Circular Dichroism Studies on New Optically Active 1,5-Benzodiazepine Derivatives Khalid Mohammed Khana, Farnaz Malikb, Mashooda Hasanb, Shahnaz Perveen3, Jodwiga Frelekc, Günther Snatzke+, Wolfgang Voelter3* a Abteilung für Physikalische Biochemie, Physiologisch-chemisches Institut der Universität Tübingen, Hoppe-Seyler-Straße 4, D-72076 Tübingen, Germany b Department of Chemistry. Quaid-i-Azam University, Islamabad. Pakistan c Institute of Organic Chemistry, Polish Academy of Scienses, Kasprzaka 44, PL-01-224 Warschau, Poland Dedicated to Prof. Ivar Karl Ugi on the ocassion o f his 65th birthday Z. Naturforsch. 51 b. 588-598 (1996); recieved September 26, 1995 CD Spectra, Benzodiazepines, Chiral 1,5-Benzodiazepine Derivatives From thirty three new optically active 1,5-benzodiazepine derivatives the UV and CD data are reported and the bands discussed related to corresponding electronic transitions respectively stereochemical features.

Introduction which all optically active Lofendazam analogs de­ In continuation of our studies on benzodiazep­ scribed in reference [9] were synthesized. These ine-based compounds, widely used as therapeutic compounds (see [9]) served also as starting materi­ agents for the treatment of anxiety and neuroses als to yield via reduction the corresponding [1-4] and for reasons of studying structure-activ- 1,3.4,5-tetrahydro-2//-benzodiazepines respecti­ ity relationships, performed by pharmacologists vely treatment with 2,4-bis-(4-methoxyphenyl)- [1-8], we described recently synthetic routes to 2,4-dithioxo-1,3,2,4-dithiadiphosphetane (La wes­ optically active l,3,4,5-tetrahydro-2//-l,5-benzodi- son reagent) [12] the corresponding 1,3,4,5-tetra- azepin-2-ones as analogs of Lofendazam [9], hydro-2//-benzodiazepin-2-thiones [10] from 1.3.4.5-tetrahydro-2//-benzodiazepines and which their UV and CD spectra were recorded. 1.3.4.5-tetrahydro-2//-benzodiazepin-2-thiones In continuation of our circular dichroism studies [10]. For the Lofendazam analogs, 4-methyl- on natural products and biologically active com­ 1.3.4.5-tetrahydro-2//-benzodiazepin-2-one with pounds like carbohydrates [13-21], nucleosides one chiral center at C-4 in the 7-membered ring and nucleotides [22-26], flavone glycosides [27- was used as a starting material which could be eas­ 29], amino acids, peptides and proteins [30-42], ily prepared by the reaction of o-phenylenedia- steroids [43,44] or benzodiazepine derivatives [9, mine with crotonic acid in 5.5 N HC1 [11]. Racemic 10, 45], the UV- and CD-spectroscopic properties 4-methyl-l,3,4,5-tetrahydro-2//-benzodiazepin-2- of thirty three new chiral benzodiazepine deriva­ one was resolved on a preparative scale via diaste- tives are discussed in this communication. reomeric salt formation with D-(+)-3-bromo- camphor-8-sulfonic acid in ethanol. After decom­ Results and Discussion position of the salt and work up, as described in [9], enantiomerically pure 4-methyl-l,3,4,5-tetra- Benzodiazepines hydro-2//-benzodiazepin-2-one was obtained from The benzodiazepines which do not contain the amide grouping neither in the ring nor in the side chain show in their UV spectra in general three * Reprint requests to Prof. Dr. Dr. h.c. W. Voelter. bands (Table I). A first one. which is found around + This communication is the result of a joint research 300 nm and of moderate intensity (e around 4000 program with the late Prof. Dr. Dr. h.c. Günther Snatzke, Lehrstuhl für Strukturchemie Ruhr-Uni- to 5000), a second one shows up between 250 and versität Bochum. Germany. 270 nm (f from 7000 to 9500) and the main absorp-

0932-0776/96/0400-0588 $06.00 © 1996 Verlag der Zeitschrift für Naturforschung. All rights reserved. D K. M. Kahn et al. • CD Studies on 1,5-Benzodiazepine Derivatives 589

tion maximum is found between 230 and 245 nm Table I. UV and CD data of compounds 1 to 33, re­ (e = 20000 to 35000). This is in good agreement corded in acetonitrile and in selected cases under acid with the well-known spectrum of the parent com­ or basic conditions. pound, orf/zo-phenylenediamine, giving rise to the Compound UV CD same type of maxima, which are, however, all ^-m ax/m in A e strongly shifted hypsochromically, 295 (e = 3200), H i 304 4888 319 + 0.43 ( e 239 = 6500), and 210 (e = 47700), with respect 279 - 1.22 to those of the benzodiazepines. These wavelength 259 +0.80 253 7497 254 + 0.83 shifts of the first two absorptions must be corre­ ca1 ' 'CHj 1 H 3 245 35869 231 -2.15 lated with the non-coplanarity of the seven-memb- 214 -5 .3 7 ered ring system with the benzene ring. In general, + H+ (1) 284 1742 283 -0.76 steric hindrance of resonance will give a blue shift 263 + 1.40 in the spectra as compared with that of the non­ 252 +3.65 substituted product, all three bands are, however, 242 8365 242 +5.60 222 -3.23 red shifted. This mentioned “general statement” 202 18927 211 -8.50 is, however, only valid if the jr-system is twisted in CHj 300 4570 319 + 0.60 such a way that the energy of the lower level is 286 - 3.64 even more lowered by twisting and that of the ex­ 260 7168 261 + 0.68 245 - 1.16 cited one is raised (e.g. the 2-3 bond of butadi­ ca2 ' CHj H 3 229 25567 230 + 3.88 ene). If, however, the contrary is the case (as e.g. 217 - 0.72 is valid for twisting around the 1-2 bond of a bu­ 205 + 0.71 tadiene), then a red shift is noticed. Without going + H+ (2) 283 3375 288 -0.94 into a more sophisticated estimation one can easily 244 9348 244 +4.35 see from the values determined for orf/zo-quinodi- 202 22154 208 -3.98 methane [46] that twisting around the N-C (aro­ 304 4557 307 - 1.03 matic) bond should indeed lead to such bathoch- 265 7283 271 + 3.69 231 23103 229 -4.95 romy. The intensities of these three absorption 225 -4.78 c6-i 'CH, bands are virtually not depending on the substitu­ J CH3 3 tion of the nitrogen atoms, since the heterocyclic system can a) never be fully conjugated with the + H+ (3) 283 3825 245 10098 aromatic ring, but b) is flexible enough to avoid 201 22154 strong twisting. Substitution in 7-position leads in CHj general to the usual red shift of the first a-band, 299 5182 310 - 1.69 270 9539 269 + 9.45 but most of the time has a negligible influence on 237 21612 235 - 1.66 219 + 1.59 the other absorption bands or may introduce even ca4 1 'CHj C H j ^ 208 -2.38 a slight hypsochromy (Table I). As discussed in our previous paper [45], in the CD spectra (Table I, Fig. 1) more bands appear, COCHj 245 +8.11 225 12357 220 -12.78 and some of them coincide practically with those 194 24640 202 +6.11 found for the UV spectra; e.g. in case of 1, posi­ ca tive Cotton effects are found at 319 and 254 nm

and a negative, stronger one, around 230 nm + H+ (5) 241 -J 1 + (A s - -2.15). In addition to this appears, however, 215 S* ÜI o OO a negative Cotton effect at 279 nm, which has no CHj 334 -0.19 analogy in the UV spectrum. Since the other just 310 4668 302 +1.34 mentioned Cotton effects coincide practically with 265 7203 264 +6.99 , - c a 233 17360 234 +3.04 f. ' CHj UV absorption bands, it is very improbable that 6 CHj 3 208 -5.33 this mentioned additional Cotton effect is caused + H+ (6) 290 2296 287 -1.16 by the presence of a conformational equilibrium 248 9554 262 -3.54 and can, therefore, be associated with a forbidden 212 23156 590 K. M. Kahn et al. • CD Studies on 1,5-Benzodiazepine Derivatives

Table I. (contiunued). Table I. (contiunued).

Compound UV CD Compound UV CD ^-max/min ^max ^max ^■max/min Ae

311 3961 309 +0.65 316 4205 318 -1.36 281 -0.26 274 6619 275 +5.70 251 5980 260 -1.30 235 22070 237 -7.85 XXX 222 24604 228 +3.22 c,'J^ X \hh 7 H 198 12875 213 -4.14 13 r CH;

+ H +(7) 283 3082 285 +0.20 195 12593 276 +0.22 + H + (13) 289 2519 273 -3.00 242 6777 246 +3.42 256 9368 243 +1.46 220 30555 222 -4.19 237 +1.53 207 +4.30 217 -3.39 207 13660 211 -3.56 ch3 310 6813 313 +0.30 259 8746 280 -1.61 CHj 307 4400 319 -1.69 jOQ. 227 24881 230 +3.92 278 8517 278 +10.36 I1 'CHj 206 17186 242 25047 238 -1.96 H ,-CQ.1 191 13678 220 +1.91 14 + H+(8) 285 2534 285 -0.72 + H + (14) 281 -1.45 244 7444 247 +3.11 252 +3.37 216 30203 222 -5.80 218 -0.99

312 4495 308 -2.61 381 -4.64 260 6305 267 +2.94 328 11172 338 +6.75 228 20765 224 -8.45 292 9724 292 (sh) +3.84 a:i256 12793 249 -10.09 9n CHj 1 CH, 3 15 217 15231 228 (sh) -8.18 203 +10.63 + H + (9) 286 2804 286 -2.00 + H + (15) 411 -0.52 250 6813 260 -5.04 392 -0.52 235 +0.44 301 2238 301 +2.39 220 20688 218 -6.38 219 3087 248 -0.51 229 -0.52 CHj 332 -0.30 196 21029 208 +2.07 307 4086 306 +1.06 264 6562 265 +8.71 JX& 230 15455 235 +4.42 389 -0.90 10 CHjCH’ 212 -5.66 358 +3.04 312 11394 313 +3.63 + H+ (10) 281 4297 286 (sh) -1.38 a:i 13926 283 +5.19 CHj 285 262 -3.77 16 3 253 11676 251 -5.38 248 7874 243 +0.72 219 13630 224 -5.79 220 30555 225 -4.14 204 +7.60 207 +2.98 313 4263 326 +0.57 + H + (16) 380 +0.46 262 6514 265 -2.37 294 27244 297 +2.86 228 26496 221 +1.59 221 13816 226 -0.70 .JÄ 288 -1.29 209 +4.01 CH, ii 246 +5.56 360 +10.01 219 -5.05 325 11114 325 +12.17 293 14054 296 +11.46 312 3574 328 +0.78 1 268 15240 266 -24.12 270 5745 288 -3.29 17 ^ 219 15730 218 +16.83 240 14624 253 -1.20

12 \ ch3 + H + (17) 300 2185 327 +7.79 296 +7.66 + H+ (12) 286 2579 290 -1.61 222 7269 265 -18.16 251 10806 248 +4.57 197 22564 218 + 10.07 212 24186 218 -5.20 K. M. Kahn et al. • CD Studies on 1,5-Benzodiazepine Derivatives 591

Table I. (contiunued). Table I. (contiunued).

Compound UV CD Compound UV CD 5 A,nax ^max /'-max/min A t /l-max ^max Amax/min Ae

CHj 362 2355 365 +10.46 + OH- (25) 365 +39.07 312 11090 311 +6.80 325 -10.12 287 18346 288 +10.75 285 6637 273 -9.97 2 h CC 266 17086 262 -25.15 251 9664 229 -13.89 2 c 00 o - X 221 13116 209 +12.93 218 20058 204 +9.06

CHj 373 2124 388 -0.73 423 -0.41 358 +6.52 395 7744 380 +1.04 314 17406 313 +3.54 323 14888 307 +0.51 290 14718 287 +6.41 274 10849 270 (sh) -0.15 J g26 ' kchj 252 13834 252 -7.69 19 242 13983 233 -4.54 212 15028 219 -2.27 221 21259 224 (sh) -3.34

324 15807 330 +23.88 CHj 377 7451 389 +5.75 300 13047 301 +19.11 319 12588 303 +2.24 270 15286 267 -45.67 270 10413 273 +1.29 CH3°' Z j r -CHj 222 16333 220 +15.57 240 15979 235 -6.91 Z/ CHj 20 h CHj 218 -8.95 202 +0.08 367 1507 364 +9.66 1 S 412 +4.97 314 10528 314 +4.65 11133 j Q r \ 291 11191 293 +6.97 366 CHjO - ^ n ^ - H 326 15337 315 +7.72 264 11316 262 -19.26 28 CHj o 2nc^ cn i 279 12151 279(sh) -6.16 225 11627 202 +7.91 21 6h3 CHj 249 13169 248 -13.38 223 (sh) -1.84 212 12759 206 +11.07 393 -0.90 363 +3.99 395 5359 403 +13.29 316 16190 316 +4.00 309 +5.23 288 19912 288 +7.88 325 12931 29 • CHj 278 11282 282 (sh) -5.63 257 19771 255 -8.97 o 2nc^ c i 249 15673 247 -19.21 220 20984 229 (sh) -4.27 22 'CH j C H * 221 (sh) +1.01 206 -7.28 204 +12.85 394 -0.56 CHj 386 -0.97 329 12374 331 +13.23 369 1464 358 +5.24 j c ö 296 13368 301 +13.73 . ! 270 18420 270 -32.92 331 (sh) +2.69 30 c & 313 15222 313 (sh) +3.13 223 14499 225 +15.62 H° 23 I ”’CH3 290 12721 288 +6.04 202 +9.34 251 12115 252 -7.43 229 (sh) -3.88 369 +11.18 219 17825 220 (sh) -3.17 318 15055 318 (sh) +7.66 208 +4.41 290 21208 294 +14.18 Cl. j ^ o a N : 269 22936 267 -29.90 CHj 326 14263 330 +13.96 31 CHj 223 15260 228 (sh) +6.22 302 11395 300 + 11.62 207 +11.10 269 13881 267 -29.90 H°CCÜ iT ^ C H 3 221 16080 222 +9.60 SCHj 352 (sh) +0.28 CHj 326 5447 338 +0.34 :x 307 -0.45 284 5160 277 (sh) +2.08 CH j 354 1787 362 +15.03 246 19520 256 +3.85 315 13206 314 +6.75 208 14447 224 -9.24 292 13831 292 +0.12 263 14614 262 -27.67 N 319 6659 322 +8.65 25 iHj CHj224 14406 222 -1.78 291 +10.64 203 +7.78 m" ^ h 261 7696 266 +5.55 3 3 i ' CHJ 250 250 21095 236 -14.03 CH, 208 16714 592 K. M. Kahn et al. ■ CD Studies on 1,5-Benzodiazepine Derivatives transition of the chromophoric system, or to a for­ does not influence the conformation of the hetero­ mally allowed one, which has, however, only a cyclic ring. The methyl group at C-5, on the other small e because of very weak overlap. Such a coin­ hand, would be in the same conformation of the cidence of band maxima is, however, not allways ring practically sy/7-periplanar to one bond of the seen; e.g. for the 7-methoxy compound 7 only the benzene ring. It is well-known from the X-ray first and third UV maximum coincides with a CD spectra of tetrahydroisoquinolines, bearing an al­ band, whereas the middle UV maximum is in be­ kyl residue in position 1, that this alkyl usually tween two Cotton effects of opposite signs. As for adopts rather a quasiaxial conformation. On the the parent compound one more additional effect other hand, this is not possible for 3, since such a can be seen arround 280 nm. A similar situation methyl group pointing into a boat conformation is holds for the chloro product 11. too overcrowded, and such a conformation is As found for the optically active benzodiazepin- never adopted in similar systems [47]. The confor­ 2-ones [45], methylation at N-l (2) leaves the mation of the heterocyclic ring of a molecule bear­ general features of the CD spectrum up to about ing a methyl group at N-5 should therefore be dif­ 250 nm, however, the Cotton effect around 230 nm ferent from that of 1 or 2, explaining the change shows up with opposite sign ( Ae = + 3.88), and the of the signs of the Cotton effects. If methylation complex system clearly becomes obvious from up at N-l does not, but at N-5 changes the conforma­ to seven Cotton effects with approximate wave­ tion, such an alteration of the 7-membered ring lengths at 319, 286, 261, 245, 230, 217 and 205 nm. should also be present in the N,N'-dimethyl deriv­ If the methyl group is positioned at N-5 (3, 4), ative 4. and, indeed, its CD spectrum closely re­ most of the Cotton effects change their signs and sembles that of 3 (Table I, Fig. 1). The N,N'-di- their wavelengths positions coincide practically methyl derivative 4, whose seven-membered with those of the UV spectra (Table I, Fig. 1). heterocyclic ring may be assumed to be relatively From these considerations it follows that methyla­ flexible, has also been investigated in MI-13 over tion at N-l does not strongly influence the CD of a wide temperature range (+ 20° to - 170°). A l­ the parent compound 1, whereas methylation at though, as usual, the magnitude of the CD in­ N-5 changes some of the Cotton effect appreciably creases, the overall shape of the curve remains (Table I, Fig. 1). This can be explained also from practically constant, which must be interpreted in molecular models. The hydrogen, attached to N-l, this way that the heterocyclic ring of the molecule points clearly out of the plane of the benzene ring, is not flexible at all. its replacement by the methyl residue obviously

Fig. 1. CD spectra of compounds 1 (----- ), 2 (------), 3 ( - - ) and 4 (------), recorded in CH3CN. K. M. Kahn et al. • CD Studies on 1,5-Benzodiazepine Derivatives 593

Benzodiazepin-2-thion band shift of the first UV band is mirrored by a similar blue shift of the corresponding CD band, The CD and UV spectra of individual thiono- and the two additional Cotton effects at 358 (As = amides indicate that the first n-Jt* band appears + 3.04 ) and 389 nm (Ae = - 0.9) have again to be around 360 nm, with an e value which is in genral ascribed to forbidden transitions. Analogous not much stronger than for any carbonyl com­ changes are seen if comparing the CD spectrum of pound. In our case (Table I), the chromophoric the parent compound 15 with those of N-5 methyl- system is, however, much more complicated, since substituted ones. The most striking differences are the system on one hand is an N-arylthionoamide, found above 330 and around 220 nm. At longest and on the other a N-thionoacyl-orf/zo-phenylene- wavelength a strong positive Cotton effect is ob­ diamine. Therfore, strong ji-k* bands appear at served around 360 nm, caused by an n-jr* transi­ long wavelengths (312 to 328 nm) with very high tion. The substituents at C-7 (24, 27, 30) do not intensity (e ~ 11000) and, at least in the UV strongly change this picture. The N-l methylated spectrum, a weak n-jz* band cannot be detected. and the N,N'-dimethylated compounds often show In the CD spectra, however, one to two rather in their UV spectra besides the before mentioned strong Cotton effects can be identified at longer four typical strong bands an additional one at wavelengths, and at least one of of them must cor­ around 375 nm (e ~ 2000), on the contrary in the respond to such an n-7i* transition. CD spectra of the dimethylated compounds one The maxima of the UV spectra depend charac­ Cotton effect has been “lost” so that now between teristically on the substitution pattern of the nitro­ longer wavelengths and 215 nm UV and CD max­ gen atoms, but in two different ways. The first ima (nearly) coincide. Some of the CD bands look strong band (325 to 328 nm) is hypsochromically not completely symmetric so that some additional shifted to about 310 nm, if N-l is methylated. The Cotton effects may be hidden, and this interpreta­ third UV band (253 to 256 nm for 15 and 16) is tion of the curves is strongly supported by those bathochromically shifted, if N-5 (17 and 18) is sub­ of some 7-substituted products (25, 28, 31), e.g. the stituted. The second UV band is slightly hypsoch­ chloro derivative 31 shows the “more usual” be­ romically shifted by methylation on N-l, and the haviour: Only a weak shoulder in the UV fourth, strongest UV band, is practically not influ­ spectrum around 375 nm, but a clearly visible first enced by the groups attached to the nitrogen CD band at 380 nm. Furthermore, around 250 nm atoms. The second, third, and fourth UV absorp­ a distinct (negative) shoulder appears in the CD tion bands correspond to three Cotton effects at spectrum (which has no equivalent in the UV practically identical wavelengths as found for the curve), and at shorter wavelengths three Cotton parent compound 15. The most characteristic dif­ effects can be identified, but only one UV band is ferences are observed at longer wavelengths: The seen. Taking into account these small but charac- first (negative) Cotton effect of 15 (Table I, Figs 2 terstic differences in band position below 260 nm, and 3) is located at 381 nm (Ae = - 4.64), and the it may be considered that the mentioned sign in­ second (positive) one does not coincide with the version of the CD band, which does not corre­ first UV maximum but is found at somewhat spond to a strong UV absorption, may actually not longer wavelength. Obviously there are (at least) be true and it is only the shift of band position two more prominent Cotton effects which have no which feigns this. eqivalent in the UV spectrum, so they should be As mentioned before, N-5-methylation (17 and assigned to either two different n-Jt* transitions 18) changes the signs of the Cotton effects corre­ (of the C=S moiety) or to one coming from dif­ sponding to the fourth UV band compared to ferent conformations. It is strongly believed that those found for 15 and 16 from negative to posi­ we are dealing here with the first case, since the tive values, which might be again caused by an al­ large difference in band positions of these first two teration of the conformation of the 7-membered Cotton effects is difficult to explain by assuming ring as discussed before for the corresponding only one single electronic transitions. 1,3,4,5-tetrahydro-2//-l,5-benzodiazepines. For 15 Methylation at N-l (16) leaves the CD spectrum (Fig. 2), the CD was also measured from + 20 rather unchanged as compared to 15. The blue down to - 170 °C in EPA, and the curves remain 594 K. M. Kahn et al. • CD Studies on 1,5-Benzodiazepine Derivatives

320 nm disappears, and a weak negative CD around 283 nm appears as expected, as the proton- ated compound should spectroscopically behave like a benzocycloheptene. for which a first band at 274 nm (tetralin) is observed. The rest of the spectrum is, however, not changed dramatically al­ though the magnitude of the Cotton effects are somewhat altered. The two N-methyl compounds 2 and 3 behave quite differently. Besides the disap­ pearance of the CD bands above 300 nm, the rest of the spectrum is also altered appreciably and be­ low 280 nm the signs of the Cotton effects are even partially exchanged. This fact must be noted with­ out any further comment since there is no experi­ Fig. 2. CD spectra of compound 15, recorded at + 20 °C ence with such optically active compounds where (------), - 2 0 °C (••••), -60 °C ( ------), -100°C (------), -140 °C (--), and -170°C (-•-) in EPA. two positive charges are located in orf/zo-position to a benzene ring. In the CD spectra of the four practically unaltered over this temperature range. different 7-methoxy compounds (7, 8, 9, and 10), It seems therefore, as if for this compound the het­ strong changes by acidification with bandshifts, erocyclic ring is very rigid, what was unexpected. sign-inversions and no simple regularities can be recognized. The only 7-hydroxy product which was Methylmercaptodihydrobenzodiazepines synthesized (6), resembles in its behaviour towards acid that of the analogous methoxy pro­ The two compounds 32 and 33 differ from all duct 10. These marked differences between the the others by being formed from the enolic form CD spectra measured in neutral and acidic solu­ of a thiolactam, therefore, their chromophore is tion prevails also for the 7-chloro compounds 12- completely different from the others. Although 14 of that type. the difference is only the presence (33) or absence Acidification of the thione analogues 15-18 (32) of the methyl group at N-5, their CD spectra brings about similar changes as already found for do not resemble each other at all (Fig. 3). Such the lactams. When N-5 is not methylated, a small a behaviour indicates different geometries (or a CD with fine structure above 400 nm is observed different equilibrium between several geometries) which can be assigned to the otherwise not dis­ for 32 and 33, which is obviously caused by stiffen­ tinctly seen first n-jz* band of the thiolactam chro­ ing of the heterocyclic ring so, that the presence of mophore. The next distincly seen Cotton effect is a methyl group on N-5 introduces already a strong present around 300 nm in a wavelength range steric interaction with the C-methyl group. where the free base shows just a minimum in be­ Furthermore, the UV maxima of both compounds tween two other Cotton effects. On the other do not coincide with those of the CD spectra and hand, if N-5 is methylated, then a stronger positive one has, therefore, to conclude that the bands in Cotton effect around 400 nm appears and this these two spectra have different parentages. Also, might perhaps overcompensate the just mentioned the relative magnitudes of the CD spectra are first n-Ji* Cotton effect for the thioamide chromo­ quite different for both compounds. Acidification phore. The relatively strong and narrow CD band of the solutions gives also no further hint about around 300 nm is, however, observed as in the case the origin of the Cotton effects since the subtances of the other two mentioned compounds 15 and 16. are quickly destroyed by acid. The three thioamidophenols 23, 24, and 25 have quite different stability towards alkali: whereas for Salt derivatives of selected benzodiazepine 23 and 25 after reacidification the orignal derivatives spectrum is obtained, this is not the case with 24, In presence of few drops of strong acid the first which obviously is destroyed under such condi­ Cotton effect of the parent compound 1 around tion. In 23 and 25, the amide proton is replaced by K. M. Kahn et al. • CD Studies on 1,5-Benzodiazepine Derivatives 595

450 A.[nm]

Fig. 3. CD spectra of compounds 15 ( ------), 32 (------) and 33 (------), recorded in CH3CN. the methyl group and is, therefore, more stable. In identification, m.p. and rotational values are com­ 24, an N-thioacylated para-aminophenol structure pared with those from the literature. is present, and (i.e. unprotected on N-l) therefore, degradation by alkali is very rapid. 4-Methyl-l ,3,4,5-tetrahydro-2H-l,5- benzodiazepine (1) Conclusion m.p. 85 °C, - [aß 0 = - 59.4 (c = 1.5, CHC13); These spectroscopic studies on thirty three chi­ ref. [10]: m.p. 86-87 °C, - [aß 0 = - 60.4 (c = ral benzodiazepine derivatives demonstrate that 1.82, CHC13). circular dichroism allows to prove experimentally the complexity of this chromophoric system in 1.4-Dimethyl-l ,3,4,5-tetrahydro-2H-l ,5- much more detail compared to UV spectroscopy. benzodiazepine (2) The CD spectra give also evidence for conforma­ oil, - [aß 0 = - 67.4 (c = 1.3, CHC13); ref [10]: tional changes in the 7-membered ring. X-ray oil, - [a]o = - 68.3 (c = 1.39, CHC13). studies which are under way for selected com­ pounds will allow to correlate the Cotton effects 4.5-Dimethyl-l ,3,4,5-tetrahydro-2H-l ,5- with the absolute configuration and also the benzodiazepine (3) adopted conformation of this series of compounds. oil, - [aß 0 = - 47.2 (c = 1.2, CHC13); ref. [10]: oil, - [aß 0 = - 50 (c = 1.28, CHC13). Experimental The CD spectra were recorded with a Jouan- 1.4.5-Trimethyl-l ,3,4,5-tetrahydro-2H-l ,5- Jobin-Yvon-ISA dichrograph Mark III, connected benzodiazepine (4) on-line with a Hobert Computing AT and a Jasco- oil, - [aß 0 = + 86.3 (c = 1.39, CHCI 3); ref. [10]: 600 dichrograph, connected on-line with a Tandom oil, - [aß 0 = + 87.2 (c = 1.48, CHC13). PCA, in acetonitrile solutions at room temper­ ature in cells of 0.01 to 2 cm path lengths. The UV 4-Methyl-l ,5-diacetyl-l ,3,4,5-tetrahydro-2-H-l ,5- spectra were performed with a PU 8740 (Philips) benzodiazepine (5) instrument in cells of 0.01 to 2 cm path lengths, from solutions prepared for the CD measurments. m.p. 190 °C (benzene), - [a]o* = + 28.4 (c = 2, The chiral 1,5-benzodiazepine derivatives were CHCI3); ref. [10]: m.p. 193 °C (benzene), - [aß 0 = synthesized according to reference [10]. For their + 29.6 (c = 2.3, CHCI 3). 596 K. M. Kahn et al. ■ CD Studies on 1,5-Benzodiazepine Derivatives

1.4.5-Trimethyl-7 -hydroxy-1,3,4,5-tetrahydro-2H-1,4,5-Trimethyl-7 -chloro-1,3,4,5-tetrahy dro-2H-l, 5- 1.5-benzodiazepine (6) benzodiazepine (14)

m.p. 45-46 °C (petroleum ether), - [a]^ = oil, - [a]^° = + 105 (c = 10.7, CHCI3); ref. [10]: + 97.1 (c = 1.5, CHC13); ref [10]: m.p. 46-47.5 °C oil, - [a]^ü = + 104.8 (c = 10.61, CHC13). (petroleum ether), - [a]^ = + 95.7 (c = 1.65, CHCI3). 4-Methyl-l ,3,4,5-tetrahy dro-2H-l,5-benzodiazepin- 2-thione (15) 4-Methyl-7-methoxy-1,3,4,5-tetrahydro-2H-l ,5- m.p. 144-145 °C, - [a ]^0 = + 125.0 (c = 2.19, benzodiazepine (7) CHCI3); ref. [10]: m.p. 145-146 °C, - [a]^° = m.p. 94-95 °C (petroleum ether), - [a]^ = + 126.2 (c = 2.23, CHCI3). - 13.1 (c = 1.23, C H C I3 ); ref. [10]: m.p. 93-94 °C (petroleum ether), - [a]o = - 11.7 (c = 1.36, 1.4-Dimethyl-l ,3,4,5-tetrahy dro-2-H-l ,5- C H C I3 ). benzodiazepin-2-thione (16) m.p. 168-169 °C (aeetone/hexane), - [ö]d° = 1.4,-Dimethyl-7 -methoxy-1,4,5-trimethyl-l,3,4,5 - + 290.0 (c = 3.13, C H C I3 ); ref [10]: m.p. 169-170 tetrahydro-2H-l,5-benzodiazepine (8) °C (acetone/hexane), - [ö]d = + 293 (c = 3.35, m.p. 42-43 °C (hexane), - [a]o = - 19.5 (c = C H C I3 ). 1.5, C H C I3 ); ref. [10]: m.p. 43-44 °C (hexane), - [a]g = - 21 (c = 1.9, C H C I3 ). 4.5-Dimethyl-l ,3,4,5-tetrahy dro-2-H-l ,5- benzodiazepin-2-thione (17) 4.5-Dimethyl-7-methoxy-1,3,4,5-tetrahy dro-2H-l,5- m.p. 162 °C (cyelohexane), - [a]o = + 425.0 (c = benzodiazepine (9) 1.56, CHCI3); ref [10]: m.p. 160-160.8 °C (cyclo- hexane), - [a ]^ = + 429 (c = 1.71, CHC13). oil, - [a]^° = - 108 (c = 2.9, C H C I3 ); ref. [10]: oil, - [ a ] o = - 111 (c = 3.15, C H C I3 ). 1.4.5 -Trimethy 1-1,3,4,5-tetrahy dro-2-H-l, 5- benzodiazepin-2-thione (18) 1.4.5-Trimethyl-7-methoxy-1,3,4,5-tetrahy dro-2H- 1.5-benzodiazepine (10) m.p. 103 °C (acetone/hexane), - [a]o 0 = + 535 (c = 1.5, CHCI3); ref [10]: m.p. 100-102 °C (ace­ oil, - [a]^ = + 103.5 (c = 2.54, CHC13); ref. [10]: tone/hexane), - [a]n? = + 533 (c = 1.62, CHC13). oil, - [a]g> = + 105.7 (c = 2.78, CHCl 3j.

4-Methyl-7-nitro-l ,3,4,5-tetrahydro-2H-l ,5- (-)-4-Methyl-7 -chloro-1,3,4,5-tetrahy dro-2H-l ,5- benzodiazepin-2-thione (19) benzodiazepine (11) m.p. 245 °C (toluene), - [a]o = + 12.0 (c = 1.54, m.p. 154-155 °C (petroleum ether), - [ajj^ = DMSO); ref [10]: m.p. 246 °C (toluene), - [a]o = - 24.0 (c = 0.5, C H C I3 ); [10]: m.p. 155.5-157 °C + 13.7 (c = 1.67, DMSO). (petroleum ether), - [a]o = - 25.3 (c = 0.79, C H C I3 ). 1.4-Dimethyl-7-nitro-l ,3,4,5-tetrahydro-2H-l,5- benzodiazepin-2-thione (20) 1.4-Dimethyl-7-chloro-l,3,4,5-tetrahydro-2H-l,5- benzodiazepine (12) m.p. 247 °C (acetone/hexane), - [a]o° = + 528 (c = 5.0, CHCI3); ref. [10]: m.p. 246 °C (acetone/ m.p. 58-59 °C (petroleum ether), - [a]o} = hexane), - [a]o = + 526 (c = 5.13, CHC13). - 24.5 (c = 2.2, C H C I3 ); [10]: m.p. 60 °C (petro­ leum ether), - [a]o = - 25.3 (c = 2.37, C H C I3 ). 4.5-Dimethyl-7 -nitro-1,3,4,5-tetrahydro-2H-l ,5- benzodiazepin-2-thione (21) 4.5-Dimethyl-7 -chloro-1,3,4,5-tetrahy dro-2H-l ,5- m.p. 190 °C (acetone/tetraehloromethane), - benzodiazepine (13) [a]o = + 675 (c = 1.08, CHC13); ref [10]: m.p. 191- oil, - [a]o = - 53.0 (c - 1 .1 1 , CHCU); ref. [10]: 192.5 °C (acetone/tetrachloromethane), - [ ö ] d = oil, - [a]b° = - 56 (c = 1.25, CHC13). + 679 (c = 1.20, CHCI 3). K. M. Kahn et al. • CD Studies on 1,5-Benzodiazepine Derivatives 597

1,4,5-Trimethyl-7-nitro-l ,3,4,5-tetrahydro-2H-l ,5- 1.4.5-Trimethyl-7-methoxy-l,3,4,5-tetrahydro-2H- benzodiazepin-2-thione (22) 1.5-benzodiazepin-2-thione (28) m.p. 45 °C (acetone/tetrachloromethane), - m.p. 102 °C (acetone/tetrachloromethane), - [a]^ = + 343 (c = 1.79, CHC13); ref. [10]: m.p. [aß0 = + 834 (c = 8.25, (CH3)2CO); ref. [10]: m.p. 43 °C (acetone/tetrachloromethane), - [a\v = 103 °C (acetone/tetrachloromethane), - [aß0 = + 342.5 (c = 2, CHC13). + 837 (c = 8.5, (CH3)2CO). 1.4-Dimethyl-7-chloro-l ,3,4,5-tetrahydro-2H-l ,5- benzodiazepin-2-thione (29) 1.4-Dimethyl-7-hydroxy-l ,3,4,5-tetrahydro-2H-l ,5- benzodiazepin-2-thione (23) m.p. 101-102 °C (acetone/tetrachlorometh­ ane), - [a]^0 = + 319 (c = 2.07, CHC13); ref. [10]: m.p. 157 °C (dichloromethane), - [a]o = m.p. 101-103 °C (acetone/tetrachlorometh­ + 228.0 (c = 1.25, CH3CN); ref [10]: m.p. 157 °C ane), - [a]g> = + 318 (c = 2.4, CHC13). (dichloromethane), - [aß* = + +229.9 (c = 1.47, 4.5-Dimethyl-7-chloro-l,3,4,5-tetrahydro-2H-l,5- CH3CN). benzodiazepin-2-thione (30) m.p. 135-137 °C (acetone/tetrachlorometh­ 4.5-Dimethyl-7-hydroxy-l,3,4,5-tetrahydro-2H-l ,5- ane), - [a]^° = + 474.0 (c = 4.1, CHC13); ref. [10]: benzodiazepin-2-thione (24) m.p. 136-137 °C (acetone/tetrachlorometh­ ane), - [a]g> = + 476.7 (c = 4.3, CHC13). m.p. 204-205 °C (dichloromethane), - [a \$ = + 230 (c = 2.07, CH3CN); ref. [10]: m.p. 204-206 1.4.5-Trimethyl-7 -chloro-1,3,4,5-tetrahydro-2H-l ,5- °C (dichloromethane), - [a]^ = + 231.5 (c = benzodiazepin-2-thione (31) 2.22, CH3CN). m.p. 71 °C (acetone/tetrachloromethane), - [a]o = + 307 (c = 3.6, CHC13); ref. [10]: m.p. 70 °C (acetone/tetrachloromethane), - [a]o = + 309 (c = 1.4.5-Trimethyl-7-hydroxy-l,3,4,5-tetrahydro-2H- 3.8, CHC13). 1.5-benzodiazepin-2-thione (25) 4-Methyl-2-methylmercaptodihydro-l,5- m.p. 116-117 °C (dichloromethane/hexane), - benzodiazepine (32) [a]o = + 563 (c = 1.21, CHC13); ref [10]: m.p. 115- oil, - [a]o = + 113.0 (c = 4.25, CHC13); ref. [10]: 116 °C (dichloromethane/hexane), - [a]^ = + 567 oil, - [a]$ = + 114.1 (c = 4.39, CHC13). (c = 1.47, CHC13). 4.5-Dimethyl-2-methylmercaptodihydro-l,5- benzodiazepine (33) 1.4-Dimethyl-7-methoxy-l ,3,4,5-tetrahydro-2H-l ,5- oil, - [a]$ = + 304.0 (c = 2.79, CHC13); ref. [10]: benzodiazepin-2-thione (26) oil, - [aß 0 = + 303.8 (c = 2.91, CHC13). m.p. 36-37 °C (acetone/tetrachloromethane), - Acknowledgements [aß 0 = + 42.0 (c = 3.99, CHC13); ref. [10]: m.p. 37- 38 °C (acetone/tetrachloromethane), - [gc]d= M. H. is grateful to the University Grants Com­ + 43.5 (c = 4.09, CHC13). mission of Pakistan and to Alexander-von-Hum- boldt-Stiftung for partial support of this work. F. M. thanks Deutscher Akademischer Austausch­ 4.5-Dimethyl-7-methoxy-l,3,4,5-tetrahydro-2H-l,5- dienst for the award of a scholarship within a Ph. benzodiazepin-2-thione (27) D. sandwich program. K. M. K. expresses his grati­ tude to the Ministerium für Wissenschaft und m.p. 155 °C (dichloromethane/tetrachlorometh- Forschung, Baden-Württemberg for granting a fel­ ane), - [a ß 0 - + 497 (c = 2.31, CHC13); ref. [10]: lowship. The support of these studies by the Fonds m.p. 154.5-156 °C (dichloromethane/tetrachloro- der Chemischen Industrie is greatly ac­ methane), - [a]o = + 492 (c = 2.5, CHC13). knowledged. 598 K. M. Kahn et al. ■ CD Studies on 1,5-Benzodiazepine Derivatives

[1] G. A. Archer, L. H. Sternbach, Chem. Rev. 68. 747 [27] W. Voelter, O. Oster, G. Jung, E. Breitmaier. Chimia (1968). 25, 26 (1971). [2] L. H. Sternbach, Prog. Drug Res. 22, 229 (1978). [28] O. Oster, E. Breitmaier, G. Jung, W. Voelter, Circu­ [3] M. C. Aversa, P. Gianetto, G. Romeo, P. Ficarra, M. lar Dichroism Studies of Flavone Glycosides, in: G. Vigorita, Org. Magn. Reson. 15, 394 (1981). Applied Physical Chemistry, Vol. 1, p. 61, Publishing [4] M. G. Bock, R. M. DiPardo, B. E. Evans, K. E. House of the Hungarian Academy of Science, Bu­ Rittle, W. L. Whitter, V. M. Garsky, K. F. Gilbert, J. dapest (1971). L. Leighton, K. L. Carson, E. C. Mellin. D. F. Veber, [29] G. Jung, M. Ottnad, W. Woelter, Eur. J. Drug R. S. L. Chang, V. J. Lotti, S. B. Freedman, A. J. Metab. Pharmacokin. 3, 131 (1977). Smith, S. Patel, P. S. Anderson, R. M. Freidinger, J. [30] G. Barth. W. Voelter, H. S. Mosher, E. Bunnenberg, Med. Chem. 36, 4276 (1993). C. Djerassi, J. Am. Chem. Soc. 92, 875 (1970). [5] T. Blair, G. A. Webb, J. Med. Chem. 20,1206 (1977). [31] G. Barth. R. Records, E. Bunnenberg, C. Djerassi, [6] F. Malik, M. Hasan, D. Rosenbaum, H. Duddeck, W. Voelter, J. Am. Chem. Soc. 93, 2545 (1971). Magn. Reson. Chem. 27, 391 (1989). [32] U. Weser, E. Bunnenberg, R. Cammack, C. Djerassi, [7] N. W. Gilman, P. Rosen, J. V. Earley, C. Cook, L. J. L. Flohe, G. Thomas, W. Voelter, Biochim. Biophys. Todaro, J. Am. Chem. Soc. 112, 3969 (1990). Acta 243, 203 (1971). [8] G. Töth, A. Levai, A. Szöllösy, Liebigs Ann. Chem. [33] E. Bayer, A. Bacher, P. Krauß, W. Voelter, G. Barth, 803 (1992). E. Bunnenberg, C. Djerassi, Eur. J. Biochem. 22, [9] F. Malik, M. Hassan, K. M. Khan, S. Perveen, G. 580 (1971). Snatzke, H. Duddeck, W. Voelter, Liebigs Ann. [34] G. Jung, U. Weser, W. Voelter, Hoppe-Seyler’s Z. Chem. 1861 (1995). Physiol. Chem. 353, 720 (1972). [10] F. Malik, M. Hassan, K. M. Khan, S. Perveen, G. [35] U. Weser, G. Barth, C. Djerassi, H. -J. Hartmann, P. Snatzke, H. Duddeck, W. Voelter, Liebigs Ann. Krauß, G. Voelcker, W. Voelter, W. Voetsch, Bi­ Chem. 127 (1996). ochim. Biophys. Acta 278, 28 (1972). 11] J. Davoll, J. Chem. Soc. 308 (1960). [36] G. Barth, W. Voelter, E. Bunnenberg, C. Djerassi, J. 12] B. Yde, I. Thomsen, M. Thorsen, K. Clausen, S.-O. Am. Chem. Soc. 94, 1293 (1972). Lawesson, Tetrahedron 39, 4121 (1983). [37] U. Weser, H. Rupp, F. Donay, F. Linnemann, W. 13] W. Voelter, E. Bayer, R. Records, E. Bunnenberg, Voelter, W. Voetsch, G. Jung, Eur. J. Biochem. 39, C. Djerassi, Liebigs Ann. Chem. 718. 238 (1968). 127 (1973). 14] W. Voelter, E. Bayer, R. Records, E. Bunnenberg, [38] U. Weser, G. -J. Strobel, H. Rupp, W. Voelter, Eur. C. Djerassi, Chem. Ber. 102, 1005 (1969). J. Biochem. 50, 91 (1974). 15] G. Barth, W. Voelter, E. Bunnenberg, C. Djerassi, J. [39] A. Attanasio, S. Fuchs, D. Gupta, H, Kalbacher, E. Chem. Soc. Chem. Commun. 355 (1969). Pietrzik, W. Voelter, Chemiker-Ztg. 100, 438 (1976). 16] W. Voelter, E. Bayer, G. Barth, E. Bunnenberg, C. [40] J. Zarbock, H. Oschkinat, E. Hannappel, H. Kal­ Djerassi, Chem. Ber. 102, 2003 (1969). bacher, W. Voelter, T. A. Holak, Biochemistry 29, 17] W. Voelter, G. Kuhfittig, (i. Schneider, E. Bayer, 7814 (1990). Liebigs Ann. Chem. 734. 126 (1970). [41] M. Czisch, M. Schleicher, S. Hörger, W. Voelter, T. 18] W. Voelter, G. Kuhfittig, E. Bayer, Angew. Chem. A. Holak, Eur. J. Biochem. 218, 335 (1993). 82, 985 (1970); Angew. Chem., Int. Ed. Engl. 9, [42] J. Bernhagen, R. A. Mitchell, T. Calandra, W. 964 (1970). Voelter, A. Cerami, R. Bucala, Biochemistry 33, 19] W. Voelter, G. Kuhfittig, O. Oster, E. Bayer, Chem. 14144 (1994). Ber. 104, 1234 (1971). [43] M. Hasan, N. Rashid. K. M. Khan, G. Snatzke, H. 20] W. Voelter, H. Bauer, G. Kuhfittig, Chimia 27, 274 Duddeck, W. Voelter, Liebigs Ann. Chem. 889 (1973). (1995). 21] W. Voelter, H. Bauer, G. Kuhfittig, Chem. Ber. 107, [44] M. Hasan, N. Rashid, K. M. Khan, S. Perveen, G. 3602 (1974). Snatzke, H. Duddeck, W. Voelter, Liebigs Ann. 22] W. Voelter, R. Records, E. Bunnenberg, C. Djerassi, Chem. 1871 (1995). J. Am. Chem. Soc. 90, 6163 (1968). [45] K. M. Khan, F. Malik, M. Hasan, S. Perveen, G. 23] W. Voelter, G. Barth, R. Records, E. Bunnenberg, Snatzke, W. Voelter, Z. Naturforsch. 50b, 1869 C. Djerassi, J. Am. Chem. Soc. 91, 6165 (1969). (1995). 24] W. Voelter, Hoppe-Seyler’ s Z. Physiol. Chem. 351. [46] E. Heilbronner, H. Bock, Das HMO-Modell und 133 (1970). seine Anwendung, Verlag Chemie, Weinheim 25] D. L. Elder, E. Bunnenberg, C. Djerassi, M. Ikeh- (1970). ara, W. Voelter, Tetrahedron Lett. 727 (1970). [47] K. Kefurt, G. Snatzke, F. Snatzke, P. Trska, J. Jary, 26] H. Vergin, H. Bauer, G. Kuhfittig, W. Voelter, Z. Coll. Czech. Chem. Commun. 41, 3324 (1976). Naturforsch. 27b, 1378 (1972).