University of Groningen

Chiroptical molecular switches de Lange, Ben

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below. Document Version Publisher's PDF, also known as Version of record

Publication date: 2006

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA): de Lange, B. (2006). Chiroptical molecular switches: synthesis and applications. s.n.

Copyright Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.

Download date: 26-09-2021 CHAPTER 5

NON-CHIRQL STERICALLY OVERCROWDED ALKENES

Temperature dependent NMR measurements have been successfully applied to obtain thermal isomerization barriers of chiral sterically overcrowded bistricyclic ethylenes functionalized with different bridging groups in the lower and upper part, which could not be resolved into their enantiomers via HPLC, as shown in the previous chapter. Introduction of an isopropyl group into the upper part of these alkenes allowed an accurate determination of the isomerization barriers using the coalescence of the NMR absorptions of the two methyl substituents in the isopropyl moiety at T, Interestingly, our investigations on bistricyclic ethylenes revealed that the chiml 10',10'- dimethyl substituted alkene 1 (Figure 5.1, R, = CH,) showed two well separated singlets in the 'H NMR and 13C NMR spectra for the two methyl substituents (see Experimental Section of Chapter 4).' Therefore, this "internally fixed" isopropyl group might also find application in temperature dependent NMR studies, because the two methyl groups may become equivalent in a rapidly inverting molecule at high temperatures as illustrated in Figure 5.1 (le * lb).

Figwe 5.1. Interconversion of folded structures la * lb

In Chapter 4, we described experiments to establish a correlation between structures and isomerization barriers of bistricyclic ethylenes. In order to determine these barriers, it was necessary to introduce substituent R, (R, = Cl-I,) in the upper part of 1 to induce chirality into the otherwise centrosymmetric structures. In contrast, application of the "internally fixed isopropyl group as a probe in dynamic NMR does not require the presence of an additional substituent R, and would therefore, theoretically, provide isomerization barriers of non-chiral alkenes (!), via interconversion of two folded structures (la * lb) as illustrated by the Newman

The anisochronous properties of these two methyl groups are clearly illustrated by the X-ray stmure of IS in Section 5.3. projections in Figure 5.1. Because functionalization of bistricyclic ethylenes with one substituent (R,) will not enhance the isomerization barrier of these compounds compared to unsubstituted derivatives? the envisaged method can be used to predict quantitatively the racemization barriers of chire1 bistricyclic ethylenes via temperature dependent NMR studies on non-chiral alkenes. The synthesis and dynamic NMR study of non-chiml ethylenes might furnish valuable information on the thermal stability (i-e. racemization barriers), which can be expected for their chiral analogues, without a prior -lengthy- synthesis of these molecules with diastereotopic substituents and extensive "trial and error" experiments to resolve these molecules (see also Discussion in Section 5.5). In order to study the feasibility of this approach, we decided to prepare the unsubstituted analogue of 1 (Figure 5.1, R, = Furthermore, the synthesis of alkenes 15 - 17 has been undertaken (see Scheme 5.1), which are functionalized with a seven-, six- or five-membered central ring in the upper part of the molecule. This decrease in ring-size should also lead to a decrease in isomerization barrier, in agreement with the results described in the Chapter 4.

5.2 Synthesis

The synthesis of alkenes 1 and 15 - 17 was achieved by the diazo-thioketone method and is depicted in Scheme 5.1. S

NrrH, C1- 2 XIS 6 - 9 3 X-CHICH 4 X-0 5 X--

11 X-S 1 X-s 12 X-CH-CH 15 X-CHICH 13 X-0 16 X-0 14 X-- 17 X-- cy. 6983% cY. 7845%

Scheme 5.1. J)*rthesirof alk-enes I and 15 - 17.

See Chapter 4, Section 4.9.1. The isomerization bamer, which can be expected for 1 (R1 = H, approximately 25 kcal.mol.', see Section 4.8) is probably slightly too high for the application of the dynamic NMR technique (A& .e 5 - 24 kcal.mol-I). 132 A Starting from hydrazones 2 - 5, which were oxidized to the corresponding diazo compounds 6 - 9 followed by addition of thioketone 10 and desulfurization of the intermediate episulfides 11 - 14, alkenes 1 and 15 - 17 were obtained in overall yields varying from 5476% see Scheme 5.1 and Table 5.1). The 'H NMR and '(3 C NMR spectra of 1, 15 and 16 showed clearly separated absorptions for the two methyl groups at position 10' and the chemical shift differences between the methyl substituents are listed in Table 5.1. In the case of the fluorene functionalized alkene 17, the methyl absorptions appeared as a single peak, indicating a rather low isomerization barrier for this compound (see Section 5.4).

Table 5.1. 'H and 13c NMR Chemical Shift DiCIemces of the Methyl Grou~sin 1 and 15 - 17.

compound X yielda 'H NMR AV~ 13cNMR hvb (%I (Hz) (Hz)

a) Isolated yields based on the amount of added thioketone 10 (see Experimental Section). b) Av values at room temperature in CDC13.

Generally, the episulfides and alkenes shown in Scheme 5.1, were obtained as slightly yellow or white solids. Crystallization of episulfide 12 and alkene 15 from ethanol, however, yielded small beautiful glittering colourless crystals for bath compounds. Because thioketone 10 was isolated as blue-purple sparkling needles (see experimental section in Chapter 3), the availability of 10, 12 and 15 as crystalline materials provides the opportunity to study the molecular structure of a thioketone, an episulfide and an alkene found in the consecutive steps of the diazo-thioketone methodology by X-ray analysis (Scheme 5.2). Apart from our desire to obtain X-ray structures of three. structural entities, namely a thiocarbonyl, a thiirane and an alkene, which are so frequently encountered throughout this thesis, several considerations have stimulated us to obtain molecular structures of 10, 12 and 15: - The opportunity to gain insight into the conformational changes of the 10',10'- dimethylanthracene part, which might occur during the thioketone + episulfide

-+ alkene conversion. - The extent of folding in the upper and lower halves of the molecules both in the episulfide as well the alkene stage. - The interatomic distances in both halves of 15, which were required for the correlation study described in Chapter 4. - The molecular structure of 15 might clearly indicate why two different methyl absorptions are obtained in this centrosymmetric compound. Before the experiments with respect to the dynamic NMR behaviour of 1 and 15 -17 . are described, we will first present X-ray structures of 10,12 and 1~.~*'

12 15 Scheme 5.2 Stnudues ofthiokemne 10, episuyidc 12 and alkene 15.

5.3 MdenJor Structures of Thioketone 10, Episw 12 and Alkene 1fl~~

Suitable crystals for an X-ray investigation of 10 were grown from n-hexane and isolated as blue-purple sparkling needles. The thioketone crystallized in the Pbca space group with 8 molecules per unit cell. The molecular structure of 10 is depicted in Figure 5.2 and consists of a nearly planar conformation with one methyl substituent above and the other below the least-squares planes defined by both aromatic rings and the thiocarbonyl moiety (max deviation 0.021 A). The C=S bond length was 1.647 which can be compared to the C=S bond lengths found in thiobenzophenone (1.636 A)' and substituted derivatives thereof?

Figun 5.2 Molecular smrcaue of 10

In Seetion 5.3, the molecular structures of 10, 12 and 15 will be briefly discussed, without comprehensive listings of bond distances, bond angles, et cetera.' The complete data of bond distances, bond angles, torsional angles and final fractional atomic coordinates are available on request. The X-ray analysis of 10 was performed by F. van Bolhuis and the X-rays of 12 and 15 were performed by k Meetsma of the Crystal Structure Centre of our department. ' Rindoff, G.; Carlsen, L Acm Clyst. 1979, B35, 1179. 13 a) Arjunan, P.; Ramamurthy V.; Venkatesan, K Acm Clyst. 1984, C40, 552, 556. @) Manojlovic, LM.; Edmunds, LG. Acta clyst. 1965,18,543. Slow crystallization from ethanol afforded 12 as small twinkling colourless crystals, which were suitable for an X-ray determination. Episulfide 12 crystallized in the Fdd2 space group, whereby the unit cell contains sixteen discrete molecules. Crystallization from ethanol afforded 15 as small calourless c~ystalssuitable for X-ray analysis. This compound crystallized in the Pbca space group with eight molecules of 15 in each unit cell. The structures of 12 and 15 including the adopted numbering schemes are pictured in Figure 5.3. The linear moieties in both halves of 12 are oriented in the same direction with the central thiirane ring like a "roof' on top of these groups (ck-configuration, see also Fi re 5.5). The C-S bond lengths of the asymmetric three-membered rin are 1.857 FS,-C,) and 1.819 A (S,-CIS) with a central ClcC15 distance of 1.550 d These values closely resemble those which have been reported for the sterically hindered 2,2di-t-butyl-3,3-diphenyl-thiirane(1.848, 1.822 and 1.550 respectively)? The corresponding C-C bond length in thiirane itself is 1.492 A as reported by Cunningham et aL1" The longer central C-C distance in 15 compared to thiirane might be ascribed to strong steric hindrance between both halves of 15. The asymmetry of the thiirane ring in 15 might be due to a more pronounced steric interaction between the nearly planar dimethyl- moiety (vide infra) and the sulfur atom compared to the steric hindrance between the folded dibenzocycloheptene tricyclic unit and the sulfur atom in the central three-membered ring (see Figure 5.5). The "back side" of the dimethyl-anthracene part of 12 (defined by the atoms C,, C6, C,, Cg, C9) deviates only slightly from planarity; the torsion angle C6C,c& is 6.3". The "front side" of the dimethyl-anthracene group of 12 (atoms q,C,, C14, C13, C,,), however, shows an increased tendency to form a folded structure (torsion angle C,C,,C,,CI2 = 19.4"), which can be ascribed to a large increase in steric hindrance between these "front side" atoms (and the hydrogens at C, and C,,) and the opposite dibenzocycloheptene moiety compared to the "back side" atoms.

Figure 5.3. Mo~CU~A~sbuctures of I2 (left) and I5 (right)

Mugnoli, A; Simonetta, M. Acta Cyst. 1976, B32,1762. lo Cunningham, G.L;Boyd, AW.;Gwinn, W.D.; Myers, RJ. J Chem. Phys. 1195,19,676. 135 The X-ray structure of 15 shows a folded configuration, whereby the tricyclic units on both sides of the central double bond are positioned in opposite directions (i.e. anti- folded)" and shows strong resemblance to the X-ray structure of 9-(2'-methyl-9'H- -9'-y1idene)-9H-xanthene described in Section 4.7 of the previous chapter. Due to the non-planar configuration of IS and, presumably, slow movement of the aromatic moieties at room temperature (i.e. process la * lb in Figure 5.1), the absorptions for the methyl substituents are found at different ppm values. Structure 15 clearly demonstrates the different environment of the methyl substituents at carbon atom 14. The central olefinic bond C,-C15 is almost planar (torsion angles: C6C7C15C16= 0.Y, C6C,ClsC, = 4.5") with a bond length of 1.338 A In Figure 5.4, the Newman projection of 15 seen along the C+-C,, bond are pictured together with the torsion angles.12

Figure 5.4. Newman projection of 15

In contrast to episulfide 12, the dimethyl-anthracene part of 15 displays a large deviation from planarity in the "front side" as well as the "back side" (vide supra) in order to diminish the severe steric hindrance between the anthracene group and the dibenzocycloheptene part of 15 (torsion angles CICl,Cl,Cl, = 41.3", C6c+2& = 43.2"). The dibenzocycloheptene tricyclic units of 12 as well as 15 revealed similar structures with comparable torsion angles in both compounds (&C,Cl5Cl6 = 69.B0 in 12 and Cl,C16ClsC, = 65.1" in IS), indicating the large folding of these moieties. Desulfurization of the episulfide is accompanied by a large change in structure, as can be clearly seen from the side views of 12 and 15 presented in Figure 5.5;13 the cis- folding of the episulfide is altered into a trans-conformation of the alkene (anti- folded).

l1 Nyburg et aL reported the X-ray structures of both the anti- as well as syn-isomer of bis- dibenzo[a,d]cycloheptenylidene: Nyburg, S.C; Dichmann, KS.; Pickard, F.H.; Potworowki, J.k Acm Cryst. 1974,830, 27. See also: SchOnberg, A; Praefcke, E, Sodtke, U. Chem Ber. 1969, 102, 1453. l2 The carbon atom C6 is used arbitrarily to define the mean plane through the central double bond and the dotted line 1 in Figure 5.4 is drawn to show the deviations from planarity observed for the CI5-C1, and CI5-Cz9bonds. l3 The molecular structure of episulfide 12 in Figure 5.5 shows some structural resemblance to the "molecular clips" described by Nolte et a[. See e.g.: Nolte, R.J.M.;Kentgens, kP.M.; Sijbesma, R.P. J. Org. Chem 1991,56, 3199. Figure 5.5. Side view of 12 (lefr) and 15 (right)

5.4 Tempemture Dependent 'H NMR studies14

The thermal behaviour of the xanthene functionalized alkene 16 was examined first because, based on the results described in the previous chapter, the isomerization bamer of this compound is expected to lie well in the range of the dynamic technique (= 5 - 24 kcal.mo1"). At room temperature the two methyl substituents were found at 1.48 and 1.58 ppm (Av = 31.3 Hz at 25 "C)using nitrobenzene-d, as a solvent. The two signals showed coalescence at 85 + 3 "C, indicating that this "internally fmed isopropyl group can also be applied in temperature dependent NMR studies of isomerization processes. Although the signals slightly broadened upon raising of the temperature, the T, was determined accurately, in contrast with the dynamic 'H NMR method used for the isopropyl derivatives described in the previous chapter. By using Equation 4.2, the isomerization bamer was calculated to be 18.2 kcal.mo1-'. Upon lowering of the temperature in CD,Cl, the single absorption which was observed for the two methyl substituents in the fluorene functionalized alkene 17 exhibited decoalescence at -25 + 3 "C and following an analysis similar to that described before, this leads to a A@ = 11.8 kcal.mo1-'. For the thioxanthene functionalized alkene 1, the initially well separated signals (Av = 10.7 Hz at 25 "C) broadened upon heating till 190 "C in nitrobenzene-d, as a solvent, but still no coalescence had occurred (Av = 9.3 Hz at 190 "C). Therefore, it can be calculated that the isomerization barrier of 1 should be at least 24.7 kcal.mo1-I). The results are summarized in Table 5.2. Unfortunately, the two methyl absorptions of 15 coincided at 25 "C in nitrobenzene-d,. This phenomenon cannot be ascribed to a fast inversion of the molecule at room temperature, because in CDCl, the resonances of the methyl substituents are clearly separated (Av = 24.7 Hz at 25 "C),but

l4 Due to the very large chemical shift differences in the 13c NMR absorptions for the two methyl substituents of alkenes 1, 15 and 16 (see Table 5.1), temperature dependent 13c NMR can only be applied for A& < 21 kcal.mol-'. Therefore, we decided to use temperature dependent 'H NMR for the experiments described here. unfortunately this solvent is not suited for the dynamic NMR meas~rements?~ The data in Table 5.2 for 1, 16 and 17 clearly indicate a comparable correlation between isomerization barriers and steric hindrance during "edge passage" (process la * lb in Figure 5.1) as found in the previous chapter: a large increase in isomerization barrier is found when the interatomic distance (C,-C,,) is enlarged, which enhances the steric hindrance during the isomerization process (see Figure 4.8 in Chapter 4)?6 Furthermore, the isomerization barriers of alkenes 1, 16 and 17 reported here lie in the same range as the barriers of the bistricyclic ethylenes functionalized with a 2- substituted thioxanthene moiety in the upper part and a thioxanthene, xanthene or fluorene unit in the lower part of the molecules (27.4, 17.1 and 12.2 kcal-mol-', respectively) as described in Chapter 4. These findings underline the value of the present approach by using the "internally fixed" isopropyl group as a probe in dynamic NMR studies.

Tabk 5.2. Isomerization Jjaniers of 1 and 15 - 17 Determined by Dynamic 'H NMR Studks.

compound X Ava ~,b isomerization barriersc C4a-C4b distanced (Hz) CC) (kcal.mo1-') A

a) Av values are calculated by extrapolating the chemical shift differences at low temperatures to T,. b) * 3 OC c) + 0.2-0.3 kcal.mo1-l. d) See Chapter 4, Table 4.5 e) Av at 190 OC in nitroben~ened~.

l5 The methyl absorptions of 15 were separated in p-xylene-dl,, as a solvent (Av = 33.7 Hz at 25 OC), but no coalescence was observed at 130 "C Therefore, A& should be at least 20.7 kdmol-'. l6 The interatomic distance C4a-C4b of 15 (3.10 A) indicates a large isomerization bamer for this compound (A& > 25 kcal.mo1-I) outside the range of the dynamic NMR technique. 138 5.5 Discusion and Corudruhg Remarks

Because resolution of a sterically overcrowded alltene can only be performed, when an isomerization barrier of approximately 20.0 kcal.morl or higher exists between the two enantiomers," the low barriers for the isomerization processes, which have been found for the non-chiml alkenes 16 and 17 indicate that their chhl analogues 16a and 17a (Figure 5.7, R, H) cannot be resolved at ambient temperatures. In contrast, the high energy barrier for the inversion process in 1 (A@ > 24.7 kcal.m01-~)denotes that the chid analogue la (R, H) might be separated into enantiomers and these enantiomers should not racemize at RT. Of course, based on the extensive research on the correlation between the energy barriers and structures of bistricyclic ethylenes described in the previous chapter, the observed barriers for 1, 16 and 17 are not surprising.

la, 1&,17a

However, several useful future applications of the present method can be envisaged and possible target structures are shown in Figure 5.8. The presence of the "internally fixed" isopropyl group in these various types of inherently dissymmetric molecules can provide reliable and quantitative data concerning the isomerization barriers of these compounds prior to extensive attempts to synthesize chiral analogues or to achieve resolution. The influence of the complexing behaviour of various metals on the inversion processes in 18 and 19 can be conveniently studied by dynamic NMR techniques. Therefore, the potential of 18 and 19 to function as asymmetric catalysts (18 and 19) and switchable crown-ethers (19) can be estimated undo~btedl~.'~

Figure 5.8. Possible applications for the 'intemal&fired" isopropyI group.

l7 Menmark, S.G. Chromatographic Enantioseparation; Methods and Applications Ellis Holwood Publishers: Chichater, 1988, Chapter 5. l8 Approaches to synthesize these types of compounds are currently under investigation in our research group. 139 For general remarks, see Section 2.7. 5H-Diinzo-[&dl--5-onehydrazone (3) was prepared as described by Taylor et a&'' The syntheses of 10,lO-dimethyl- 9(10H)-anthracene-9-thione (10) and hydrazones 2, 4 and 5 are described in Chapter 3.

Crystal structure determination of 10~~~ The crystal structure determination of 10 was performed on a blue-purple crystal of dimensions 0.50 x 0.50 x 0.20 mm obtained by crystallization from hexane. Ctystal data: q,H ,S, Y = 238.35, orthorhombic, Pbca, a = 11.729(1), b = 15.463(2), c = 13.611(32 V = 2468.6 A3, Z = 8, Dx = 1.283 gmJ, A(Mo&) = 0.70930 p = 2.24 cm- , F(000) = 1008, T = 293 K, R, = 0.048 for 1425 unique observed reflections with I r 3.0 o(1) and 154 parameters.

Episulfides 11 - 14 were prepared following the general method described for dispiro[2-rnethyl-9H-thioxanthene-9,2'-thiirane-3', 9"-(9"H)-thioxanthene] (25) in Section 4.11.

Dispim[10,10-dimethyl-9(1OH)-anthracene-9, 2'-thiirane-3', 9"-(9"H)-thioxanthene] (11) Starting from 9H-thioxanthene-9-one hydrazone (2, 452 mg, 2.0 mmol) and thioketone 10 (476 mg, 2.0 mmol), episulfide 11 precipitated from the solution and was isolated by filtration as a white powder (690 mg, 1.60 mmol, 80.0%, based on the amount of added thioketone): mp 193.9-194.7 "C (dec); 'H NMR (300 MHz) S 1.49 (s, 3H), 1.97 (s, 3H), 6.73 (ddd, J = 8.8, 7.3, 1.5 Hz, 2H), 6.93-7.05 (m, 4H), 7.08 (dd, J = 8.0, 1.1 Hz, 2H), 7.16 (ddd, J = 8.8, 7.3, 1.1 Hz, 2H), 7.31 (dd, J = 8.1, 1.5 Hz, 2H), 7.51 (dd, J = 8.1, 1.5 Hz, 2H), 7.83 (dd, J = 7.7, 1.1 Hz, 21-i); 13C NMR S 32.79 (q), 36.83 (q), 38.44 (s), 60.00 (s, C-S), 67.80 (s, C-S), 123.55 (d), 125.17 (d), 126.00 (d), 126.35 (d), 126.50 (d), 126.90 (d), 129.14 (d), 129.46 (s), 130.88 (d), 135.01 (s), 136.07 (s), 146.22 (s); HRMS Calcd for C29H22Sz:434.116, found 434.116.

Dispiro [SH-dibenzo-[a,d] -cycloheptene-5, 2'-thiirane-3', 9"-(10",10"-dimethyl)- 9"(1OWH)-anthracene] (12) Oxidation of hydrazone 3 (880 mg, 4.00 mmol) to the dark red diazo compound followed by addition of 10,lO-dimethyl-9(10H)-anthracene-9-thione (10, 610 mg, 2.56 mmol), yielded 12 as a white glittering solid. This episulfide precipitated from the ether solution and was isolated by filtration (904 mg, 2.11 mmol, 82.5 %, based on the amount of added thioketone) mp 188.3-190.2 "C (dec); 'H NMR (200 MHz) S 1.57 (s, 3H), 2.00 (s, 3H), 6.39 (s, 2H), 6.53-6.62 (m, 2H), 6.86 (d, J = 7.7 Hz, 2H), 7.04-7.16 (m, 6H), 7.34-7.42 (m, 4H), 7.97 (d, J = 7.8 Hz, 2H): 13C NMR S 32.10 (q), 37.56 (q), 37.92 (s), 58.01 (s, C-S), 69.49 (s, C-S), 123.44 (d), 126.29 (d), 126.30 (d), 126.69 (d), 126.81 (d), 127.91 (d), 128.84 (d), 129.88 (d), 130.32 (s), 131.82 (d), 134.30 (s), 137.63 (s), 145.60 (s); HRMS Calcd for C3,H,S: 428.160, found 428.159.

Crystal structure determination of 12~~" The crystal structure determination of 12 was performed on a colourless transparent

-- I' Taylor, G.A.; Falshaw, C.P.; Hashi, N.A.; I. Chem Soc., Perkin Trans. I 1985, 1837. " Numbers in parentheses are estimated deviations in the least significant digits. 140 parallelepiped crystal of approximate size 0.10 x 0.35 x 0.42 obtained by crystallization from ethanol. Crystal data: q,H,S, M, = 428.59, orthorhombic, Fdd2, a = 24.076(2), b = 36.059 2), c = 10.468 V = 1497.5(5) A3, Z = 16, Dx= 1.253 g~m'~,A(MoKa) = 0.71073 k p = 1.5 cm-', F(000) = 3616, T = 130 K, R, = 0.028 for 2013 unique observed reflections with I 1 2.5 u(1) and 385 parameters.

Dispiro[10,1O-dimethyl-9(I0H)-anthrscene-9, 2'-thiirane-3', 9"-(9"H)-xanthene] (13) 9H-Xanthene-9ane hydrazone (4, 420 mg, 2.00 mmol) was oxidized to the green diazo compound and after addition of thioketone 10 (300 mg, 1.26 mmol), the episulfide separated from the solution. Crystallization from ethanol (150 rnL) afforded 13 as small slightly yellow crystals (420 mg, 1.00 mmol, 79.7%, based on the thioketone): mp 189.0-189.4 "C (dec); 'H NMR (300 MHz) S 1.19 (s, 3H), 1.71 (s, 3H), 6.87-6.93 (m, 4H), 7.04-7.16 (m, 6H), 7.29-7.33 (m, 2H), 7.70 (dd, J = 7.7, 1.5 Hz, 2H), 7.97-8.01 (m, W);13C NMR 6 26.52 (q), 31.63 (q), 39.40 (s), 57.98 (s, C-S), 66.56 (s, CS), 115.38 (d), 120.50 (s), 121.91 (d), 123.14 (d), 124.46 (d), 127.26 (d), 128.41 (d), 130.01 (d), 130.22 (d), 133.46 (s), 147.57 (s), 154.10 (s); HRMS Calcd for C#,OS: 418.139, found 418.139. Dispiro[lO,lO-dimethyl-9(10H)-anthracene, 2'-thiirane-3', 9"-(9"H)-fluorene] (14) 9H-Fluorene-9-one hydrazone (5, 388 mg, 2.00 rnmol) was oxidized to purple red diazo compound followed by addition of 10 (360 mg, 1.51 rnmol). The solvent was removed under reduced pressure and after crystallization from ethanol (200 mL), the episulfide 14 was obtained as yellow crystals (420 mg, 1.05 mmol, 69.2 %, based on the thioketone): mp 148.4-150.2 "C (dec, red melt); 'H NMR (300 MHz) 6 0.52 (s, 3H), 1.68 (s, 3H), 6.90 (ddd, J = 8.4, 8.1, 1.1 Hz, 2H), 7.11 (d, J = 8.1 Hz, 2H), 7.18 (ddd, J = 7.7, 7.3, 1.1 Hz, 2H), 7.21-7.32 (m ,4H), 7.37 (dd, J = 7.7, 1.5 Hz, 2H), 7.55 (dd, J = 7.7, 0.7 Hz, 2H), 8.22 (dd, J = 7.7, 1.5 Hz, 2H); 13C NMR 6 26.03 (q), 28.% (q), 39.51 (s), 58.80 (s, C-S), 62.10 (s, C-S), 119.18 (d), 123.57 (d), 124.63 (d), 124.87 (d), 126.28 (d), 127.42 (d), 127.79 (d), 128.95 (d), 136.50 (s), 141.06 (s), 142.98 (s), 148.16 (s); HRMS Calcd for W,S: 402.144, found 402.143. Alkenes 1 and 15 - 17 were obtained from the corresponding episulfides as described for 2-methyl-9-(9'H-thioxanthene-9'-ylidene)-9H-thionthene (11) in Section 4.11.

9-(10',1(Y-Dimethyl-9'(10'H)-anthracene-9-1idene)-9H-thionthene (1) Desulfurization of episulfide 11 (434 mg, 1.00 mmol) yielded alkene 1 as a white solid (380 mg, 0.95 mmol, 94.5%) after crystallization from absolute ethanol (400 mL): mp 260.5-262.1 "C;'H NMR (300 MHz) G 1.92 (s, 3H), 2.02 (s, 3H), 6.89 -7.22 (m, 12H), 7.57 (d, J = 7.7 Hz, ZH), 7.64 (d, J = 8.1 Hz, 2H); 13C NMR S 24.11 (q), 30.91 (q), 40.19 (s), 122.72 (d), 124.49 (d), 125.56 (d), 126.35 (d), 126.60 (d), 127.24 (d), 128.95 (d), 129.41 (d), 134.14 (s), 135.72 (s), 136.67 (s), 136.85 (s), 136.87 (s), 147.02 (s); HRMS Calcd for C&H,S 402.144, found 402.145.

5-(10',1(Y-Dimethyl-Y(lO'H)-anthmcene-9'-ylidene)-5H~ibenm[a,d]cycloheptene (15) Starting from episulfide 12 (428 mg, 1.00 mmol), 15 was obtained as a white crystalline solid (309 mg, 0.78 mmol, 78.0 %) after crystallization from ethanol (200 mL): mp 260.2-262.4 "C; 'H NMR (300 MHz) S 1.83 (s, 3H), 1.91 (s, 3H), 6.47 (dd, J = 7.8, 1.2 Hz, 2H), 6.72 (ddd, J = 7.8, 7.3, 1.2 Hi, 2H), 6.93 (dd, J = 7.7, 0.7 Hz, 2H), 7.02-7.11 (m, 4H), 7.17 (s, 2H), 7.21 (ddd, J = 7.7, 7.3, 1.5 Hz, 2H), 7.40 (dd, J = 7.8, 1.2 Hz, 2H), 7.44 (dd, J = 7.8, 0.7 Hz, 2H); 13cNMR S 24.06 (q), 31.31 (q), 39.91 (s), 122.38 (d), 124.44 (d), 126.11 (d), 126.47 (d), 127.97 (d), 128.02 (d), 128.31 (d), 128.89 (d), 131.36 (d), 134.08 (s), 135.00 (s), 135.66 (s), 136.86 (s), 139.37 (s), 146.54 (s); HRMS Calcd for GIH, 396.188, found 396.188.

Crystal stn~cturedetermination of 1563 The crystal structure determination of 15 was performed on a transparent plate shaped crystal of dimensions 0.20 x 0.40 x 0.50 mm obtained by slow crystallization from ethanol. Cystal data: = 396.53, orthorhombic, Pbca, a = 7.046(1), b = 17.062(1) c = 36.599(2) A3, Z = 8, Dx= 1.197 gcm-3, A(MoKa) = 0.71073 k p = 0.6 cm-l, F(W) = 1680, T = 295 K, & = 0.063 for 2318 unique observed reflections with I r 2.0 u(1) and 377 parameters.

9-(lW,lW-Dimetayl-Y(1WH)gllthcene-9 (16) Starting from l3 (259 mg, 0.62 mmol), alkene 16 was obtained as a white solid (211 mg, 0.55 mmol, 88.2%) after crystallization from ethanol (50 mL): mp 235.1-236.7 "C; 'H NMR (200 MHz) S 1.80 (s, 3H), 1.94 (s, 3H), 6.83 (ddd, J = 7.7, 7.3, 1.5, 2H), 6.89-6.94 (m, 2H), 7.01 (dd, J = 7.7, 1.1 Hz, 2H), 7.13-7.22 (m, 6H), 7.27 (dd, J = 8.1, 1.1 Hz, 2H), 7.51 (d, J = 7.7 Hz, 2H); 13c NMR S 24.21 (q), 30.27 (q), 40.39 (s), 116.78 (d), 120.46 (s), 122.14 (d), 123.04 (d), 124.66 (d), 124.82 (s), 126.59 (d), 127.74 (d), 127.% (d), 128.94 (d), 131.23 (s), 137.62 (s), 147.48 (s), 154.96 (s); HRMS Calcd for C&Iz0: 386.167, found 386.166.

10,10-Dimethyl-9-(9'H-fluorene-9'-ylidene)-9(lOH)nthmne (17) Starting from episulfide 14 (240 mg, 0.60 mmol), crystallization from ethanol afforded 17 as a slightly yellow solid (172 mg, 0.47 mmol, 77.9%): mp 223.9-225.1 "C; 'H NMR (300 MHz) S 1.74 (s, 6H), 6.98-7.04 (m, 2H), 7.19 (ddd, J = 7.7, 7.3, 1.5 Hz, 2H), 7.24-7.32 (m, 4H), 7.56 (dd, J = 7.7, 1.1 Hz, 2H), 7.71 (dd, J = 7.7, 0.7 Hz, 2H), 7.75 (d, J = 8.1 Hz, 2H), 7.87 (dd, J = 7.7, 1.5 Hz, 2H); 13c NMR S 27.29 (q), 41.05 (s), 119.21 (d), 123.60 (d), 124.58 (d), 125.42 (d), 125.77 (d), 127.43 (d), 127.80 (d), 128.34 (d), 129.51 (s), 138.67 (s), 138.99 (s), 140.09 (s), 140.65 (s), 147.61 (s); HRMS Calcd for wz:370.172, found 370.172.