 8.10 (d, J = 8.6, 1H), 7.65−7.34 (m, 3H), 4.00 (s, 3H), 3.83 (s, 3H). 13C NMR (75 MHz, CDCl3), δ: 166.7, 158.3, 142.3, 139.1, 138.7, 135.5, 134.7, 129.1, 128.5, 126.0, 125.7, 125.3, 124.0, 123.0, 119.3, 115.7, 52.7, 30.3. Photochemical Synthesis of 6-Oxo-5-(3-(9-oxo-9H-thioxanthen-2-yl)propyl)-5,6-dihydrobenzo[4,5]thieno[2,3-c]quinoline-9-carboxylic Acid (11). A solution of 210 mg (0.36 mmol, 7.12 × 10−3 M) of thioxanthone-linked acid 9a in 50 mL of nitrogen saturated dioxane was irradiated with a 450 W Hanovia medium pressure mercury lamp with a Pyrex filter for 4 h. The cloudy mixture was then concentrated in vacuo, and the precipitate was washed with water to give 195 mg (99% yield) of photoproduct 11, mp >290 °C. HRMS (ESI-TOF), m/z: [M + H]+ calcd for C32H22NO4S2, 548.0985; found, 548.0954. 1H NMR (400 MHz, DMSO-d6), δ: 13.28 (br s, 1H), 8.82 (d, J = 8.7 Hz, 1H), 8.71 (d, J = 8.2 Hz, 1H), 8.65 (d, J = 1.6 Hz, 1H), 8.34 (d, J = 8.2 Hz, 1H), 8.17 (s, 1H), 8.00 (dd, J = 8.7, 1.6 Hz, 1H), 7.81 (d, J = 8.7 Hz, 1H), 7.73−7.62 (m, 3H), 7.56−7.40 (m, 4H), 4.54 (t, J = 6.9 Hz, 2H), 2.92 (t, J = 7.2 Hz, 2H), 2.16 (p, J = 7.2, 6.9 Hz, 2H). 13C NMR (75 MHz, DMSO-d6), δ: 180.7, 169.1, 159.4, 33128.7, 128.6, 128.4, 128.2, 127.8, 127.5, 126.4, 125.0, 120.7, 118.4, 110.0, 44.2, 34.3, 30.0. Photochemical Synthesis of Methyl 6-Oxo-5-(3-(9-oxo-9Hthioxanthen-2-yl)propyl)-5,6-dihydrobenzo[4,5]thieno[2,3-c]quinoline-9-carboxylate (12). A solution of 205 mg (0.34 mmol, 6.9 × 10−3 M) of thioxanthone-linked ester 10a in 50 mL of nitrogen saturated dioxane was irradiated with a 450 W Hanovia medium pressure mercury lamp with a Pyrex filter for 3 h. The solution was

then concentrated in vacuo, and the solid product was washed with water. The product was chromatographed on a silica gel (60−200 mesh) column eluting with 5% ether in DCM to give 189 mg (98% yield) of photoproduct 12, mp >280 °C. HRMS (ESI-TOF), m/z: [M + H]+ calcd for C33H24NO4S2, 562.1141; found, 562.1147. 1H NMR (400 MHz, CDCl3), δ: 8.69−8.46 (m, 4H), 8.35 (s, 1H), 8.11 (d, J = 8.3 Hz, 1H), 7.68−7.53 (m, 2H), 7.52−7.35 (m, 6H), 4.53 (s, 2H), 4.00 (s, 3H), 2.96 (s, 2H), 2.27 (s, 2H). 13C NMR (75 MHz, CDCl3), δ: 180.7, 169.1, 159.4, 143.5, 142.0, 140.3, 139.4, 138.7, 136.5, 136.3, 136.0, 135.2, 134.8, 131.5, 131.2, 130.5, 130.2, 130.2, 128.7, 128.6, 128.4, 128.2, 127.8, 127.5, 126.4, 125.0, 120.7, 118.4, 110.0, 44.2, 34.3, 30.0. General Procedure for Quantum Yield Determinations. A semimicro optical bench was used for quantum yield determinations, which was similar to the apparatus described by Zimmerman.37 Light from a 200 W high-pressure mercury lamp was passed through an Oriel monochromator, which was set to a wavelength of either 310 or 395 nm. The light was collimated through a lens. A fraction of the light was diverted 90° by a beam splitter to a 42.5 mL side quartz cylindrical cell containing an actinometer. The photolysate was contained in a 37.5 mL quartz cylindrical cell placed in the light path. Behind the photolysate was mounted a quartz cylindrical cell containing 42.5 mL of actinometer. Light output was monitored by ferrioxalate actinometry using the splitting ratio technique. All quantum yields reported were the average of two or more independent runs. Thioxanthone Sensitized Photolysis with Various Concentrations of Benzothiophene-2-carboxanilide 7. The general procedure for product quantum yields was used. Thioxanthone (98%) was purified by recrystallization from methanol and drying through vacuum filtration. A stock solution of 2.12 g (9.99 mmol, 0.02 M) of thioxanthone was prepared in 500 mL of dioxane containing 15% aqueous potassium phosphate buffer (v/v). The anilide 7 was dissolved in 37.5 mL of the stock solution to obtain a series of photolysates with anilide concentrations of 0.002, 0.003, 0.004, 0.005, 0.006, 0.009, 0.0012, and 0.015 M. After being flushed in the dark with nitrogen for ca. 30−60 min, each sample was irradiated at 395 nm for 1−2 h. Light absorbed during photolysis was determined by ferrioxalate actinometry. At the end of each photolysis, ca. 0.5 mL of 1 M HCl was added to the exposed photolysate followed by concentration in vacuo in the dark. Each concentrate was freezedried for 2 h and further dried overnight under vacuum in the dark. The dried samples were dissolved in DMSO-d6, and a known amount of DMF was added to each as an internal standard. The yield of photoproduct 8 was determined by 1H NMR analysis by integration of the N−CH3 signal of the photoproduct at 3.67 ppm relative to the DMF methyl groups. Photolysis with Various Concentrations of Thioxanthone Sensitizer. The general procedure for product quantum yields was used. A stock solution of 0.52 g (1.50 mmol, 0.003 M) of benzothiophene carboxanilide 7 was prepared in 500 mL of dioxane containing 10% aqueous potassium phosphate buffer and kept in the dark. Thioxanthone was dissolved in 37.5 mL of stock solution to give concentrations of 0.01, 0.02, 0.04, and 0.06 M. After being flushed in the dark with nitrogen for ca. 30−60 min, each sample was photolyzed at 395 nm for 1.5 h, followed by ferrioxalate actinometry. To each photolysate was added 0.5 mL of 1 M HCl followed by concentration in vacuo in the dark. The concentrates were freezedried for 2 h and then dried under vacuum overnight in the dark. The dried samples were dissolved in DMSO-d6, and a known amount of DMF was added to each as an internal standard. The yield of photoproduct 8 was determined by 1H NMR analysis by integration of the N−CH3 signal of the photoproduct at 3.67 ppm relative to the DMF methyl groups. Quantum Yield for 9a at Various Temperatures for Ea Determination. The general procedure for product quantum yields was used. A stock solution of 0.336 g (0.61 mmol, 1.92 × 10−3 M) of thioxanthone-linked acid 9a in 300 mL of dioxane containing 20% aqueous 100 mM phosphate buffer was prepared. For each photolysis, a 37.5 mL sample was flushed with nitrogen for 30−60 min in the 9004


Article

The Journal of Organic Chemistry dark. Temperature was controlled by circulating water from a constant temperature bath through the jacketed photolysis cell containing the photolysate. Photolysis times ranged from 1 to 0.4 h, depending on the temperature. The temperatures for photolyses were 10, 15, 20, 25, 30, 35, and 40 °C. The photolysis wavelength was 395 nm, and ferrioxalate actinometry was performed to determine the light absorbed by each sample. After photolysis, ca. 0.5 mL of 1 M HCl was added followed by concentration in vacuo in the dark. The concentrate was then freeze-dried for 2 h and dried under vacuum overnight in the dark. Photoproduct 11 was dissolved in DMSO-d6, and a known amount of benzyl methyl ether (CH2 and CH3 peaks) was added as an internal standard. The photoproduct was quantified by 1H NMR analysis by integrating each of the three CH2 signals of 11 at 4.54, 2.92, and 2.16 ppm and averaging. Quenching of the Direct Photolysis of 9a by Cyclohexadiene. The general procedure for determining product quantum yields was used. A stock solution of 0.336 g (0.61 mmol, 1.92 × 10−3 M) of thioxanthone-linked acid 9a in 300 mL of dioxane containing 20% aqueous 100 mM phosphate buffer was prepared. Samples of 37.5 mL were flushed with nitrogen for ca. 30−60 min in the dark followed by addition of cyclohexadiene via syringe to give final concentrations of quencher of 1.12 × 10−3, 3.37 × 10−3, 5.06 × 10−3, 5.61 × 10−3, 7.30 × 10−3, and 9.56 × 10−3 M. Samples were sealed and photolyzed at 395 nm for 30−90 min. Light absorbed was determined by performing ferrioxalate actinometry. After photolysis, ca. 0.5 mL of 1 M HCl was added and the photolysate was concentrated in vacuo in the dark. The concentrate was then freezedried for 2 h and dried under vacuum overnight in the dark. Photoproduct 11 was dissolved in DMSO-d6, and a known amount of benzyl methyl ether (CH2 and CH3 peaks) was added as an internal standard. The photoproduct was quantified by 1H NMR analysis by integrating each of the three CH2 signals of 11 at 4.54, 2.92, and 2.16 ppm and averaging. Quenching of the Direct Photolysis of Anilide 7 by Cyclohexadiene. A stock solution of 3.0 × 10 −3 M of benzothiophene carboxylic acid in dioxane containing 17% aqueous 100 mM phosphate buffer at pH 7 was prepared. The photolysis procedure was the same as that for the quenching of 9a by cyclohexadiene, except that the wavelength was 310 nm and the photolysis times were 1−2 h. Light absorbed was determined by ferrioxalate actinometry. Concentrations of cyclohexadiene used were 8.1 × 10−5, 1.62 × 10−4, 2.16 × 10−4, and 3.24 × 10−4 M. At the end of each photolysis, ca. 0.5 mL of 1 M HCl was added followed by concentration in vacuo in the dark. The concentrate was freeze-dried for 2 h and dried under vacuum overnight in the dark. Photoproduct 8 was dissolved in DMSO-d6, and DMF was added as an internal standard. The yield of photoproduct 8 was determined by 1H NMR analysis by integration of the N−CH3 signal of the photoproduct at 3.67 ppm relative to the DMF methyl groups. Quantum Yields for Acids 9a,c−e. A 1.92 × 10−3 M solution of each of the thioxanthone-linked acids 9a,c−e in 20% aq dioxane containing 100 mM phosphate buffer in a 37.5 mL cell was flushed with nitrogen for ca. 30−60 min in the dark. Each sample was sealed and photolyzed at 395 nm for 30−90 min. Light absorbed was determined by performing ferrioxalate actinometry. At the end of each photolysis, 0.5 mL of 1 M HCl was added to the solution, which was concentrated in vacuo in the dark. The concentrate was then freezedried for 2 h and dried under vacuum overnight in the dark. The photoproduct was dissolved in DMSO-d6 and quantified by 1H NMR analysis using benzyl methyl ether (CH2 and CH3 peaks) as an internal standard by averaging the integrals of the three CH2 signals of the photoproduct at 4.54, 2.92, and 2.16 ppm. Quantum Yields for Esters 10f−h. A 1.5 × 10−3 M solution of each of the thioxanthone-linked esters 10f−h in 10% aq dioxane containing 100 mM phosphate buffer in a 37.5 mL cell was flushed with nitrogen for ca. 30−60 min in the dark. Each sample was sealed and photolyzed at 395 nm for 30−90 min. Light absorbed was determined by performing ferrioxalate actinometry. At the end of each photolysis, 0.5 mL of 1 M HCl was added to the solution, which was concentrated in vacuo in the dark. The concentrate was freeze-dried

for 2 h and dried under vacuum overnight in the dark. The photoproduct was dissolved in DMSO-d6 and quantified by 1H NMR analysis using benzyl methyl ether (CH2 and CH3 peaks) as an internal standard by averaging the integrals of the three CH2 peaks of the photoproduct at 4.54, 2.92, and 2.16 ppm. Percent Yield Determination for the Photorelease of Leaving Groups from Compounds 9a, 9c, 9d, and 9e. The respective concentrations for 9a and 9c−9e of 8.57 × 10−3, 7.81 × 10−3, 5.33 × 10−3, and 5.11 × 10−3 M in 20% aq DMSO-d6 containing 100 mM phosphate buffer (prepared from D2O) in NMR tubes were flushed with nitrogen for ca. 30 min in the dark followed by the addition of 20 μL of 2.19 × 10−1 M PhCH2OCH3 as an internal standard to each tube. The samples were then irradiated with a 450 W Hanovia medium pressure mercury lamp with a Pyrex filter. The reactions were monitored for completion by taking the NMR readings at 30 min intervals. Compound 9a showed 100% conversion to photoproduct 11 after 1.5 h, compounds 9d and 9e showed 98% and 97% conversion, respectively, to photoproduct 11 after 2.5 h, and compound 9c showed 98% conversion to photoproduct 11 after 3.5 h. Percent Yield Determination for the Release of Compounds 10f, 10g, and 10h. Compounds 10f, 10g, and 10h with concentrations of 5.11 × 10−3, 5.16 × 10−3, and 5.81 × 10−3 M, respectively, in 10% aq DMSO-d6 containing 100 mM phosphate buffer (prepared from D2O) were put in different NMR tubes. The NMR tubes were flushed with nitrogen for about 30 min in the dark followed by addition of 20 μL of 2.19 × 10−1 M PhCH2OCH3 as an internal standard. The samples were then irradiated with a 450 W Hanovia medium pressure mercury lamp with a Pyrex filter. The reactions were monitored for completion by taking NMR readings at 30 min intervals. Compound 10f showed 97% conversion to photoproduct 12 after 1.5 h, compounds 10g and 10f showed 98% and 97% conversion to photoproduct 12 after 2 h, and compound 10h showed 98% conversion to photoproduct 12 after 2.5 h.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.joc.8b01173. X-ray crystallography of compound 7 (CIF) X-ray crystallography of compound 10a (CIF) X-ray crystallography of compound 12 (CIF) X-ray crystallography of compound 20 (CIF) Absorption spectra of 10a, 12, 7, and 8, thioxanthone and anilide 7, 9a, methyl esters of 7 and 8; Stern− Volmer quenching plot of 7 with cyclohexadiene; plot of Φ−1 vs [S]; energy diagrams and kinetic scheme and related equations; original copy of 1H and 13C NMR spectra of all new compounds; X-ray crystallography of compounds 7, 10a, 12, and 20 (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*Email: [email protected]. ORCID

Mark G. Steinmetz: 0000-0002-2044-2522 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors thank the National Science Foundation for financial support of this work (Grant No. CHE-1055339). We thank Prof. Shama P. Mirza and Dr. Anna Benko of the Shimadzu Laboratory for Advanced and Applied Analytical Chemistry at the University of Wisconsin, Milwaukee for the HRMS analyses. We also thank Dr. Sergey Lindemann, 9005


Article


Resolved Thermal Lens Spectroscopy: Solvent and Population Lens Effects. Chem. Phys. Lett. 2000, 322 (6), 483−490. (20) Zander, M.; Jacob, J.; Lee, M. L. Empirical Quantitative Relationship between Molecular-Structure and Phosphorescence Transition Energy of Polycyclic Aromatic Thiophenes. Z. Naturforsch., A: Phys. Sci. 1987, 42A (7), 735−738. (21) Closs, G. L.; Piotrowiak, P.; MacInnis, J. M.; Fleming, G. R. Determination of Long Distance Intramolecular Triplet Energy Transfer Rates. A Quantitative Comparison with Electron Transfer. J. Am. Chem. Soc. 1988, 110, 2652−2653. (22) Closs, G.; Johnson, M. D.; Miller, J. R.; Piotrowiak, P. A Connection between Intramolecular Long-Range Electron, Hole, and Triplet Energy Transfers. J. Am. Chem. Soc. 1989, 111 (10), 3751− 3753. (23) Klan, P.; Wagner, P. J. Intramolecular Triplet Energy Transfer in Bichromophores with Long Flexible Tethers. J. Am. Chem. Soc. 1998, 120 (9), 2198−2199. (24) Wagner, P. J.; Klán, P. Intramolecular Triplet Energy Transfer in Flexible Molecules: Electronic, Dynamic, and Structural Aspects. J. Am. Chem. Soc. 1999, 121 (41), 9626−9635. (25) Vrbka, L.; Klán, P.; Kríz, Z.; Koca, J.; Wagner, P. J. Computer Modeling and Simulations on Flexible Bifunctional Systems: Intramolecular Energy Transfer Implications. J. Phys. Chem. A 2003, 107 (18), 3404−3413. (26) Chen, Z.; Mocharla, V. P.; Farmer, J. M.; Pettit, G. R.; Hamel, E.; Pinney, K. G. Preparation of New Anti-Tubulin Ligands through a Dual-Mode, Addition-Elimination Reaction to a Bromo-Substituted α,β-Unsaturated Sulfoxide. J. Org. Chem. 2000, 65 (25), 8811−8815. (27) Olah, G. A.; Husain, A.; Singh, B. P.; Mehrotra, A. K. Synthetic Methods and Reactions. 112.1 Synthetic Transformations with Trichloromethylsilane/Sodium Iodide Reagent. J. Org. Chem. 1983, 48 (21), 3667−3672. (28) Wagner, P. J. In Creation and Detection of the Excited State; Lamola, A. A., Ed.; Marcel Dekker, Inc.: New York, 1971; pp 171− 212. (29) Omota, L.; Iulian, O.; Ciocîrlan, O.; Nita, I. Viscosity of Water, 1, 4-Dioxane and Dimethyl Sulfoxide Binary and Ternary Systems at Temperatures from 293.15 to 313.15 K. Rev. Roum. Chim 2008, 53 (11), 977−988. (30) In principle, laser flash photolysis could be used to determine the lifetimes of the T1 and T2 triplet excited states in Scheme 3, which in turn would provide estimates of k12 and k21, as pointed out by a reviewer. An example is given by the following study: Woll, D.; Laimgruber, S.; Galetskaya, M.; Smirnova, J.; Pfleiderer, W.; Heinz, B.; Gilch, P.; Steiner, U. E. On the Mechanism of Intramolecular Sensitization of Photocleavage of the 2-(2-Nitrophenyl)propoxycarbonyl (NPPOC) Protecting Group. J. Am. Chem. Soc. 2007, 129, 12148−12158. (31) Neumann, M. G.; Gehlen, M. H.; Encinas, M. V.; Allen, N. S.; Corrales, T.; Peinado, C.; Catalina, F. Photophysics and Photoreactivity of Substituted Thioxanthones. J. Chem. Soc., Faraday Trans. 1997, 93 (8), 1517−1521. (32) The behavior of quantum yields Φ for photolysis of solutions containing a constant 0.003 M anilide 7 in the presence of various (0.01−0.06 M) concentrations of thioxanthone (TX) at 395 nm was examined. Φ was observed to decrease with increasing [TX], and the fit of Φ−1 vs [TX] to a line gave slope = 34.9 and intercept 3.54 (R2 = 0.9613) (Figure S6). Treating TX as a quencher, the Stern−Volmer slope would correspond to (k21/kr)ΦiscΦet where Φisc pertains to Tx while Φet is the efficiency of triplet energy transfer from Tx to anilide 7. The intercept would be (kr + kd2)(krΦiscΦet)−1. If kd2 ≪ kr, the intercept would give ΦiscΦet = ca. 0.28, while the slope with kr = 9.6 × 106 s−1 would give k21 = 9.6 × 106 M−1 s−1, which is quite low for 4 kcal mol−1 exothermic intermolecular triplet energy transfer. The reasons for the low k21 are not understood. (33) Saltiel, J.; Charlton, J. L.; Mueller, W. B. Nonvertical Excitation Transfer. Activation Parameters for Endothermic Triplet Energy Transfer to Stillbenes. J. Am. Chem. Soc. 1979, 101 (5), 1347−1348.

Department of Chemistry at Marquette University, for the Xray structure determinations and Dr. Sheng Cai, Department of Chemistry at Marquette University, for assistance with NMR spectroscopy.

■

REFERENCES

(1) Klán, P.; Š olomek, T.; Bochet, C. G.; Blanc, A.; Givens, R.; Rubina, M.; Popik, V.; Kostikov, A.; Wirz, J. Photoremovable Protecting Groups in Chemistry and Biology: Reaction Mechanisms and Efficacy. Chem. Rev. 2013, 113 (1), 119−191. (2) Dynamic Studies in Biology. Phototriggers, Photoswitches and Caged Biomolecules; Goeldner, M., Givens, R., Eds.; WILEY-VCH Verlag GmbH & Co. KGaA: Weinheim, 2005. (3) Mayer, G.; Hechel, A. Biologically Active Molecules with a “Light Switch. Angew. Chem., Int. Ed. 2006, 45 (30), 4900−4921. (4) Pelliccioli, A. P.; Wirz, J. Photoremovable Protecting Groups: Reaction Mechanisms and Applications. Photochem. Photobiol. Sci. 2002, 1 (7), 441−458. (5) Cui, J.; Miguel, V. S.; Del Campo, A. Light-Triggered Multifunctionality at Surfaces Mediated by Photolabile Protecting Groups. Macromol. Rapid Commun. 2013, 34 (4), 310−329. (6) Herrmann, A. Using Photolabile Protecting Groups for the Controlled Release of Bioactive Volatiles. Photochem. Photobiol. Sci. 2012, 11 (3), 446−459. (7) Ma, C.; Chen, Y.; Steinmetz, M. G. Photochemical Cleavage and Release of Para-Substituted Phenols from α-Keto Amides. Org. Lett. 2003, 5 (1), 71−74. (8) Ma, C.; Steinmetz, M. G.; Kopatz, E. J.; Rathore, R. TimeResolved pH Jump Study of Photochemical Cleavage and Release of Carboxylic Acids from α-Keto Amides. Tetrahedron Lett. 2005, 46 (6), 1045−1048. (9) Ma, C.; Steinmetz, M. G. Substituent Effects on Competitive Release of Phenols and 1, 3-Rearrangement in α-Keto Amide Photochemistry. Org. Lett. 2004, 6 (4), 629−632. (10) Ma, C.; Chen, Y.; Steinmetz, M. G. Photochemical Cleavage and Release of Para-Substituted Phenols from α-Keto Amides. J. Org. Chem. 2006, 71 (11), 4206−4215. (11) Jia, J.; Steinmetz, M. G.; Shukla, R.; Rathore, R. Photochemical Electrocyclization of α,β-Unsaturated Anilides to Give Zwitterionic Intermediates Which Eliminate Carboxylate and Phenolate Leaving Groups. Tetrahedron Lett. 2008, 49 (31), 4621−4623. (12) Jia, J.; Sarker, M.; Steinmetz, M. G.; Shukla, R.; Rathore, R. Photochemical Elimination of Leaving Groups from Zwitterionic Intermediates Generated via Electrocyclic Ring Closure of α,βUnsaturated Anilides. J. Org. Chem. 2008, 73 (22), 8867−8879. (13) Ma, C.; Steinmetz, M. G.; Kopatz, E. J.; Rathore, R. Photochemical Cleavage and Release of Carboxylic Acids from αKeto Amides. J. Org. Chem. 2005, 70 (11), 4431−4442. (14) Sarker, M.; Shahrin, T.; Steinmetz, M. G. Photochemical Eliminations Involving Zwitterionic Intermediates Generated via Electrocyclic Ring Closure of Benzothiophene Carboxanilides. Org. Lett. 2011, 13 (5), 872−875. (15) Sarker, M. I.; Shahrin, T.; Steinmetz, M. G.; Timerghazin, Q. K. Photochemical Electrocyclic Ring Closure and Leaving Group Expulsion from N-(9-Oxothioxanthenyl)Benzothiophene Carboxamides. Photochem. Photobiol. Sci. 2013, 12 (2), 309−322. (16) Herkstroeter, W. G.; Lamola, A. A.; Hammond, G. S. Mechanisms of Photochemical Reactions in Solution. XXVIII. Values of Triplet Excitation Energies of Selected Sensitizers. J. Am. Chem. Soc. 1964, 86 (21), 4537−4540. (17) Merkel, P. B.; Dinnocenzo, J. P. Thermodynamic Energies of Donor and Acceptor Triplet States. J. Photochem. Photobiol., A 2008, 193 (2−3), 110−121. (18) Villnow, T.; Ryseck, G.; Vidisha, R.-C.; Marian, C. M.; Gilch, P. Chimeric Behavior of Excited Thioxanthone in Protic Solvents: I. Experiments. J. Phys. Chem. A 2014, 118, 11696−11707. (19) Allonas, X.; Ley, C.; Bibaut, C.; Jacques, P.; Fouassier, J. P. Investigation of the Triplet Quantum Yield of Thioxanthone by Time9006


Article

The Journal of Organic Chemistry (34) Niu, S.; Schneider, R.; Vidal, L.; Balan, L. Thioxanthone Functionalized Silver Nanorods as Smart Photoinitiating Assemblies to Generate Photopolymer/Metal Nano-Objects. Nanoscale 2013, 5, 6538−6544. (35) Nicolaou, K. C.; Estrada, A. A.; Zak, M.; Lee, S. H.; Safina, B. S. A Mild and Selective Method for the Hydrolysis of Esters with Trimethyltin Hydroxide. Angew. Chem., Int. Ed. 2005, 44 (9), 1378− 1382. (36) Wright, S. W.; Hageman, D. L.; Wright, A. S.; McClure, L. D. Convenient Preparations of t-Butyl Esters and Ethers from t-Butanol. Tetrahedron Lett. 1997, 38 (42), 7345−7348. (37) Zimmerman, H. Apparatus for Quantitative and Preparative Photolysis. The Wisconsin Black Box. Mol. Photochem. 1971, 3 (3), 281−292.

9007


Reversible Triplet Excitation Transfer in a ... - ACS Publications

Recommend Documents