Lumisterol to Tachysterol Photoisomerization in ... - ACS Publications

Subscriber access provided by University of Newcastle, Australia

Article

Lumisterol to Tachysterol Photoisomerization in EPA Glass at 77 K. A Comparative Study Malgorzata Bayda, Christopher E. Redwood, Shipra Gupta, Olga Dmitrenko, and Jack Saltiel J. Phys. Chem. A, Just Accepted Manuscript • DOI: 10.1021/acs.jpca.6b12843 • Publication Date (Web): 24 Feb 2017 Downloaded from http://pubs.acs.org on March 9, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

The Journal of Physical Chemistry A is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Lumisterol to Tachysterol Photoisomerization in EPA Glass at 77 K. A Comparative Study Malgorzata Bayda,† Christopher E. Redwood, Shipra Gupta, Olga Dmitrenko,‡ and Jack Saltiel* Department of Chemistry and Biochemistry, Florida State University, Tallahassee, FL 323064390 Corresponding Author *[email protected]

ABSTRACT: We present a comparative study of the photoisomerizations of lumisterol (Lumi), previtamin (Pre) and provitamin D3 (Pro) to tachysterol (Tachy) at 77 K in EPA (5:5:2 ether, isopentane and ethanol by volume) glass. Fluorescence, fluorescence excitation and UV spectra, measured in the course of these reactions were analyzed using singular value decomposition with self-modeling (SVD-SM). This represents an extension of previous work that led to the conclusion that in the EPA glass Pre exists as an s-cis,s-cis-conformer (cZc-Pre) which gives, exclusively, an unstable s-cis,s-cis-conformer of Tachy (cEc-Tachy) and Pro gives mainly the tEc-Tachy, that corresponds to a stable s-trans,s-cis-conformer.1 The surprising result was that the major Pre photoproduct from Pro also has a tZc-Pre conformation instead of the expected cZc-Pre conformation. Accordingly, the Pre to Tachy cis-trans photoisomerization proceeds via a conformer specific one-bond-twist (OBT) process as proposed by Havinga.2-4 The role of the

ACS Paragon Plus Environment

1


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 40

EPA glass in controlling conformer distributions and reaction outcomes is further explored by the extension of the studies to Lumi, whose structure differs substantially from that of its stereoisomer, Pro. Initially, the light induced conrotatory ring openings of Pro and Lumi are expected to give cZc-Pre conformers that differ in the relative orientation of the double bond dihedral angles that define the chiral axis of the triene moiety: (-)cZ(-)c-Pre and (+)cZ(+)c-Pre, respectively. In the case of Pro, much of the cZc-Pre proceeds to tZc-Pre, the precursor of tEcTachy. In contrast, we show that under the same conditions most cZc-Pre formed from Lumi retains the cZc-conformation and isomerizes to cEc-Tachy. cZc-Pre from Lumi was not detected by fluorescence, but UV absorption measurements establish its formation as an essential intermediate to Tachy. Aided by theoretical calculations of conformer UV and CD spectra, we conclude that fluorescent thermodynamic Pre and non-fluorescent Pre from Lumi are both (+)cZ(+)c-Pre conformers. They differ in the orientation of the OH in the A ring, pseudoequatorial in the former and pseudoaxial in the latter. The most likely major photochemical sequences starting from Pre and Lumi are (+)cZ(+)c-Pre-eq-OH  (+)cE(+)cTachy-eq-OH and Lumi  (+)cZ(+)c-Pre-ax-OH  (+)cE(+)c-Tachy-eq-OH.

INTRODUCTION Havinga’s principle of nonequilibration of excited rotamers (NEER) is based on the reversal of double/single bond order on excitation of polyenes and related molecules to lowest excited states.5,6 The NEER principle explains the excitation wavelength (λexc) dependence of previtamin D (Pre) and tachysterol (Tachy) photochemistry, Scheme 1.5-10 The (+)/(-) designations in the Scheme give the sense of the chirality axis defined by the triene moiety. Following the rules of Klyne and Prelog: plus (+) refers to clockwise front to back double bond dihedral angle deviation


2

Page 3 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


from 0o being planar s-cis, minus (- to counterclockwise deviation.11 Conformers have different absorption spectra allowing for different excited conformer compositions to be accessed by varying λexc. Control of photoresponses, including product yields, is afforded because light absorption converts freely interconverting conformers in the ground state to distinct excited isomers. A key experiment Scheme 1. Pre photoreactions in solution.

confirming the NEER principle was conducted by

Havinga and co-workers.2-4 It showed that excitation of the equilibrium conformer mixture of Pre in an EPA (diethyl ether, isopentane, ethanol in 5:5:1 ratio by volume) glass at 92 K gives a weakly structured Tachy UV spectrum, which, on warming to 100-105 K in the dark and recooling to 92 K, develops a much better resolved vibronic progression. Assuming that the tEcTachy conformer is favored at thermal equilibrium, they assigned their initial spectrum to an scis,s-cis-conformer (cEc-Tachy) and reasoned that it derived from an excited s-cis,s-cisconformer of Pre (cZc-Pre).2-4 By irradiating Pro in EPA (usual 5:5:2 ratio) at 90 K, Fuß obtained Tachy by a Pro → Pre → Tachy two-step photochemical sequence.12 On the assumption that irradiation of Pro in the viscous medium would ensure retention of the nascent cZc-Pre geometry, Fuß and coworkers proposed the sequence: Pro → cZc-Pre → tEc-Tachy because the spectrum of the latter resembles the spectrum of Havinga’s thermally equilibrated Tachy. Havinga’s results were then reinterpreted by assuming that tZc-Pre rather than cZc-Pre is the thermodynamically favored Pre conformer. Because the proposed Pre → Tachy reactions involve both conformational and configurational isomerization, Fuß presented them as the first examples


3


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 40

of products predicted by Liu’s Hula-twist (HT) cis-trans photoisomerization mechanism for volume constraining media.13,14 Wide acceptance of the Fuß interpretation accounts for the ensuing avalanche of claims of cis-trans HT photoisomerizations involving simultaneous rotation in S1 of two adjacent bonds in a polyene chain.15-17 While there is general agreement that tEcTachy is the most stable Tachy conformer, as assumed by Havinga and Fuß, theoretical calculations at different levels of theory predict complex mixtures of Pre and Tachy conformers in close energetic proximity in the gas phase.18-25 In addition, the crucial role of the medium in determining conformer distributions was already established by comparing the behavior of Pre in EPA glass to its behavior in a hydrocarbon glass.2-4 Because conformers have different absorption and fluorescence spectra, the ability to change the distribution of conformer responses by varying excitation and monitoring wavelengths (exc and mon) renders fluorescence spectroscopy with the aid of singular value decomposition with self-modeling (SVD-SM) analysis much more powerful in resolving conformer specific responses than UV-vis spectroscopy. In previous work we have provided descriptions of the mathematical operations involved in SVD-SM and in the related principal component analysis with self-modelling26-28 that was first described by Lawton and Sylvestre.29,30 Briefly, treatment of a spectral set as a matrix of spectral vectors ideally yields a set of n orthonormal eigenvectors, where n is the number of pure spectra that contribute to the experimental spectra (the rank of the matrix). The experimental spectra with reduced random noise can then be faithfully reproduced as linear combinations of the eigenvectors using an easily derived set of combination coefficients. Self-modelling involves the search of the pure component combination coefficients in the space of the combination coefficients of the experimental spectra.


4

Page 5 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


We tested the claim that the Tachy products obtained on irradiation of Pro and Pre D3 in EPA are those predicted by the HT motion using fluorescence spectroscopy.1 We found that equilibrated Tachy exists as a mixture of three conformers in glassy EPA at 77 K, the major of which is tEc-Tachy. In contrast, Pre exists exclusively as an s-cis,s-cis-conformer (cZc-Pre) and undergoes cis → trans photoisomerization to cEc-Tachy.1 To our surprise, light induced ringopening of Pro gives three Pre conformers, the major contributor to the subsequent reaction to Tachy being tZc-Pre instead of the expected cZc-Pre. SVD-SM curve resolution yielded the fluorescence spectra of the Pre conformers and of their Tachy products.1 Structures were assigned based on experimental and theoretical evidence showing that s-cis,s-cis-conformers of Pre and Tachy absorb UV light to the red of s-trans-s-cis-conformers. tZc-Pre and cZc-Pre photoisomerizations in EPA proceed by the OBT mechanism to give tEc-Tachy and cEc-Tachy, respectively.1 In this regard, the importance of the earlier photochemical and spectroscopic observations in solution and in EPA glass in guiding the structure assignments cannot be overstated. Havinga’s NEER principle was based on the recognition that Pre exists as a mixture of conformers in solution whose selective excitation controls the composition of the Pro/Lumi/Tachy photoproduct mixture, Scheme 1.5,6 The systematic quantitative studies of Dauben and coworkers showed that excitation of Pre in ether at 0 oC in the 285-295 nm region, where the Pre UV spectrum shows significant absorption, gives Tachy as the main photoproduct,

Tachy = 0.42.7-9 Only on excitation at the onset of the UV spectrum in the 305-325 nm region does photocyclization to Lumi and Pro gradually overtake cis-trans photoisomerization to Tachy.7-9 A roughly 3:1 Lumi:Pro ratio is maintained throughout the entire exc range with Lumi ≈0.05 for exc = 285-300 nm, increasing to 0.21 at 325 nm and Pro ≈ 0.02 for exc = 287-300 nm increasing to 0.08 at 325 nm. The logical explanation of these observations is that the UV


5


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 40

absorptions of the minor cZc-Pre conformers that are responsible for photocyclization extend to the red of the UV absorption spectrum of the dominant tZc-Pre conformer that gives Tachy. Moreover, the preference of Lumi over Pro formation suggests that (+)cZ(+)c-Pre is about three times more abundant than (-)cZ(-)c-Pre. Analogous results were obtained for Tachy in fluid solution.10 The Tachy  Pre quantum yield increases from 0.1 for exc = 254 nm to 0.4 for exc = 313 nm. Based on the theoretical conformer distribution and the red shifted UV spectrum,18-25 it was concluded that preferential excitation of cEc-Tachy at 313 nm gives Pre 4 times more efficiently than preferential excitation of tEc-Tachy.10 The efficient formation of Pre from cEcTachy was predicted by recent non-adiabatic molecular dynamics calculations.25 Starting from Pro in glassy EPA, the remarkable result is that following its ring opening to cZc-Pre, much of it proceeds to tZc-Pre (Scheme 2). The reasonable expectation that the viscous medium would trap the initial cZc-Pre geometry is not borne out. It is reassuring that the most advanced theoretical calculations on the Pro ring Scheme 2. Ring-opening of Pro in EPA at 77 K.

opening to date had predicted the cZc-Pre

(S1) to tZc-Pre (S0) isomerization.31 Excited Pro should initially undergo conrotatory electrocyclic ring opening to give (-)cZc(-)-Pre. The failure of the highly viscous medium32 to trap the initial cZc-Pre ring opening product from Pro indicates that the cavity afforded to Pro by EPA is large enough to permit the cZc- to tZc-Pre transformation. We wondered whether the cavity afforded to Lumi, the stereoisomer of Pro, under the same conditions, would also allow an


6

Page 7 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


s-cis-  s-trans-Pre conformer transformation. Figure 1 shows B3LYP/6-311+G(d,p)-optimized structures of Pro and Lumi. Obvious differences are in the C9C10 junctions, in the orientation of the 3-OH which is pseudoequatorial in Pro but pseudoaxial in Lumi, and the C ring which

Pro

Lumi

Figure 1. B3LYP/6-311+G(d,p) optimized structures of Pro and Lumi. The hydrophobic surfaces shown were calculated using the Chimera program.33 adopts a chair conformation in Pro but is a twist-boat in Lumi. Differences in calculated S0 molecular volumes for the expected initial ring opening reactions, Pro*  (-)cZ(-)c-Pre and Lumi*  (+)cZ(+)c-Pre, are substantial. The changes, 317 to 361 cm3/mol for Pro and 329 to 350 cm3/mol for Lumi, suggest that Lumi, beginning from a larger molecular volume and requiring a smaller volume change to reach its Pre product, may be less influenced by the medium than Pro. We now show that cEc-Tachy is the major photoproduct from Lumi. In contrast to Pro, the EPA medium effectively traps cZc-Pre conformers from Lumi.

EXPERIMENTAL SECTION Materials. Cholesta-5,7,9(11)-trien-3-ol (trienol) contained as a minor, highly fluorescent impurity in 7-dehydrocholesterol (Pro, United States Biochemical Corp.) was destroyed by irradiation at 313 nm in ethanol (absolute anhydrous, 200 proof, ACS/USP grade, PharmcoAAPER). A Hanovia medium pressure 450-W Hg lamp was used and the 313 nm filter solution has been described.33 At 313 nm Pro is almost transparent but the trienol absorbs strongly. The


7


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 40

two step 254/313 or >300 nm (Ace Glass 12-W low pressure Hg lamp through quartz, followed by Hanovia medium pressure 450-W Hg lamp through Pyrex) irradiation sequence of trienol-free Pro in distilled tetrahydrofuran (Sigma-Aldrich, ACS grade) gave Pre,10 which on further irradiation was converted to a Lumi rich photoproduct mixture.7-9 Lumi and Pre were purified by semi-preparative high performance liquid chromatography (HPLC). Dichloromethane, isooctane and ethyl acetate (Aldrich, HPLC grade, ≥99%), and ethanol (absolute anhydrous, 200 proof, ACS/USP grade, Pharmco-AAPER) were used as received. THF (Fisher Scientific, ACS grade) was freshly distilled prior to use. Irradiation Procedures. Irradiations were carried out directly in the fluorimeter using its 150 W Xe lamp, or, externally, using low pressure or medium pressure Hg lamps.1 Samples were in quartz NMR tubes (4.2 mm i.d. Norell, S-5-600-QTZ-8) immersed in liquid N2 in the quartz Dewar of the phosphorescence accessory of the fluorimeter. Analytical Procedures. UV-vis spectra were measured at room temperature using a Varian Cary 300 Bio spectrophotometer. Fluorescence spectra were recorded at 77 K with a Hitachi F4500 spectrofluorometer (excitation and emission slits 2.5 nm, scan speed 240 nm/min, PMT voltage 700 V) and are not corrected for instrumental nonlinearity. Reaction progress was followed in situ by fluorescence spectroscopy. HPLC measurements were performed on a Beckman Coulter HPLC (PDA UV-Vis detector, module 168). Analytical measurements were carried out on a BHK SAL-Silica column (250 x 4.6 mm), flow rate: 1 mL/min, Vinj = 20 μL. Lumi used in initial fluorescence measurements was isolated with 99.75% CH2Cl2/0.25% isopropanol as eluent on a BHK Silica column (20 x 250 mm) provided with a BHK Silica guard column (250 x 10 mm), flow rate: 7 mL/min, Vinj= 1 mL. Lumi used in UV absorption experiments was isolated using 15% ethyl acetate/85% hexanes (both solvents Sigma-Aldrich,


8

Page 9 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


HPLC grade, ≥99%) as eluent on the SAL-Silica column (250 x 4.6 mm), flow rate: 3 mL/min, Vinj= 0.5 mL. Calculations. Pre-treatment of spectra and SVD-SM calculations were performed in the MATLAB 7.12.0.635 (R2011a) environment using standard routines.35 For self-modelling calculations, spectra were normalized to unit area. Quantum mechanical calculations were carried out using the Gaussian 09, Revision A.02 system36 utilizing gradient geometry optimization.37-39 All geometries were fully optimized at the B3LYP 6-311+G(d,p) basis set level, except as otherwise indicated.40-43

RESULTS AND DISCUSSION Fluorescence Observations. Successive fluorescence spectra starting with Lumi in EPA at 77 K were measured in the fluorometer using exc = 290 nm, Figure 2a. For every tenth spectrum,

exc = 310 nm was employed, Figure 2b. Curve resolution of the matrix was achieved by SVDSM.26-28 The initial spectrum in Figure 2a is assigned to Lumi and, as expected from the Lumi absorption spectrum, is much less structured than the fluorescence spectrum of Pro under the b

aa

Figure 2. a) Fluorescence spectra from the irradiation of Lumi excited and monitored with λexc = 290 nm and b) every tenth fluorescence scan monitored with exc = 310 nm. Tachy formation is reflected in increased fluorescence intensity with irradiation time.


9


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 40

same conditions. The analogous 290/310 nm spectral matrix from Pro is a seven component system (Pro, and three Pre/Tachy pairs).1 Two of the three conformers of Tachy that contribute to fluorescence in EPA at 77 K, initially designated TachyA and TachyB, were assigned tEc- and cEc-Tachy structures and the third, TachyC, was tentatively assigned as cEt-Tachy1 (reassigned to a different cEc-Tachy conformer below). In contrast to Pro, the Lumi spectra in Figure 2 define a 3 component system. The Lumi spectrum gives one triangle corner in combination coefficient space. The other two corners were identified by inserting each of the three previously resolved Tachy conformer spectra,1 in turn, into the matrix, Figure 3b. The rank of the matrix remains 3 on inclusion of the spectra of cEc-Tachy (the photoproduct of pure Pre) and TachyC,1 but increases to 4 when the tEc-Tachy spectrum is included. The Lumi photoproducts give mainly cEc-Tachy emission. In contrast to Pro, Lumi does not give tEc-Tachy. Varying λexc, after Lumi conversion to Tachy, gave the spectra in Figure 3a whose inclusion in the matrix did not augment its rank. a

b

Figure 3. a) Fluorescence spectra of Tachy from Lumi as a function of λexc in the 270-320 nm range and b) the combination coefficients for the spectra in Figures 2a (●), 2b (●) and 3a (●); see text for corners. Lumi (♦), cEc- (♦) and TachyC (♦) define a triangle in eigenvector combination coefficient space (Figure 3b). The spectra in Figure 2a begin at the Lumi corner of the triangle and progress


10

Page 11 of 40

with time toward a combonation coefficient on the Tachy edge that is close to the cEc-Tachy corner. The spectra in the inset of Figure 3a fall on the cEc-/TachyC edge of the triangle progressively shifting toward the TachyC corner as exc is increased. We expected the spectra recorded at 310 nm, where Lumi is transparent, to show the Pre  Tachy steps, as observed for Pro.1 Instead, the 310 nm spectra in Figure 2b define the cluster of red points on the Tachy edge. We also carried out the Lumi  Tachy photoreaction using λexcs in the 280-300 nm range. Combination coefficient trajectories vary slightly with longer λexc favoring TachyC, but in the main, all trajectories approach the cEc-Tachy corner of the normalization triangle (see Figure S1 in Supporting Information). Differences in fluorescence excitation spectra measured in the course of the photochemical ring openings of Pro and Lumi are striking, Figure 4. The higher Pro fluorescence quantum yield, aa

b

150 Intensity →

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


100 50 0

280

300

320 λ/nm →

340

Figure 4. Fluorescence excitation spectra, mon = 370 nm, for 290 nm Pro and Lumi photoisomerization in a) and b) panels, respectively. Tachy formation is reflected in increased fluorescence intensity with irradiation time.

f ~ 0.2,44 Figure 4a, relative to Lumi, Figure 4b, and the better resolved vibronic progression in the Pro fluorescence spectrum account for initial differences in appearance. However,


11


persistence of vibronic structure in the Tachy products from Pro and diminished structure as Tachys form from Lumi are consistent with tEc-Tachy 1 formation from Pro and not from Lumi. In contrast to observations starting from Pre or Pro,1 there is no discernible fluorescence from Pre conformers formed from Lumi. A possible explanation for this absence of Pre fluorescence is adiabatic excited Pre formation followed by its efficient conversion to Tachy: Lumi*  Pre*  Tachy. This requires that Pre* from Lumi* be predisposed to give Tachy and thus differ from fluorescent Pre* formed by direct excitation of Pre. Another possibility is that the Pre conformers formed from Lumi do not fluoresce under our excitation conditions. UV Spectra. To determine whether there is a competing path of ground state Pre formation followed by its photoconversion to Tachy, we monitored the reaction using UV spectroscopy. UV absorption spectra measured in the course of the 290 nm irradiations of Pro and Lumi in EPA at 77 K are shown in Figure 5. Pro and Lumi irradiations at 254 nm give similar spectral evolutions to those in Figure 5 (see Figure S3 for Lumi). Comparison of Figures 4 and 5 reveals that there is an initial decrease in absorbance, more clearly evident for Lumi that is followed by progressively increased absorbance due to Tachy formation. However, whereas the fluorescence

a Absorbance →

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 40

b

2

1.5 1

0.5 0

240

260

280 300 λ/nm →

320

340

Figure 5. Photoisomerization of a) Pro and b) Lumi with exc = 290 nm. The bold black spectra in the two panels are Pro and Lumi.


12

Page 13 of 40

excitation and UV absorption spectra in the Pro reaction bear the expected strong resemblance, the same does not hold for Lumi. Missing from the Lumi fluorescence excitation spectra is the vibronic structure that develops in the UV spectra of the Tachy products from Lumi. The evolution of fluorescence spectra, starting from Lumi, is accounted for entirely by cEc-Tachy and TachyC formation, Figure 3b. Since the cEc-Tachy spectrum is almost structureless, Figure 7a, the vibronic structure in Figure 5b is that of TachyC. Loss of intensity in the initial product spectra in Figure 5b shows that Pre forms from Lumi as a ground state intermediate. Following the initial drop in absorbance due to ground state Pre formation, the absorbance of the UV spectra in Figure 5b increases signalling formation of the Tachy products. SVD analysis of the spectra in Figure 5b reveals a four component matrix three

a

b

3 Absorbance →

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


2.5 2

1.5 1

0.5 0

240

260

280 300 λ/nm →

320

340

Figure 6. a) Molar absorptivity spectra of Pro (black), Pre (green) and Lumi (blue) in EPA at 298 (dashed) and 77 K (solid). b) Photoisomerization of Pre, λexc = 254 nm, Pre, the Tachy 77 K photoproduct at the end of this reaction and the thermal Tachy mixture are shown in bold green, black, and red, respectively. of which must be the spectra of Lumi, cEc-Tachy and TachyC and the fourth is either a Pre conformer or a mixture of Pre conformers. The spectrum of Lumi is the 0-time spectrum in


13


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 40

Figure 5b. UV spectra of pure Pro, Lumi and Pre samples in EPA at 298 and 77 K are shown in Figure 6a. The 77 K molar absorptivities in Figure 6a are not corrected for the 22.9% contraction of EPA on cooling from room temperature to a glass at 77 K,45 accordingly, to obtain actual molar absorptivities at 77 K the values in Figure 6a must be multiplied by 0.771. In addition, the UV spectra are not corrected for EPA background absorption. There was some variability in consecutive measurements of the baseline (see Figure S2 in the Supporting Information) and we wished to avoid introduction of systematic errors stemming from the use of the identical baseline for all spectra. The evolution of the fluorescence spectra obtained on excitation of Pre in EPA at 77 K was assigned exclusively to the cZc-Pre cEc-Tachy photoreaction.1 Concomitant changes in UV spectra during the course of this reaction are complete in minutes, Figure 6b, and readily yield the UV spectra of cZc-Pre andcEc-Tachy. The initial cEc-Tachy UV spectrum is almost structureless but on prolonged irradiation it develops the weak vibronic structure that was

a

b

Figure 7. a) Absorption spectra of Lumi (blue), Pre (green), cEc-Tachy (purple) and TachyC (red). b) Combination coefficient trajectory of the spectra in Figure 5b; diamonds defining the corners of the tetrahedron are color coded as in a).


14

Page 15 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


observed by Havinga and co-workers.2-4 Use of the UV spectra of Lumi andcEc-Tachy as two of the four components in the spectral matrix in Figure 5b and substituting the UV spectrum of cZcPre for Pre formed from Lumi, allows an approximate SVD-SM resolution of the spectral matrix by limiting it to the identification of the UV spectrum of TachyC. Use of the cZc-Pre spectrum as a pure component in the matrix is an approximation because it is fluorescent,1 whereas Pres formed from Lumi were not detected by fluorescence. Below, we rely on calculated UV spectra to conclude that the UV spectrum of the major Pre conformer from Lumi is expected to be significantly blue shifted relative to the spectrum we employed in Figure 7a. Normalized UV spectra of the four components are shown in Figure 7a. They define the corners of the tetrahedron in Figure 7b. The blue points in that Figure show an approximate trajectory of the combination coefficients of the spectra in Figure 5b. Beginning at Lumi (♦), the reaction proceeds rapidly toward Pre (♦) and then turns toward a cEc-Tachy/TachyC (♦/♦) mixture that favors cEc-Tachy, but gradually become richer in TachyC. When exc = 254 nm is employed, where strong absorption from Pre is expected, the trajectory turn from Pre toward the Tachys occurs faster (see Figure S4 in the Supporting Information). A significant difference between the Pro and Lumi reaction sequences is the much larger initial Pre contribution in reaction mixtures starting from Lumi than starting from Pro. The derived TachyC absorption spectrum is obviously a combination of at least two transitions: a weak structureless transition with an onset at ~330 nm and an allowed transition whose vibronic structure coincides almost exactly with that of tEc-Tachy (see the spectrum of thermodynamic Tachy in Figure 6b in which tEc-Tachy is the major component). Both transitions could belong to a single conformer or each could be due to different Tachy conformers. Assuming that a single conformer is involved, the long wavelength onset of the UV


15


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 40

spectrum of TachyC is consistent with the fact that its fluorescence is preferentially excited by employing exc in the longer wavelength region. In agreement with the trajectory of the fluorescence spectra in Figure 3b, the trajectory of the absorption spectra in Figure 7b shows that cEc-Tachy, the sole Tachy product starting from Pre, is the major initial Tachy product starting from Lumi. The vibronic structure of the strong transition in the absorption spectrum of TachyC is absent in the fluorescence excitation spectra in Figure 4b. That apparent discrepancy is resolved if TachyC fluorescence occurs only following direct excitation into the forbidden absorption band. Assuming that the forbidden and allowed transitions in TachyC correspond to excitation into A- and B-like polyene states, respectively, this explanation finds strong precedent in low T observations on Pre (note that forbidden polyene doubly excited A states gain transition probability by vibronic coupling with the allowed singly excited B states46-48).9 The large discrepancy between Pre absorption and fluorescence excitation spectra was attributed to fluorescence exclusively, from a forbidden A-like lowest excited state, not accessed from the allowed B-like excited state.9 Structure Assignments. Reminiscent of observations on biphenyl,49-51 the sharpening of the vibronic progression in the UV spectrum of the Tachy product from Pre in Figure 6b could be due to an annealing of the initial cEc-Tachy structure as it assumes a more planar form. Alternatively, it could be due to a hot ground state reaction leading to a different Tachy conformer.52,53 To distinguish between these two possibilities, we measured fluorescence spectra of the final Tachy product in Figure 6b by varying λexc from 250 – 320 nm in 10 nm increments, Figure 8a. SVD treatment of the spectral matrix in Figure 8a reveals a two component system that retains the rank of 2 when the fluorescence spectrum of TachyC1 is included in the matrix, Figure 8b, but increases to rank 3 when it is replaced by the fluorescence spectrum1 of tEc-


16

Page 17 of 40

Tachy. Longer λexc favor TachyC, such that λexc = 320 nm gives almost pure TachyC emission (yellow spectrum in Figure 8a, note the blue circle point that overlaps the black diamond in

a

−3

x 10

b

−3

10

2

x 10

5 β

Normalized Intensity →

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1

0 300

0

350

400 450 λ/nm →

500

−5 −0.0425

−0.042

−0.0415 α

−0.041

−0.0405

Figure 8. a) Photoisomerization of Tachy from Pre, λexc = 250-320 nm. b) Normalization line for the spectra in 7a, the red and black diamonds are the combination coefficients of the spectra of cEc-Tachy and TachyC, respectively. Figure 8b). It follows that slow cEc-Tachy TachyC conversion occurs, probably as a hot ground state process, analogous to the photoconversion of s-trans- to s-cis-1,3-butadiene.52,53 The fact that those two Tachys are common products from Pro and Lumi is significant. Molecular orbital symmetry considerations predict that the conrotatory ring openings of Pro and Lumi should give (-)cZ(-)c-Pre and (+)cZ(+)c-Pre, respectively.54 Those Pre conformers differ in the helicity of the triene moiety and are expected to photoisomerize to two Tachy conformers, namely, (-)cE(-)c-Tachy and (+)cE(+)c-Tachy. (-)cE(-)c-Tachy and (+)cE(+)c-Tachy are likely candidates for cEc-Tachy and TachyC, but not necessarily in that order. Reversal of double/single bond character on excitation could provide the driving force for their torsional interconversion, consistent with the principle of least motion. Assignment of TachyC as one of the s-cis,s-cisconformers requires re-examination of the conclusion that the EPA medium fails, in large part, to trap the s-cis,s-cis-Pre conformer from the ring opening of Pro.1 First, the vibronic bands in the


17


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 40

UV spectrum of TachyC coincide with those of tEc-Tachy and second, comparison of Figures 4b and 5b shows that excitation into those bands yields no fluorescence. Therefore, in addition to tEc-Tachy, a major Tachy product from the ring-opening of Pro in glassy EPA is probably a cEcTachy conformer and has (-)cZ(-)c-Pre as its precursor. Calculated structures of (-)cZ(-)c- and (+)cZ(+)c-Pre and the (-)cE(-)c- and (+)cE(+)c-Tachy conformers in which a methyl group was used at C17 instead of the alkyl group are shown in Figure 9. It follows that the expectation that the EPA medium will trap the cZc-Pre ring opening photoproduct,12 is realized for Lumi and is also realized, to a significant extent, for Pro. In both cases, Pre OBT photoisomerizations give the same cEc-Tachy conformers. A partial sequence of initial events is shown in Scheme 3. For Lumi, the second photon in Scheme 3 is required, because, the UV spectra establish that ground state Pre is a necessary intermediate to Tachy formation. The major pathway from Lumi to Tachy follows ring opening to c(+)Zc(+)-Pre that may or may not be excited. For Pro, Scheme 3 is an elaboration of the minor path in Scheme 2 with the caveat that it is more important than was revealed by our fluorescence measurements. a

b

c

d

Figure 9. Top views (left) and central double bond eclipsed views (right) of calculated structures for (-)cZ(-)c-, (+)cZ(+)c-Pre, (-)cE(-)c- and (+)cE(+)c-Tachy are shown in a), b), c), d), respectively. All structures are with CH3 at C17.


18

Page 19 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Provided below is an estimate of the relative contribution of the Tachy conformers that are obtained from Pro. Figure 10a shows spectra measured in Scheme 3. Initial OBT photoisomerizations starting from Pro and Lumi.

the course of 254 nm

irradiation of Pro in EPA at 77 K. They are analogous to those shown in Figure 5a for exc = 290 nm, except that the reaction in Figure 10a was taken to completion. Molar absorptivities for the different Tachy conformers were obtained by projection of SVD-SM derived normalization

a

b

Figure 10. a) Photoisomerization of Pro with exc = 254 nm. The bold black spectrum is that of Pro and b) molar absorptivity spectra in EPA at 77 K of Pre (green), cEc-Tachy (purple) and TachyC (red). Extinction coefficients in b) must be multiplied by 0.771 due to EPA contraction.


19


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 40

spectra to the corresponding stoichiometric surface.55,56 This is illustrated in Figure 10b which shows molar absorptivity UV spectra for cZc-Pre, cEc-Tachy and TachyC in EPA at 77 K, obtained by projection of the normalized spectra to the stoichiometric plane defined by the spectra from the irradiation of Pre, Figure 6b. As in Figure 6a these UV spectra are not corrected for EPA background absorption and for contraction on cooling from room temperature to the EPA glass at 77 K.45 Corrected values of actual molar absorptivities at max are 19,660, 11950, 12640, 8590 and 6150 M-1cm-1 for Tachy A, B, C, Pro and Lumi, respectively. The ratio of the final to the initial maximum absorbance in the Pro irradiation in Figure 10a is 1.88 and the max molar absorptivity ratios of TachyA/Pro, TachyB/Pro and TachyC/Pro are 2.29, 1.39 and 1.47, respectively. Neglecting, initially, the contribution of TachyB, if x = fraction of TachyA, then 2.29x + 1.47(1-x) = 1.88. This gives x = 0.50. The absorbance of the Tachy product mixture at 304 nm where only TachyB and TachyC absorb (molar absorptivities of 7,920 and 2,360 M-1cm-1, respectively) corresponds to an effective molar absorptivity of 2,550 M-1cm-1, which is much larger than half the molar absorptivity of TachyC. It follows that the contribution of TachyB is not negligible. The absorbance changes in Figure 10a indicate that ~50% of the photoproduct from Pro is tEc-Tachy and of the remaining 50%, ~30% is the cEc-Tachy conformer that corresponds to TachyC and ~20% is the cEc-Tachy conformer that corresponds to TachyB. The large TachyC contribution in the Pro reaction mixtures was also evident in the analysis of the fluorescence spectra that was provided as Supporting Information for ref 1. In view of significant conformational interconversion evident in the formation of tEc-Tachy from Pro, Scheme 2, we emphasize that the initially expected +/- chirality designations in Scheme 3 may not correspond to those of the final intermediates and products.


20

Page 21 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Theoretical Considerations. Theoretical conformational abundances of model compounds of Pre were evaluated by one of the authors using molecular mechanics (MMX),19 and semiempirical methods (AM1 and QCFF),21 HF/6-31G(d) and B3LYP/6-31G(d).22,23 Analogous calculations were also performed on the conformational abundance of Tachy.21,24 In view of considerable theoretical advances since then, we thought it prudent to repeat those calculations. The Pre conformational landscape was examined using B3LYP/6-311+G(d,p) on Pre D3 itself, and those results were compared to predictions using wB97XD/6-311+G(d,p) and M06-2X/6311+G(d,p) gas phase calculations that include dispersion force corrections. Pre conformer excited state properties were calculated using TD CAM-B3LYP/6-31G(d)//B3LYP/6311+G(d,p). Tachy calculations were performed at the B3LYP/6-31G(d) level with TD CAMB3LYP/6-31G(d) for excited state calculations. The primary motivation for this work is the use of theoretical conformer UV and CD spectra as guides to structural assignments to conformers with known experimental spectra. Results for Pre are summarized in Table 1. The shortcomings of DFT B3LYP calculations to accurately predict relative hydrocarbon energy differences have been highlighted in several recent publications.57,58 In our experience, very good agreement between experimental results and DFT B3LYP theoretical predictions was realized for conformer and isomer energetics of a simple butadiene,59 diphenylpolyenes,60-62 and 1,2-diarylethenes35,63,64 in their S0 and T1 states. The calculations predict a plethora of Pre and Tachy conformers in close energetic proximity. As before, conformers can be grouped in two families based on the orientation of the 3--OH group, pseudoequatorial vs pseudoaxial in half-chair conformations of the A-ring.23 The B3LYP calculations in Table 1 are in reasonable agreement with those in ref 23. Substitution of Me for the alkyl group at C17 does not significantly alter conformer geometries or distributions.


21


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 40

Predicted conformer distributions are consistent with the interpretation of the exc dependence of photoisomerization and photocyclization quantum yields in ether that reveal mainly photoisomerization to Tachy, generally attributed to tZc-Pre conformers, and the presence of cZc-Pre conformers in low abundance, whose selective excitation at the red edge of the Pre Table 1. Pre Conformational Distributions in % Using Different Theoretical Approaches Conform er

B3LYP/ 6-31G* (%)a

B3LYP/ 6311+G( d,p), gas phase

B3LYP/6311+G(d,p ), gas phase, R = Me

CAMB3LYP/631G(d)//B3 LYP/6311+G(d,p) & scrf = EtOH. cpcm

wB97XD /6311+G(d, p), gas phase

M062X/ max/n 6m (sign 311+G( CE) d,p), gas phase, R = Me

3-OH-eq (-)cZ(-)c

3.6

2.70

2.97

3.37

0.963

2.42

263 (-)

(+)cZ(+)c 12.3

12.7

12.3

14.2

1.72

1.48

266 (+)

(-)tZ(+)c

29.9

35.0

31.6

25.74

1.15

0.49

262 (-)

(+)tZ(-)c

24.8

18.0

20.5

14.7

0.82

0.76

271 (+)

(-)cZ(+)t

2.04

2.63

2.96

18.88

10.1

251 (+)

0.02

0.02

0.058

0.84

0.21

221 (-)

(+)cZ(-)t

3-OH-ax (-)cZ(-)c

a

2.1

2.6

2.86

3.9

0.96

1.19

273 (-)

(+)cZ(+)c 2.1

2.1

2.1

4.3

2.2

2.6

255 (+)

(-)tZ(+)c

6.6

13.1

13.1

7.8

0.67

0.59

267 (-)

(+)tZ(-)c

18.6

9.6

9.3

19.5

40.5

46.3

262 (+)

(+)cZ(-)t

0.003

0.003

0.038

0.68

0.29

220 (-)

(-)cZ(+)t

2.15

2.51

3.48

30.8

35.7

257 (+)

From ref 23.


22

Page 23 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


absorption spectrum gives Pro and Lumi.7-9 A good measure of the predicted Pre conformer geometries at the different levels of theory are the magnitudes of the dihedral angles between the three double bond planes (D1 and D2, A ring to central and C ring to central double bond, respectively). The plot of D2 vs D1 shows significant variation between conformers, but very good agreement between theoretical approaches for each conformer (see Figure S7 in the Supporting Information for the D1/D2 plot). The predicted major role of (-)cZ(+)t-Pre conformers, especially for the axial 3--OH case, in the 6th and 7th columns of Table 1 may indicate that inclusion of dispersion forces in such gas phase calculations overestimates the importance of specific intramolecular CH/C=C and/or OH/C=C interactions. A more detailed comparison of results obtained using different computational methods, including the ab initio MP2/6-31G(d) approach, is provided in the Supporting Information (see Tables S1-S3 and Figures S8 and S9). Major contributing conformers are highlighted in red in Table 1. Predicted conformer distributions for Tachy are shown in Table 2. Previous results based on force field calculations are included for comparison.24 Major contributing conformers are again Table 2. Tachy Conformational Distributions Using Different Theoretical Approaches Conformer

MMX and B3LYP/6MMP2, gas 31G(d) (%), phase, R = Me a gas phase, R=Me

CAM-B3LYP/6max/nm (sign CE) 31G(d)//B3LYP/631G(d) & scrf = EtOH. cpcm , R=Me

3-OH-eq (-)cE(-)c

3

0.08

0.16

280

(+)cE(+)c

1

0.27

0.55

281 (-)

(-)cE(+)c

1

0.04

0.12

271 (+)

(+)cE(-)c

5

0.1

0.33

277 (-)


23


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 40

(+)tE(-)c

13

15

20.1

288 (+)

(+)tE(+)c

5

12.7

18.3

292 (-)

(-)tE(+)c

5

(+)tE(+)t

8

29.2

28.5

281 (+)

(+)cE(+)t

1

0.35

0.57

276 (-)

No spectrum

3-OH-ax

a

(-)cE(-)c

3

0.24

0.32

280

(-)cE(+)c

1

0.12

0.21

277 (+)

(+)cE(-)c

3

0.03

0.06

271 (-)

(+)cE(+)c

1

0.07

0.094

278 (-)

(-)tE(-)c

12

9.4

8.3

291 (+)

(-)tE(+)c

4

10.6

9.4

290 ‐

(-)tE(+)t

5

20.9

12.0

290 (+)

(+)tE(-)c

10

Not found

Not found

No spectrum

(+)cE(+)t

1

0.1

0.1

271 (-)

(-)cE(+)t

2

0.61

0.53

278

From ref 24.

highlighted in red. A common feature is the predicted prominent role for tEc-Tachy conformers. However, the B3LYP calculations also predict similarly significant contributions from tEt-Tachy conformers. In their original study, Havinga and coworkers carried out the Pre  Tachy photoreaction in EPA at 92 K and measured both UV and CD spectra.2 A positive Cotton effect (CE) was observed for Pre, max = 270 nm, with a slightly negative onset. On excitation it gave Tachy with a strongly negative CE, max = 284 nm. Warming the Tachy product to 105 K followed by


24

Page 25 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


recooling to 92 K gave thermally equilibrated Tachy with a positive CE that shows vibronic structure, max = 283 nm, consistent with the Tachy UV absorption spectrum. We carried out the Pre  Tachy photoreaction in EPA at 77 K, a much more viscous medium.32 Based on the SVD-SM analysis of fluorescence spectra measured in the course of the reaction we concluded that, under our conditions, Pre is a single conformer that on excitation gives a single Tachy conformer. That Tachy conformer, on warming to ambient T and recooling to 77 K, gives the thermodynamic Tachy conformer mixture whose exc dependent fluorescence reveals three components, including a small contribution from the Tachy obtained from Pre.1 Consistent with Havinga’s original assignments,2 the Pre reaction was assigned as cZc-Pre  cEc-Tachy and the major Tachy conformer in the thermodynamic mixture was assigned to tEcTachy.1 The fluorescence excitation and UV absorption measurements reported here are consistent with this interpretation. We rely on theoretical conformer UV and CD spectra as guides in assigning conformer structures with correct helicity to Havinga’s experimental spectra. The theoretical spectra34 for conformers with pseudoequatorial 3--OH are shown in Figure 11. The analogous set of spectra with 3--OH pseudoaxial is Figure S11 in the Supporting Information. The maxs and the signs of the CEs for the lowest energy transitions are given in the last columns of Tables 1 and 2. We can eliminate (-)cZ(-)c-Pre as the structure of the thermodynamic Pre conformer obtained in EPA because it is predicted to have negative CEs for OH axial (ax-OH) or equatorial (eq-OH). The two (+)cZ(+)c-Pres are predicted to have positive CEs with max = 255 nm (ax-OH) and 266 nm (eq-OH). It follows that (+)cZ(+)c-Pre with OH equatorial is a likely candidate as the structure of pseudoaxial is Figure S11 in the Supporting Information. The maxs and the signs of the CEs for


25


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 40

the lowest energy transitions are given in the last columns of Tables 1 and 2. We can eliminate ()cZ(-)c-Pre as the structure of the thermodynamic Pre conformer obtained in EPA because it is predicted to have negative CEs for OH axial (ax-OH) or equatorial (eq-OH). The two (+)cZ(+)cPres are predicted to have positive CEs with max = 255 nm (ax-OH) and 266 nm (eq-OH). It follows that (+)cZ(+)c-Pre with OH equatorial is a likely candidate as the structure of thermodynamic Pre in EPA. Among Tachy conformers with gas phase energies no higher than 2

Figure 11. Predicted conformer UV and CD spectra for Pre in panels a) and b), respectively, and for Tachy in panels c) and d), respectively; the OH is pseudoequatorial throughout.


26

Page 27 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


kcal/mol above the global minimum, strongly negative CDs are predicted only for (+)cE(+)cTachy, max = 278 nm (ax-OH) and 281 nm (eq-OH). Thus the calculations predict that the experimental spectra are most consistent with the (+)cZ(+)c-Pre  (+)cE(+)c-Tachy photoreaction with eq-OH giving a better max match than ax-OH. Examination of Figure 3b shows that the Tachy from Pre is also the major Tachy photoproduct from Lumi. If the axial 3OH of the A ring of Lumi is maintained in Pre, the conrotatory ring opening of Lumi should initially give (+)cZ(+)c-Pre with ax-OH, Figure 9b . The 11 nm blue shift of the calculated max in the UV spectrum of the ax-OH (+)cZ(+)c-Pre conformer relative to the max of the corresponding conformer with eq-OH nicely explains why using exc = 290 nm fluorescence is observed from thermodynamic Pre but not from Pre formed from Lumi in EPA at 77 K. The much lower absorbance at 290 nm coupled with a low Pre fluorescence quantum yield should lead to a much weaker fluorescence response from (+)cZ(+)c-Pre with ax-OH than from (+)cZ(+)c-Pre with eq-OH. The low Pre absorbance is also consistent with the product trajectory in Figure 7b that indicates a significant initial build-up of Pre. We conclude that the most probable sequence for the formation of the major Tachy from Lumi is: Lumi  (+)cZ(+)c-Preax-OH  (+)cE(+)c-Tachy-eq-OH. Because the UV measurements in the course of the Lumi photoreaction reveal substantial ground state Pre formation, a second photon is required for the Pre  Tachy step. In the gas phase, the B3LYP results in Table 2 predict that (+)tE(-)c-, (+)tE(+)c-, (+)tE(+)tTachy with eq-OH and (+)tE(-)c-, (-)tE(+)c- and (-)tE(+)t-Tachy with ax-OH should each contribute significantly (8 – 29%) to the thermodynamic conformer distribution. Of these only (+)tE(-)c- and (+)tE(+)t-Tachy with eq-OH are predicted to have weak positive CEs, but with


27


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 40

max = 288 and 281 nm, respectively, they are about 10 nm to the red of the experimental spectrum.2 Nonetheless, of the tEc-Tachy conformers, (+)tE(-)c-Tachy appears to be the only viable possibility for the major Tachy in the EPA glass. The inescapable conclusion is that the predicted multi-conformer Pre and Tachy distributions in the gas phase are significantly altered by the EPA medium to exclusively a single conformer in the case of Pre and a single major conformer in the case of Tachy. The critical effect of the medium in controlling Pre conformer distribution is further demonstrated by the experimental CD spectrum of Pre in a 1:5 methylcyclohexane to isopentane glass at 92 K.3 Instead of the positive CE that was observed in EPA,2 the CE in the hydrocarbon solvent is strongly negative with max = 267 nm. Examination of the Pre conformer CD spectra in Figure 11a indicates that while EPA glass traps the (+)cZ(+)c-Pre conformer, the hydrocarbon glass traps the (-)cZ(-)c-Pre conformer instead.

CONCLUSION The photochemistry of Pro and Lumi in EPA glass at 77 K was monitored by fluorescence, fluorescence excitation and UV spectroscopy and the spectral matrices were analyzed by SVDSM. Our results and previous observations were interpreted with the aid of theoretical calculations of the structures of Pro, Lumi and the conformers of Pre and Tachy and calculated UV and CD spectra of the photoproducts. The substantial difference in the sequence of events starting from Pro and Lumi can be attributed to large structural differences between the two diastereomers. As previously described, on 290 nm excitation the photochemical ring-opening of Pro gives three fluorescent Pre conformers that undergo OBT cis-trans photoisomerization to give three fluorescent Tachy conformers.1 The Tachy photoproducts include the Tachy formed from Pre as a minor contributor and two major Tachys that are present in similar yields. The


28

Page 29 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


spectra of one of the latter two Tachys are consistent with the major conformer in thermally equilibrated Tachy. Excitation of Lumi gives a much simpler three component fluorescence spectral matrix consisting of Lumi and two of the Tachys observed from Pro. The Tachy formed from Pre is the major Tachy photoproduct from Lumi, but in contrast to Pro, Lumi gives none of the thermodynamic Tachy. Comparison of experimental and theoretical conformer UV and CD spectra leads to the following structural assignments for the Pre photoreaction and for the major photochemical sequences in the Pro and Lumi photoreactions in EPA at 77 K: (+)cZ(+)c-Pre-eq-OH  (+)cE(+)c-Tachy-eq-OH

(1)

Lumi  (+)cZ(+)c-Pre-ax-OH  (+)cE(+)c-Tachy-eq-OH

(2)

Pro  (-)cZ(-)c-Pre-eq-OH  (-)cE(-)c-Tachy-eq-OH

(3)

This interpretation is consistent with the expected conrotatory ring opening reactions and OBT instead of HT photoisomerizations. The calculated UV spectra explain the absence of fluorescence from the Pre intermediate in the Lumi photoreaction sequence. The pseudoaxial OH in Lumi is maintained in the Pre photoproduct leading to an 11 nm blue shift and less effective excitation relative to the corresponding thermodynamic Pre that has a pseudoequatorial OH. A major photoreaction sequence starting from Pro leads to the major Tachy conformer obtained upon rapid cooling of an EPA solution of Tachy to 77 K. This Tachy has a blue shifted Pre as its precursor.1 Its structure assignment is less secure, but if it is one of the tEc- conformers as has been generally assumed, its positive CE indicates that it is (+)tE(-)c-Tachy. The likely photochemical sequence for its formation is given in eq 4. Theory predicts that the Pre conformer isomerization in eq 4 may


29


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Pro  (-)cZ(-)c-Pre-eq- OH  (+)tZ(-)c-Pre-eq-OH  (+)tE(-)c-Tachy-eq-OH

Page 30 of 40

(4)

occur as a hot ground state reaction.31 In this work we observed slow (+)cE(+)c-Tachy-eq-OH  (-)cE(-)c-Tachy-eq-OH conversion on prolonged irradiation of (+)cE(+)c-Tachy-eq-OH. AUTHOR INFORMATION Corresponding Author Jack Saltiel *Email: [email protected] The authors declare no competing financial interests. †

On leave from Faculty of Chemistry and Center of Advanced Technologies, Adam Mickiewicz

University, Umultowska 89 b/c, 61-614 Poznan, Poland. ‡

Department of Chemistry and Biochemistry, University of Delaware, Newark, Delaware 19716.

ACKNOWLEDGMENT This research was supported by the National Science Foundation, most recently by Grant No. CHE-1361962. JS thanks Professor E. F. Hilinski for helpful discussions. REFERENCES (1) Redwood, C.; Bayda, M.; Saltiel, J. Photoisomerization of Pre- and Provitamin D3 in EPA at 77 K: One Bond-Twist Not Hula-Twist. J. Phys. Chem. Lett. 2013, 117, 716-721.


30

Page 31 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(2) Maessen, P. A.; Jacobs, H. J. C.; Cornelisse, J.; Havinga, E. Photochemistry of Previtamin D3 at 92 K – Formation of an Unstable Tachysterol3 Rotamer. Angew. Chem. Int. Ed. Engl. 1983, 22, 718-719. (3) Maessen, P. A.; Jacobs, H. J. C.; Cornelisse, J.; Havinga, E. Photochemistry of Previtamin D3 at 92 K – Formation of an Unstable Tachysterol3 Rotamer. Angew. Chem. Int. Ed. Engl. 1983, 22(Suppl.) 994-1004. (4) Maessen, P. A. The Formation of Toxisterols from Previtamin D – Mechanistic Studies. Ph. D. Thesis, State University at Leiden, 1983. (5) Havinga, E.; Schlatmann, J. L. M. A. Remarks on the Specificities of the Photochemical and Thermal Transformations in the Vitamin D Field. Tetrahedron 1961, 16, 146-152. (6) Havinga, E. Über Einige Photochemische Reactionen. Chimia 1962, 16, 145-172. (7) Dauben, W. G.; Phillips, Wavelength-Controlled Production of Previtamin D3. J. Am. Chem. Soc. 1982, 104, 355-356. (8) Dauben, W. G.; Share, P. E.; Ollmann, R. R., Jr. Triene Photophysics and Photochemistry: Previtamin D3. J. Am. Chem. Soc. 1988, 110, 2548-2554. (9) Dauben, W. G.; Disanayaka, B.; Funhoff, D. J. H.; Kohler, B. E.; Schilke, D. E.; Zhou, B. Polyene 2 1Ag and 1 1Bu States and the Photochemistry of Previtamin D3. J. Am. Chem. Soc. 1991, 113, 8367-8374.


31


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 32 of 40

(10) Saltiel, J.; Cires, L.; Turek, A. M. Conformer-Specific Photoconversion of 25Hydroxytachysterol to 25-Hydroxyprevitamin D3: Role in the Production of Vitamin Ds. J. Am. Chem. Soc. 2003, 125, 2866-2867. (11) Klyne, W.; Prelog, V. Description of Steric Relationships across Single Bonds. Experientia, 1960, 12, 521-568. (12) Müller, A. M.; Lochbrunner, S.; Schmid, W. E.; Fuß, W. Low-Temperature Photochemistry of Previtamin D. A Hula-Twist Isomerization of a Triene. Angew. Chem. Int. Ed. 1998, 37, 505507. (13) Liu, R. S. H.; Asato, A. E. Photochemistry of Polyenes. 22. The Primary Process of Vision and the Structure of Bathorhodopsin - A Mechanism for Photoisomerization of Polyenes. Proc. Natl. Acad. Sci. U.S.A. 1985, 82, 259-263. (14) Liu, R. S. H.; Mead, D.; Asato, A. E. Application of the H.T.-n Mechanism of Photoisomerization to the Photocycles of Bacteriorhodopsin: A Model Study. J. Am. Chem. Soc. 1985, 107, 6609-6614. (15) Liu, R. S. H.; Hammond, G. S. The Case of Medium-Dependent Dual Mechanisms for Photoisomerization: One-Bond-Flip and Hula-Twist. Proc. Natl. Acad. Sci. U.S.A. 2000, 97, 11153-11158. (16) Liu, R. S. H. Photoisomerization by Hula-Twist: A Fundamental Supramolecular Photochemical Reaction. Acc. Chem. Res. 2001, 34, 555-562.


32

Page 33 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(17) Liu, R. S. H.; Hammond, G. S. Hula-Twist: A Photochemical Reaction Mechanism Involving Simultaneous Configurational and Conformational Isomerization. In Handbook of Organic Photochemistry and Photobiology; 2nd Edition, Horspool, W. M.; Lenci, F., Eds., CRC Press: London, 2004, Ch. 26, pp 1-11. (18) Dauben, W. G.; Funhoff, D. J. H. Theoretical Evaluation of the Conformations of Previtamin D3. J. Org. Chem. 1988, 53, 5070-5075. (19) Dmitrenko, O.; Reischl, W. Molecular Mechanics-Based Conformational Analysis of Previtamin D and its A-Ring Analogues. Monatsh. Chem. 1996, 127, 445-453. (20) Cholinski, J.; Kutner, A. Molecular Modeling of Three-Dimensional Structure of 1,25Dihydroxycholecalciferol and its A-Ring Analogs. Polish J. Chem. 1997, 71, 1321-1328. (21) Dmitrenko, O.; Vivian, J. T.; Reischl, W.; Frederick, J. H. Theoretical Studies of the First Strongly Allowed Singlet States of 3-Desoxy Analogs of Previtamin D, Vitamin D, and their EIsomers. J. Mol. Struct. (Theochem.) 1999, 467, 195-210. (22) Dmitrenko, O.; Frederick, J. H.; Reischl, W. Ab Initio Study of Conformational Stability in Previtamin D, Vitamin D and Related Model Compounds. J. Mol. Struct. (Theochem.) 2000, 530, 85-96. (23) Dmitrenko, O.; Frederick, J. H.; Reischl, W. Previtamin D Conformations and the Wavelength-Dependent Photoconversions of Previtamin D. J. Photochem. Photobiol. A: Chem. 2001, 139, 125–131.


33


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 34 of 40

(24) Dmitrenko, O.; Reischl, W. Ground-State Conformational Equilibrium of Previtamin D and its E Isomer: Effect on the Absorption Characteristics. Res. Chem. Intermed. 1997, 23, 691-702. (25) Cisneros, C.; Thompson, T.; Baluyot, N.; Smith, A. C.; Tapavicza, E. The Role of Tachysterol in Vitamin D Photosynthesis – a Non-Adiabatic Molecular Dynamics Study. Phys. Chem. Chem. Phys. 2017, Advance Article DOI. 10.1039C6CP080648 (26) Sun, Y.-P.; Sears, D. F., Jr.; Saltiel, J. Three-Component Self-Modeling Technique Applied to Luminescence Spectra. Anal. Chem. 1987, 59, 2515-2519. (27) Saltiel, J.; Sears, D. F., Jr.; Choi, J.-O.; Sun, Y.-P.; Eaker, D. W. The Fluorescence, Fluorescence-Excitation and UV Absorption Spectra of trans-1-(2-Naphthyl)-2-phenylethene Conformers. J. Phys. Chem. 1994, 98, 35-46. (28) Zimányi, L.; Kulcsár, Á.; Lanyi, J. K.; Sears, D. F., Jr.; Saltiel, J. Singular Value Decomposition with Self-Modeling Applied to Determine Bacteriorhodopsin Intermediate Spectra. Analysis of Simulated Data. Proc. Natl. Acad. Sci. USA 1999, 96, 4408-4413. (29) Lawton, W. H.; Sylvestre, E. A. Self Modeling Curve Resolution. Technometrics 1971, 13, 617-633. (30) Lawton, W. H.; Sylvestre, E. A.; Maggio, M. S. Curve Resolution Using a Postulated Chemical Reaction. Technometrics 1974, 16, 353-368. (31) Tapavicza, E.; Meyer. A. M.; Furche, F. Unraveling the Details of Vitamin D Photosynthesis by Non-Adiabatic Molecular Dynamics Simulations. Phys. Chem. Chem. Phys. 2011, 13, 20986-20998.


34

Page 35 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(32) Greenspan, H.; Fischer, E. Viscosity of Glass-forming Solvent Mixtures at Low Temperatures. J. Phys. Chem.1965, 69, 2466-2469. (33) Pettersen, E. F.; Goddard, T. D.; Huang, C. C.; Couch, G. S.; Greenblatt, D. M.; Meng, E. C.; Ferrin, T. E. UCSF Chimera – A Visualization System for Exploratory Research and Analysis. J. Comput. Chem. 2004, 25, 1605-1612. (34) Hammond, G. S.; Saltiel, J.; Lamola, A. A.; Turro, N. J.; Bradshaw, J. S.; Cowan, D. O.; Counsell, R. C.; Vogt, V; Dalton, C. Photochemical Cis-Trans Isomerization. J. Am. Chem. Soc. 1964, 86, 3197-3217. (35) Redwood, C.; Kumar, R. V. K.; Hutchinson, S. Mallory, F. B.; Mallory, C. W.; Dmitrenko, O.; Saltiel, J. Four Component Fluorescence of trans-1,2-Di(1-methyl-2-naphthyl)ethene at 77 K in Glassy Media. Conformational Subtleties Revealed. J. Phys. Chem. A. 2014, 118, 10575-10586. (36) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A. et al. Gaussian 09, Revision A.02, Gaussian, Inc., Wallingford CT, 2009. (37) Schlegel, H. B. Optimization of Equilibrium Geometries and Transition Structures J. Comput. Chem. 1982, 3, 214-218. (38) Schlegel, H. B. Optimization of Equilibrium Geometries and Transition Structures. Adv. Chem. Phys. 1987, 67, 249-286. (39) Schlegel, H. B. Modern Electronic Structure Theory, D. R. Yarkony, Ed.; World Scientific: Singapore, 1995; p 459-500.


35


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 36 of 40

(40) Becke, A. D. Density-Functional Exchange-Energy Approximation with Correct Asymptotic Behavior. Phys. Rev. A 1988, 38, 3098-3100. (41) Lee, C.; Yang, W.; Parr, R. G. Development of the Colle-Salvetti Correlation-Energy Formula into a Functional of the Electron Density. Phys. Rev. B 1988, 37, 785-789. (42) Becke, A. D. Density‐Functional Thermochemistry. III. The Role of Exact Exchange. J. Chem. Phys. 1993, 98, 5648-5652. (43) Stephens, P. J.; Devlin, F. J.; Chabalowski, C. F.; Frisch, M. J. Ab Initio Calculation of Vibrational Absorption and Circular Dichroism Spectra Using Density Functional Force Fields. J. Phys. Chem. 1994, 98, 11623-116237. (44) Nakashima, N.; Meech, S. R.; Auty, A. R.; Jones, A. C.; Phillips, D. Fluorescence Properties of Ergosterol. J. Photochem. 1985, 30, 207-214. (45) Passerani, R.; Ross, I. G. An Apparatus for the Ultra-Violet Spectroscopy of Solutions at Low Temperatures. J. Sci. Instr. 1953, 30, 274. (46) Hudson, B. S.; Kohler, B. E. Polyene Spectroscopy: The Lowest Energy Excited Singlet State of Diphenyloctatetraene and Other Linear Polyenes. J. Chem. Phys. 1973, 59, 4984-5002. (47) Birks, J. B.; Tripathy, G. N. R.; Lumb,M. D. The Fluorescence of All-trans Diphenyl Polyenes. Chem. Phys. 1978, 33, 185-194. (48) Hudson, B. S.; Kohler, B. E.; Schulten, K. Linear Polyene Electronic Structure and Potential Surfaces. In Excited States; Lim, E. C., Ed.; Academic Press: New York, 1982; Vol. 6; pp 1-95.


36

Page 37 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(49) Hughes, E., Jr.; Wharton, J. H.; Nauman, R. V. The Effect of Varied Excitation on the Fluorescence Spectra of 2-Phenylnaphthalene and Biphenyl: Photoselection of Conformers. J. Phys. Chem. 1971, 75, 3097-3103. (50) Baca, A.; Rossetti, R.; Brus, L. E, Structure and Dynamics of the Biphenyl Ring Torsion in Solid Neon and Argon. J. Chem. Phys. 1979, 70, 5575-5581. (51) Fuju, T.; Komatsu, S.; Suzuki, S. The Influence of Molecular Conformation on the Electronic Relaxation Processes. Biphenyl in Solution. Bull. Chem. Soc. Jpn. 1982, 55, 2516-2520. (52) Squillacote, M. E.; Sheridan, R. S.; Chapman, O. L.; Anet, F. A. L. Planar s-cis-1,3Butadiene. J. Am. Chem. Soc. 1979, 101, 3657-3658. (53) Arnold, B. R.; Balaji, V.; Michl, J. Structure of Matrix-Isolated s-cis-1,3-Butadiene from Polarized IR Spectra: A Closer Look. J. Am. Chem. Soc.1990, 112, 1808-1812. (54) Woodward, R. B.; Hoffmann, R. The Conservation of Orbital Symmetry. 1971, Verlag Chemie, GmbH, Weinheim/Bergstr. (55) Kulcsár, Á.; Saltiel, J.; Zimanyi, L. Dissecting the Photocycle of the Bacteriorhodopsin E204Q Mutant from Kinetic Multichannel Difference Spectra. Extension of the Method of Singular Value Decomposition with Self-Modeling to Five Components. J. Am. Chem. Soc. 2001, 123, 3332-3340. (56) Redwood, C. E. Advances in Mechanistic Photochemistry: Dienes and Trienes Ph.D. Dissertation, Florida State University, Tallahassee, Florida, 2016.


37


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 38 of 40

(57) Wodrich, M. D.; Corminboeuf, C.; Schreiner, P. R.; Fokin, A. A.; Schleyer, P. v. R. How Accurate Are DFT Treatments of Organic Energies? Org. Lett. 2007, 9, 1851-1854. (58) Schreiner, P. R. Relative Energy Computations with Approximate Density Functional Theory – A Caveat! Angew. Chem. Int. Ed. Engl. 2007, 46, 4217 – 4219. (59) Saltiel, J.; Dmitrenko, O.; Reischl, W.; Bach, R. D. The Triplet Potential Energy Surface of s-trans-2,4-Hexadiene. A Comparison of Theory and Experiment. J. Phys. Chem. A 2001, 105, 3934-3939. (60) Turek, A. M.; Krishnamoorthy, G.; Sears, D. F., Jr.; Garcia, I.; Dmitrenko, O.; Saltiel, J. Resolution of Three Fluorescence Components in the Spectra of all-trans-1,6-Diphenyl-1,3,5hexatriene under Isopolarizability Conditions. J. Phys. Chem. A 2005, 109, 293-303. (61) Saltiel, J.; Dmitrenko, O.; Pillai, Z. S.; Klima, R.; Wang, S.; Wharton, T.; Huang, Z.-N.; van de Burgt, L. J.; Arranz, J. Triplet and Ground State Potential Energy Surfaces of 1,4-Diphenyl1,3-butadiene: Theory and Experiment. Photochem. Photobiol. Sci. 2008, 7, 566-577. (62) Saltiel, J.; Klima, R.; van de Burgt, L. J.; Wang, S.; Dmitrenko, O. Temperature Dependence of the 1,6-Diphenyl-1,3,5-hexatriene Triplet Lifetime in Solution and Theoretical Evaluation of Triplet Conformer Interconversion. J. Phys. Chem. B 2010, 114, 14480-14486. (63) Catalán, J.; Saltiel, J. On the Origin of Nonvertical Triplet Excitation Transfer. The Relative Role of Double-Bond and Phenyl-Vinyl Torsions in the Stilbenes. J. Phys. Chem. A 2001, 105, 6273-6276.


38

Page 39 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(64) Saltiel, J.; Hutchinson, S. R.; Chitwood, K.; Dmitrenko, O. Photoisomerization of cis-1-(3Methyl-2-naphthyl)-2-phenylethene in Glassy Methylcyclohexane at 77 K. J. Phys. Chem. A 2012, 116, 5293-5298.

TOC GRAPHICS

Lumi

(+)cZ(+)c-Pre-ax-OH

(+)cE(+)c-Tachy-eq-OH


39


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Lumi

(+)cZ(+)c‐Pre‐ax‐OH

(+)cE(+)c‐Tachy‐eq‐OH


Page 40 of 40

Lumisterol to Tachysterol Photoisomerization in ... - ACS Publications

Recommend Documents