A-J‘ 4:132 w '3 $.13? 3.3 "-\‘$~\MJ 5'52if’1‘6y1 rfiL‘i.3.‘§‘3f3 , 3, 3 X I 3, ¥¢§333. I“. TEES" f», *3 1" ES." "3'; , . V' I I "5““ < p3-3 . .3 3W5,- .¢3, 3 ,3.3.3,33..3 3.- L V.,).» .3 45-3 ‘3, 3 FE. mtg-'33 33,333,3 3.33 ~I333 . 3‘ I 3 '1', '3 r3- 3 3 3‘,‘ 73,131,- ‘7; 3.. I , a. I1_¢.‘CL33 31' 3‘“ :L-“ )234 3 ‘ 33 3373.336fl3 .L ' [3| ' ' : 3 (I? (...‘L'A‘HIH‘ ' _ 3 23t' ””333 33333 If)?“ “3‘49"?" "'M ”'1. "3:393'3343'“ ‘ 33‘. .3.' FE". 3 53"“:31-‘I {:dfé‘ 3i ". ' ”Raw U3"- '\r3.' 3' ‘ 333%.“, 34:1,. "‘333";3I,I\ 3‘.“ n .--: : ... -:+;3'3‘.‘: '-.'".‘3337"-,""“3i" “‘1,- - ' “" " 33 3,3 .3-.L'. . 33.333: 3 ”33.1.3. .:-,3 3 335333.33, ,. 'Hx Yfi'}, '1'? ' {d 33:"). 3'A":?:bz"""q '5‘“ ' 31 "'5' \ "l‘ll.l'3\::¢'¢ ‘fi'fib $4M ‘ }|:;3:3353‘J 3:313:33: ‘r‘R ‘ 3331‘). '3 . 3 --‘ 2“ 3' “ “"3 ~""“23«3‘3'3.§3~.333~ -- 3:33: . .3 III‘IAIH . 3,, ,. ,3”, 6"" -. 153‘: """:',w..13r,‘.'f . k ‘ “33")LY :nug‘v‘ $33.3“ 'U¢:I‘3.3‘:':33 9" u .‘ 3'3”" . v.33 "1' 3‘3 3153‘ .- 3,3533%”? 553363 33:3». .‘3l "3' “3::313‘ . .. 323- 3333'fi1?§'¢3,1"" 33 .33..3,3/3._ "\3 3.,“ 3‘3 F}. 3 ‘ {'3' - . . 33 ),\3}-3 3 ' 3,33v33lh3fi 3 - ... ,3, 4.3363 3,3 3 3,3" 3.3%“? 35“ , 1" 3:,‘L,3'3~,:L 0, r 'lezu 3 3 3' Aug" 3:93:31 1.33“ “ah-33$?" 03.39.3(11 3 '3..33Q33,333‘ ‘53:; ‘3‘33 (‘AI 33! 'L'Q‘fl'r'v.‘ ‘3‘ {ficfiéfi'g} '~: . .3. 3 'Qki‘“ - . "323333.34; . '3 ' (fix: I 3 ‘ ~ . ‘ 4, " "33$“? 2, 1“,,33: 3» . 23.23.33‘3- .7. .33 3*. \'Ir rmfifi‘figfi» #:3333333 5‘2:— 1;: , yzafigg— F3- . .-:-.~ - ‘3. A W}? 3 I ’3 .33 . D .‘ ..3... . . - "I ' r . . k' .3..- 33:33; ... 3,... 3.3-3. .... . ., .3 . 3' «w. . . 3 333.33. -- 33mm 3 3 . . . 3'33. 3 '. 3 . 3 h - . . '. r. .. Eut- - . ._~ .. 3.3. 333.3:-'-3-~ 3:33 =3. 333' =3.“ " ‘3 . . ..3 .: _ .. 3..., 33.3..3....3....3..,._ ..‘ ‘ ' ,3 j3,I"“-‘-.-.'3'. )‘.,3 3353.3 33?; “M35 .,,.3“"‘” ”3.33.!“ 31 ‘.3 1 , 3W.‘m':: ’1 ,3, R»: '3 ..' ‘ ' I‘ I" " ' '-' "I' ‘I;/ 3-}. ‘,‘>\ 1. 'I'I'f',’ ' \ (”7' I v .3 l 3 '5‘ x“ If: ....3 3'3 ‘3‘3 t" ”3..., '7 33 '33 “v.“ I 35 fig“; ,9- ..- 3 ,3»,- . I 3 , 3 3 “,va \3.,, .,3r,‘3‘13,33g5u,, 3 vv'gmfin 3. .3 33 ,3. L 3. . 3 3 3... 333..., lf'g'."'r‘\,i"'"13*. 3 953:“ .. , ,3!" at no. -3 3' ‘7' .Z"3:.3‘:'-“:*,1"' "‘ . ‘33‘332333 333%...» ' 3 ‘ ' ' --'.3'.'3 '3 ‘ . 3733,:IHI3-rc.‘ 33\' ‘ A“ . .9 3- - ', ’> . - ' I 3" ',-f',)’l ”(‘3‘ .‘f 3. ";‘ V3333: ' Yuk no '- 0. . I ' .' 'OA.J.- 3.3.37 '3‘3 Q.‘I3",}‘7{' J. A 3 . 3.3:?! $561: 3 ' :ir.v,:.~:3333 . ‘ (,3 .1... 3.3.32” § ‘7‘?! “3‘23 23 V~ w: 3 3 3 " "'x3'31-33‘33 “'4“ fig'fi-P‘fi'g‘ v""fi‘ “Hi I C‘hfi 'ty'.:%.:" 3:33.33 .. 3333,... 3., '-V. --' 3 lil- . . ...—~— " . q (3‘ ' 3 n A 3‘3‘ ‘ 9 ‘3 3‘ 'r‘ \ - I I‘ . , .33.... ,;,. 333 ...,33 3:3». . 2'” 3'3‘ 3‘" {It . 3 75%;" nun '3‘? '3‘: “RR '3'?" 'I "' ':‘::4;'3- 'Wr' 'q‘{- 5'41“" .3 \e'. ' 39 I rtfl-"Kfiué (£3. wVA’I 23* ~ .- 7' 1" ' ""33" '3 " "~"\"':1‘3‘3'Q3\)"v::‘ In!“ J.” """' MP?“ .3-3 ,3 3““ _,‘,:3,,37,333,« 3:3«3 - 3 33333 :7 ' ..3‘ “‘ 33.13; '1'?..'.‘¢.,'.‘ H;- .3. .3; 3- a “ ‘ .iz‘x . " - 3 . 'i' “ ' :.3"- ' ". . ".37.: 1" - .2'.‘ 3 ".‘7 3‘: {3h xfifig‘, '2' .‘ ' . ‘ "3} “:3... ‘3‘. 33. , ‘. 17 'L 'J' C3 I‘ Q! ' . 3 '1. ' run '5 I ,3. . "I N, . '33 333 .. 3 3 .3 ,3. ,33 3 ‘ 33 3 3 .3.'3‘33313333v' ”7,! ",I'X3wy312‘n'” g5} 1' '3 3 ‘ 3 3 ' -- ‘ 3 3 3 3. _"3 ‘3 , 3 .' j 3-, ‘3 '3' 3 .‘.‘ ' ' 3 3":3‘ fibe ‘..|." 3\. ,0) :" '_' ‘ ‘ 3 ' 3 3'f;-.' .1 ‘ 3 -. 3.. .. ‘3: “33333 3 3,3333. 3. 3‘ '. 3 ‘ ' ' ' ' :" - I -'3 33» .. .j : 4 . I- .3 '3'3,‘ GIVE" l"? ”div "utxv'. 3‘3' .3‘3 333. 33 3 333 3 _ , 3 3 ~'-33 3 33.3 13.3. -.'3,'3 .3 3 3 3.3..A3‘. 3 , \3 .3333 3,-- \ 3 3 3 3.3 I ’ V'I'Ii)\4‘3fl ”53'?“ ‘- 33.»: 3-...‘-. -' 3 ‘..I-"-3 "cu: 3333393.- . ... 33 '3 :3 ,3 .3. .3_ "3 .3 ,_ ' ~- ..3'3 . 3 . ' - 3-3} 3 ' . “3.3". 7:33 iifi'fih: . . . . .3 .. 33 *3, ..., , 3 . ,3 '31'3'3547'" 33.3, . .. , . 3-33 3 3, .3 .. . . . 3 _ 3 . .. . _ g .3 3' _,3' .3 3, 3 3 . 3 3 . ,2,’ 3 -. 3 ,-'-_,' ~ . 3‘, 33 3 3”...” €333};3.3‘3 . 2 » ' . ' ' - "J I ‘3'" '2 h ' ". ‘ -‘ ' ' ' ' ' .- 13"“? "" .7 3. "3 3 33 " 3’. . 3 ‘3 'L"'. 3 33. , 'hcui' ’ :31qu n7"; ‘ .,_ .‘ 3 _ ' , ,‘ ,. '_‘ ., " ' . . 3‘ ,. . ...,. :3"; ‘ . . ,3 I 3 ‘. ' ' ' ' I, 3 3-3 I 3, , ,';3' 3-3', ’53 ”3’7"." I ARV" :“ . '4' H ' ' ' ' "'I ' " " "'- ' I' L" "'I' .' '33“ 1' '3"OO'II\:"' “viz": , 7": 4 if»? "'-,3le""" , ' «;3- 3' ' , 3,« ‘3'," ,, ,,‘3,~,.',33 ~., 3-13,. 3.3.334.- ..g, .3333: :33? £33. . " .3" 3 .- ' 3_3 ,3. ..3'H3H331'3 33'3, ' '.‘3'&..‘vf~l:'l ',.'33~ Dr: 3333'. 3. ' '3'.3.'r . '|_' ' ' 3 ‘ ' " " ' “7"" .3313? 3'3.'3 3333l33 ,'3§3,. . ~33. 39' 3" 7.7.333 V" "'l H } "‘u'f'fiI'WI" , ,3 33 3 3 3.. '3. ' ' 3333 31,}: .".3l‘3 ’3‘, "333.3 '-° "(IW/ 3\%'.' 1 a. (‘3. "33, ‘ 3 , . an ,33 , ‘ ' 3- 3 ,3 “I !,' ‘3', . .3 ' 3 333 3 am” I'm ' 3'"? I ' 3 3 ‘ ' r ' , . . ' ' I" " " 3 '< ‘ 3 I' . 3. 3" ‘ .. "' h ' ‘ 'u 333'3333L3'3... .‘.' " ":'.- 3 .3313). 2' .343-Arfm‘3 .31..‘.'....'3.. -..‘. 3.3333133333391713"'.L'-’-'.\¢.'3'.3'I'3'-'3 . 3533..“ ”Willi? lMIfll'II‘IIIlilWWWflfllWflflMM L 3 1293 01096 7192 I. ~ : ‘W‘b Q . . -0 o PART 1 HETEROCYCLIC PHOTOCHEMISTRY: THE PHOTOCHEMICAL SYNTHESIS OF B-LACTAMS PART 2 THE 13c NMR SPECTRA OF NAPHTHALENE CROWN ETHER COMPLEXES: FIELD INDUCED fl POLARIZATION AND CROWN ETHER CONFORMATIONAL CHANGES BY Mark R. Johnson A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1978 Sim: Star hove Prod Mark R. Johnson \E}C>j\ ABSTRACT CA ’ PART 1 HETEROCYCLIC PHOTOCHEMISTRY: THE PHOTOCHEMICAL SYNTHESIS OF B-LACTAMS PART 2 THE 13C NMR SPECTRA OF NAPHTHALENE CROWN ETHER COMPLEXES: FIELD INDUCED N POLARIZATION AND CROWN ETHER CONFORMATIONAL CHANGES By Mark R. Johnson The first part of this thesis describes attempts to photochem- ically synthesize B-lactams. Two proposed methods were studied, one of them being successful. The first method involved a hoped for 1,3-acy1 migration of a 3,6-dihydro-2(1H)-pyrazinone. A general synthesis of this little-known heterocyclic system was developed, starting from an a-amino acid and an a-aminoketone. Photolysis, however, led to slow decomposition and formation of no identifiable products. I l H HN N + several 89 * , NH2 '9‘ R , N/ R l 'W no ——-> Ph N/ identifiable products Chem R dn Mark R. Johnson The second method involved sulfur dioxide extrusion from 1,1- dioxo—é-thiazolidinones. These were synthesized by literature methods. Stereochemical assignments for sulfones i: and 32 were made by x-ray crystallography. An attempt to use standard NMR techniques to assign the stereochemistry of the corresponding sulfoxides resulted in com- pletely erroneous assignments. Photolysis of cis sulfone $2 in t-butyl alcohol/acetonitrile resulted in formation of both cis and trans B-lactams, £3 and £3, with the cis isomer the major product. Photo- lysis of the trans sulfone also gave the B-lactams, with the trans B- lactam the major product. Photolysis in isopropanol gave reduction products 32 and 3%, along with some B-lactams, indicating radical intermediates. 17° T3 S "I N hu 3 02 CH3 5 14353 + 24 O c ’I "a 2 “3 t Ph CH3 25 29 28 CH3 Pb hv " (12P11 34,0; 23 "’ 29 ”CH, 4- PhCstoa-i-Pr 02 H3 2 CH3 3] 30 In the course of the first study, the photochemistry and thermal chemistry of 2,4,4-trisubstituted Az-oxazolin-S-ones were examined,.and a dramatic substituent effect of the trifluoromethyl group noted. ethe WETE Mark R. Johnson P hi «AA «— . fl / ..., ph CF3 IH:=#( CE3 (322CFk3 (3QZCFE; Pb A N 5r" H CH3 Ph 3 In Part 2 of this thesis the 13C NMR spectra of naphthalene crown ethers 3% through 3% and their alkali and alkaline earth metal complexes were studied. The complexation induced shifts of the aromatic carbons are dependent on only the cation's charge, consistent with field induced H polarization of the naphthalene, and correlate reasonably well with INDO calculated charge density changes. The chemical shift changes of the ether carbons are cation dependent and provide information about conformational changes in the crown ring. ”:Wb 42-44, n=3,4,5 . CHC o C 2H2 " O—(-CH2CH20 46-48,n=3,4,5 45 di th: pat Nat ACKNOWLEDGMENTS I would like to thank Lynn Sousa for an excellent job of directing my graduate education. No better a job could have been done. Thanks also go to the other members of the research group for their friendship and fellowship, and to my wife, Cheri, who not only patiently waited for me to finish, but typed most of this thesis. Financial support from Michigan State University and the National Science Foundation is also gratefully acknowledged. ii HETEROCYCLIC PHOTOCHEMISTRY: THE PHOTOCHEMICAL SYNTHESIS OF B-LACTAMS PREFACE . . . . . . SYNTHESIS AND PHOTOCHEMISTRY OF 3,6-DIHYDRO-2(1H)- PYRAZINONES . . . . Introduction . TABLE OF CONTENTS PART 1 Synthesis of 3,6-Dihydro-1(1H)-pyrazinones . . Photochemistry of 1,3,5,6,6-Pentamethyl-3-pheny1- 2(1H)-pyrazinone . Experimental . SYNTHESIS AND CHEMISTRY OF 1,1-DIOXO-4-THI- AZOLIDINONES. . . . Introduction . Synthesis of 1,1-Dioxo-4-thiazolidinones O Stereochemistry of Substituted 4-Thiazolidinones . Photolysis and Thermolysis of 1,1-Dioxo-4-thi- azolidinones: Photolysis and Thermolysis of 1,1-dioxo-4-thi— azolidinones: Experimental . Results. Discussion . iii Page 12 14 23 23 25 28 39 43 48 TABLE OF CONTENTS (continued) Page PHOTOLYSIS AND THERMOLYSIS 0F 2,4,4-TRISUBSTITUTED Az-OXAZOLIN-S—ONES: SUBSTITUENT EFFECT OF A TRI- FLUOROMETHYL GROUP . . . . . . . . . . . . . . . . . . . . . . 59 Introduction . . . . . . . . . . . . . . . . . . . . . . 59 Results: Photolysis of AZ-Oxazolin-S-ones 2 and,3%. . . . 62 Results: Thermolysis of Az-Oxazolin-S-ones 2 and.%% . . . 63 Discussion: Thermolysis of Az-Oxazolin-S-ones 2 and.%%° . 64 Discussion: Photolysis of AZ-Oxazolin-S-ones 2 and 3% . . 65 The Effect of the Trifluoromethyl Group . . . . . . . . . 66 Experimental. . . . . . . . . . . . . . . . . . . . . . . 70 PART 2 THE 13C NMR SPECTRA 0F NAPHTHALENE CROWN ETHER COMPLEXES: FIELD INDUCED n POLARIZATION AND CROWN ETHER CONFORMATIONAL CHANCES Introduction . . . . . . . . . . . . . . . . . . . . . . . 76 Methods and Results . . . . . . . . . . . . . . . . . . . 78 Aromatic Carbon Shifts . . . . . . . . . . . . . . . . . . 88 Ether Carbon Shifts . . . . . . . . . . . . . . . . . . .106 Conclusions . . . . . . . . . . . . . . . . . . . . . . .109 Experimental . . . . . . . . . . . . . . . . . . . . .111 REFERENCES e e e e e e e e e ' 113 iv LIST OF TABLES TABLE Page I. Aromatic Solvent Induced Shifts in Sulfoxides . . . . . 31 2. Eu(fod)3 Induced Shifts in Sulfoxides . . . . . . . . . 33 3. Calculated and Observed Eu(fod)3 Induced Shift Ratios . 37 4. Yields from Photolysis of cis Sulfone 25 in Different Solvents . . . . . . . . . . . . . .th. . . . . . . . 41 5. Correlation of Chemical Shift Changes with Charge Density Changes . . . . . . . . . . . . . . . . . . . . 100 LIST OF FIGURES FIGURE Crown Ethers Studied by 13C NMR Spectroscopy . . Chemical Shift Changes from Titration of 2,3-Naph- 20-Cr-6 with Alkali Metal Salts in Deuteromethanol at Room Temperature. Crown Ether Concentration is No.2 M. O O O O O O O O O O O O O I O O O O 0 Chemical Shift Changes from Titration of 2,3-Naph- 20-Cr-6 with Barium and Calcium Salts in Deuterometh- anol at Room Temperature. Crown Ether Concentration is No.2 M. . . . . . . . . . . . . . . . . . . . Chemical Shift Changes from Titration of 2,3-Naph- l7-Cr-5 with Alkali Metal Salts in Deuteromethanol at Room Temperature. Crown Ether Concentration is mOZ M. O O O O O O O O O O O O O O O O O O O O 0 Chemical Shift Changes from Titration of 2,3-Naph- 17-Cr-5 with Barium and Calcium Salts in Deuterometh— anol at Room Temperature. Crown Ether Concentration is NOIZ M. O O O O O O O O O O O O O O O O O O O 0 Chemical Shift Changes from Titration of 1,8-Naph- 21—Cr-6 with Alkali Metal Salts in Deuteromethanol at Room Temperature. Crown Ether Concentration is No.2 MO I I O O O O O O O O O O O O O O O O O O O 0 Chemical Shift Changes from Titration of 1,8-Naph- 21-Cr-6 with Barium and Calcium Salts in Deuterometh- anol at Room Temperature. Crown Ether Concentration is No.2 M. . . . . . . . . . . . . . . . . . . . Chemical Shift Changes from Titration of 2,3-Naph- 14-Cr-4 with Alkali Metal Salts in Deuteromethanol at Room Temperature. Crown Ether Concentration is No.2 M. O O O O O O O O O O O I I I O O I O O O 0 Chemical Shift Changes from Titration of 1,5-Naph— 22-Cr-6 with Alkali Metal Salts in Deuteromethanol at Room Temperature. Crown Ether Concentration is WOZ M. O O O O O O O O O O O O O O O O O O O O 0 vi Page 77 79 . 80 . 81 . 82 83 . 84 . 85 86 LIST OF FIGURES (continued) FIGURE Page 10. 13C NMR Spectra of 2,3-Naph-20-Cr-6 and 1:1 Complexes in Deuteromethanol at Room Temperature. Crown Ether Concentrations are No.2 M. . . . . . . . . . 89 11. 13C NMR Spectra of 1,8-Naph-21-Cr-6 and 1:1 Complexes in Deuteromethanol at Room Temperature. Crown Ether Concentrations are No.2 M. . . . . . . . . . 90 12. 13C NMR Spectra of 2,3-Naph—l7-Cr-5 and 1:1 Complexes in Deuteromethanol at Room Temperature. Crown Ether Concentrations are No.2 M. . . . . . . . . . 91 13. 13C NMR Spectra of 2,3-Naph-14-Cr-4 and 1:1 Complexes in Deuteromethanol at Room Temperature. Crown Ether Concentrations are No.2 M. . . . . . . . . . 92 14. 130 NMR Spectra of 1,5-Naph-22-Cr-6 and 1:1 Complexes in Deuteromethanol at Room Temperature. Crown Ether Concentrations are No.2 M. . . . . . . . . . 93 15. 13C NMR Spectra of 1,8-Naphthocrowns and other 1,8-Disubstituted Naphthalenes in Deuteromethanol at Room Temperature. . . . . . . . . . . . . . . . . . . 94 16. 13C NMR Spectra of 2,3-Naphthocrowns and other 2,3-Disubstituted Naphthalenes in Deuteromethanol at Room Temperature. . . . . . . . . . . . . . . . . . . 95 17. Correlation of INDO Calculated Charge Density Changes with Measured Chemical Shift Changes for Potassium and Barium Complexes of 2,3-Naph-20-Cr-6 and 1,8-Naph-21- Cr-6 in Deuteromethanol. x and + are C(9) of Potassium and Barium Complexes of 1,8-Naph-21-Cr-6, Respectively. 99 18. INDO Calculated Charge Density Changes for Aromatic Carbons of 2,3-Crowns with Plus One Mbnopole (MI) in Indicated Positions. Charge Density Change is Pro- portional to Distance of Dashed Line from Carbon. . . .102 19. Chemical Shift Changes of Aromatic Carbons of 2,3- Crowns upon Complexation with Potassium. Chemical Shift Change is Proportional to Size of Arrow. . . . . .103 20. INDO Calculated o, H, and Total Charge Density Changes for Complexation of 2,3-Naph-20-Cr-6 with Plus One Mbnopole and Measured Chemical Shift Changes for 2,3-Naph-20-Cr—6 with Potassium Salt. . . . . . . .104 vii PART 1 HETEROCYCLIC PHOTOCHEMISTRY: THE PHOTOCHEMICAL SYNTHESIS OF B-LACTAMS Preface The potential usefulness of photochemistry in organic synthesis has been often noted in recent years. Despite this "potential", little use has been made of photochemistry as a synthetic tool on a laboratory scale. A variety of reasons have been suggested for this, the most common being the general unpredictability of photochemical reactions, and the sensitivity of photochemical reactions to modest changes in substitution. This apparent capriciousness is due to our difficulty in dealing with molecules on potential energy surfaces other than that of the ground state. Another principal difficulty with photo— reactions is the frequent intermediacy of radicals of various sorts. A brief examination of useful laboratory synthetic reactions reveals that most of them involve ionic species rather than radical inter- mediates, although many industrial reactions involve radicals. With the possible exception of reactions involving transition metals, rad- ical reactions are not very useful synthetically, on a laboratory scale. Despite these difficulties, there is at least one area in which photochemical reactions have found considerable use, the preparation of strained ring systems. In particular, four mem- bered rings of various sorts, both carbocyclic and heterocyclic, have frequently been synthesized by photochemical means. These 1 2 include bicyclobutanes, oxetanes,3 and lactams. cyclobutanes, Perhaps the two most important reasons for the use of photochemical reactions in synthesizing these systems are the difficulty in using ground state reactions in making four membered rings, and the sensitivity of these strained systems to many reagents, making them unusable. In contrast, there is reason to believe that photochemical reactions may generally prefer to give high energy ground state systems such as strained rings," and these strained rings generally show no special instability under photo- chemical reaction conditions, mainly because they usually don't absorb light. One of these ring systems, whose synthesis by photochemical means has attracted only limited attention, is the 2-azetidinone, or B-lactam ring, I. B-Lactams are extremely important compounds, as this ring system is found in all of the penicillin, 2, and cephalosporin antibiotics, 3.5 Currently, all of the penicillins 2-a zetidinone I Penicillins 2 / Hzofic H3 (qu g‘“ Cepl,a|osporins 3 6 'APA 4 F" bi SO: arc and cephalosporins in use are produced either entirely by bacteria or are synthesized from 6-amino penicillanic acid, 4, which is also made microbially. The total synthesis of the penicillins and cephalosporins is exceptionally difficult, until recently there being only one of penicillin6 and one of‘cephalosporin,7 neither of which has had much impact on the commercial production of any useful drugs.8 Contrary to earlier opinion, the minimum structural requirement for biological activity is considerably less than the multitude of functionality surrounding the B-lactam rings of the penicillins and cephalosporins. The B-lactam is, of course, essential, but the minimum structural requirements for biological activity seem to be those described by structure 3. Some examples of compounds which show signifigant biological activity are also shown below.9’1° H O‘NO 2.2-: TAr’ H o I o H’COZH Nocardicin A Q ‘- 0 c1120" 1! Bl ° 1 oz” x HZ. Clavulinic Acid With this in mind, we have chosen to attempt to devise a practical photochemical synthesis of B-lactams, which would be applicable to the formation of biologically important compounds. A handful of other photochemical syntheses of B-lactams are known, none of which, however, have any practical importance. The most common photochemical method is the photolytic decomposition of diazo compounds, followed by either C-H bond insertion11 or 2 Wolf rearrangement.1 In these cases, it is the carbene chemistry, not the photochemistry, which is of interest. Other reactions 3 include the cyclization of substituted acrylamides,1 and the ring contraction of 3-oxo-pyrroline 1-oxides.1“ The two methods which we have proposed and studied illustrate the two extremes to which one might proceed. The first proposed method involved an unprecedented photochemical migration in a complex and virtually unknown heterocyclic system. The second method studied involved a simple reaction with ample precedent, using a known and readily available heterocyclic precursor of the B-lactam. Perhaps significantly, the former method was unsuccessful, and the latter successful. In the course of the first investi- gation the photochemistry of another heterocyclic system was explored, and an interesting substituent effect on both photolytic and thermal reactions observed. The above three investigations comprise the heterocyclic photochemistry portion of this thesis and will be discussed in the above order. Synthesis and Photochemistry of 3,6-Dihydro-2(1H)-pyrazinones Introduction The first proposed photochemical method for the synthesis of B-lactams involves a 1,3-acyl shift in a 3,6—dihydro-2(1H)-pyrazinone. R. I 1 .. . 2 N o W" (SP—NR —)- ' ——-)- 3 /5 - ’ . I \~’ 1 =.( ‘ The proposed reaction would be analogous to that frequently ob- served with B,y-unsaturated ketones.15 Of course, our system is a very "hetero" version of this, the carbonyl group being an amide, and the double bond one between carbon and nitrogen, rather than two carbons. This makes any predictions based on B,y-unsat- urated ketones tenous. Another objection to the scheme is that the reaction involves two functional groups, an amide and an imine, which are not very reactive photochemically. While these are serious problems (hindsight is always 20/20), the scheme does have some nice features. If the reaction were to occur, the B-lactam would have a nitrogen attached to C(3), where a heteroatom sub- stituent is generally required for biological activity. Further, the nitrogen substituent at C(3), being an imine, would be man- ipulable, so that one could choose any side chain amide that was desired. Hydrolysis of the imine group in the product would dispose of C(3) of the starting material and the groups attached to it, allowing a choice of substituents at C(3) which would enhance cleavage of the C(2)-C(3) bond in the pyrazinone. These would then function as disposable photo-activating groups. Another nice feature of this scheme also involves the imine group of the product. Photochemically, acyclic imines are generally inert, due to energy dissipation by syn-anti isomerization. This would protect the B-lactam product from undergoing secondary photoreactions, a frequent problem in preparative organic photochemistry. With this in mind we chose as our initial synthetic target, 1,3,5,6,6- pentamethyl-B-phenyl-Z(1H)-pyrazinone,é. The substituents at Ph / 6 C(3) and C(5) were chosen so to enhance cleavage of the C(2)-C(3) bond, and the other methyl groups chosen so as to minimize possible photochemical hydrogen abstraction or migration reactions. Synthesis of 3,6-Dihydro-2L1H)-pyrazinones16 A search of the literature revealed only one method for the synthesis of the desired heterocyclic system. This method involved O-alkylation of a diketopiperazine, and hence, was limited to the 17 We have developed production of S-alkoxy substituted derivatives. a more general method for the synthesis of this system, which does not limit the choice of substituents on the ring. This approach H R (CH ) 0' ii R 173+ ill " ’ CH H H 3 enViSi acid 5 bonds prote; 0f thf amino CODdE] ketoni aCCOII iSOpr. methy alSo SYntT envisions the product arising from the condensation of an a-amino acid and an a-amino ketone, with formation of the amide and imine bonds as key steps. The necessary steps in this sequence are protection of the amino group and activation of the carbonyl group ii thl C) hi ‘1’ 1++rj H2| C)” “1” of the amino acid, coupling of the activated amino acid with the amino ketone, deprotection of the remaining amino group, and condensation of the amine with the ketone. The required amino ketone, 3-methyl-3-(methylamino)-2-butanone, 7, was prepared according to literature procedures by chlorination of methyl isopropyl ketone with sulfuryl chloride,la followed by reaction with methylamine.19 The required amino acid, a-phenylalanine, 2, was also prepared by literature methods, using the Strecker amino acid synthesis.20 >’fi~ 502: 12 Al c2252" (xs) 4’8‘ c1 NHCH3 7' KCN NH Cl Pb PhC CH3 ___‘__). "C' CH3 029 NH3 H20 9 Ecol-I NH3 yiel prot stit ketc dry . carr coup otheI amine aQDe and e 336~d was p of al Protection and activation of a-phenylalanine was accomplished by treatment with trifluoroacetic anhydride to give the N-trifluoro- acetyl amino acid, followed by cyclization using thionyl chloride to 4-methyl-4-phenyl-2-trifluoromethyloxazolin-S-one, 2, in 76% 1 Other methods for activation and yield from a-phenylalanine.2 protection of amino acids fail when applied to hindered a-disub- stituted amino acids. Coupling of the oxazolinone and a-amino ketone were accomplished in good yield simply by mixing the two in dry acetonitrile overnight. Deprotection and cyclization were carried out in one step by treatment of a methanol solution of the coupling product {a with dry hydrogen chloride for seven hours at 500 C. The crude product contained some starting material, and was purified by silica gel chromatography, eluting with 2% methanol in methylene chloride, which gave the pure 3,6-dihydroé2(1H)~pyraZinone g in 49% yield (69% based on recovered starting material, which was recycled). Higher reaction temperatures resulted in cleavage of both amide bonds of 10, giving the methyl ester of a-phenylalanine along with unreacted 10. As expected from literature reports,21 mm other deprotection methods failed for this protected a-disubstituted amino acid. kg'was found to be stable to sodium hydroxide in aqueous dimethyl sulfoxide, sodium borohydride in refluxing ethanol, and even lithium aluminum hydride in refluxing tetrahydrofuran. To demonstrate the generality of this synthetic scheme, the 3,6-dihydropyrazinone from an a-monosubstituted amino acid, alanine, was prepared, using the same a-amino ketone as before. Reaction of alanine with excess trifluoroacetic anhydride at 140° C for 9+7 two } In t% more olinc The Q cYtli hYdro with glyci: trifll 10 0 Ph coo (CF co) 0 cc’2" soc: CPh H0 CF3C02H HNHcoc1=3 N — ””3 cps 14 1 5 Ph CH CN HCI 9+7 —3—> oi" CH3... NHCOCF3 SO'C 10 two hours gave the Aa-oxazolinone,,bl, directly, in 57% yield.22 In the case of 4-monosubstituted oxazolin-S-ones the A3 isomer is more stable. As this isomer is less reactive than the Az-oxaz- olinones, coupling with,z required 15 hours in refluxing acetonitrile. The coupling product,’£%, was extremely difficult to purify, and was generally deprotected without purification. Deprotection and cyclization to‘IQ could be affected by treatment with methanolic hydrogen chloride as before, or, more conveniently, by treatment with a mixture of aqueous potassium hydroxide and ether. An attempt to prepare the 3,6-dihydropyrazinone from a,a-diphenyl- glycine and ,7, was msuccessful. Treatment of the amino acid with trifluoroacetic anhydride gave the N-trifluoroacetyl amino acid, lg, ll “'3me (cracogzo CH3 \ __ NH3 A 0:3 11 k 4"? + HCl/MeOH NHCOCF3 CHaCN 12 which was cyclized as before with thionyl chloride to give the oxazolinone,l£5. Coupling of the oxazolinone with l was effected by mixing at 500 C for 48 hours to give 13. Attempts to deprotect i3 with methanolic hydrogen chloride at 400 C gave the methyl ester of diphenylglycine along with unreacted 16. Lower temper- mm atures resulted in no reaction. PI: Pgfiozo (crgcohoa " C°2H socn Ph "Hag “3332" NHCOCF3 N -— I‘3 i4 15 C N I HCI Pb COZCH3 43L» .. ——..->- ...>r NHcoc1=3 C“3" ""2 Ain't 16 12 This synthetic route to 3,6-dihydro-2(1H)—pyrazinones seems general except when extremely hindered a-disubstituted a-amino acids are used. While only one a-amino ketone was used, there is no reason to expect difficulties with others. Yields are quite good, 8 being formed in 20% yield from a—phenylalanine, and kg in 41% yield from alanine. Photochemistry of l,3,5,6,6-Pentamethy1-3jphenyl-2(1H)jpyrazinone, Q Photolysis of Olin a variety of solvents failed to produce any detectable amounts of any B-lactam. Photolyses were conducted in acetonitrile, methanol, hexane, acetonitrile with trifluoro- acetic acid, and acetonitrile with acetophenone. In each case except the last, the only materials obtained were starting material and polymeric material. Photolysis through pyrex with acetophenone afforded the pinacol of acetophenone as the only identifiable product. Decomposition of starting material under all conditions was fairly slow. This leaves us then with the questions of what went wrong, and what can we do about it. The problem seems to be one of lack of reactivity, rather than another process going on faster than the desired reaction. In order to "heat up" the C(2)-C(3) bond even more, the attempt to synthesize the 3,3-diphenyl pyrazinone, was made. As already noted this was unsuccessful. In more general terms the failure of this method is perhaps not unexpected, and the mistake made was a strategic one. The mis- take was made in attempting to try and predict the photochemical behavior of a fairly complex heterocyclic compound, with little 13 precedent for the desired photoreaction. Considering our relatively primitive ability for predicting photochemical reactions it is perhaps wiser to choose a well documented and common photochemical reaction to use in B-lactam synthesis. The chemistry described in the next section of this thesis was our response to our first unsuccessful method and the system and reaction chosen were based on the above considerations. me we 1 spe the Chl (lia was mate tami 301E extr Chle Unde EXPERIMENTAL General. Proton NMR spectra were recorded on a Varian T-60 spectro- meter with tetramethylsilane as an internal standard. Infrared spectra were obtained on a Perkin-Elmer 237B grating spectrophotometer. Mass spectra were taken on a Hitachi Perkin-Elmer RMU-6D spectrometer. Microanalyses were performed by Instranal Laboratory, Rensselaer, N.Y. Melting points are uncorrected. Preparation of 3-chloro-3-methyl-Z-butanone18. Over a period of 1.5 h 65 g (0.5 Mole) of sulfuryl chloride was added to 43 g (0.5 Mole) of 3-methyl-2—butanone in an ice-cooled round bottom flask. Upon comple- tion of the addition the solution was allowed to warm to room tempera- ture and stirred for 48 h. The mixture was distilled directly from the reaction mixture and then redistilled to yield 35 g, 60%, of 3- chloro-3-methy1—2-butanone; NMR (CD013) 6 1.65 (6H,s),2.35 (3H,s); ir (liq film) 1720 (vs),1100 (s),1125 cm-1 (3). Preparation of 3-methy1-3-methylamino-Z-butanone (1)19. This material was prepared by a slight modification of the published procedure. The material obtained from the literature method contained an unknown con- taminant, which was removed by extraction of a methylene chloride solution of the ketone with hydrochloric acid, neutralization of the extract with solid sodium hydroxide, and reextraction into methylene chloride. The extract was dried (MgSOu), filtered, concentrated under reduced pressure, and redistilled to yield the desired ketone. 14 Th: at Pre wet cya 150 tur eth wa t1 W8 S bot dil mov $01 the con} pyri prod nine 1670 15 The ketone is stable when stored in the cold, but slowly decomposes at room temperature; NMR (CD013) 6 1.2 (6H,s), 2.15 (3H,s), 2.20 (3H,s). Preparation of q:phenyla1anine (8)20. In a 2 1 round bottom flask a. were placed, in the order mentioned, 66 g (1,0 Mo1e) of potassium cyanide in 100 mL of water, 59 g (1.1 Mole) of ammonium chloride in 150 ml of water, and 134 mL of aqueous ammonia (sp. gr. 0.9). The mix- ture was shaken and 120 g (l MOle) of acetophenone in 300 mL of 95% ethanol was added. The flask was stoppered and heated for 5 h in a water bath maintained between 600 and 80°C. After 5 h the solution was cooled and carefully added to 800 mL of 12 M HCl in a 5 1 round bottom flask which was immersed in an ice bath. The solution was diluted with 1 1 of water and refluxed for 2 h. Solvent was then re- moved by distillation, and the solid mass taken up in 600 mL of ab- solute ethanol. The solution was filtered to remove inorganic salts, the salts washed with another 600 mL of ethanol, and the filtrates combined. The filtrate was then concentrated to 750 mL, 100 mL of pyridine added, and the solid product collected by filtration. The product was dried under vacuum to yield 35.5 g, 21%, of a-phenylala- nine; NMR (TFAC) 6 1.9 (3H,s), 7.0 (5H,s); ir (nujol) 1590 (m),and 1670 cm"1 (m). Synthesis of N-trifluoroacetyl-a-phenylalanine21. A solution of 11.1 g (67.3 mmol) of a—phenylalanine and 10 mL (68 mmol) of trifluoroacetic anhydride in 30 mL of trifluoroacetic acid was stirred at room tem- perature for 8 h under a nitrogen atmosphere. The trifluoroacetic acid and anhydride used were dried by distillation from phosphorous pentoxide. Solvent was removed under reduced pressure to yield a 16 brown solid which was purified by filtration through a short silica gel column (Mallinckrodt CC-7), eluting with methylene chloride. Solvent removal yielded 14.6 g, 83% of N-trifluoroacetyl-a-phenylalanine, (mp 9 131.5-132.5 °C (Tit.21 126-128°Cxx NMR (CDC13) 5 2.0 (3H,s), 6.5 (2H, br. 3.), 7.3 (5H,m); ir (nujol) 1710 (vs),1550 cm‘1(s). Preparation of N—trifluoroaetyl-a,a-diphenylglycine (151; To an ice mm bath cooled solution of 543g(22 mmol) of a,a-diphenylg1ycine (Aldrich) 3120 mL of trifluoroacetic acid, 4.6 g (22 mmol) of trifluoroacetic anhydride was added over a period of 1 h. The solution was stirred overnight at room temperature, and then solvent removed under reduced pressure. The residue was treated with 100 mL of ether, and the un- reacted starting material removed by filtration. Solvent removal from the filtrate gave a white solid which was recrystallized from chloro- form/hexane to yield 4.2 g, 62%, of N-trifluoroacetyl-a,a-diphenylgly- cine, mp(l83-185°C);ir (nujol) 3300 (m) 3250-2900 (s),1740 (s),1710 cm-1(s). Anal, Calcd. for C16H12F3N03: C, 59.45; H, 3.74; N, 4.33. Found: C, 59.47; H, 3.98, N, 4.35. Preparation of 4-methyl-4-pheny1-2-trifluoromethyl-Az-oxazolin-5- one (9)21 'b phenylalanine in 30 mL of thionyl chloride (purified by distillation . A solution of 14.6 g (60 mmol) of N-trifluoroacetyl-a- from triethyl phosphite) was heated to 60°C and maintained at that temperature for one h. Excess thionyl chloride was removed at room temperature using aspirator vacuum, and the residue distilled at reduced pressure to yield 12.6 g, 92%, of 4-methy1-4-pheny1-2-tri- f1uoromethy1-A2-oxazolin-5-one,(bp 520 C (0.5 mm) lit.21 53-570 C) in til: Wit SOlL Pres yell Chlc ed U: meth: % 3‘0xC methy f 0f 3- Stirr 17 NMR (CDC13) 6 1.9 (3H,s), 7.2-7.6 (5H,m); ir (neat) 1850 (vs),1680 (s), 1370 cm'1 (vs). Preparation of4,4-diphenyl-2-trifluoromethyl-Az-oxazolin-S-one (15). IUD A solution of 3,0 g (9.3 mmol) of N-trifluoroacetyl-OL,CL-diphenyl glycine in 12 mL of thionyl chloride was heated at reflux for 2 h. The solu- tion was allowed to cool and the remaining thionyl chloride removed with aspirator vacuum. The residue was distilled under vacuum to yield 2.37 g, 84%, of 4,4-dipheny1-2-trif1uoromethyl-A2-oxazolin-5-one, (bp 87°C, 0.4 mm); NMR (CDC13) 6 7.1-7.4 (m); ir (neat) 3050 (m), 1850 (vs), 1690 (s),1370 (vs),1225 (vs),1170 cm'l (vs); mass spectrum (70 eV) m/e (rel. intensity), 305 (<1), 209 (20), 208 (100), 180 (13), 165 (13), 152 (30), 77 (12). Preparation of 4wmethy1-2-trif1uoromethyl-Az-oxazolin-S-one (11).22 A solution of 4.32 g (54.7 mmol) of alanine in 15 ml of trifluoroacetic anhydride was refluxed in an oil bath maintained at 140°C for 2 h. The solution was allowed to cool, volatile materials removed under reduced pressure, and the residue distilled at atmospheric pressure to give a yellow liquid (bp 140°C). This material was taken up in methylene chloride, washed with saturated sodium bicarbonate, dried, and concentrat- ed under reduced pressure to yield 4.7 g, 51%, of 4-methyl-2-trif1uor- methyl-AZ-oxazolin-S-one. Preparation of 2-pheny1-2-trifluoroacetamido-N-methyl-N-(2-(2-methyl- 3-oxo)butyl)propanamide (10). A.solution of 7.7 g (31 mmol) of 4-'- mm methy1-4-phenyl-2-trifluoromethyl-Az-oxazolin-S-one and 3.6 g (31 mmol) of 3-methyl-3-methylamino-Z-butanone in 200 mL of dry acetonitrile was stirred at room temperature for 24 h. Solvent removal under reduced 18 pressure gave an oil which was crystallized by addition of a little ether. Recrystallization from ether gave 8.8 g, 78%, of the desired amide,,19 (mp 108-109°C); NMR (CDC13) 0 1.30 (3H,s), 1.35 (3H,s), 2.05 (3H,s), 2.10 (3H,s), 2.50 (3H,s), 7.10 (5H,s), 8.8 (1H,br.s, re- moved by D20); ir (nujol) 3300 (m),3050 (s),1710 (s),1625 cm"1 (m); mass spectrum (70 eV) m£g_(re1. intensity), 360 (<1), 315 (21), 216 (48), 181 (20), 169 (20), 131 (32), 119 (34), 103 (47), 72 (99), 69 (100). Anal; Calcd. for C17H21F3N203: C, 56.98; H, 5.91; N, 7.69 Found: C, 56.70; H, 5.84; N, 7.69. Preparation of 2,2-dipheny1-2-trifluoroacetamido-N—methyl-N-(2-(2- methyl-3-oxo)-buty1)acetamide (12). A solution of 1.93 g (6.33 mmol) of 4,4-dipheny1—2-trif1uoromethyl-A2~oxazolin-5-one and 0.728 g (6.33 mmol) of 3~methy1-3-methy1amino-2-butanone in 125 mL of dry acetonit- rile was heated at 50°C for 42 h. Solvent was removed under reduced pressure, the residue taken up in methylene chloride, washed with 0.1 M HC1(aq), dried (MgSOH), filtered, and solvent removed to yield an oil Which crystallized from ether/pentane, 66%, (mp 139-141°C); (CDC13) 6 1.2 (6H,s), 2.0 (3H,s), 2.3 (3H,s), 7.0-7.4 (5H,m), 8.8 (1H,br.s); ir (nujol) 3250 (m), 1720 (s), 1660 cm"1 (3). “Anal; Calcd. for C22H23F3N203: C, 62.85; H, 5:51, N, 6.66. Found: C, 62.65; H, 5.19; N, 6.55. On one occasion the material crystallized as a monohydrate, mp 154°-156° C (-H,0). Anal, Calcd. for C22H25F3N20“: C, 60.27; H, 5.75; N, 6.39. Found: C, 60.42; H, 5.93; N, 6.38. 19 Preparation of 1,3,5,6,6-pentamethyl-3,6-dihydro-2(1H)1pyrazinone(l3). mm A solution of 2.40 g (14.2 mmol) of 4-methyl-2-trifluoromethyl-A3- oxazolin-S-one and 1.65 g (14.2 mmol) of 3-methyl-3—methylamino-2— butanone in 125 mL of dry acetonitrile was heated at reflux for 10 h. The solution was allowed to cool and solvent removed under reduced pressure to yield a dark oil, which consisted mainly of N-methyl-Z- trifluoroacetamido—N-(Z-(2-methy1-3—oxo)-buty1)propanamide, in ~85% yield (determined by NMR). This material was extremely difficult to purify and was generally used as obtained. A small amount of material was purified by column chromatography on silica gel (Silicar CC-7), eluting with methylene chloride/methanol (1/1), and had the following data, mp 78-80°C; NMR (DMSO) 6 1.2 (6H,s), 1.3 (3H,d), 1.9 (3H,s), 3.0 (3H,s), 4.6 (1H,q), 9.0 (1H,br.d). The crude amide product was taken up in a mixture containing 75 mL each of l M KOH (aq) and diethyl ether, and stirred at room tempera- ture for 13 h. The ether layer was removed, and the aqueous layer extracted four times with methylene chloride.) The combined methylene chloride extracts were dried (MgSOH), filtered, and solvent removed to yield an oil which crystallized on cooling in a dry ice/acetone bath. Recrystallization from hexane/chloroform afforded 0.8 g (33% based on azlactone 11)of the desired diazine; NMR (CDC13) 1.45 (6H,s), 1.50 (3H, d, J = 6 Hz), 2.05 (6H,d, J = 2 Hz), 2.90 (3H,s), 4.10 (1H,q of d, J = 6 Hz, 31 - 2 Hz); ir (neat) 2960 (m1 1640 (vsL,1450 (s1 1380 cm-1 (8); mass spectrum (70 eV) gig (rel. intensity) 169 (3.3), 168 (26), 153 (14), 127 (25), 125 (10), 111 (13), 110 (23), 99 (14), 93 (31); 13C NMR (CD013) 6 169.32 (s), 166.43 (s), 60.09 (3), 55.88 (d), 26.49 (m), 24.98 (m), 23.80 (m), 23.02 (m), 20.39 (m). 19 20 ‘Anal. Calcd. for C9H16NO: C, 64.25; H, 9.59; N, 16.65. Found: C, 64.40; H, 9.25; N, 16.47. Preparation of l,3,5,6,61pentamethyl-3ephenyl-3,6-dihydro-2(1H)1pyra- zinone (6). A solution of 500 mg (1.39 mmol) of the keto amide 10 in m mm 60 mL of dry methanol was treated with gaseous hydrogen chloride for a period of 7.5 h. The solution was cooled in an ice bath for the first 0.5 h, and then heated at 50°C for the remainder of the time. The solution was stirred at room temperature for another 2.5 h after the hydrogen chloride was turned off. Solvent was removed under re- duced pressure to give a gummy residue, which was taken up in NZO mL of methanol. This solution was made basic by addition of triethylamine (pH > 12), and the triethylamine hydrochloride precipitated by addition of 80 mL of ether. The solution was chilled in an ice bath, filtered, and the filtrate dried over sodium sulfate and filtered again. Sol- vent removal under reduced pressure gave an oil which contained a mix- ture of the desired product and starting material. This material was chromatographed on silica gel (Mallinckrodt CC-7) eluting first with methylene chloride, which brought off 145 mg of starting material, and then 3% methanol/methylene chloride, which brought off 167 mg (49%) of 1,3,5,6,6-pentamethy1-3-phenyl-3,6-dihydro—2(lH)-pyrazinone. An analytical sample was prepared by recrystallization from ether/ pentane, mp 98-99OC; NMR (CDC13) 6 1.2 (3H,s), 1.45 (3H,s), 1.8 (3H,s), 2.2 (3H,s), 2.9 (3H,s), 7.0-7.4 (5H,m); ir (CHC13) 2900 (s) 1680 cm‘1 (vs); mass spectrum (70 eV) mlg_(re1. intensity), 245 (16), 244 (100, 229 (24), 201 (12), 187 (20), 172 (52), 146 (78), 145 (28), 127 (10), 104 (48), 103 (26); uv (hexane),}.max - 264 nm (e - 400) 257 nm (E = 542). 21 Anal. Calcd. for C12H20N20: C, 73.74; H, 8.25; N, 11.47. Found: C, 73.55; H, 8.26; N, 11.56. Unsuccessful attempts to deprotect ijhenyl-Z-trifluoroacetamido—N- methyl-N-(2-(2-methyl-3-oxo)-buty1)propionamide (10). Samples of 2 mm 200 mg or 250 mg of 10 were submitted to each of the following con- mm ditions: a) 100 mL of 0.2 M sodium hydroxide in 9/1 dimethylsulfoxide/ water at 65°C for 42 h; b) 110 mL of dimethylformamide stirred over 0.8 g of sodium hydroxide at 100°C for 40 h; c) 15 mg of sodium borohydride in 50 mL of ethanol at 65°C for 8 h; and d) 15 mg of lith- ium aluminum hydride in 100 mL of tetrahydrofuran at reflux for 24 h. In each case standard work-up provided only unreacted starting materi- al. Attempted deprotection- of 2,2-diphenyl-Z-trifluoroacetamido-N-methyl- N-(2-(2-methy1-3-oxo)-buty1)acetamide (16). A solution of 250 mg (0.59 mm mmol) of keto amide 16 in 50 mL of methanol was treated with HCl (g) mm continously for 8 h, with the temperature maintained at 0°C for the first 30 min and at 40°C for the remainder. The solution was allowed to cool, solvent removed under reduced pressure, the residue taken up in 10 mL of methanol, and then treated with 5 mL of triethylamine. The solution was mixed with 100 mL of ether, chilled in an ice bath, and filtered to remove the salt. Solvent removal under reduced pres- sure gave a residue of mainly starting material and a little methyl diphenylglycinate. Exploratory photolyses of 6. Solutions of 6 in hexane, acetonitrile, m m nwmhanol, and acetonitrile with trifluoroacetic acid were irradiated through either vycor or quartz filters, for varying time~periods. Photolysis in acetonitrile with acetophenone was done through a pyrex 22 filter. In each case solvent was removed under reduced pressure, the crude photolysate was examined by NMR spectroscopy, and the residue chromatographed on silica gel. In no case was anything other than starting material or polymer obtained. 23 Synthesis and Chemistry of 1,1-Dioxo-4-thiazolidinones Introduction Perhaps the most common and easily predicted photochemical reactions are those that involve extrusion of a small stable molecule. Examples include the loss of nitrogen upon photolysis of diazo “ and extrusion compounds,23 loss of carbon monoxide from aldehydes,2 of sulfur dioxide from sulfones.2S This last reaction frequently occurs when sulfones are photolyzed, and was chosen as a suitable candidate for use in B-lactam synthesis. Since the sulfone functional group is not a chromophore in accessible regions of the ultraviolet spectrum, the photochemistry involves the interaction of the sulfone group with nearby excited states. Irradiation of appropriate cyclic sulfones gives loss of sulfur dioxide and olefin formation. In some instances this reaction is thought to be a concerted chelotropic reaction. Examples of this 26 27 reaction include the photoreactions of episulfones, sulfolenes, and.dihydrothiepin dioxides.28 Irradiation of simple acyclic C) (Si by ——_" I’I'ICHICH2 *502 Ph 02 S ec mfg) H3 M” M + 502 (moior isomer) ’/’ SC) bu -_- 2 ——y-- | +so2 \ _.. 24 benzylic or aryl sulfones also leads to sulfur dioxide loss with 29 formation of products expected from radical intermediates. More 1: 21150291. E2137) Ph—ph hp (p-CH3C6H4)2 so2 PM p-CH3C6H4-Ph hu PITCH2 $02CH2Ph ——* PhCHzCHzPh complex sulfones can also lose sulfur dioxide photochemically.30 There are some cases where other reaction occur, unaffected by the presence of the sulfone. These include some 2+2 cycloadditions31 and pinacol formation.32 0 Ph Ph 11 )KD’ ___.)"" 1’th 02 We have chosen to investigate a method involving ring contraction of a five membered ring sulfone to a B-lactam via loss of sulfur dioxide. 25 The system chosen for study was the 1,1-dioxo-4-thiazolidinone system. This choice was based on the desire to place the sulfone group away from the carbonyl function, which might complicate the photochemistry, and the accessibility of the ring system. Synthesis of 1,1-Dioxo-4-thiazolidinones The desired sulfones are readily prepared by oxidation of the 33 The sulfides can be prepared by conden- corresponding sulfides. sation of the appropriate aldehyde, a-thiol carboxylic acid, and ammonium carbonate, with azeotropic removal of water.3“ These RCHO + RYCO 2" +(NH Woo —P——>"" N-unsubstituted compounds can be N-alkylated with sodium hydride and the appropriate alkyl halide. Alternatively the N-substituted compound can be prepared by condensation of the pre-formed aldimine ‘with the a-thiol carboxylic acid with azeotropic water removal.35 ‘Yields using this method are exceptionally good. Synthesis of 3~methyl-2-phenyl-4-thiazolidinone, 17, was achieved in better than.99% yield from benzylidene methylamine and thioglycolic acid. Condensation of benzylidene methylamine with thiolactic acid afforded the 3,5-dimethy1-2-pheny1-4-thiazo1idinones,‘18 an.L2’ in 95% yield as an 8/1 mixture of the cis and trans isomers. These samua compounds may be prepared by alkylation of 5-methyl-2-phenyl- 4-thiazolidinone with sodium hydride and methyl iodide. This method .afforded a 1/1 mixture of the two isomers in 84% yield. Separation 26 was accomplished by silica gel chromatography, eluting with 1/1 ether/pentane. The stereochemical assignments for these compounds will be discussed in detail in the next section. H 1'" R 1) NoH, THF Rf" ., ) 2) R I R' s . NA" R' coH Y 2 A R e R 17, R=PI1I R'=H, [(1043 180ml 19 18 and I9 R=Ph,R'-R"=CH3 8/1 The sulfides could easily and controllably be oxidized to the corresponding sulfoxides and sulfones. The sulfoxides were prepared to assist in assigning the stereochemistry of the sulfides and sulfones. Treatment of sulfides 17, 18, and 12 with sodium periodate in aqueous methanol afforded sulfoxides 20” 21, and 22, respectively.36 l3 I 8 NolO4 Pb . - HZO/CH30H of, CH3 2] Ta N N Io Ph 19 a 4 a)» HZO/CH30H 17 KIO4 27 C“: H20 /CH 3CH1 °20 Treatment of acetic acid solutions of the same sulfides with aqueous potassium permanganate afforded the desired sulfones in good yields. to sulfones,37 the sulfones in even higher yields. A large number of other reagents also oxidize sulfides and it is likely that one or more of these will give No attempts were made to investigate this, as the reaction with potassium permanganate was 17V KAAHCLQ HOAc/ H20 HOAC/ H20 HOAC / H20 ‘1 H ”Kg? 67% 2 r\"3 N "'45 I 58% 02 "a 60% 28 exceptionally clean and easy to work up. The epimeric sulfones ’24 and 25 were easily isomerized to a 1/1 mixture of the two isomers by treatment with base, florisil, or silica gel. All attempts to separate these mixtures were unsuccessful, and it was necessary to separate the isomers at the sulfide stage. The sulfones may be alkylated at C(S) by treatment with base and methyl iodide. In this way 3,5,5-trimethyl—1,17di9x0é2—phenyl-4-thi- azolidinone, 26, was prepared in 66% yield. Ph.H0m oaumfiou< 32 magnetic axis of the complex and the lanthanide ion/hydrogen axis."3 [—3cosze - I] .3 l _. .4 A6 ll 7: The principal magnetic axis is usually assumed to be the lanthanide - oxygen bond (in sulfoxides). This equation is frequently used by assuming 9 to be similar for all hydrogens and using just the 1/r3 relationship. This is often adequate for making stereochemical assign- ments. Another approximation often used is to assume an average con- formation of oxygen-lanthanide bond when including the angular depen- dence. This occasionally leads to erroneous assignments, which can be corrected by averaging the shifts predicted by the appropriate confor- ‘mations.““ This method has frequently been used, although errors have been made using the above approximations. The shifts induced in sulfoxides 20, 21, and 22 by addition of one mm mm mm equivalent of Eu(fod)3 are indicated in Table 2. At the crudest level of use of the McConnell-Robertson equations, these results agree with the assignments previously made. The hydrogens on the same side of the ring experience similar shifts, which are different from those on the other side of the ring. Inclusion of the angular term in calcula— tions also predicts the same shifts for the conformations shown below, “2]” 33 uaowmou uwfism mo ucoam>wavo moo mo cofiuwoom ou mvcoawmunoo umwnw some .vaofimcsoo mumwnm Ham .aoo 6% @< Am 6.6a 6.6 I. a.6 ww a.6 N.6 N.m m.m6 6.6 mm 6.6 . 6.66 6.6 6.6 em amuuzu (mommo Lmuamo Lmoem avaxomaam :owouvzm aaaaaxo6aam :6 666666 666:66H maaoCVam N QHQMH 34 which seem to be the most reasonable choices on steric grounds. Distances and angles were measured from molecular models and all calculations were done by hand. Other conformations, including those with the europium complexed to the amide, failed to rationalize the observed shifts. At this stage, the stereochemical assignments seemed quite cer- tain. Nevertheless an x-ray crystal structure determination was per- formed on the sulfone which had been assigned the trans configuration, 23. The results, much to our surprise, indicated it was cisz not 5 At this stage, several possible explanations for these trans.“ seemingly contradictory, results presented themselves. The most ob- vious suggestion was a mix-up of samples somewhere between the sulfides and sulfoxides. This was checked by oxidizing the sulfoxides to the corresponding sulfones with potassium permanganate. In each case, the sulfone formed was exclusively the predicted one. The other most likely possibility was that the crystal used for the x—ray study was a stray crystal of the other isomer. This is not entirely unreasonable, as crystallographers choose the exceptional crystal, not the average crystal, to do their experiments with. This was checked by collecting x-ray diffraction data on another crystal of the same sulfone, from an extremely high purity sample. The data were virtually identical to those of the original crystal. This left the two least likely explanations (from our point of view at the time), that either the NMR results were .gll anomalous, or a single inversion of configuration occurred on oxida- tion of the sulfoxide to the sulfone, which would appear to be extreme- ly unlikely. This question was settled by having the crystal structure 35 determined for one of the sulfoxides. The x-ray crystal structure determination was performed on a crystal of what had been previously assigned as the trans sulfoxide, 22. This sample was carefully re— crystallized, and after removal of several crystals for the x-ray study, the remainder checked by 1H NMR with and without shift reagent to insure it was indeed the “trans" isomer. Once again, the results of the x-ray crystal structure determination indicated that the phenyl and methyl groups were not trans, but cis, and that the sulfoxide oxygen was trans to both of them, 2%.“5 9‘3 Ph 8‘ CH3 220 b, non, 1. A66 agancaxtly clear that the NMR results were anoma- 22 lous. A variety of explanations and rationalizations can be offered. For the europium induced shift experiments, one possibility, which had earlier not seemed very likely based on McConnell-Robertson calcula- tions, was that the europium might prefer to complex with the amide rather than the sulfoxide. This was checked by doing a competition experiment, in which dimethyl sulfoxide and dimethyl formamide competed for the shift reagent. It was found that addition of one equivalent of dimethyl sulfoxide to a sample containing one equivalent each of dimethyl formamide and Eu(fod)3 reduced the shift of the amide to about one half of its original value. This result suggests that for sulfoxides 22,‘%1, and 2%, an averaging of shifts due to the europium complexed with the amide and with the sulfoxide might give rise to the 36 observed shifts. In fact, the calculations give a reasonably good fit to the observed shifts, which is a little better than that from the initially assigned configuration, with the europium exclusively com- plexed to the sulfoxide, (Table 3). A similar combination was tried to explain the shifts of the other sulfoxide isomer, assuming the phenyl and methyl groups were trans rather than cis. However, no simple combination of amide complexation and sulfoxide complexation (for both possible isomers) was able to reproduce the observed shifts. It is quite likely that some reasonable linear combination of conformations and sites of complexation does agree with the ob‘ served shifts, but an exhaustive search is not practical. The matter of the aromatic solvent induced shifts is somewhat more difficult to explain. One possibility is that the sulfoxides are too hindered for the normal approach of the solvent, such that opposite results are obtained in some undefined way. There is at least one report in the literature of anomalous results of this sort which have been attributed to steric interference."6 The C(5) and C(6) hydrogens in the phthalimido penicillin sulfoxide 22, both show relatively large upfield shifts on changing solvents from deuterochloro- form to deuterobenzene. This was attributed to interference to approach of the solvent to the 8 face by the large phthalimido group. However, the method has been applied to the dihydrobenzothiophene oxide isomers, 2;, which closely resemble the thiazolidinones sterically."7 The stereochemical assignments in this case are in agreement with those based on trifluoroacetic acid induced shifts in the 1H NMR spectra. cowkxo mowemnbu ooxoaaaoo Esfioouoo unooxo www no cowumuswamcoo meow An 6mo=o pom o.H mo umwsm m>fiumaou m ou nonmanoo mamanm HH< Am 37 o.~ N.H ~m.0 ~o.on e.m 6wWIZI o.H o.~ . o.H o.~ o.H 6ww=01 6.6 6.6 66.6 6.6- 6.6 amumw- n.~ n.m oa.o m.w1 ~.m :MWWI oo>ummno wm‘wwm,www MW QQW :ww: cowouohn some Now coHuMustwaoo 6666666 66666 6666666 6A6666=6 66>ammao 666 6666Haoaao 6 66666 38 "II «I 12 y“ [Mo 3 C) num 021! Pb 2 6 27 While it is perhaps, not unusual that the aromatic solvent effect should be influenced by steric factors, it i§_unusua1 for these factors to result in different shifts for hydrogens cis to each other, and for directly opposing results to be obtained with the hydrogens trans to each other. The question of the stereochemistry of the sulfidestb§ and $2 is still somewhat puzzling. By analogy with the sulfoxide and Sulfone assignments, sulfide 18 would be assigned the trans configuration, and 12 the cis configuration. The alternative assignments would require that each oxidation occur with a single inversion of configuration at carbon, an unreasonable suggestion. This leaves the question of the "W" coupling in the 1H NMR spectrum °f.%§’ a question for which no obvious answer presents itself. The most important feature of these results is the ability of standard NMR techniques, of three different kinds, to all produce completely misleading predictions of the stereochemistry of the thiazolidinones. In gagh case, each experiment gave exactly the results expected for a different isomer. While some of these results can be rationalized, others are still left unexplained. P1 39 The most important lesson to be learned from this relates to whether one can use these methods, at all, for assigning stereochemistry in all but the simplest molecules. If it had not been for the x-ray crystal structure determination, there is no way we could have or would have made the correct stereochemical assignments. This is not a caution pertaining to the care necessary in interpreting results, but a caution pertaining to the applicability of the method for molecules with more than one functional group or modest steric requirements. Photolysis and Thermolysis of 1,1-Dioxo-4-thiazolidinones: Results Irradiation of 3-methy1-1,1-dioxo-2-pheny1-4-thiazolidinone, ’23, in methanol or acetonitrile leads to rapid decomposition, but no identifiable products were obtained. We then turned to the hi Phfl-‘r' (D hV i- No CH3CN Identifiable Products Of <°2 CffifDH Imare highly substituted 3,5-dimethyl-1,1-dioxo-2-pheny1-4-thiazol- idinones, 2’4, and 2,5. Photolysis of the cis isomer, 2,5, in a mixture of'tr4butyl alcohol and acetonitrile resulted in formation of the isomeric B-lactams, cis and trans 1,3-dimethyl-4-phenyl-2-azetidinone, 2,8 and a. The cis B-lactam was obtained in 31% yield, the trans isomer in 8% yield, and 28% of starting material recovered unchanged. The combined yield of B-lactams based on recovered starting material 40 was 55%. Photolysis of the trans sulfone, 24, in the same solvent system also afforded B-lactams 28 and 29. In this case the trans «A. mm 9'3 c ii 3 H P *_ bv 3 12836 +OH/CH351) I I 25 CH "I ”’/,C 3 8 "a o2 c113 Pb 25 28 31% 29 8% $“3 Ph N o bv .s ,, +6.47%? 28 , 7% + 29, 14% + 25, 17% O ’2 2 CH3 24 lactam was the major product, 14% with a 7% yield of the cis lactam. The recovered sulfone, 21%, however, was completely cis. The products were identified by comparison of their 1H NMR spectra to those re- ported in the literature."8 The appropriate carbonyl stretching fre- quency was observed in the infrared spectra. Yields were determined by adding known amounts of a reference material to the NMR samples. When the cis sulfone 22 was photolyzed in isopropanol, two additional products were formed. They were identified as N-benzyl- N-methyl propionamide, ’32, and isopropyl benzenemethanesulfonate, '01? Tina sulfonate was identified by comparison of its NMR and infrared spectra to those reported in the literature."9 The amide 22 was identified from its NMR spectra, which was similar to that reported for N-methyl-N-benzylacetamide, with appropriate changes.50 These Innatolyses were also done in several different mixtures of isopropanol and t-butyl alcohol, and the yields for the products of these reactions are summarized in Table 4. Photolysis of the trans sulfone in 41 .ucm>Hom pom unmoxo mcowufiwcoo wEmm mnu zHuomxm Hows: uso wwwuumo musoEHumdxm HHm :uw3 .Hmwumums wawuumum mo nuances Hmfluwcfi co momma muamo you mum mvafim HH< Am n x H 8 mm | | m R moamutzommo m x H 00 RN m m N «H 305mlu\mOHmlH H \ H me n mH OH m CH mODMIu\mOHm|fi mm «H NH 0H m NH SOHAIH N» S HmuOH cmuo>oomu Hm Q” ww, WW uco>aom uosvoum mmuco>Hom ucoummwww ca WW odomasm mHo mo mfithouoLQ Scum mchH» s «Hams 42 isopropanol also afforded reduction product 30. mm CH3 ' 28,179; + 29 8% + 14% Ph.1(’ hv "'---iI- S i-PrOH 02 CH3 PhCstoai-Pr PhCH2 \Nk CH2 (HS 31 ,17% CH3 30, 16% Several attempts to sensitize and quench the photoreaction were made. Irradiation of the cis sulfone in acetone or in acetonitrile with acetophenone present through a pyrex filter gave only starting material. Attempts to quench the photo decomposition of‘%% with piperylene in both isopropanol and the t-butyl alcohol/acetonitrile mixture were unsuccessful, although the photolysates were much messier in these cases. Irradiation of the B-lactams in t-butyl alcohol/acetonitrile re- sulted in decomposition. The trans B-lactam gave none of the cis isomer, but the cis B-lactam gave a small amount of the trans B-lactam. No reduction products were formed on photolysis of the trans B-lactam in isopropanol. Thermolysis of sulfones a: and 22 also leads to B-lactams. When either sulfone isomer is heated at 200° C, the trans B-lactam 23 is formed in good yield. $“3 Ph 200°C ('53 o 02 CH3 P 24 or 25 43 Photolysis and thermolysis of 3,5,5-trimethyl-1,l-dioxo- 2-phenyl-4-thiazolidinone, 32, gives 1,3,3-trimethyl-4-phenyl- 2-azetidinone, £2.51 The photochemical reaction gives the B-lactam in only 10% yield, while the thermal reaction gives a 66% yield, based on unreacted starting material. fi“3 CH3 Ph~1<1::2E:> hv ‘jt::j:;: S or 3| H3 02 01:3 A n‘ ”3 26 32 The photochemistry of sulfide 13 was also briefly examined. Photolysis of the cis sulfide in acetonitrile resulted in formation of some of the trans isomer. As desulfurization with ring con- traction by photolysis of sulfides in triethyl phosphite has been reported,52 these results suggested use of this method. Photolysis of the cis sulfide in triethyl phosphite or triethyl phosphite/ acetonitrile (1/1) gave only mixtures of the two isomeric sulfides, with very slow decomposition. Photolysis and Thermolysis of 1,1-Dioxo-4-thiazolidinones: Discussion The data do not allow a complete description of the mechanism of the photochemical reaction, but some possibilities can be ruled out. The failure to sensitize or quench the reaction suggests reaction occurs from the excited singlet state. The mechanism of the reaction could in principle be either a concerted expulsion of sulfur dioxide with concurrent bond formation between carbons two and five, or a stepwise process involving radical or ionic intermediates, 44 or a combination of concerted and stepwise processes. For a concerted process, orbital symmetry predicts that the process would occur with retention of configuration at both carbons two and five, assuming sulfur dioxide to be a linear leaving group.53 For the other extreme of free diradical intermediates we would expect to find mainly the trans B-lactam, whose formation would be suppressed by addition of a radical trap. Neither of these descriptions fit the data, but some combinations or modifications of the two will. The observation that both sulfone isomers give B-lactams in which the major, but not exclusive product has retained its stereo- chemistry, rules out long-lived equilibrating diradicals. The formation of both B-lactam isomers from each sulfone isomer does not rule out all or some concerted sulfur dioxide expulsion, as the product lactams could photochemically interconvert (not very likely at a fast rate) or the sulfones could interconvert. It was observed that the recovered sulfone from photolysis of the trans isomer was all cis. The fact that the lactams formed in this photoly- sis are mostly trans eliminates the possibility that the lactams arose mainly from the cis sulfone. The formation of reduction products upon photolysis of both sul- fones in isopropanol implies the intermediacy of radicals. This does not, however, require that the radicals are intermediates in the formation of B-lactams, although it does seem likely. The fact that some B-lactams are still formed in isopropanol requires either contribution from a concerted mechanism, or the intermediacy of short—lived diradicals. 45 Perhaps the most likely description is shown in Scheme 1. Photolysis of either sulfone results in cleavage to a diradical, which can either reclose to the sulfone, lose sulfur dioxide, rotate and reclose to the alternate sulfone, or rotate and then lose sulfur dioxide. The diradicals from sulfur dioxide loss can then either close to the B-lactam or rotate and then close to the other B-lactam isomer. For this scheme to work, isomerization of the diradicals must be slow compared to the rate of closure to B- lactams, or the stereoselectivity wouldn't be observed. The difficulty with this scheme is that it predicts increasing stereoselectivity when radical traps are present. In fact, little difference is observed. This objection may be circumvented by postulating an alternate pathway for inversion of stereochemistry, one that doesn't involve the above diradicals. Such a pathway might be isomerization of the B-lactams photochemically, or a hydrogen abstraction recombination in isopropanol. Whatever the details of the reaction, the important feature is the predominant formation of the cis 3,4-disubstituted B-lactam. This is the stereochemistry required for biological activity in nearly all of the penicillin and cephalosporin antibiotics. The yield of cis 1,3-dimethyl-4-phenyl-2-azetidinone from benzyl- idene methylamine and thiolactic acid is 21%. Clearly, a sequence that could be used to provide biologically active B-lactams in comparable yields would be extremely useful. The next question is then will the photoreaction be applicable to thiazolidinone dioxides with the proper substituents for biological activity, Scheme 1 CH3 46 /H3 H I ”’CH3 r— --1 CH ‘QESI <3 ‘\3 (3 hv j S and .<: i ‘-'— . or ”I . P" s CH3 0' CH3 02 2 .,... .7 CH3 h" h and /0t i <—' thu Ph .00, ° CPfiB °2 '50 2 L-PI'|)’-_="**2 — 2% [234.5842] _> -co [5}; 1><:2. :' _’_. L62 1 m e g ,n, 'R-c=H=c —> 61 There are two reports of AZ-oxazolin-S-one photolysis. Slates et al report that photolysis of a 2,4,4-trialkyl-A2-oxazolin-5- one, followed by acidic work-up, yields a ketone derived from C(4) and its substituents.55 Loss of carbon dioxide to give an N-acyl imine followed by hydrolysis was postulated, but not demonstrated. Padwa and Wetmore report that photolysis of an Az-oxazolin-S—one in the presence of an electron-deficient olefin dipolarophile trap gave no Al-pyrroline product.56 However the oxazolinone was not specified, nor were the identity of any products reported. This behavior contrasts with that of Aa-oxazolin-S—ones, which usually lose carbon dioxide photochemically to give nitrile ylide inter- mediates, which may be trapped with electron-deficient dipolarophiles (1.56 57 to give Al-pyrrolines in good yiel An exception to this behavior is the loss of carbon monoxide upon photolysis of a 2-trif1uoromethyl substituted Aa-oxazolin-S-one.58 Attempted thermolysis of 2,4,4-trisubstituted Az-oxazolin- 5-ones in refluxing xylene is reported to give no reaction,59 but higher temperatures do cause reaction. In these cases, loss of carbon dioxide occurs to give products expected from nitrile ylide intermediates.5° Carbon monoxide loss is reported to occur only when a 2-trifluoromethyl and 4-thiophenoxy group are present.61 Contrary to our results (see below) compound 2 was reported to be stable to 2000 0.62 Thermolysis of 2,4-disubstituted Az-oxazolin- 5-ones with dipolarophiles present, gives Al-pyrrolines also,59 but this reaction involves oxazolium ion intermediates, which are trapped by the dipolarophile and then lose carbon dioxide. Thermolysis of A3-oxazolin-5-ones also gives loss of carbon 62 dioxide and formation of products expected from nitrile ylides.60 Results: Photolysis ofzsz-Oxazolin-S-ones 9 and 33.63 Photolysis of 3 and methyl acrylate in acetonitrile gives modest yields (26% and 17%) of cis and trans-Z-trifluoromethyl- 4-carbomethoxy-S-methyl-S-phenyl-Al-pyrrolines,,géa and gab respective- ly, along with 7% acetophenone. Spectral and elemental analysis + PI! COCCH3 C5 _ .1] 9 + CHz-CHCCDZCH3 W 3!:331-0 ””013 C02CH3 co2 CH3 4:: 34b support the structures proposed. The stereochemical assignments are based on 1H NMR spectra. Pyrroline 31a is assigned the structure with phenyl and carboxymethyl groups cis since the ester methyl group is more shielded inlaéa than in 3&b.6” Apparently none of the 3-carboxymethy1 substituted pyrroline is formed as there is no higher field multiplet present as would be expected from a C(4) unsubstituted Al-pyrroline. Starting material was recovered unchanged from a dark control sample worked up by the same procedure, indi— cating that all products have a photochemical origin. Irradiation Of.%% in either hexane or acetonitrile gave a good yield of N-(1-methy1benzylidene)acetamide,’35. The structure assignment was based on the 1H NMR spectrum (singlets at 6 2.1 and 2.3 and the aromatic region signals expected of a C-phenyl imine) and chemical transformations. Acid catalyzed hydrolysis immediately gave acetamide and acetophenone, while reduction with sodium borohydride gave N-(1-phenylethyl)acetamide, 36. Silica gel 63 chromatography gave a mixture of acetophenone, acetamide, and N-(1-phenylvinyl)acetamide, 27. Compound 3} was identified by itsJTINMR spectrum and by comparison to an authentic sample, synthesized by literature methods.65 Since enamide 37 is easily PhCOCHa a.) '13 it + “cows + f H3 4- CH3CONH2 (H2 37 CH3CONH2 hydrolyzed to acetophenone and acetamide it is possible, but not necessary that they are formed from gl'via 22. Irradiation of 3% in acetonitrile with methyl acrylate gave, as the only additional product, polymethyl acrylate. The photoreactions of 2 and 22 were not quenched by added piperylene in concentration sufficient that excited states living for 10 7 seconds would have been 90% trapped. Results: Thermolysis of Az-Oxazolin-S-onesz and 23. When a is refluxed in dry xylenes a good yield (66%) of N-(l-phenylvinyl)trifluoroacetamide, 28: is obtained, along with a small amount of acetophenone. The identification of 32 is based on its spectral properties. The 1H NMR spectrum showed an exo methylene, an amide hydrogen and a phenyl group. The mass spectrum, in addition to the correCt parent and parent plus one peaks, showed fragments from the loss of trifluoromethyl and trifluoroacetamide fragments. Hydrolysis of ’38" gave acetophenone and presumably trifluoroacetamide. Though 32 may have formed from 64 N-(l-methylbenzylidene)trifluoroacetamide, no evidence for it was found in the reaction mixture. When.%% was refluxed in xylenes for periods up to 15 hours, with or without methyl acrylate, it was recovered unchanged. flux '0' PhCOCH 9 f. H 3 xYlonos “2r:8/8\(F3 fl 0 i + 2 CF3CONH2 reflux 33 ——> N.R. xylenes Discussion: Thermolysis of Az-Oxazolin-S—onesfighand43%. The most important feature of both the thermal and photochemical reactions is the profound effect of the trifluoromethyl group. In the thermal reaction, the trifluoromethyl group both activates the ring and diverts the normal course of the reaction from loss of carbon dioxide to loss of carbon monoxide. This means that the trifluoromethyl group has drastically lowered the activation energy for loss of carbon monoxide, with unknown effects on the alternate reaction pathway. It is important to note that this effect is different from that of a phenyl group at C(2), which accelerates 6° 51 A consideration of the thermal carbon dioxide expulsion. stability of diradicals which might be generated as intermediates or which may resemble species on a concerted reaction pathway, helps rationalize the thermal results. The fact that the reaction of 2 is faster requires that the trifluoromethyl group enhance either 4,5 bond cleavage or 1,5 bond cleavage, or both. Cleavage of the 4,5 bond would produce a 65 n—type allyl radical, which would be stabilized by a 2—trif1uoro- methyl group, relative to a methyl group. How the trifluoromethyl group would effect the diradical from 1,5 bond cleavage is not apparent. The results require that 1,5 bond cleavage be enhanced relative to 1,2 bond cleavage by the trifluoromethyl group. Cleavage of the 1,2 bond would produce an iminoyl o-type radical at carbon 2.66 Trifluoromethyl groups are known to destabilize acyl radicals relative to methyl and phenyl.67 One particularly attractive rationale which is consistent with the preceding discussion is enhanced 4,5 bond cleavage in a rate-determining step, followed by 1,5 bond cleavage rather than 1,2 cleavage in the product determining step. It is not possible to comment on the reported thermal stability Of.8 as the conditions were not described. Discussion: Photolysis of AZ-Oxazolin-S-ones 9 and 33. V M. The very different photochemical behavior of 2 andlse shows that the trifluoromethyl group can also strongly affect excited state behavior. The photochemical expulsion of carbon monoxide inlee evidently has precedent in the work of Slates et al.55 However this is the first instance in which the postulated N-(methylidene)acetamide or the tautomeric enamide have been observed before hydrolysis. Substitution of the trifluoromethyl group leads to loss of carbon dioxide to evidently give a nitrile ylide which is trapped by a dipolarophile. This is in contrast to previous reports where carbon dioxide loss was not observed from Az-oxazolin-S-ones. As is usually observed, the nitrile ylide shows high regioselectivity in its reaction with methyl acrylate. 66 The direction of this cycloaddition may be rationalized by the work of Caramella and Houk, who have shown that the direction of dipolar cycloadditions may be predicted on the basis of orbital coefficients for the HOMO and LUM0.68 The reaction pathways of R’and 32 may not be completely ex- clusive, since the acetophenone formed in the photolysis of 9’ could have arisen from carbon monoxide expulsion. However, it could also arise via rearrangement of the nitrile ylide to an enamine, followed by hydrolysis. Reaction of both a and 3% probably involve singlet or very short-lived triplet states, since the reactions were not quenched by piperylene. The Effect of the Trifluoromethyl Group. The means by which the trifluoromethyl group influences the photoreaction of 9,13 not clear. The fact that a number of points at which the trifluoromethyl group could control the reaction exist, and that we cannot say which is correct is perhaps indicative of our still rudimentary knowledge of photochemical reactions. However, it is certainly useful to consider these potential points of control. The first potential point of control is, in a gross sense, the localization of the excited state energy. One might consider ’3 andtae to have one, two or three excited states, depending on the interaction of the phenyl, carbonyl, and imino groups. This localization could be determined by the state formed first on absorption of light, or by subsequent energy transfer from one "chromophore" to another. Examination of the UV spectra of,9 and 33 shows the same maxima in each, with no shifts in those maxima mm as solvent polarity is changed. The spectra are like that of 67 toluene, although extinction coefficients in the 225-250 nm region are bigger. The extinction coefficient of 2 between 250 and 280 nm is approximately twice that of 3%. These small changes suggest that control of the photochemical reaction is determined at another point. Intramolecular energy transfer from the phenyl group is perhaps a more likely point of control in the reaction. a-Phenyl to ester carbonyl energy transfer has been reported in studies of ester photochemistry by Morrison et al.69 With methyl present as in 3%, energy transfer to the carbonyl may dominate, followed by a-cleavage to give diradical 33, like that postulated for the photo-Fries reaction, which may lose carbon monoxide to give the observed product 852° The substitution of trifluoromethyl for methyl would lower the energy of the fl-fl and n-0* levels of the imine,71 and energy transfer to the imine followed by a-cleavage, as postulated for azirines, could lead to diradical 40. Loss of carbon dioxide would give nitrile ylide 4% which when trapped by a dipolarophile would give the observed Al-pyrrolines. If the P ‘1 Ph hv ‘ . .1? L. H3 4 39 P“ co —Ph 0 e 1 h - 9 —."—). , -—L twat-ca, —> \' C as _ _ 41 L .5 68 carbonyl and imine form a single excited state, the trifluoromethyl group could increase the importance of the "imine" description (by changing orbital weighting coefficients) and thus aid carbon dioxide expulsion. Another potential point of control closely related to this involves electron transfer rather then energy transfer. Electron transfer from phenyl to carbonyl might be postulated forlee, and transfer from phenyl to imine forte. These species may then lose carbon monoxide and carbon dioxide, respectively. The importance of electron transfer in systems like 2 and.3% is not known, and it is possible that electron transfer and energy transfer compete. Electron transfer, may, for example, be important only forlg. The means by which the trifluoromethyl group affects the photo— chemistry may be more subtle. In the points of control discussed so far, trifluoromethyl and methyl cause major perturbations of the excited state. The different behavior of 3 and 3% may be due to small changes in appropriate regions of the excited state surface.” For example, trifluoromethyl perturbation of the "imine" 2 and minima region of the surface might lower activation barriers7 to favor 1,2 bond lengthening and eventually cleavage, by making the excited state more zwitterionic. Finally, a different potential point of control involves decay to the ground state surface, in a region near the geometry of the product. This decay process is aided by a close approach of the ground and excited state surfaces.“ This can be accomplished by increasing the energy of the ground state or decreasing the energy of the excited state. The trifluoromethyl destabilization 69 of the iminoyl O-type radical, may make 1,2 bond cleavage photo- chemically accessible by raising the energy of the ground state. The methyl group, which would stabilize a o-type iminoyl radical, would not force the ground state surface high enough for carbon dioxide loss to occur, resulting in decay to the ground state at a geometry corresponding to carbon monoxide loss. The large number of possible points of control make it difficult to determine which is most important. The concurrence of several of these points of control may rationalize the sharp difference in reactivity of’g andlgé. Whatever the mechanism by which the trifluoromethyl troup perturbs the reactions of the A2 and A3-oxazol- inones, the effect is striking. In every case studied thus far, both thermal and photochemical, the trifluoromethyl group has induced the molecule to choose an alternate reaction pathway, despite the fact that this group survives all of these reactions intact. EXPERIMENTAL General. Compound 9 was prepared as described in the first section m of this thesis. Ultraviolet spectra were recorded on a Cary 17 spectrophotometer. All other instrumentation and procedure were as described in earlier portions of this thesis. Preparation of 2,4,-dimethyl-4-phenyl-A2-oxazolin-5-one (33).73 A solution of 2.00 g(12.1 mmol) of a-phenylalanine in 15 mLmzf acetic anhydride was heated at reflux under a nitrogen atmosphere for l h. The solution was allowed to cool and excess acetic anhydride removed under reduced pressure. The residue was distilled under reduced pres- sure to yield 1.62 g (71%) of 2,4,-dimethyl-4-phenyl-A2-oxazolin-5-one (33); bp 62°C (0.12 mm); NMR (0001,) 5 1.7 (3H,s), 2.2 (3H,s), 7.1—7.6 mm (5H,m). Absorption Spectra of (9) and (33). UV absorption spectra of 9 and 33 m mm m Iv» were recorded in both hexane and acetonitrile, and are summarized here: A (e) (an asterisk indicates maximum, A in nm) , 9 (hexane), 268* (243), a. 263* (429), 261* (484), 257* (629), 251* (828), 245 (1036), 240 (1172), 230 (1930), 220 (3502), 210 (8138); 9 (acetonitrile), 268* (209), 263 n. (359), 261 (443), 257* (585), 251*, (802), 245 (969), 240 (1130), 230 (1595), 220 (3174), 210 (7643); 33 (hexane), 268* (137), 263* (269), 'VM 261 (288), 257* (428), 251* (607), 245 (786), 240 (866), 230 (995), 220 (3731), 210 (8407); 33 (acetonitrile), 268* (98), 263* (197), 261 (208), 'vu 257* (384), 251 (519), 245 (711), 240 (909), 230 (1081), 220 (3091), 210 (8547). 70 71 Photolysflsof 2,4-dimethyl-4-phenyl-A2-oxazolin-5-one (33). A solution of 205 mg (1.08 mmol) of 2,4-dimethyl-4-phenyl—A2-oxaz:lin-5-one, 33, in 330 mL of dry hexane was irradiated through a vycor filter for 2“ period of 2.5 h. Solvent removal under reduced pressure gave a yellow liquid (presumably N-(l-methylbenzylidene)acetamide), which had an NMR spectrum containing singlets at d 2.1 (3H), and multiplets at 5 7.0-7.4 (4.5H) and 7.6-7.8 (2H, in addition to broad high-field signals attri- buted to polymeric material. No acetophenone or N-(l-phenylvinyl)aceta- mide were observed. Treatment of an NMR sample of this material with slightly wet trifluoroacetic acid resulted in immediate formation of acetophenone and acetamide, identified by addition of authentic samples. Chromatography of 182 mg of the initial photolysis product on silica gel (Mallinckrodt CC-7), with methylene chloride elution, yielded a mix- ture containing 63 mg (54%) of acetophenone and 23 mg (15%) of N-(l- phenylvinyl)acetamide which had the following spectral data: NMR (CDC13) d 2.1 (3H,s), 5.1 (lH,br s), 5.8 (lH,br s), 6.8-7.2 (lH,br s), 7.0-7.2 (5H,m). Irradiation of 2,4-dimethyl-4-phenyl-A2voxazolin-S-one in the presence of methyl acrylate gives, as the only additional pro- duct, polymethyl acrylate. Similar irradiation of 167 mg of 33 in the presence of 86.6 mg (1.28 mmol) of trans-piperylene for 2.5mh followed by the same workup showed no starting material to be present. The combined yield of ace- tophenone and 37 was 53%. The concentration of piperylene was suffi- mm cient to reduce the quantum yield to 10% of its original value for a reactive lifetime of 10'7 3, assuming k = 2.7 X 1010 for hexane. diff 72 Photolysis of 2,4-dimethyl-45phenyl-A2-oxazolin-5-one and reduction of products. A solution of 192 mg (1.015 mmol) of 2,4-dimethy1-4-phenyl- Az-oxazoline-S-one,,33, in 320 mL of dry hexane was irradiated through a vycor filter for a period of 2 h. Solvent removal under reduced pres- sure gave 196 mg of a yellow oil. The oil was dissolved in 50 mL of dry tetrahydrofuran, 110 mg (2.9 mmol) of sodium borohydride was added, and the solution was heated at reflux temperature for 18 h. The solu- tion was allowed to cool and quenched with water, and then methylene chloride and more water were added. The organic layer was removed, washed once with 0.1 M HCl, dried over sodium sulfate, and filtered, and the solvent was removed to yield 165 mg of a yellow oil. The oil was chromatographed on a silica gel column (Mallinckrodt CC-7) using 2% methanol/methylene chloride elution. One band came off, which con- sisted of a mixture of N-(l-phenylethyl)acetamide (identified by its NMR spectrum) and polymeric material. NMR analysis of the mixture using dioxane as an internal standard indicated a yield of 41% of N- (1-phenylethyl)acetamide, which had the following NMR spectrum; (CDC13) 6 1.5 (3H,d), 1.9 (3H,s), 4.95 (1H,quintet), 6.5-7.0 (lH,br s), 7.1 (5H,m). Photolysis of 4-methy1-4-phenyl-2-trifluoromethy1-A2-oxazolin-5-0ne (9). A solution of 205 mg (0.843 mmol) of 4-methyl-4-phenyl-2-trifluorometh- yl-Az-oxazoline-S-one, 2, and 2 mL of methyl acrylate in 230 mL of dry acetonitrile was irradiated for 8 h through a vycor filter. Solvent removal at reduced pressure followed by silica gel chromatography (Mallinckrodt CC-7) with methylene chloride elution gave 105 mg of a mixture containing, by NMR analysis, 52% and 34% of the cis and trans-2- 73 fluoromethyl-4-carbomethoxy-S-methyl-S—phenyl-Al—pyrrolines, respective- ly. The remainder of the material, 14% was acetophenone. By repeated silica gel chromatography, eluting with methylene chloride, a pure sam- ple of the cis pyrroline was obtained, having the following spectral data: NMR (CDC13) 6 1.85 (3H,s), 3.1 (3H,s), 2.9-3.7 (3H,m), 7.0-7.2 (5H,m); ir (neat) 1740 (vs), 1440 (m), 1200 (vs), 1150 cm"1 (vs); mass spectrum (70 eV) m/e (rel. intensity) 285 (28), 270 (11), 266 (6), 254 (6), 226 (42), 199 (57), 198 (25), 104 (100), 103 (88), 91 (15), 77 (57). .éflél: Calcd for C14H14F3N02: C, 58.95; H, 4.95; N, 4.91. Found: C, 58.92; H, 5.07; N, 4.98. By the same method a small portion of the pure trans isomer was also obtained, having the following spectral data: NMR (CDC13) 6 1.48 (3H,s), 2.8-3.6 (3H,m), 3.7 (3H,s), 7.0-7.2 (5H,m); mass spectrum (70 eV) m/e (rel. intensity) 285 (50), 270 (18), 266 (11) 254 (16), 226 (76), 199 (100), 198 (46), 179 (21), 104 (92), 103 (82), 91 (17), 77 (58). Photolysis of 177.5 mg of 9 in the presence of 152.5 mg (2.24 mmol) of trans-piperylene in 230 mL of acetonitrile for 6.25 h followed by the same workup gave the same products in essentially identical yields. The concentration of quencher was sufficient to reduce the quantum yield to 10% of its original value if the reactive lifetime were 10-7s, assuming = 10 k diff 1 X 10 for acetonitrile. Thermolysis of 4-methy1-4-phenyl-2-trifluoromethyl-Az-oxazolin-5-one (9). a. A solution of 143 mg (0.588 mmol) of 4-methy1-4-phenyl-2-trifluoromethyl- AZ-oxazolin-S-one,,g, in 5 mL of dry xylenes was heated at reflux tem- perature for a period of 20 h under a nitrogen atmosphere. Solvent re- moval under reduced pressure yielded a mixture containing, by NMR 74 analysis, 6.7 mg (9.6%) of acetophenone, 21 mg (15%) of starting mater- ial, and 83 mg (66%) of N-(l-phenylvinyl)trifluoroacetamide, which had the following spectral data: NMR (CDC13) 6 5.3 (lH,br s), 5.8 (lH,br s), 6.8-7.6 (lH,br s), 7.2 (5H,s) (signals assigned to acetophenone and starting material were also present); mass spectrum (70 eV) m/e (rel intensity)7” 216 (11.5), 215 (100), 146 (75), 120, 118 (13), 105, 104 (38), 103 (96), 91 (58), 77, 69 (52), 51; ir (neat) 3300 (s), 3050 (m), 1720 (vs), 1310 (vs), 1160 cm.1 (vs). On standing for several weeks, the enamide hydrolyzed completely to acetophenone. PART 2 THE 13C NMR SPECTRA OF NAPHTHALENE CROWN ETHER COMPLEXES: FIELD INDUCED n POLARIZATION AND CROWN ETHER CONFORMATIONAL CHANGES Introduction In recent years crown ethers have found a wide variety of uses, ranging from catalysis in organic reactions to models for ion trans- port in biological systems.75 Information about the properties of crown ethers, such as conformation and complexation constants has therefore become important. Recently a series of naphthalene crown ethers has been used to study photo-excited state behavior, with specific emphasis on the perturbation of excited states by complexed cations.76 With the hope of obtaining information which might assist in interpretation of photophysical behavior I undertook a study of the 13C NMR spectra of these crowns and their complexes. The crown ethers studied are shown in Figure 1. An important structural feature of these crown ethers is the one—carbon bridge between the naphthalene ring and first ether oxygen. This bridge min- imizes direct resonance interaction of the naphthalene 0 system and the crown ether and any complexed ions. The kinds of information which might be obtained from this study fall into two classes. The first relates to the perturbation of the aromatic ring (mainly the 0 system) by a complexed ion. Changes in chemical shifts of aromatic systems have been extensively studied, and these changes are frequently related to charge density changes.78 Crown ethers 4% through 48 should offer a unique opportunity to study these perturbations due to the known location of the perturbing cation, and its insulation from the aromatic ring by the sp3 carbon atom. The second kind of information is the conformational changes in the crown ether ring of both the complexed and uncomplexed crown ether. Since 76 77 42 n=3 , 2,3 'Cr'4 43 n=4 , 2,3-Cr-5 44 n=5 , 2,3-Cr'6 o—(CHZCHZO 46 n= 5,1,8-Cr ' 6 47 n= 4,1,8-Cr - 5 45,1,5'Cr'6 48 n= 3, 1,8-Cr '4 Figure 1. Crown Ethers Studied by 13C NMR Spectroscopy. 78 saturated carbon chemical shifts are sensitive to conformational changes, there is hope that some questions relating to conformation might be answered. Methods and Results The 13C NMR spectra of crown ethers 4% through an were measured with increasing concentrations of alkali and alkaline earth metal salts in deuteromethanol. Chemical shifts initially were measured relative to the deuteromethanol signal at 5 47.00 ppm. It was dis- covered, however, that this chemical shift depended on salt concen- tration at high concentrations of some salts. An external aqueous acetic acid reference line was then used. For the alkali metals, the acetate salts were used as they were quite soluble in methanol and did not shift the methanol reference. For calcium and barium, the chloride and bromide salts were used, respectively, for solubility reasons, although they did shift the internal reference signal. Titrations of the crown ethers with most salts show a clear bend near the equivalence point, and in most cases complexation is complete when the salt/crown ratio is > 2-3. Titration curves for crown ethers 4% through.ég, are shown in Figures 2 through 9. Spectra of the crown ethers with less than an equivalent of salt show only one signal per carbon, indicating the rate of complexation is fast on the NMR time scale.79 As expected, the crown 6's, 4% and.é£, complex well with all the alkali metal and calcium and barium salts (Figures 2,3,6, and 7).80 One spectrum of 2,3-Cr-6 with a 5-fold excess of lithium chloride indicated no complexation. The smaller crown, 2,3—Cr-5, 4%, also complexed well with the same ions except for calcium 79 .2 «.0? ma coHumuucoocoo Honum nacho .musumuwaEmH Boom um Hosmnuoaoumuaoa ca muHmmaHmuoz HHHAH< nufia cutouo~unamz-m.~ to cogumuufie scum mmwcmeu unfiem HmuHamnu .N muawfim .mllllu_c>>nzunrxcnsqu om! ow ohm o_~ on. 9... .20: .88. fi 40.»! I I . 4 lllllllll 4 lllllll .1 I AVCU O [Od/ JON... x llllllllllllll xnl III] / I xzz//./4_ .1. 2.9 O i. . xn/ or legs (1 I . . [.../".70 2 VS \u4 AM.NV U ”x” d , & \QI . 40.— 10M OmUu C 6.0m" I OX.” 0 002 I x O.” meow £3 mLo..om-fioz-nm to 8:3: sec $820 new .8520 80 ox Noo). mH QOfluQHuGQUGOU ECHO .0H3UQHOQBUH. 800% an HOGBUOn—OHOUH—g GH OUHQm 5HOH8 can .523 5? o-uouo~..fiaznn.~ .3 8333a sea finance ”.35. H8256 .n 35E All Z30mu\._.n.ZOwnu\LH:Sm Ofi Oé. Qfl AQN O; _ . _ d . 10.0.: 03: x 830. .06! I: lllllllllll In! . I . I I I A? —v0 [III 10”. ‘ lllllllll ‘II III I, I, u/ //ZOmfiu\LH:%m 00— 0.6 06 0M ON 0.— — u — — d - loan DUI I 9 Act: 0 I: I l/// Odl / // I /o/ 0.»! I / 0.Nl . . DI lllll I, ? linnanu .wAV 0.0 N+OU I D .II' N+Om I Cal 0.— ‘ \ ‘O‘II‘|I . \\\\\\\ am An NV U *0 “““ 0’ 0.” £8 £3 mLukLucaoth .6 cozotc. Eot mmocozu Ecm 62,55 83 .2 «.9. m... ceaumuucoucoo .553. 56.5 .ousuauoaaus Boom um Hocunuoaououaoo a... 3.3m H30: .3934 p.33 oluonalnnflTwJ mo 60:93:. Beam Iowans“. amasm Hanan—26 .0 «Hanan AIII 230m0\......>0m0\._.u.>>1N_h nun/sKWN. mm. mw_ om. Nm_ 3...: JOIN—Fl] a.» 3 ha m o._ o. EoEcBmmo n 53 ¢o.vm ....on _ r h g _ _ Io a. III!!! \ .\ u .. x x. _ — ~ 58 0%. . :0... X \ :0 Ion \ x7, x x 5838519 “588.8. (~wa ” L _ fix _ nIBfo- \\\\\.\ . I \ _ _ _ .r $100“an + . r, . x 2305 , . _ x __ . _ fl w _ _ mmzuofe. 1’ I . s u 230.8 _ _ _ _ _ w~100~xu+ , 2305 :23. 59.3 35¢ .o co_om...u=oEo..o.o m g 8266.. can 83885802.. m. _ Lo $22 on. m m 95 .ouamuoaefia Boom um Hocasumaouuuaon ca moccamnunnmz vmuauwumnsmHQUn.~ umnuo new mcaouuonunamznm.u uo_muuumam mzz umH .oH unamam mm , Ob >nN_ ON. hm. mm. .m. . mm. mm. 50. . . «J 4!? .3h 1n . _ o_.m _ new EmEchmo n EQQ.ma-_.m_ . MIDI FI _. —’ FI fill 10 II _ ’ Ir L I f . E \— 1%qu EQQ mm._m . 'NT—UII I 7 \ \\ . can 38510.. “can an: ...Nxo- . fl h aloof? . . a I _ _ _ _ _ _ $03 + L. b . _ 2308 _ _ _ __ _ $83 f f 23010 - _. I/ _ _ _ _ _ wmxuofov. . 230mm 868 contoo 35¢ .n :23. 2660.66 m m g 826%.. nco mczoboczaoZumN .6 $22 on. m 96 unimportant mechanism for the transmission of substituent effects. However, two recent studies offer evidence to the contrary. The first of these studies involved examination of the 13C NMR spectra of benzene rings substituted with alkyl groups with dipoles attached 7 The second involved several atoms removed from the aromatic ring.8 13C NMR spectra of phenyl substituted amino acids in which the chemical shifts of the aromatic carbons were measured at several different pH's.88 Crown ethers 42 through 48 offer several unique advantages for the study of this effect. The naphthalene crown ethers offer a more extensive and perhaps more general fl system to be perturbed than substituted benzenes. The location of the monopole (the complexed ion) is known quite well, and can be varied such that polarization from either the C(2,3) side, the C(1,8) side, or the face can be effected. The distance of the ion from the aromatic system can be varied by changing the crown ring size, and the charge on the ion can also be varied. These chemical shift changes in aromatic systems are often cor- related with calculated charge density changes.78 Frequently, good linear correlations are obtained, though not always. There is some theoretical justification for a linear charge density/chemical shift relationship.89 However, the theoretical expressions for chemical shift changes are too complicated and not well enough understood to be useful. For this reason, most workers in this area attempt to explain chemical shifts in terms of charge density changes. Examination of the NMR spectra of 2,3-Cr-6 and its complexes (Figure 10), reveals several interesting features. Except for C(1,4), 97 the aromatic shifts are the same for all of the plus one ions. These are, in turn, different from those of the plus two ion complexes. This indicates that the aromatic chemical shift changes are induced by the presence of a nearby charge, rather than some other property of the ion (C(1,4) shifts will be discussed later). The NMR spectra of 1,8-Cr-6 and its complexes (Figure 11), show the same behavior. Except for C(2,7) and C(9), the aromatic shifts are largely indepen— dent of the identity of the complexed plus one ion. Note that in each case the ipso carbon is shifted upfield, while all the other aromatic carbons are shifted downfield by complexation. This is what might be expected for field induced n polarization, which would build up electron density at the ipso carbons. These shifts were correlated with charge density changes cal- culated at the INDO level. As the crown ether itself contained more atoms than the INDO program was equipped to handle, calculations were performed on the bis(methoxymethy1) compounds, £2 and QQ, with and without ions present. For the 2,3-diether, 3?, coordinates were OCH 0 CH3 49 so taken from the x-ray crystal structure determination of the 2,3-Cr-6/K+ complex.9° For the 1,8-diether, 28’ coordinates were taken from a preliminary x-ray crystal structure determination of 1,8-Cr-6.91 Calculations were first done on 2,3-diether, g3, using standard 98 parameters for the plus one ion (Li+). However, attempts to use a plus two ion (Be+2), resulted in the SCF calculation diverging. The parameters for the ions were then adjusted to mimic a plus one or plus two monopole (see experimental section), and this allowed the calculations to work. 1 A plot of the calculated charge density changes on complexation versus the chemical shift changes for the 2,3-Cr-6 with a plus one ion is linear for both the conventional plus one ion and the adjusted plus one ion. Inclusion of both plus one and plus two ions for either or both the 2,3-Cr-6 and 1,8—Cr-6 results in satisfactory correl- ations also (Figure 17). Slopes and correlation coefficients for various plots are summarized in Table 5. Considering the quality of the calculations and the expected sensitivity of ortho carbon chemical shifts to steric effects (see below), the correlation is as good as might be expected. Note that C(9) of the 1,8-Cr-6 is way out of line, and is not included in the calculations of slope or correlation coefficient. The slopes of these plots mean very little in this case, as the use of "adjusted" ions, and the diethers rather than the entire crown ethers, should cause considerable changes in the magnitude of the charge density changes, although the relative magnitudes should stay nearly the same. If the shifts of the aromatic carbons in the naphtho crown ethers are due to field induced fl polarization, as predicted, we would expect the INDO calculations to show little n charge transfer into or out of the naphthalene ring. In fact, for 2,3-Cr-6 and 1,8-Cr-6, the n charge transfer is small. For 2,3-Cr-6 calculations the charge transfer is 15 millielectrons (me-) out of the naphthalene, 99 .>Ho>wuouomom .wluolfimunomZIw.~ mo moxoaoeou eswumm can Esammmuom mo Amvo one + can x .Hocunuoa Iouousua ca otuou-t:om2tm.~ can onuolowlsooZIm.~ mo moXmHosou aswuon vow Esammouom How nowoocu uuwnm Hoowaonu consume: now: newsman muamsua uwuono vuumaouaoo oozH mo aowuoamuuoo .NH «woman Aline x mom/Eu 439 z_ 6245 on on ma ow m. o. m o m: o... 9: own 8.. JIJIIIII. _ _ _ _ _ _ _ _ _ 4 + no.7 D I x [QMI O o low. 0 I 10.... o . . sag lo 0 o o m 1 oo m< . l o ..o.~ oltolmlfiozlo; c. 8&3 .. mltolwlfiozlm; 5 co. . ..o no,» ontolouafiozun.~ c. cam..- o mltoaomlfioz ..n.~ 5 :2 « ..- o.o >.:mzwo momdlo It>> Kim Jade/BIO mo ZOPdJMwEOU 100 oluotw.H oqo.c mumam m . mofi o .p ++mm +M mam. oluolm N oqm.o so wmumsnwm ++mm .+x ouuonm.~ osa.o «as emumshcm ++nm .+u ouuonm.H mom.o mm pmumafivm ++mm .+mo cluolm.m HNm.o @NH pmumnflvm +mU oIHoIm.N mNm.o mm mumcwwuo +mu cIHUIm.N ucwwoflmmmoo me\Emmv Amzzv oowumamuuoo mooam :aoH oazH: AmvcoH :3ouo Amvuozum :3ouo mowcmno mufimomn omumno sows mowcmzo umwfim HmOflEmso mo wcowumamuuoo m manna 101 as compared to the total change in W charges at all carbons (absolute values) of 136 me-. For 1,8-Cr-6, the corresponding numbers are 28 92 me" and 143 me'. This is in the range expected based on previous 78 ab initio calculations on substituted benzenes. As expected from other calculations,78 the changes in a charge density are in the direction opposite those of the N charge changes, and are smaller than the W charge changes, indicating that 0 charge changes are in response to the W charge changes (Figure 20). As the crown ring is made smaller, the distance of the complexed ion from the naphthalene ring should decrease, and the polarization of the W system should be increased. Examination of the data for 2,3—Cr-5 (Figure 12), reveals this to be the case. All of the aromatic carbon shifts increase by a modest amount compared to 2,3-Cr-6. INDO calculations in which the ion is moved 1 K closer to the aromatic ring are also in agreement with this, as all of the charge density changes become a little larger (Figure 18). One would then have expected the shifts to further increase in magnitude as the crown ring was shrunk again to the crown-4 size. This, however, is not the case. The aromatic carbon shifts for 2,3-Cr-4 and its complexes are smaller than those of 2,3-Cr-S, and like those of 2,3-Cr-6 (Figures 13 and 19). One possible explanation for this was that the crown was not the crown-4, but a dicrown, such as é&.93 Ghana/jg @@ 09 SI 102 CALCULATED (INDO) CHANGES IN VALENCE ELECTRON DENSITIES carbons 6.7 5.8 9.IO l,4 2,3 /‘\ O— / c/v “4+ Crown-6 s 5 __§_ 3 r’j’mgc . . x’"'”' / - \J \o_ (+ greater e' density) A I \ II ’ /°' + Crown-5 6 c C C I4: ...-C M ,-----...\ 1' ""C\ . v’ \ I, o... \\ I oM" Ih\ I I 71’ O I _ a, n (""{y’() Crown 4 o ’_3___fi-‘+’—c, \C/ ..l \\v’ Figure 18. *INDO Calculated Charge Density Chan es for Aromatic Carbons of 2,3-Crowns with Plus One Monopole ( ) in Indicated Positions. Charge Density Change is Proportional to Distance of Dashed Line from Carbon. 103 EXPERIMENTAL CHANGES IN 13C CHEMICAL SHIFT THE EFFECT OF DISTANCE carbons 6,7 5.8 9,IO IA 2.3 o.- T c/ l‘ r- ”1“". Crown-6 L + V V w W K 1. CK. c». (~65 (+ upfield) (y. Crown-5 e—e—j—H (_ K+ I \c\ 0" J, o... Crown-4 C " P ' ILW'C/ 7 + G; J; " ‘FV c\. (“—4( O- J. Figure 19. Chemical Shift Changes of Aromatic Carbons of 2,3- Crowns upon Complexation with Potassium. Chemical Shift Change is Proportional to Size of Arrow. 104 EXPERIMENTAL CHANGES IN 13c CHEMICAL SHIFT 0F 2.3-NAPH-2o-CR-6 HITH K+ carbon 6.? 5.8 9.!0 l,4 2.3 o- T "H'C/ '\ ((5 6“ (..fi C) I CALCULATED (INDO) CHANGES IN VALENCE ELECTRON DENSITY carbon6,7 5.8 9.0 L4 2.3 I" O- , / cr-Tr €—~ C —e—-——£+7+—<§2:;g “I ,—_.——~\\ I \0- \el . ’~ I I I I w I {/0- .- .- I n\\"' Fifi: \- \. Ij“(\ "/ \‘__N I (D- \l \ ’ \vl - (3— .\ cs /’ a. \ ’ 00°C C - \ III A CK. \ O- \ \V K‘I" (T upfier) hflT .(+ greater “4+ “4+ e' density) Field Ind. n Polariz. "a response to n pol." Figure 20. INDO Calculated 0, fl, and Total Charge Density Changes for Complexation of 2,3-Naph-20-Cr-6 with Plus One Monopole and Measured Chemical Shift Changes for 2,3-Naph-20-Cr-6 with Potassium Salt. 105 This would have the ion placed further away from the naphthalene ring. This possibility was ruled out by doing a Rast molecular weight determination on the crown ether.91+ Another possible explanation is that the ion, on the average, is located out of the plane of the naphthalene. This would be expected to reduce the D polarization. INDO calculations with the ion moved out of the plane of the naph- thalene agree with this prediction, as charge density changes are reduced somewhat. Agreement, of course, is not perfect, as the exact location of the ion is not known, and additional Y effects are expected as discussed below. Further evidence that the complexed ions in 2,3-Cr-4 sit out of the plane of the naphthalene ring may be obtained from examination of the spectra of 1,5-Cr-6, fig, (Figure 14). In each case all of the aromatic carbon chemical shift changes are small. This is what would be expected for field induced fl polarization effects, as the ion is brought over the face of the aromatic system.95 The chemical shift changes of C(2,7) and C(9) of 1,8-Cr-6, and, to a lesser extent C(1,4) of the 2,3-crowns display a dependence on the identity of the cation. This dependence may be explained on the basis of Y effects of second-row heteroatoms. Ths chemical shift of a carbon with a second-row heteroatom Y to it is moved to higher field, with the effect largest for syn and anti conformations. These upfield shifts are typically in the range of 1-3 ppm. This has been shown to be the case for both saturated carbons96 and aromatic carbons.97 Curiously, the upfield shifts aren't observed for other non-second-row heteroatoms, such as sulfur. The aromatic carbons for which we see the cation dependence are all Y to a crown ether oxygen atom. The large chemical shift change differences for these 106 carbons in 1,8-Cr-6 suggests a crown ring adjustment, as the cation size is changed, by changing the dihedral angle C(2): C(1): C(11): 0(12). As yet, no satisfactory explanation for such Y effects has been offered in the literature. Several INDO calculations with the 2,3-diether,‘%9, were performed with the appropriate dihedral angle varied. No simple charge density variations that could be correlated with chemical shift changes were observed. Of course, this is not the best model system for an investigation of such a possible correlation. It is quite possible (even likely) that these effects will not correlate at all with charge density changes. Ether Carbon Shifts The chemical shift behavior of the ether carbons of the crowns contains information about the conformations and conformational changes of the crown ring, as 13C shifts are quite senstitve to con— formational changes. Information about conformation in solution is frequently obtained from examination of 1H NMR spectra. However, this technique, which is applicable to a few kinds of crown ethers, is not applicable when most of the hydrogen signals coincide, as is the case for crown ethers i£ through :3. In the absence of this tool, other sources of conformational change information must be examined. In the case of these naphtho crown ethers, no specific information about solu- tion conformation is available from the 13C NMR spectra, but more gen- eral information about flexibility and conformational changes upon complexation is available. Examination of the ether carbon regions of the spectra of the uncomplexed crown ethers (Figures 15 and 16) reveals chemical shift 107 degeneracies for the larger crown ethers, which disappear as the crown ring size diminishes. We would expect that as we went down the side chain away from the naphthalene ring, the effect of the naphthalene would decrease until at some point the chemical shifts would be the same. This, of course, will be complicated by conformational changes. For ether carbons held rigidly in a particular conformation, chemical shifts are more likely to be different. The chemical shift degenera- cies in 2,3-Cr-6 and 1,8-Cr-6, then, suggest a rapid averaging of con- formation on the NMR time scale. On going from 2,3-Cr-6 to 2,3-Cr-5, a dramatic change occurs, there being no shift degeneracies for the smaller crown. This change may be reasonably attributed to a more rigid crown ether ring which diminishes conformational averaging on the NMR time scale, or a change in the nature of the conformational changes which results in different average conformations for the ether carbons. This behavior is also seen for the 1,8-crowns, although to a lesser extent. Cation complexation for each of the crown ethers results in shift changes for the ether carbons. For some crowns it is possible to follow the individual carbon shift changes. In these cases (for instance 2,3-Cr-5) shifts may change by as much as 2 ppm, and may go upfield or downfield. These shifts are probably due to two kinds of effects, electron polarization by the ion, and conformation- al changes due to ion complexation. In each case, complexation tends to break chemical shift degeneracies (or leave them unchanged). This effect may be due to the conformational changes induced by ion com- plexation. It is perhaps more likely due to a slightly non-symmetri- cal placement of the ion in the crown ring, such that different 108 carbons are different distances away from the ion, with different alignments of bonds, resulting in different chemical shifts due to simple polar effects. Comparison of the ether carbon shifts of a given crown with different plus one ions reveals a remarkable similarity in most cases. This suggest that crown conformations are similar for the different ions, and that the ion is located in nearly the same place in each complex. The most notable exception to this is the 2,3-Cr- 6/Na+ complex, where a conformational change might be expected due to the poor fit of the small Na+ in the fairly large crown ring. 9 Finally, perhaps the most obvious pattern in the ether carbon shifts is that of the naphthylic carbons. For gzg£y_crown ether studied, the chemical shift of the naphthylic carbon proceeds upfield as the cation is changed from sodium through cesium. The shifts of the calcium and barium complexes show the same pattern, with the shift of the carbon at the lower field in the calcium complex. While the pattern is striking and persists through changes in crown size, point of attachment, and ion charge, no obvious explanation presents itself. The center of the complexed ion would be expected to be similar for all of the ions, so polarization effects should not be responsible. A consistent conformational change that was responsible for such a change would be most surprising, since the effect is con- stant for a variety of changes in crown structure. Since the effect persists through a variety of crown ether changes, it might be reason- able to associate it with some property of the ions, rather than the crowns, though what property that might be is still not clear. 109 Conclusions In the preceding sections a number of observations about the pro- perties and behavior of these crown ethers have been made, along with some observations on the nature of field induced n polarization. These observations are summarized here: 1) Complexation constants can be roughly measured from titration plots of chemical shift changes Kg. mole-fraction of added salt; 2) Rates of complexation of all the crowns and ions studied are fast on the NMR time scale (>103 sec-1); 3) The small crown, 2,3-Cr-4, complexes even large cations such as cesium, and holds them out of the plane of the naphthalene ring; 4) The chemical shift changes of the aromatic carbons correlate reasonably well with INDO calculated charge density changes, for both plus one and plus two ions; 5) The aromatic chemical shift changes are nicely rationalized by a combination of field induced fl polariza- tion and the 0 response to n polarization; 6) The aromatic carbons Y to oxygens display chemical shift changes which are frequently ion dependent, as expected for a variation in the appropriate dihedral angle; 7) In the absence of complexed cations, the larger crown rings are more flexible than the smaller ones, judging from chemical shift degeneracies; 8) Cation complexation tends to break chemical shift degeneracies and to shift the ether carbons; 9) The ether carbon shifts may be either upfield or downfield, as large as 2 ppm, and mostly unpredictable and uninterpretable; 10) A curious and still unexplained pattern of naphthylic carbon shifts is observed, which, when explained, might offer more information about the behavior of crown ethers and their complexes in solution. 110 The 13C NMR spectra of these crown ethers have been a useful tool for the understanding of both crown ether behavior as well as substituent effects in aromatic systems (as has been often demon- strated). Application of 13C NMR spectroscopy to the study of other crown ethers may offer more insight into the behavior of these important compounds. EXPERIMENTAL Materials. The crown ethers used in this work were available from a previous study.98 Deuterium labeling of the napthylic hydrogens was accomplished by treatment of the crown ether in dimethyl sulfoxide-d6 with butyl lithium. All salts used were commercially available and were used as received. NMR Spectra. All 13C NMR spectra were obtained on a Varian Associates OFT-20 operating at 20 MHz, with broad-band proton decoupling. Spectra were obtained using either an 8 mm or 10 mm probe. With the 8 mm probe, a 5 mm tube was placed in an 8 mm tube, and the sample solution placed in the inner tube and the reference solution in the outside tube. With the 10 mm probe, an 8 mm tube was placed in a 10 mm tube, with the reference solution in the inside tube and sample solution in the out- side tube. Sample solutions were approximately 0.2 M in crown ether. The reference solution was 1.295 M.aqueous acetic acid, with the methyl carbon signal at 406.8 Hz taken as the reference line. INDO Calculations. Atomic coordinates for 2,3-bis(methoxymethyl)- naphthalene were taken from the x-ray crystal structure of the potassium thiocyanate complex of 2,3-Cr-6.9° All carbon-hydrogen bond lengths were adjusted to standard values. The atomic coordinates for 1,8-bis (methoxymethyl) napthalene were taken from.a preliminary crystal struct- ure of l,8-Cr-6,99 with carbon-hydrogen bond lengths adjusted as appropriate. Coordinates for other conformations were determined by 111 112 application of simple trigonometry to these coordinates, without chang- ing any bond lengths. All bond lengths were checked using a computer program which measures interatomic distances given atomic coordinates.100 INDO calculations were performed using a computer program written by Professor J.F. Harrison of this department}01 Calculations were done on Michigan State University's CDC 65000 computer. Monopoles were simulated in these calculations by setting a - 10, B = 10, G = 0, and F = O, for both berrylium and lithium, resulting in charges of + 0.97 and + 1.97 for "lithium" and "berrylium", respectively in the calcula- 2 tions performed}o Calculation of W charges was accomplished using a computer program which squared and summed appropriate atomic orbital coefficients in each occupied molecular orbital.100 REFERENCES (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) 113 REFERENCES W.G. Dauben and J.S. Ritscher, J. Am. Chem. Soc., 2%, 2925 (1970). H. Yamazaki and R.J. Cvetanovic, J. Am. Chem. Soc.,gl, 520 (1969). D.R. Arnold, Adv. Photochem., 6, 301 (1968). ’b (a) W.G. Dauben, L. Salem. and N.J. Turro, Acc. Chem. Res.,g§, 41 (1975); (b) J. Michl, Mol. Photochem., 4, 243, 257 (1972); (c) J. Michl, Top Curr. Chem., 46, 1 (1974). mm For a general review of important B-lactams see "Cephalosporins and Penicillins: Chemistry and Biology", E.H. Flynn, Ed., Aca- demic Press, New York, N.Y., 1972. J.C. Sheehan and K.R. Henery-Logan, J. Am. Chem. Soc., 8%, 3089 (1959). R.B. Wacdward et al., J. Am. Chem. Soc., 88, 852 (1966). D. Lednicer and L. Mitscher, "The Organic Chemistry of Drug Syn— thesis" Wiley-Intersecience, New York, N.Y., 1977, p. 418. T. Kamiya in "Recent Advances in the Chemistry of B-Lactam,Anti- biotics", Chem. Soc. Spec. Pub. No. 28, Burlington House, London, 1977, p. 281. A.G. Brown, et al. in "Recent Advances in the Chemistry of B-Lactam Antibiotics" Chem. Soc. Spec. Pub. No. 28, Burlington House, London, 1977, p. 295. (a) E.J. Corey and A.M. Felix, J. Am. Chem. Soc., 87, 2518 (1965); (b) R.J. Earle, D.T. Hurst, and M. Viney, J. Chem.~§gg; C, 2093 (1969); (c) G. Lowe and J. Parker, Chem. Commun., 577 (1971); (d) D.M. Brunwin, G. Lowe, and J. Parker, ibid., 865 (1971); (e) B.T. Golding and D.R. Hall., J. Chem. Soc., Perkin Trans 1, 1517 (1975). (a) G. Lowe and D.D. Ridley, ibid., 2024 (1973); (b) C. Lowe and H.W. Yeung, ibid., 2907 (1973); (c) J.R. Hlubecek and G. Lowe, Chem. Commun., 419 (1974); (d) G. Stork and R.P. Szajewski, g:_ Am. Chem. Soc., 23, 5787 (1974). (a) O.L. Chapman and W.R. Adams, ibid., , 2333 (1968); (b) T. Hasegawa, H. Aoyama, and Y. Omote, Tetrahe ron Lett., 1901 (1975). (14) (15) (16) (17) (18) (19) (20) (21) (22) (23) (24) (25) (26) (27) (28) (29) (30) (31) (32) 114 D.S.-C. Black and A.B. Boscacci, Chem. Commun., 129 (1974). For several examples see: (a) P.S. Engel et al., J. Am. Chem. Soc., 96, 924 (1974); (b) C. Lam and J.M. Mellor, J. Chem. Soc., Perkianrans. 2, 865, (1974). M.R. Johnson and L.R. Sousa, Synthesis, 317 (1978). (a) P.G. Sammes, K.W. Blake, and E.A. Porter, J. Chem. Soc., Per- kin Trans. 1, 2494 (1972); (b) P.J. Machin and P.G. Sammes, ibid., 698 (1974); (c) S. Rajapa and B.G. Advani, Tetrahedron, 29, 1299 (1973); (d) s. Rajapa and B.G. Advani, J. Chem. Soc., Pe'ic’kin Trans. 12 2122 (1974). P. Delbaere, Bull. Soc. Chim. Belg., 51, 1 (1942). mm W. Pfleiderer and H. Zondler, Chem. Ber.,22, 3008 (1966). R.B. Steiger, "Organic Syntheses", Collect. Vol. III, E.C. Horning, Ed., Wiley, New York, N.Y., 1955, p. 176. P.S. Jones, G.W. Kenner, J. Preston, and R.C. Sheppard, J. Chem. Soc., 6227 (1965). F. Weygand, W. Steglich, and H. Tanner, Liebegs Ann. Chem., 658, 128 (1962). ”m“ H. Zollinger, "A20 and Diazo Chemistry", Interscience, New York, N.Y., 1961, p. 170. (a) K. Schaffner, Chimia, 19, 575 (1965); (b) N.C. Yang, E.D. Fest, M.H. Hoi, N.J. Turro,~3nd J.C. Dalton, J. Am. Chem. Soc., 92, 6974 (1970). 'VD E. Block, Quart. Rep. on Sulfur Chem., 4, 236 (1969). P.G. Bordwell, J.M. Williams, E.D. Hoyt, and 3.8. Jarvis, J. Am. Chem. Soc., 90, 429 (1968). Mb J. Saltiel and L. Metts, J. Am. Chem. Soc., 89, 2232 (1967). mm W.L. Mock, private communication cited in ref. 25. (a) N. Kharasch and A.I.A. Khodair, Chem. Commun., 98 (1967). (b) T. Sato, Y. Goto, T. Tohyama, S. Hayashi and K. Hata, Bull. Chem. Soc., Japan, 42, 2975 (1967). O.L. McIntosh, P.De Mayo, and K.W. Yip, Tetrahedron Lett., 37, (1967). ' ' W. Davies and F.C. James, J. Chem. Soct, 315 (1955) I.W.J. Still and M.T. Thomas, J. Org. Chem., 32, 2730 (1968). (33) (34) (35) (36) (37) (38) (39) (40) (41) (42) (43) (44) (45) (46) (47) (48) 115 A.R. Surrey, J. Am. Chem. Soc.,,zg, 4262 (1948). A.R. Surrey and R.A. Cutler, J. Am. Chem. Soc., 18, 578 (1954). H.D. Troutman and L.M. Long, J. Am. Chem. Soc., 72, 3436 (1948). N.J. Leonard and C.R. Johnson, J. Org. Chem.,,gz, 282 (1962). Reid "Organic Compounds of Bivalent Sulfur", Vol. 2, Chemical Publishing Co., New York, N.Y., 1960, p. 64-66. Michael Fazio, unpublished results. L.M. Jackman and S. Sternhell, "Applications of Nuclear Magnetic Resonance Spectroscopy in Organic Chemistry", Pergamon Press, New York, N.Y., 2nd Ed., 1969, p. 334—341. (a) W. Amann and G. Kresze, Tetrahedron Lett., 4909 (1968); (b) R. Lett and A. Marquet, Tetrahedron. Lett., 2855 (1971); (c) R.D. G. Cooper, P.V. Demarco, C.F. Murphy, and L.A. Spangle, J. Chem. Soc. C, 340 (1970); (d) R. D. G. Cooper, P. V. Demarco, J. C. Cheng, and N. P. Jones, J. Am. Chem. Soc., 91,1408 (1969); (e) D. H. R. Barton, F. Comer and P. G. Sammes, J. WAm. Chem. Soc.,m9l, 1529 (1969). C.R. Johnson and D. McCants, Jr., J. Am. Chem. Soc., 87, 1109 (1965). (a) R.R. Fraser, T. Durst, M.R. McClory, R. Viau and Y.Y. Wigfield, Int. J. Sulfur Chem. A, l, 133 (1971); (b) R. Lett and A. Marquet, Tetrahedron1_30, 3379 (1974); (c) R. Lett, S. Bory, B. Moreau, and A. Marquef? Bull. Soc. Chim. Fr., 2851 (1973); (d) M. Kishi, K. Tori, and T. Komeno, Tetrahedron Lett., 3525 (1971); (e) U. Folli, D. Iarossi, and F. Taddei, J. Chem. Soc. Perkin 2, 933 (1974); (f) G. Barbarella, A. Garbesi, and A. Fava, J. Am. Chem. Soc.,_97, 5883 (1975). (a) H.M. McConnell and R.E. Robertson, J. Chem, Phys., 29, 1361 (1958),; (b) C.C. Hinckley, M.R. Klotz, and F. Patil, J’Y"Am. Chem. Soc., 2%, 2417 (1971). R..M. Wing, J. J. Uebel, and K. K. Anderson, J. Am. Chem. Soc., 95, 6046 (1973). All x—ray crystal structure determinations were done by D.L. Ward and assistants of the Michigan State University crystallography service. R.D.G. Cooper, P.V. Demarco, and D.O. Spry, J. Am. Chem. Soc., 91, 1528 (1969). mm 1. Sataty, Org. Mag. Res.,2, 8 (1974). K.D. Barrow and T.M. Spotswood, Tetrahedron Lett., 3325 (1965). (49) (50) (51) (52) (53) (54) (55) (56) (57) (58) (59) (60) (61) (62) (63) (64) (65) (66) (67) (63) (69) 116 J.F. King and T. Durst, J. Am. Chem. Soc., 81, 564 (1965). ’VD A.B. Levin and M. Frucht, Org. Mag. Res., 7, 206 (1975). m J.C. Martin, K.C. Brannock, R.D. Burpitt, P.G. Gott, and V.A. Hoyle Jr., J. Org; Chem., 33’ 2211 (1971). (a) F.W. Hoffmann, R.J. Ess, T.C. Simmons and R.S. Hanzel, J. Am. Chem. Soc., 82, 6414 (1956); (b) E.J. Corey and E. Block, J. Org. Chem., 34, 1233 (1969). W R.B. Woodward and R. Hoffmann, "The Conservation of Orbital Sym- metry", Verlag Chemie, GmbH, Weinheim/Bergstrof, West Germany, 1970, p. 159. A. Wexler and J.S. Swenton, J. Am. Chem. Soc,, 38, 1602 (1976). H.L. Slates, D. Taub, C.H. Kuo, and N.L. Wendler, J. Org. Chem. 22, 1424 (1964). A. Padwa and S.I. Wetmore, Jr., J. Am. Chem. Soc., 96, 2414 (1974). mm N. Gakis, M. Marky, H.J. Hansen, H. Heimgartner, H. Schmid, and W. E. Oberhanski, Helv. Chim. Acta, 59, 2149 (1976). W W. Steglich and R.A. Wellenstein, unpublished work cited in W. Steglich, Fortschr. Chem Forsch, 12 77 (1969). H. Gotthardt, R. Huisgen, and H.O. Bayer, J. Am. Chem. Soc., 92, 4340 (1970). ”m W. Steglich, P. Gruber, H.D. Heininger, and F. Kneidl, Chem. Ber., 18%, 3816 (1971). P. Gruber, L. Muller, and W. Steglich, Chem. Ber., 122, 2863 (1973). G. Hofle, unpublished work cited in ref. 61. M.R. Johnson and L.R. Sousa, J. Org; Chem., 42, 2439 (1977). A. Padwa, M. Dharan, J. Smolanoff, and S.I. Wetmore, Jr., J. Am. Chem. Soc., 22, 1945 (1973). A. Padwa and W.P. Koehn, J. Org. Chem., 40, 1896 (1975). mm W. C. Danen and C.T. West, J. Am. Chem. Soc., 95, 6872 (1973). mm P.J. Krusic, K.S. Chen, P. Meaken, and J. Kochi, J. Phyg, Chem. 78, 2036 (1974). P.(hramella and K.N. Houk, J. Am. Chem. Soc,, 98, 6397 (1976). W R. Brainard and H. Morrison, J. Am. Chem. Soc., 93, 2685 (1971). mm (70) (71) (72) (73) (74) (75) (76) (77) (78) (79) (80) (81) (82) (83) (84) (85) (86) 117 V. Stenberg in "Organic Photochemistry", Vol. 1, O.L. Chapman, Ed. Marcel Dekker, New York, N.Y., p. 127. K. Yates, S.L. Klemenko, and I.G. Csizmadia, Spectrochim. Acta. Part A, 25, 765 (1969). F.D. Lewis and C.E. Hoyle, J. Am. Chem. Soc., 27, 5950 (1975). P.A. Levene and R.E. Steiger, J. Biol. Chem., 22, 581 (1931). The signals at m/e 120, 105, 77, and 51 were due to acetophenone, present as an impurity in the sample. For review see (a) J.J. Christenson, D.J. Eatough, and R.M. Izatt, Chem. Rev” 74, 351 (1974); (b) C.J. Pederson and H.K. Frendsorff, AngewIVChem., Int. Ed. Engl., 1, 16 (1972); (c) G.W. Gokel and H.D. Durst, Synthesis, 168 978). (a) L.R. Sousa and J.M. Larson, J. Am. Chem. Soc., 99, 307 (1977); (b) J.M. Larson and L.R. Sousa, J. Am. CheMY Soc., 100, 1943 (1978). ”m” (a) L.R. Sousa and M.R. Johnson, J. Am. Chem. Soc., 100, 344 (1978). Other 13C NMR studies of crown ethers includXVM(b) D. Live and S.I. Chan, J. Am. Chem. Soc., 98, 3769 (1976); (c) M.-C. Fedarko, J. Mag. Res.,llg, 30 (1973YYV W.J. Hehre, R.W. Taft, and R.D. Topsom, Prog. Phys. Org. Chem., 1e, 159 (1976) and references cited therein. D.H. Hayes, A. Kowalsky, and B.G. Pressman, J. Biol. Chem., 244, 502 (1965). ”m“ R.M. Izatt, L.D. Hansen, D.J. Eatough, J.S. Bradshaw and J.J. Christenson, Jerusalem Symp. Quantum Chem. Biochem., 9, (1977). I» D. Doddrell and P.R. Wells, J. Chem. Soc. Perkin 2, 1333 (1973). N.K. Wilson and J.B. Stothers, J. Mag. Res.,llé, 31 (1974). O.A. Gansow, P.A. Loeffler, R.E. Davis, R.E. Lenkinski, and M.R. Willcott III, J. Am. Chem. Soc., 33, 4250 (1976). R.D. Topsom, Prig, Phys. Org. Chem.,'%%, 1 (1976). For a discussion of terminology see W. Kitching, M. Bullpitt, D. Gartshore, W. Adcock, T.C. Khor, D. Doddrell and I.D. Rae, J;. Org. Chem°z.$%’ 2411 (1977). (a) P.J. Mitchell and L. Phillips, J. Chem. Soc. Perkin 2, 109 (1974); (b) E.M. Schulman, K.A. Christensen, D.M. Grant, and C. Walling, J. Org. Chem".32’ 2686 (1974). (87) (88) (89) (90) (91) (92) (93) (94) (95) (96) (97) (98) (99) (100) (101) (102) 118 R.T.C. Brownlee, G. Butt, M.P. Chan, and R.D. Topsom, J. Chem. Soc. Perkin 2, 1486 (1976) W.P. Reynolds, I.R. Peat, M.H. Freedman, and J.R. Lyerla, Jr., Can J. Chem., 51, 1857 (1973). M. Karplus and J. A. Pople, J. Chem. Phys., 38, 2803 (1963). See ref. 88 for a good discussion. D.L. Ward, H.S. Brown, and L.R. Sousa, Acta Crystallogr., Sect. B, 33, 3537 (1977). ‘——————’ mm Several bond lengths and bond angles were adjusted to more reason- able values. These values are for the "adjusted monopole", not an ordinary ion. However, calculations with the ordinary Li+ and 2,3-diether yield a charge transfer of only 23 me” out of the ring. These values might even decrease for a calculation on the entire crown ether. Dicrowns are a frequent by-product of the crown ether preparations, and, at times, are the major product. D.J. Pasta and C.R. Johnson, "Organic Structure Determination" Prentice-Hall, Englewood Cliffs, N.J., 1969, p. 74. It is interesting to note that in the photophysical studies, the most dramatic effects are observed for 1,5-Cr-6, with much smaller ones for the 2,3 and 1,8 crowns (ref. 76). E.L. Eliel, W.P. Bailey, L.D. Kopp, R.L. Willer, D.M. Grant, R. Bertrand, K.A. Christensen, D.R. Dalling, M.W. Duch, E. Wenkert, F.M. Schell, and D.W. Cochran, J. Am. Chem. Soc., 97, 322 (1975). W. Adcock, B. D. Gupta, and W. Kitching, J. Org. Chem. ., 41,1498 (1976) L.R. Sousa, H.S. Brown, and C. Pshock, to be submitted for publica- tion. D.L. ward, H.S. Brown, and L.R. Sousa, unpublished results. Program written by H.S. Brown, Michigan State University. This program uses the method of J.A. Pople, D.L. Beveridge and P.A. Dobosh, J. Chem. Phys., 47, 2026 (1967). mm J.A. Pople and D.L. Beveridge, "Approximate Molecular Orbital Theory", McGraw-Hill, New York, N.Y., 1970. "I11111111111111.1111“