Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The study of the thermal decomposition of Tetracyclo[[sup 3,7].0[sup 1,3]]decane in octane and… McIntyre, Brian William 1976

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-UBC_1976_A6_7 M22_6.pdf [ 3.2MB ]
JSON: 831-1.0061038.json
JSON-LD: 831-1.0061038-ld.json
RDF/XML (Pretty): 831-1.0061038-rdf.xml
RDF/JSON: 831-1.0061038-rdf.json
Turtle: 831-1.0061038-turtle.txt
N-Triples: 831-1.0061038-rdf-ntriples.txt
Original Record: 831-1.0061038-source.json
Full Text

Full Text

THE STUDY OF THE THERMAL DECOMPOSITION OF TETRACYCLO^.S.I.I^.O^JDECANE IN OCTANE AND CUMENE AT 195° BY BRIAN WILLIAM M INTYRE C  B.Sc. (Hons) University of British Columbia, 1972 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in the Department of CHEMISTRY We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA June, 1976  ©  Brian William Mclntyre, 1976  In  presenting  this  thesis  an a d v a n c e d  degree at  the L i b r a r y  shall  I  f u r t h e r agree  for  scholarly  by h i s of  this  written  the U n i v e r s i t y  make  that permission  It  of  The U n i v e r s i t y o f B r i t i s h  for  Columbia  the  requirements  Columbia,  I agree  r e f e r e n c e and copying of  that  not  copying  or  for  that  study.  this  thesis  by t h e Head o f my D e p a r t m e n t  f i n a n c i a l gain shall  2075 W e s b r o o k P l a c e V a n c o u v e r , Canada V6T 1W5  British  is understood  permission.  Department  of  for extensive  p u r p o s e s may be g r a n t e d  for  fulfilment of  it freely available  representatives. thesis  in p a r t i a l  or  publication  be a l l o w e d w i t h o u t  my  -iiABSTRACT  Tetracyclo[ ' ^ » ]decane, also called 3  7  1  1,3-dehydroadamantane  3  or DHA, was heated at 195° C in n-octane and cumene solutions, with and without added 2,6-di-t-butyl-p-cresol, hereafter referred to as cresol. The kinetic order of the decomposition of DHA at 0.015 to 0.15 M solutions varied from 1.7 in octane to 1.5 in cumene and was 1.0 in solutions containing cresol. The overall rate of reaction increased on changing solvent systems in the order octane, cumene, octane with cresol and cumene with cresol. The f i r s t order rate constants for the reactions in the octane and cumene solutions containing cresol were 7.9 x 10- .min u  and 14.3 x 10" min 3  -1  3  _1  respectively. By comparison, the rate constants  from a f i r s t order calculation using the data for the reactions in octane and cumene were 2.2 x 10~ min 3  -1  and 4.6 x 10" min 3  -1  respectively.  The products of reaction in octane solution were l,l'-biadamantane, ca. 70%, adamantane, ca. 3%, polyadamantanes, ca_. 15%, and three unidentified liquids, ca. 10% total. With the addition of the cresol, the products were l,l'-biadamantane, ca. 70%, adamantane, ca. 17%, and the three unidentified products, ca. .10% total, with no polyadamantanes observed. In cumene solution, the products of reaction were 1,1'-biadamantane, 29%, 2-(1-adamantyl)~2-phenylpropane,  20%, 2,3-dimethyl-2,3-di phenyl butane,  37%, adamantane, 8%„ and an unidentified product, 5%. The presence of dimeric and polymeric products and the change in rate on addition of cresol suggest a radical chain mechanism. The initiation step was considered to be most likely the attack of the vibrationally excited DHA molecule on the solvent, the dissolved cresol i f present, or  another DHA molecule. Some brief, unsuccessful attempts were made to prepare DHA from 1-bromoadamantane using n-butyllithium and using KOH in methanol and from 1,3-dibromoadamantane using n-butyl Grignard reagent. No DHA was observed. The unusual bonding present in DHA results in a UV spectrum having four local maxima at 282-285 nm., 274 nm., 265 nm., and 254-257 nm., with the maximum at 274 nm., e = 13.  -1  V-  TABLE OF CONTENTS Page INTRODUCTION  1  RESULTS  15  Synthesis  15  Kinetics  18  Reaction Products  22  DISCUSSION  26  Alternate Preparation of DHA  26  Kinetics  28  Mechanism of DHA Thermal Decomposition  32  EXPERIMENTAL  40  Synthesis  41  Preparation of Kinetic Samples  47  Kinetic Runs  48  Product Studies  49  Preparation of Solvents  54  REFERENCES APPENDIX  ' •  57 59  -V-  LIST OF FIGURES Figure  Page  1  Inverted Bridgehead Geometry  2  UV Spectra of DHA and MBA in Octane Solution  17  3  General Mechanism of DHA Reaction in Solution  32  4  Reactions in Octane Solution  33  5  Reactions in Cumene Solution  33  6  Correlation Diagram for the Opening of [3.2.Upropellane to 1,3-dimethylenecyclohexane . . .  36  Sublimation Apparatus  44  7  3  _vi_  LIST OF TABLES Table  Page  1  Data from the Kinetic Runs  19  2  Mean Values of Kinetic Data from Table 1  20  \ \  -vi i -  ACKNOWLEDGEMENT  I would like to extend my most sincere thanks to Dr. R.E. Pincock for his help and guidance during the time of this work. I also wish the National Research Council of Canada and the H.R. MacMillan family for financial support during this time.  -1-  1NTR0DUCTI0N In recent years, interest in the bonding and properties of strained carbocyclic systems has led to the investigation of small ring tricyclo[m.n.p.O]alkanes, which are also called [m.n.plpropellanes . For 1  large values of m, n and p, the molecule is not expected to be significantly different in reactivity or bonding from a hexasubstituted ethane , but when m, n and p become small, such that the number of atoms 2  in the framework is  less than or equal to about ten, the physical size  of the cage structure forces unusual bonding at the bridgeheads of the zero bridge. Decreasing thermal stability and increasing reactivity toward electrophiles, radicals and to hydrogenation  with breaking of the  central bond become more pronounced as the sum of m, n and p decreases. Some of the larger of the small ring propellanes, such as [4.2.1]-, [4.2.2]-, [3.2.2]- and [3.3.1]propellane, have been investigated - . The 3  5  reactions reported for these and related compounds are shown below. None of these compounds are thermally reactive to any extent at 160° or 180° with the exception of [4.2.1]propellane, II,  for which no comment was made.  All reacted quickly with bromine, and [4.2.1]propell-3-ene, III, propellane, V, and [4.2.1]propellane, II,  [3.3.1]-  are all reported as reacting  with acetic acid with halflives of 1.6 to 19 hours. No appreciable hydrogenation was reported  using hydrogen and  platinum or Adam's catalyst.  -2-  160<  No reaction  H 1 atm. —2 > Adam's cat. 0  "no significant reaction"  3  Br  IM Br /CCl„ 9  "complete in 10-20 minutes"  25° C  3  Br H /Pt/Et 0 2  No reaction  2  HOAc  OAc  50° C  j^U kx^J  B  II  V  C H  2 2  >  C 1  -78° C  t  h  - 1.6 hour  k = 1.2±.05 x 10" "instantaneous"  4  sec*  1  k  Br 180°, 22 hr.  hypt/Etpp r  —r  as in  HOAc  No reaction  ^  50° C.  k = 1.0±.05 x 1 0 ° sec"  Br /CH C1 2  OAc  2  -78° C  k  1 ->  Br'instantaneous" No reaction  h  Br  1  4  -3-  160  c  H  No reaction  1 atm.  0  Adam's cat.  IV  IM  Br /ca ?  "No significant reaction"  Br "complete in 10-20 minutes"  A  25°  3  3  Br No reaction  180° 22 hours H /Pt/Et0H o  40 psi,6 hr. Br /CH C1o 0  No reaction No reaction  Br  "immediate"  0  -78°  V HOAc  100°  5  Br OAc  t,^ 8 h r .  OAc 46% With a ring contraction to [3.2.1]propellanes, the bridgehead carbon atoms become "inverted", where the geometry is such that the four interatomic vectors from the bridgehead atom to the four atoms bonded to i t all l i e on one side of a plane containing the bridgehead atom, as shown in Figure 1. The [3.2.1Jpropellanes that have been prepared are  r~p—7  Figure 1. Inverted bridgehead geometry  5  -4-  8,8-dichloro[3.2.1]propellane , VI, [3.2.Upropellane " ? 5  S-oxafS^.llpropellane . ^!!, 9  11  7  VII,  1  and tetracyclo[ . .0 , ]decane ,  IX. The structure of compounds VI, VII  2  5  1  6  12  and VIII were all determined by  electron diffraction or X-ray diffraction to have inverted bridgehead geometry . The central bond in VI is long 9  6  (1.57 A), as compared to the  o  normal value for a C-C single bond (1.53 A), and is typical of the long central bonds seen in small propellanes. The reactions reported for these compounds are given below.  VI  VII  VIII  IX  OAc  HOAc  <5 minutes, "rapi d  RT  VII 195°  > diphenyl ether  bridgehead coupled polymer, t. = 20 hours  O H  H  n~9  -5-  Br BrCCl , dark 3  V  CCl  VII  3  H /Pd 0  —  >  *  Q"  W%  C  H  3  7U tj = 4 hours  VIII Cl  <#  Cl  0  2  >  "relatively stable"  6  VI  318° gas phase  Br /CrLBr 0  •50  c  >  Br 0  ->  0.6 M HOAc in hexane at 25.6°  [WI  "rap"'  Br OAc v->_W  1.5 times as fast as VII,extrapolated to pure HOAc, t, = 1 sec.  -6-  The rapid reaction with bromine found with-the larger propellanes is s t i l l observed, although very l i t t l e chemistry of VI and VIII has been reported. The parent compound VII, and IX continue the observed trend however. The reaction of VII and IX with acetic acid is faster than with the larger propellanes, IX having an extrapolated h a l f - l i f e in pure acetic acid of one second . Compound VII begins to show thermal instability 12  at 195°, with a h a l f - l i f e of around 20 hours,  giving a polymer  with  bridgehead-bridgehead coupling , and VI rearranges at 190° with a half9  l i f e of four hours . One important reaction of VII is the reaction with 11  oxygen to give a peroxide copolymer of molecular weight of around 1700 . 10  Wiberg, Lupton and Burgmaier conclude that the strain energy of VIII is around 60 kcal mole very s i m i l a r  10  and that the parent [3.2. Upropell ane must be  -1  The central bonding arrangement is later described as  an essentially pure p-p a bond with the bridgeheads being sp  2  hybridized . 9  The conclusions that have been drawn for the [3.2.1]propellane bonding by Wiberg et a l .  6  include centre bonds consisting of s p  2 9 , l t  hybrids,  essentially pure p orbitals. They state that although the bridgehead is geometrically inverted, the bond hybrids at the bridgehead are not inverted, but that the six interhybrid angles range from 101° to 116°. The result of this is that although there is an internal bond, there is considerable electron density outside of the bridgehead, which results in the high affinity for and rapid reaction of [3.2.1]propellane with Lewis acids. Stohrer and Hoffmann  13  have done calculations on the [3.2. Upropell ane  system and they have examined the energetics of the system, including the energy difference between the HOMO and LUMO, which they predict to be around 4 eV, or 100 kcal mole" . The reactivity of [3.2. Upropellane to 1  -7-  acid, base and radical attack is calculated, with the conclusion that the compound should be stable to base, but highly reactive to acids and radicals. This conclusion is consistent with the known reactivity of [3.2.1]propellane. The isomer of [3.2.1]propellane, [2.2.2]propellane, X, has also been prepared. The f i r s t preparation was of the propellane amide X I , followed li+  by the preparation of X propellane s y s t e m ' ' 1 3  1 7  1 5  1 8  '  1 6  . Theoretical treatments of the [2.2.2]-  suggest  inverted bridgehead geometry with a  2i  X o  centre bond of 1.54A , longer than the C2 to C3 length 17  Newton and Schulman  17  18  of around 1.50A.  have calculated that the central bond hybrids should  be sp . 9  Since the reported preparations of X have been in low concentrations and have been verified by treating the reaction mixture with halogen with no isolation of X, most of the reported chemistry of [2.2.2]propellanes is of the amide XI. The reactions of [2.2.2]propellanes are given below.  X  Cl  -8-  % /  0  X  Br 25° CON(CH , 3  2  XI  f  .  r s t  o  r  d  e  r  k=4 x 1 0 - sec-i  t, = 28 min. 'a  ^90%  ^10%  -CONICH ) 3'2 C O N ( C H  E = 22 kcal mole" over 7°-35° 1  a  !l2 -70°  >  "instantly" -CON(CHo)  9  >  "does not react .... at an appreciable rate"  The chemical reactivity of the system to bromine and chlorine is consistent with the series of propellanes. Unlike [3.2.1]propellane which is reactive to oxygen, the [2.2.2]propel lane amide is relatively unreactive. The low thermal stability of XI is in agreement with  the  theoretical analyses of the [2.2.2]propellane system which show that X can isomerize to the singlet diradical XII.  XII  The barrier to this  XIII  isomerization is calculated to be around 29 kcal mole"*, which compares 17  well to the observed activation energy of XI of 22 kcal mole"*. Diradical XII  has a symmetry allowed pathway  for  the rearrangement  to  1,4-dimethylenecyclohexane, XIII. The t h e o r e t i c a l analysis of the [2.2.2]propellane  system by Stohrer and Hoffmann  13  p r e d i c t s that the  system should be stable to base a t t a c k , and should react r a p i d l y with Lewis acids and r a d i c a l s , but i n s u f f i c i e n t work has been reported to v e r i f y t h e i r c a l c u l a t i o n s . They also p r e d i c t that the d i r a d i c a l isomer of [2.2.2]prope11ane, XII,  has no counterpart in the  system, but that a d i r a d i c a l of the [3.2.1]  [3.2.Upropellane  system i s only a stretched  form of the parent compound and relaxes i n t o  [3.2.Upropellane.  The compound of i n t e r e s t in t h i s thesis i s t e t r a c y c l o [ 3 . 3 . 1 . I - . 0 > ] 3  decane. hereafter r e f e r r e d to as dehydroadamantane  or DHA, XIV.  19-22  7  ]  3  This  compound i s a bridged [3.3.Upropellane containing inverted bridgeheads, a: A = B = CHb: A = H, B - CN c: A = H, B = Br  XV  XIV  as has been v e r i f i e d by X-ray c r y s t a l l o g r a p h y on XVb, in w h i c h  2 3  the  o  central bond length i s 1.64 A. The pne carbon bridge between C5 and C7 changes [3.3. U p r o p e l l ane, which i s i t s e l f not extremely r e a c t i v e , to 5  DHA, which i s a very much more r e a c t i v e compound. The reason f o r t h i s large d i f f e r e n c e is postulated by Warner et a l .  5  to be due to the increased  s t r a i n on the cyclopropyl r i n g due to the pinching together of C5 and C7, and also to an e c l i p s i n g i n t e r a c t i o n between C1-C9 and C2-H r  [3.3.Upropellane  in exo  J  lacking in DHA. They conclude that DHA "resembles, in  geometry and r e a c t i v i t y , [3.2.Upropellane, rather than The reactions of DHA are very s i m i l a r to those of which also has inverted geometry, and are shown b e l o w  [3.3.Upropellane." [3.2.Upropellane,  1 9 - 2 2  .  -10-  130°-160°  polymer  -11-  DHA is extremely reactive to electrophilic and free radical attack. The reaction with electrophiles breaks the central bond and presumably forms a tertiary cation at the other bridgehead, which commonly reacts with the complementary anion. DHA was also found to react immediately with galvinoxyl free radical and the conclusion, possibly invalid, was drawn that DHA is either able to form carbon radicals or else is very radical-like in nature ; The reaction of DHA with free radicals similarly 24  breaks the central bond to leave a radical centre at the other bridgehead. DHA is highly reactive to oxygen, resulting in the precipitation of a white oxygen~DHA copolymer, which has an explosion point of about 160°  1 9  .  Dimethyldehydroadamantane, XVa, has also been made and is similar to DHA in its chemistry. DHA is stable at room temperature under nitrogen or vacuum either as the solid or in solution. As the solid, DHA polymerizes at approximately 140°. Dimethyl DHA is similar, melting before polymerization, however. Some 5-substituted DHA compounds have been made, XVb, XVc. Isolable  derivatives of XV are obtained when B has neither electron  donor nor solvolytic properties, and when B does not meet this requirement, the adamantane cage structure fragments to yield  bicyclo[3.3.1]nonyl  products . 22  DHA was f i r s t prepared by the reaction of a sodium-potassium alloy with dibromoadamantane, DBA, in heptane , but with the production of 19  adamantane as a side product. DHA has also been made from DBA using sodium-naphthalene , but the preferred method involves the reaction of 20  n-butyllithium with DBA in ether solution in the presence of HMPA at approximately - 40 °. Attempts to make DHA photolytically in this lab were  -12-  not successful , but a report has been made of making DHA from DBA 25  electrochemi c a l l y . 2 6  The reaction of DHA with oxygen has been studied by a previous worker in this l a b . A solution of DHA in octane or xylene under an 2 4  oxygen atmosphere at room temperature yielded a white peroxide copolymer. The f i r s t order reaction had a h a l f - l i f e of 125 minutes in octane and 50 minutes in xylene. A free radical reaction mechanism was suggested for the reaction of oxygen and DHA under these conditions, as shown below. In  DHA  Oo  -Ad-0 2  Oo  M > -0 -Ad-0 . 2  2  DHA Oo >—%>  DHA  >  the presence of a free radical inhibitor, the precipitate was not formed, indicating that the polymer was not formed. The reaction rate dropped to 25% and 40% in octane and xylene respectively with added free radical inhibitor, which was concluded to be due to the interruption of the DHAoxygen chain reaction by reaction with the inhibitor, terminating the growing chains. The ratio of the number of moles of oxygen taken up by the solution to the number of moles of DHA was around 1.25 with or without inhibitor. For a long polymer, a value of one would be expected, and for a one unit compound, a value of two would be seen. The reaction below is the suggested  DHA + 20  2  2 h  reaction reaction path for the inhibited reaction,  >  -00-Ad-OO-  inhibitor >  HOOAdOOH  The initiation of the oxygen-DHA reaction was thought to have been by way of one of two possible mechanisms, shown below. One mechanism involves the direct reaction of DHA with the oxygen diradical. The other  -13-  considers the possibility that DHA exists in equilibrium with the adamantane diradical, and that i t is the diradical that reacts with the oxygen molecule, leaving a diradical to react further. No conclusions were drawn as to which mechanism is responsible for the initiation step of the chain reaction of DHA and oxygen to give the DHA-oxygen copolymer.  DHA DHA  ^> ^  ^  -Ad-uy  ^  -Ad-  -0 -Ad-0 2  2  -Ad-0 2  %>-0 -Ad-0 2  2  The observed products of the DHA reaction with oxygen in the presence of 2,6-di-t-butyl-p-cresol, XVI, were 1,3-dihydroxyadamantane, formed from the diperoxy-  or dihydroperoxyadamantane in workup,  l,2-bis-(3,5-di-t-butyl-4-hydroxyphenyl)ethane, XVII, 3,3',5,5'-tetra-tbutylstilbene-4,4'-quinone, XVIII, and l-hydroxy-7-methylenebicyclo[3.3.1]nonan-3-one, XIX. Compounds XVII and XVIII were derived from XVI by loss of hydrogen atoms and subsequent dimerization. A mixture of crystals of XVII and XVIII was observed to be yellow. In an effort to further Understand the reactions and reactivity of DHA arising from its unusual bonding, this present study was undertaken. The thermal decomposition of DHA in solution was studied at 195°, and both the kinetics and product of decomposition were determined in order  to i n v e s t i g a t e the i n i t i a t i o n and possible chain reactions of DHA. The solvents used were n-octane and cumene> with and without added 2 , 6 - d i - t b u t y l - p - c r e s o l , XVI. This allows a comparison of systems of varying hydrogen donor a b i l i t y in order to learn about t h e mode of formation of free r a d i c a l s from DHA.  -15-  RESULTS Synthesis Preparation of 1,3-Dehydroadamantane 1,3-Dibromoadamantane, DBA, was prepared aluminum tribromide catalyst  in  a  24  solution of  by  bromination  using  adamantane in refluxing  bromine. The product was worked up in an ice-water-CCl^ system using NaHSOg and NaHCOg to destroy the remaining bromine and HBr. The solution was dried over ^ S O ^ , evaporated and taken up in hexane. After passing the solution through an alumina column, the DBA was recrystallized to high purity in 70% yield in hexanes. The DBA collected melted at 109-110.5° (corr) sealed tube, and was identified by nmr and glc comparison to an authentic sample.  27  Due to the great sensitivity of the reaction rate to  the amount and quality of the catalyst, i t was found necessary to monitor this reaction by glc. 1,3-Dehydroadamantane, DHA, was prepared from DBA in ether solution containing HMPA at either -35 to -25° or -50 to -45° by the addition of a 15% solution of n-butyllithium in hexane . The reaction mixture was 24  warmed to room temperature and washed with water. After drying over ^ S O ^ the solvent was removed and the DHA was sublimed under vacuum. The DHA collected was used immediately and was identified by its glc retention time and its reaction with oxygen. Halfway through the kinetic studies, the DHA preparation was improved by lowering the reaction temperature. The colder reaction temperature was found to give a higher yield of  -16-  DHA, and during the sublimation step, the product could be seen to be less contaminated by liquid reaction sideproducts, which sometimes left the sublimed DHA sticky or with a definite wet look. The DHA produced was stable indefinitely under nitrogen or vacuum in the washed and dried reaction mixture, or as the solid. The major impurities in the sublimed product were 1-bromoadamantane and 1-hydroxyadamantane, both in quantities less than 5%. Adamantane was not normally detected in the sublimed product. An attempt was made to make DHA from DBA reacting with n-butyl Grignard reagent at room temperature, in analogy with the n-butyl1ithiurn reaction but was not successful. No DHA was observed by glc in the reaction mixture. A few attempts to prepare DHA from monobromoadamantane, MBA, using n-BuLi or KOH in methanol, by elimination of HBr, were not successful. The use of n-BuLi did not give DHA, but gave adamantane and unreacted MBA. The use of KOH in methanol was not successful. Unreacted starting material plus two very small peaks, not DHA, were seen on the glc trace after 22 days of reaction. No adamantanol was seen. To gain information on the bonding in DHA, the UV spectrum was taken in n-octane solution versus n-octane in 1 cm. c e l l s , shown in Figure 2. The spectrum observed had four local maxima before cutoff, at 282-285 nm., 274 nm., 265 nm., and at 254-257 nm., with the maximum at 274 nm., e = 13. The spectrum of MBA, also seen in Figure 2, relates to the shoulder at 262 nm., the shoulder at 272 nm. and the small maximum at 282 nm., to which the presence of MBA in the solution, less than 5% of the total solid dissolved, most likely contributes. The spectrum attributed to the of MBA however, since MBA was found to have  presence  e = 1.7 at 274 nm. Small  -17-  250  300  350  WAVELENGTH IN Nm.  -18  amounts of adamantane and 1-adamantanol were also present in the sample, but have no appreciable absorption above 230 nm. The absorption is concluded to be due primarily to the DMA in solution. Due to the difficulty in handling DHA, the concentrations are only approximate since the reaction of DHA with oxygen produces suspended white particles which make the taking of U'V spectra impossible due to light scattering. Thus the DHA solution was prepared with as l i t t l e exposure to air as possible, without weighing the DHA.  Ki netics n-0ctane Solution The results of the kinetic runs of DHA in n-octane at 195-195.5°, obtained by following the DHA concentration by glc, may be seen in Table 1, runs 1 through 3. The apparent order of the decomposition of the DHA in n-octane was approximately 1.7, this being determined by comparing plots of the data to the f i r s t , three halves and second order integrated rate expressions. These and further plots of kinetic data may be seen in the Appendix. None of the plots has much curvature and in Tables 1 and 2, the rate constants calculated from each plot by least squares analysis are given for comparison. A condensation of Table 1 is given in Table 2. The total reaction time until the DHA concentration became too small to give reliable integration on the glc trace was around 1100 minutes.  Cresol Run Vials Solvent cone.  Initial Apparent DHA order of cone, reaction  1  21  octane  .032  2  2  18  octane  .014  1.7  3  33  octane.  .036  4  11  octane  5  17  octane  6  11  7  Half Order First Order 3/2 Order Second Order Rate const. Rate Rate const. Rate const. const. xlO" xlO" in XlO' 1 wV^in" min" min. M ^ ^ i n " M-^-min3  2  3  1  2.6±.2  1  262  1  Time in min.  .34±.01  4.5±.2 xlO"  2.2+.1 314  3.01.1  .45±.04  1020  1.7  1.9±.l  367  1.96±.04  .22±.003  1140  .015  -  3.8±.4  184  3.7±.4  .37±.04  .077  1.0  7.9±.2  87  6.8±.3  cumene  .025  1.0  4.6±.l  151  6.9±.4  1.2±.l  660  9  cumene  .023  1-1.5  4.41.2  157  5.6±.4  0.8±.l  540  8  23  cumene  .14  1.5  4.6±.2  149  2.45+.05  .137±.004  440  9  21  cumene  .093  1.5  4.8±.2  142  2.8±.l  .165±.008  400  10  11  cumene  .084  .11  1-1.5  2.4±.2  13.9±.4  50  9.2±.4  200  11  10  cumene  .084  .086  1.0  2.3±.2  14.7±.4  47  10.7±.7  180  .114  1.04±.09  3  630  200 360  The error values quoted are in determining the slope of the line by least squares analysis only, and do not represent the total error of the values from the true values. Concentrations are in moles per l i t r e .  -20-  TABLE 2. Mean values of kinetic data from Table 1 Half First Order 3/2 Order Second . , Order k t, . _ Order 2 . Apparent . , _ Jg k x 10 Time Cresol order of ^ xlO" , in Solvent cone, reaction M ^ ' ^ i n " min" min. M"^io^min M - ^ m i n - min. T  n ™  3  v .  a  „ +  n  K  x1  octane  3  1.7 .114  cumene cumene  1.0  1.04±.09  1.5 .084  1.0  1  -1  2.2±..6 314  2.5±.6  7.9±..2  6.8±.3  87  4.6±..4 149 2.4±.2  u  3  U  1  octane  2 2  i n  14.3±.,7  49  1  .34±.12  1140 360  4.4±2.5  .13-1.2  600  9.7±1.0  200  Concentrations are in M. The errors are derived from the observed scatter of values for the individual runs in each solvent/cresol system.  The result of the kinetic run of DHA plus 0.114 M 2,6-di-t-butylp-cresol, XVI, may be seen in Table 1, run 5, and in Table 2. The run in octane plus added cresol was of an apparent order of close to 1.0 by comparison of the half, f i r s t and three halves order plots in the Appendix. The total time of the reaction until the DHA concentration was too small to be reliably determined was 360 minutes. The addition of cresol increases the observed rate constants by a factor of 3 or 4 and reduces the total reaction time by a similar factor. In the octane solutions containing cresol, the solutions yellowed during the reaction, this becoming very noticeable after 150 minutes of reaction. This is thought to be due to the formation of XVIII by the loss of four hydrogen atoms and dimerization of the c r e s o l .  24  XVII and  XVIII are known products of the dimerization of the cresol radical X X * 28  produced by radical hydrogen abstraction from the cresol, and were also seen in the DHA-oxygen reaction studies done previously . 24  29  -21-  Cumerie Solution The results of the kinetic runs of DHA in cumene (isopropylbenzene) at 195-195.5° may be seen in Table 1, runs 6 to 9, and in Table 2. The apparent order of the reaction was approximately 1.5, by comparison of the plots of the data to half, first,, three halves and second order integrated rate expressions, which may be seen in the Appendix. The total reaction time was approximately 600 minutes. The f i r s t order rate constants are very consistent, but the second order rate constants vary over a factor of ten, as seen in Table 1. The change in solvent from octane to cumene resulted in an approximate doubling of the rate constants, and in the reduction of the reaction time by a factor of two. The results of the kinetic runs of DHA in cumene containing 0.084 M 2,6-di-t-butyl-p-cresol, XVI, may be seen in Table 1, runs 10 and 11, and in Table 2. The apparent order of the reaction was close to 1.0 by comparison of the plots of the data to the half, f i r s t , and three halves order integrated rate expressions, which may be seen in the Appendix. The total reaction time was approximately 200 minutes. The addition of cresol increased the rate constants by a factor of two to three and decreased the total reaction time by a factor of three.  Unlike the reaction of DHA plus cresol in octane the similar reaction in cumene did not produce a yellow colour. This is presumeably due to the lack of formation of the dimers XVII and XVIII of the cresol, indicating that the cresol radicals XX are not of sufficiently long lifetime and concentration to react with each other in this system.  Reaction Products n-Octane Solution 20 ml. of an approximate 0.2 M DHA solution in octane were heated at 185° for 5 days in a Carius tube. The reaction mixture was concentrated down to give a white solid and an oil containing adamantane  and four  main components at retention times 4.75 minutes, 5.25 minutes, 5.75 minutes and the white solid at about 40 minutes on a Carbowax column at 200°. A white deposit, very insoluble in all tried solvents, including CHCl^, CCl^ and CSg, was scraped from the inside of the reaction vessel. The oil was separated by preparative glc into about 25 mg. amounts of the  4.75 and 5.75 minute peaks, both being liquid. 100 MHz nmr data was  consistent with the components being monosubstituted adamantanes. The very small amounts of liquid obtained prevented further study. The white solid was identified as l,l'~biadamantane, XXI, m.p. 272-280° (corr) sealed tube, l i t  3 0  288-290°, the depressed and broadened  melting point being due to impurities in the solid. Identification was  -23-  made by mass spectrometry, analysis, IR, and nmr comparison to a genuine sample, retention time 40 minutes, m.p. 285-290° (corr) sealed tube, available in this lab and prepared by other means . The IR spectrum of 31  the white solid in a KBr pellet was in good agreement with the expected absorptions of monosubstituted adamantanes . 32  The white solid scraped from the inside of the Carius tube was identified as containing higher polymers of DHA, including the tetramer and pentamer, by high resolution mass spectrometry. The extreme insolubility of this solid in all tried solvents prevented the obtaining of nmr and solution IR spectra, but an IR spectrum was taken in a KBr pellet. The melting point of the solid was consistent with the solid being a high polymer of DHA, the material only browning in air at 325°, sealed tube. The only other product seen in any quantity in octane solution was adamantane, which was identified by its glc retention time of h minute later than DHA on Carbowax at 80°. The major product of the reaction in octane at 195° was biadamantane, XXI, about 70%. Other identified products were adamantane, 3%, and polyadamantanes, about 15%, containing the trimer, tetramer, and pentamer of DHA in quantities sufficient to show on a mass spectrum. Three unidentified products, about 10% total, were seen, giving nmr spectra like monosubstituted adamantanes, probably due to the reaction of intermediates with the octane solvent. In octane solution with added 2,6-di-t-butyl-p-cresol, XVI, the yield of biadamantane was around 70%, with a higher yield of adamantane at 17%. No polyadamantanes were seen beyond biadamantane. The three  -24-  unidentified peaks were again seen at around 10% total. The addition of cresol increased the amount of adamantane produced and seemed to eliminate the higher polymers of DHA. Cumene Solution 20 ml. of an approximate 0.2 M DHA solution in cumene were heated at 185° for five days in a Carius tube. The reaction mixture  was concentrated  down to give a white solid and an oil which partially solidified on drying in air at 30° for one week. The oil contained adamantane plus four major components by glc, retention times of 9 minutes, 13.5 minutes, 25 minutes and 39.5 minutes on a Carbowax column at 200°. The white precipitate was identified as biadamantane., XXI, having glc retention time 39.5 minutes and m.p. 269-275° (corr) sealed tube. Identification was made by comparison to a genuine sample , glc retention 31  time 40 minutes, using nmr, melting point, and analysis. Three fractions were separated out of the solution concentrate by glc, with retention times of 9, 13.5, and 25 minutes. The 39.5 minute peak was biadamantane and is omitted from further discussion.  XXII  XXIII  The 13.5 minute peak was identified by IR, nmr, melting point, mass spectrometry, and analysis as 2,3-dimethyl-2,3-diphenylbutane, XXII, called bicumyl henceforth, m.p. 114.5-116.5° (corr) sealed tube, l i t lit  3 4  3 3  115°,  118°. Heating a control sample of cumene for one week at 185° gave a  -25-  very small but noticeable amount of a compound with a glc retention time the same as bicumyl, presumeably bicumyl. The 25 minute peak was identified as 2-(l-adamantyl)-2-phenylpropane, XXIII, hereafter called cumyladamantane. The white hairlike crystals, m.p. 70.5-72.5° (corr) sealed tube, were identified by IR, nmr, mass spectrometry and elemental analysis. The only other products seen in any quantity were adamantane, identified by its glc retention time, and the 9 minute glc peak, which was not identified due to the very small amount collected. The products of reaction in cumene were therefore biadamantane, 29%, cumyladamantane XXIII, 20%, bicumyl XXII, 37%, adamantane, 8%, and the unidentified product at 5%. The DHA decomposed to yield adamantane, 17%, biadamantane, 62%, and cumyladamantane, 21%. The products of the reaction of DHA in cumene plus added cresol XVI were not investigated, except that the data from the kinetic runs showed that 34% of the DHA decomposed to adamantane, a substantial increase over the reaction in cumene alone.  -26-  DISCUSSION  Alternate preparation of DHA Some brief attempts were made to produce DHA by other means than the DBA and n-butyllithium reaction, both to compare the reaction  of  the analogous n-butyl Grignard reagent as well as to better understand the formation of and bonding in DHA. In the reactions with monobromoadamantane, the goal was to carry out the internal displacement reaction shown below, similar to the reaction of n-butyllithium with DBA, also shown. The success of the  »  BH + DHA + Br"  n-BuBr+ DHA + Br"  n-butyllithium reaction with DBA is no doubt due to the effect seen by Schleyer et a l .  3 5  in their nmr studies on the adamantyl cation, in which  the bridgehead hydrogen atoms' are the most deshielded in the molecule, rather than the nearest hydrogen atoms as expected. They consider that this  -27-  unusual effect is most probably due to the inside lobe interactions of the outward projecting bridgehead sp  3  orbitals. These reactions with MBA  were done to test i f a strongly basic system could be used to prepare DHA •by an internal displacement reaction aided by this interaction. Monobromoadamantane was treated with KOH in methanol for 22 days, but no DHA was detected. The reason for the failure of this reaction was most probably the insufficient basicity of KOH in methanol. The starting material was unchanged after 22 days with the exception of two very small unidentified glc peaks which appeared. Monobromoadamantane was reacted with n-butyllithium to see i f DHA would be a product. No DHA was seen. Only adamantane and unreacted starting material were detected in the glc trace of the reaction mixture. The failure of these attempts to make DHA from MBA was most likely due to the insufficient  basicity of the attacking  anion. The attempt to prepare DHA from DBA using a Grignard reagent was undertaken in an attempt to discover i f the Grignard reagent was capable of producing DHA. The reaction was carried out at room temperature by addition of DBA to the freshly prepared n-butyl Grignard reagent. A very wide variety of products was seen, indicating that the reaction conditions were such that the DHA, i f produced, reacted immediately with other species produced by side reactions or with the intermediates. These brief attempts at forming DHA by means other than those previously determined are only indicative of the very large amount of work to be done in investigating the chemistry and reactions of the relatively unknown bonding and reactivity in DHA.  -28-  Kinetics The difficulty of quantizing and handling DHA was the primary problem experienced in this study. The handling of DHA was greatly complicated by the rapid reaction of DHA with oxygen. As a result, DHA could not be exposed to air, either in solution or in the solid form since the peroxides or hydroperoxides formed are good free radical sources at elevated temperatures, a very highly undesirable  characteristic for these  kinetic studies. The DHA-oxygen copolymer at the temperature used is around 50° over its explosion temperature in the solid form , and at 195° 20  would be highly unstable. As a result, the DHA used was not weighed, but was used with as l i t t l e exposure to air as possible in the transfer from the sublimation apparatus to the flask from which the kinetic samples were made. The concentration values were determined by glc comparison to a standard solution of adamantane and to an internal standard during the actual kinetic runs. Extreme care had to be taken to avoid  peroxide and hydroperoxide  contamination in the kinetic vials, since the free radicals produced by the decomposition of the peroxides  and hydroperoxides at the temperature  used may attack the DHA present, giving lowered DHA concentrations. The varying amounts of peroxides and hydroperoxides in the different vials may also contribute to the scatter seen in the kinetic plots. In the case of the solutions of DHA in octane without cresol, the white oxygen-DHA copolymer was seen forming in visible amounts in almost all preparations and a syringe f i l t e r was used to remove the bulk polymer, but undoubtedly small amounts of the more soluble oxygen-DHA reaction products, such as dihydroperoxyadamantane, and the smaller particles  of copolymer remained  -29-  in the solution. In the case of solutions in cumene and solutions containing cresol XVI, however, the reaction products of DHA with oxygen were not separable since no precipitate results. The reaction occurring is most likely that shown below, in which the chain reaction neccessary to  to produce the polymer is interrupted by the reaction of the radical with the easily removed hydrogen atoms of the solvent or the cresol. This results in a soluble product from the DHA-oxygen reaction. An additional complication in the solutions in cumene is that the oxygen-DHA reaction would also initiate the autoxidation of cumene ' , 36  37  as shown below, and would result in the presence of cumene hydroperoxide  which not only has been shown to be a good source of free radicals, but has also been used as a free radical initiator in the plastics industry  3 8  .  The DHA-oxygen reaction in solutions containing cresol XVI is shown below . This results in a peroxide and a hydroperoxide, in which R is the 36  -30-  adamantyl, 3-hydroperoxyadamantyl or similar group. These reactive compounds are similarly not removable from the solution, and as a result the exposure to the air of DHA or solutions of i t was kept to a minimum in order to have acceptable solutions. The large scatter seen in the kinetic runs, sometimes making the results unuseable, such as in run 4, given as an example of the degree of scatter occasionally observed was considered to be primarily the result of the d i f f i c u l t y in quantizing DHA by glc. Although the glc column used for separating the DHA solutions was used only for this one purpose, the detection and reproduceability of the DHA peaks varied from day to day. Since each analysis took 40 to 50 minutes, a long run of ten or twenty vials took many hours to analyze, which could result in errors arising from the glc response changes over that time period. Two successive injections of the same solution of DHA could also at times vary in response by as much as 20% or could give nearly identical results. The errors quoted in Table 1 are the error generated by the scatter of points in the individual plots only. In a more systematic fashion, the average response of the chromatograph varied from day to day, presumeably due to the column, and this is the probable cause of the variation  of  values for the rate constants in Table 1. The errors given in Table 2 are estimated from the errors and values in Table 1.  -31-  A further source of scatter in the plots was thought to be the varying amounts of peroxides and hydroperoxides in the individual vials formed from the DHA-oxygen-solvent reactions during the preparation of the vials. In an attempt to reduce the effect of the presence of these compounds, time zero of the kinetic runs was chosen arbitrarily at about fifteen minutes after the immersion of the vials in the heating bath, thus giving a better value for the i n i t i a l concentration. The d i f f i c u l t y in quantizing DHA probably arises from a combination of polymerization in the injector and from reaction of DHA with other species on the column. Thermal polymerization on the surface of the injector is unavoidable since the injector is normally kept at a temperature well above the around 140° at which DHA spontaneously polymerizes. A similar condition exists in the detector. The loss, however, does not seem to be great due to this factor. The greater and changeable loss is attributed to acidic or radical species and catalysts produced on the column support by thermal and oxidative degradation, since the losses are more noticeable after the column has remained unused for a long period of time, and in general increase with the age of the column. The results of the kinetic study are shown in Table 2. The general trend noted is that the overall rate of reaction increases and the order of the reaction decreases to a value of around 1.0 as the hydrogen atom donor ability of the system increases from octane to cumene to octane plus cresol XVI to cumene plus cresol XVI.  -32-  Mechanism of DHA Thermal Decomposition The postulated general mechanism of reaction of DHA in solution is shown in Figure 3. From the products of reaction in cumene, especially bicumyl XXII, which is a known product of free radical reactions in c u m e n e ' ; , and also from the increased yield of adamantane in 33  39  40  solutions containing cresol XVI, a free radical mechanism is deduced, as shown in Figure 3.  INITIATION  DHA -— > DHA + SH -— > DHA + DHA -— >  -Ad-AdH + S-Ad-Ad-  (1) (2) (3)  PROPAGATION  R- + DHA -— > R- + SH -— >  R-AdRH + S-  (4) (5)  TERMINATION  R. + R*- -— > R- + XH -— >  R-R' RH + X- (unreactive)  (6) (7)  Figure 3. General mechanism of DHA reaction in solution The reactions most likely responsible for the observed products in octane solution are shown in Figure 4, with the exception of the initiation steps, which will be discussed later. The octane radicals produced  and their further reactions  unidentified  are the postulated source of the  products seen. In cumene solution, the reactions of Figure 5  are most likely responsible for the observed products. Due to the extreme reactivity of DHA toward radicals and the thermal stability of DHA at 195°, having a h a l f - l i f e measured in hours, the concentration of free radical species in the DHA solution must be very small at any one time. The slow step of the reaction must therefore be the initiation step.  -33-  PROPAGATION  (-Acl-4-H + DHA -  n •AdH + SH -(-Ad-^H + SH — -> S- + DHA - -> S-Ad- + SH - ->  >Ad-(-Ad4fjH  n = 1 to 4  (1)  2-5  (2) (3)  HAdH + S<  H4-Ad-^H + SS-AdS-AdH + S-  (4) (5i  TERMINATION  H-4-Ad-)^ + -(-Ad-^H $. + •(_Ad-4^H S- + SR- + cresol 2 cresol•  -> H-4-Ad-4jppjH m+n = 2 to 5  (6)  -> S-(~Ad-4-H n = 1 to 4 n S-S RH + cresol• coloured products, etc.  (7) (8) (9) (10)  Figure 4. Reactions in octane solution. Reactions (9) and (10) are in solutions containing cresol XVI. SH is n-octane.  PROPAGATION  TERMINATION  •AdH •AdH CuCuAd•AdAdH  + + + + +  DHA CuH DHA CuH CuH  •AdH + Cu•AdH + -AdH Cu- + CuR- + cresol cresol• + Cu •  -> -» -> -> -»  -Ad-AdH HAdH + CuCuAdCuAdH + CuHAdAdH + Cu'  (1) (2) (3) (4) (5)  •» -> •> ->•  CuAdH HAdAdH CuCu RH + cresol• cresol-Cu  (6) (7) (8)  (9) (10)  Figure 5. Reactions in cumene solution. Reactions (9) and (10) are in solutions containing cresol XVI. CuH is cumene, CuCu is bicumyl. The initiation step in solution is most likely one or both of reactions (1) and (2) of Figure 3, f i r s t order reactions in DHA similar to the postulated initiation reactions in the DHA-oxygen reaction . Reaction 24  (3) does not seem to be a major initiating step, although the possibility of i t contributing cannot be ruled out, since the presence of adamantane as a product argues for the presence of a f i r s t order initiating reaction  -34-  being present. The high order of reaction in octane and cumene, around 1.7 and 1.5 respectively, is consistent with a f i r s t order initiation followed by an induced reaction, that of the attack on DHA by radical species  produced by the initiation reaction. The induced reaction  introduces higher order terms in the kinetics, increasing the observed order. When switching to a system with a better hydrogen donating capacity, the order lowers, such as observed on changing from Octane to cumene where the order drops from 1.7 to 1.5. The yield of adamantane also increases in cumene, as does the production of chain termination products not containing DHA. In going to solutions containing cresol XVI, the order drops to close to 1.0, presumeably due to the elimination of chain reactions by the enhancement of chain transfer to form the relatively unreactive cresol radical XX. Since the order of around 1.0 argues for the elimination of significant amounts of induced reactions, the order of the overall reaction should reflect the order of the initiating reaction or reactions under these conditions. The f i r s t order reactions (1) and (2) in Figure 3 are the likely reactions involved and they represent the two extremes possible. Reaction (1) and the resulting reactions consider the possibility that the DHA molecules become thermally excited to form a highly reactive diradical species which then reacts further with other molecules by hydrogen abstraction or by attacking other DHA molecules. Reaction (2) considers the possibility that the DHA molecule is not a diradical species, but reacts to abstract a hydrogen.atom from the solvent system to produce two discrete radicals, with reaction (3) being the attack of this excited DHA molecule on another DHA molecule.  -35-  The diradical species postulated by reaction (1) would be a thermally excited species, presumeably with the promotion of one electron to the LUMO to produce a diradical. For [3.2.Upropellane, for which the theory has been worked o u t  13  and to which DHA is similar in its reactions and  bridgehead geometry, the thermal decomposition also gives the bridgehead coupled product with a h a l f - l i f e at 195° of around 20 hours , around five 9  times that of DHA. The calculations of Stohrer and Hoffmann  13  for [3.2.1]-  propellane show a LUMO - HOMO energy gap of around 100 kcal mole" at the 1  equilibrium bridgehead-bridgehead distance  which drops to about 20 kcal o  mole" or less when the bond is stretched by around 0.4 A, the stretching 1  requiring around 30 kcal mole . Correspondingly, a weak UV absorption is -1  seen  in  DHA, with  the maximum at 274 nm., which  corresponds  to  an  absorption of 104 kcal mole . Since DHA is a larger system, stretching -1  the bridge should be easier due to the lesser deformation of the cage involved. This could account for the faster reaction of DHA at 195°, i f this mechanism is responsible for the initiation. An energy of at least around 50 kcal mole to promote an electron to the LUMO,  -1  therefore seems necessary  and the activation energy for the  abstraction of a hydrogen atom must be added to this. Considering that the activation energy for hydrogen abstraction from a hydrocarbon by, for example, methyl radical  is  around  7 kcal  mole , - 1  4 1  and  that  abstraction by a tertiary radical like DHA would have a higher activation energy, an overall activation energy of around 60 kcal mole  -1  would be  needed to have abstraction by.a diradical DHA species. Due to this very large activation energy, i t would seem that a true diradical species is not responsible for the initiation step  resulting  in  the  ultimate  -36-  formation of adamantane. The only adamantane  reasonable mechanism  producing  from DHA involves the abstraction of two hydrogen  atoms  consecutively, reaction (2) of Figure 3 followed by reaction (5). The  bridgehead  coupled  product of the thermal decomposition  of  [3.2.llpropellane provides additional information as to the characteristics of the initiating species. As stated by Stohrer and Hoffmann, the fragmentation of the [3.2.1]propellane to 1,3-dimethylenecyclohexane, XXIV,  03 # VII  & XXIV  is a symmetry forbidden reaction due to the level ordering shown below, since the [3.2.1]propellane molecule correlates to a doubly excited 1,3-dimethylenecyclohexane molecule. Examination of the correlation diagram leads to the conclusion that a symmetry allowed pathway exists for the singlet, excited diradical of [3.2.1]propellane to fragment to the excited  Figure 6. Correlation diagram for the opening of [3.2.1]propellarie to 1,3-dimethylenecyclohexane, from reference 13.  -37-  1,3-dimethylenecyclohexane. Since upon heating the [3.2.1]propellane in diphenyl ether at 195°, Wiberg and Burgmaier did not find 1,3-dimethylene9  cyclohexane, the conclusion can be drawn that the [3.2.1]propellane diradical did not exist in appreciable quantities, but that the bridgehead coupled product observed must arise from another intermediate. Although DHA has no analogous fragmentation reaction due to the lack of a two carbon bridge, the similarity in geometry, reactivity and products formed to those from [3.2.1]propellane leads to the conclusion that a discrete diradical species is not responsible for the initiation of the thermal decomposition of DHA in solution. In a similar fashion the presence of a diradical of DHA in the initiation of the oxygen-DHA reaction previously studied  24  is also not  indicated, especially since a DHA ^ DHA-diradical equilibrium should be very greatly shifted to favor the diradical by the temperature increase of around 170° of this present study over the previous study. As this effect is not seen, a diradical is effectively eliminated as an initiating species in the room temperature oxygen-DHA solution reaction. The initiating reaction in the-oxygen-DHA reaction is most likely the direct attack of the t r i p l e t oxygen diradical on the DHA molecule, giving rise to •00-Ad« diradical. This is consistent with the high reactivity of DHA to free radicals, as DHA reacts with the relatively stable Galvinoxyl r a d i c a l . 24  The initiating species in the DHA thermal decomposition  seems to be  the stretched DHA molecule. As the bridgehead-bridgehead distance increases as the molecule is stretched, the bonding interaction would be reduced due to the lesser overlap of the essentially pure p orbitals involved in the bonding. This would result in a higher reactivity of the central bond and an increase in free radical type character, especially since there is a  -38-  substantial portion of the orbital electron density outside of the cage structure , which also leads to the high reactivity of inverted bridgehead 6  carbon atoms with electrophiles. Upon the stretching of the molecule associated with the higher vibrational levels, the two bridgeheads should exhibit mere and more radical-1 ike character, with the relief of strain in the DHA molecule providing the driving force for any reactions taking place with the stretched DHA molecule, as shown in reactions (2) and (3) of Figure 2. Whether any energetically reasonable stretching of the DHA molecule could result in an increase of the radical character of the bridgehead orbitals sufficient to result in the occurrence of free radical type reactions is not known. The size of the DHA cage, and the one carbon bridge, effectively limit the stretching of the bridgeheads to not more o  than about 2.5 to 2.6 A. The faster rate of DHA versus [3.2.Upropellane decomposition is again in agreement with the larger and more bendable cage structure of DHA. Theoretical calculations of the behavior of the bridgehead orbitals as the bond is stretched are needed to examine the validity of this theory of the initiation of the thermal' decomposition of DHA and [3.2.Upropellane. The DHA system shows very great promise in the preparation of adamantyl systems' and polymers. Further study of the chemistry of DHA is warranted in an attempt to understand and more fully use the chemical properties of DHA. In continuing these studies, a study of the thermal decomposition of DHA in a fluorocarbon solvent may give  information  in the rate and order of reaction without the presence of available hydrogen atoms. A study into the photochemistry, i f any, of DHA is also  -39-  indicated. The thermal polymers of DHA show promise* having an upper useful temperature limit greatly in excess of tetrafluoroethylene, especially in the absence of a i r , and this is an area of possible commercial application of DHA and its derivatives.  \  -40-  EXPERIMENTAL  Proton nmr spectra were taken at 60 MHz on a Varian T-60 spectrometer unless specified as 100 MHz, in which case the spectra were taken on a Varian XL-100 spectrometer. Infrared spectra were taken on. a Perkin-Elmer 137 Sodium Chloride Spectrophotometer, principally of solutions using one set of matched 0.5 mm c e l l s , consistently using one of the cells for the sample. The solvent used or any other method of sample preparation will be mentioned. Classification used to classify absorption strengths are for example very strong (vs), strong (s), medium (m), or weak (w). Analytical gas-liquid chromatography (glc) was done on a PerkinElmer 900 gas chromatograph with flame ionization detection using either a 0.125" x 6' Carbowax 20M 10% on Chromosorb W 80/100 mesh  column,  abbreviated K-20M, or a 0.125" x 6' OV-17 8% AW-DMCS Chromosorb W 80/100 mesh column, abbreviated 0V-17. The helium flow rate was approximately 30 ml min" . The temperature programming of the machine, i f used, is 1  abbreviated as, for example, 80°/7 min ^ M l L ^ O O , meaning after 7 0  minutes at 80°, the temperature programs up at 32°/minute and upon reaching 200°, remains at 200°. Preparative glc was done on a Varian Aerograph, model 90-P, using a V  -41-  x 5' 10% Carbowax 20M on Chromosorb W 80/100 mesh column, with a helium flow rate of about 60 ml min" . The peaks were collected in dry ice and 1  . isopropanol cooled vessels. Melting points were performed on a Thomas-Hoover Unimelt capillary, melting point apparatus. Mass spectrometry was low resolution unless specified as high resolution. The high temperature bath used for kinetic runs was a stirred silicone oil bath, with temperature regulation using a Philadelphia Micro-set thermal regulator and an observed temperature range of 195° to 195.5°. UV spectra were taken on a Carey 15 visible, UV spectrophotometer at a scan'rate of approximately one or two nanometers per second, using the automatic s l i t setting. Matched 1 cm. quartz cells were used for the sample and reference, consistently using one cell for the reference sample.  Synthesis of 1,3-Dibromoadamantane 1,3-Dibromoadamantane, DBA, was made in varying amounts at three different times by the general method  24  given here.  To a stirred mixture of 10.0 grams of adamantane (Aldrich 10,027-7) in 40 ml. of dry bromine (previously shaken twice with H2S0^(conc)) in a 200 ml. three neck flask fitted with a magnetic stirring bar, a condenser, and two stoppers, at 0°, was added as catalyst 100 mg. aluminum bromide (BDH 27071). After the vigorous evolution of HBr had moderated (about five minutes), the mixture was refluxed for one half hour. The reaction mixture was cooled to 0° and a further 200 mg. of aluminum bromide was added, after which the reaction mixture was allowed to stand at room  -42-  temperature for one hour and was then refluxed for one half hour. Crude glc analysis (OV-17, 200°) showed 98+ % dibromoadamantane, traces of mono- and tribromoadamantane and traces of two compounds seen as yellow impurities. It was discovered that i t was important to follow this reaction by glc, as the rate of reaction was very dependant on the quality of catalyst used, and since this is a sequential reaction, f i r s t forming MBA, followed by DBA and then tribromoadamantane. The correct moment for stopping the reaction was determined for each preparation separately so that adequate results could be consistently obtained. The aluminum tribromide bottles used contained both white powder and rose coloured lumps. The rose coloured lumps only gave the desired results. The reaction mixture was worked up by cooling to room temperature and pouring into a 3 l i t r e separatory funnel containing 200 ml. CCl^ and 500 grams of ice and water. Sufficient solid NaHSOg was added in a fume hood to reduce the excess bromine, ice being added to keep the mixture at 0°. The CCl^ layer was separated, 200 ml. of water were added and the mixture was neutralized with solid NaHCO^. The CCl^ layer was washed with water twice more, dried over Na^^O^ overnight and evaporated to dryness. A large part of the yellow colour was removed by passing a concentrated hexane solution of the product through a 5 cm. by 2 cm. column of reagent grade alumina (Shawnigan 6250). The product was then recrystallized to high purity from hexane. The product was identified by melting point, 109-110.5° (corr) sealed tube, nmr and by glc comparison to an authentic sample . 27  -43-  Synthesis of  1,3-Dehydroadamantane  The method used was basically that of a previous worker in this l a b , 2 4  except that the reaction conditions were improved. A 100 ml. round bottom flask containing 2.0 grams of dibromoadamantane and a I V  magnetic  stirring bar was fitted with a rubber septum and was three times evacuated by an aspirator and flushed with dried L grade nitrogen. Deoxygenated anhydrous ether (Mallinckrodt 0848), 40 ml., were added using a syringe, and similarly 5 ml. of deoxygenated hexamethylphosphoramide (Fisher H-343) were added. The resulting solution was further degassed by bubbling nitrogen through using syringe needles for one minute, with stirring. The flask was then cooled in an isopropanol-dry ice bath to either -35° to -25° or -50° to -45°, the cooler temperature being discovered to give a better yield of cleaner product half-way through these studies. Upon cooling to the specified temperature, a small amount of dry L grade nitrogen was added to remove the pressure differential between the flask and the room pressure, and 10 ml. of 15% n-butyllithium (Foote Mineral Co.-MCB) were added dropwise over a period of 5 to 10 minutes with the formation of a white precipitate, giving at the end of addition a solution of porridgelike consistency. After 5 minutes the reaction mixture was allowed to warm to room temperature during which time the precipitate dissolved. About 30 ml. of deoxygenated, d i s t i l l e d water were added for washing. Failure to wash the reaction mixture resulted in eventual (2 or 3 days) loss of any DHA present. The organic layer was washed three additional times with the deoxygenated water, and was dried over Na S0 with the organic layer o  d  -44-  always under nitrogen. The solution was then transferred to the sublimation apparatus, Figure 7, containing isopropanol in the cold finger. Immediately after the addition of the organic layer to the flask and addition of the flask to the apparatus, dry ice was added to the isopropanol, the stirrer was turned on and the stopcock to the vacuum pump was momentarily opened a few times and was used to control the boiling of the ether. Liquid nitrogen was then added into the Dewar flask. Using hand heat and a warm water bath along with control of the vacuum, the solvent was removed to the liquid nitrogen trap. After the removal of the solvent, the system was flooded with nitrogen and the adapter with stopcock at A was replaced with a 25 ml.  Nitrogen  I  isopropanol dry iceB  vacuum connected here i n i t i a l l y  44— 1iquid nitrogen  Figure 7. Sublimation apparatus.  vacuum p u m p  -45-  round bottom flask, the vacuum pump connection was moved from point B to point C and stopcock D was used to control the vacuum. A 70° water bath was used on the flask containing the organic layer residue to complete the sublimation (about 20 minutes). After the sublimation was complete, the assembly was flooded with nitrogen and was disconnected at E with stopcock D closed. The cold finger was emptied and warmed up with warm tap water to prevent condensation of water vapour on the cold finger when the apparatus was opened. The apparatus was then opened and the DHA was used directly. After sublimation, the major impurity was monobromoadamantane, maximum 5%, with a trace of adamantanol and no adamantane. The DHA was analyzed by glc (K-20M, 80°/7. min  3 2  °  / m i n  > 2 0 0 ° ) , and has a usual  retention time of around 8 minutes, % minute later than adamantane. DHA prepared by a similar technique  has been characterized in this lab  previously , so no further characterization was made. 19  The UV spectra of Figure 2 were taken in deoxygenated octane and the dilutions were only approximate, due to the d i f f i c u l t y in handling DHA. The solid peroxide polymer suspended in the solution after the reaction with even a trace of oxygen made the taking of a UV spectrum impossible, as was discovered the f i r s t time the spectra were taken. The DHA solution used in these spectra was filtered through a syringe f i l t e r before use and was handled under nitrogen at all times.  Attempted Preparation of DHA from Monobromoadamantane 1) Using n-Butyllithium To 120 mg. of monobromoadamantane, MBA, in 5 ml. dry ether  -46-  (Mallinckrodt 0848) and 0.6 ml. HMPA (Fisher H-343) at -25° under nitrogen was added 1.0 ml. 15% n-butyl1ithium in hexane (Foote Mineral Co. - MCB). A small amount of white precipitate was seen. After a few minutes this was warmed to room temperature. Not all of the precipitate present dissolved. The reaction mixture was washed with deoxygenated water and run on the glc (K-20M, 80°/7 min  3 2  °  / m 1 n  > 2 0 0 ° ) . No DHA was seen,  but a very large MBA peak was present and a very large adamantane peak was seen as well. 2) Using KOH in Methanol To a solution of 60 mg. KOH in 1.5 ml. of methanol was added 100 mg. of MBA. A glc check (K-20M, 80°/7 min  3  2  °  /  m  n  >200°) of the solution  was made after 5 minutes, 18 hours, 36 hours, and 22 days. No DHA was seen in any check. Two small peaks, not DHA, were seen in the 22 day trace, but the MBA peak was not noticeably affected.  Attempted Preparation of DHA from DBA and n-Butyl Grignard Reagent Magnesium turnings, 0.7 grams, n-butylbromide, 1.85 grams, 25 ml. of dry ether, a small crystal of iodine and 0.5 ml. of 1,2-dichloroethane were refluxed for 90 minutes, after which 0.5 grams of DBA were added and the mixture was allowed to stand for 15 minutes. The reaction mixture was filtered, washed 5 times with 20 ml.of deoxygenated water and checked by glc (K-20M, 80°/7 min  3 2  °  / m i n  > 2 0 0 ° ) . No DHA was observed,  although a very wide variety of unknown products was seen.  -47-  Preparation of Kinetic Samples Due to the extreme reactivity of DHA with oxygen and the subsequent peroxide product which acts as a radical source under the kinetics conditions, no attempt was made to weigh the amount of DHA used. The DHA was immediately washed from the sublimation apparatus cold finger into the flask used to prepare the kinetics solution with a measured amount of the solvent used. The concentration of the DHA used was determined by glc comparison to an adamantane standard solution of similar concentration. The flask was quickly fitted with a rubber septum, evacuated and flushed with nitrogen three times. Nitrogen was also then bubbled through the solution for a few minutes using a long syringe needle. A measured amount of an internal standard (n-dodecane in the case of n-octane solutions, and para-xylene in the case of cumene solutions) was put into the solution using a 50 or 100 wi syringe. In the case of runs containing 2,6-di-t-butyl-p-cresol, XVI, the weighed solid was placed in the flask before the addition of the DHA solution. After purging, the flask was always kept under a positive pressure of nitrogen to avoid oxygen contamination. One half mi 11ilitre of the prepared solution was placed in each of ten one ml. ampoules (Kimble 12012-L) using a graduated 5 ml. syringe. In the case of the octane solutions without cresol, this was done through a syringe f i l t e r , since most of the peroxide forms a precipitate ("snows") and can be filtered out. In the case of cumene solutions and octane solutions with added cresol, this f i l t e r i n g was not done, since the peroxide formed under these conditions does not precipitate.  -48-  The ampoules were immediately connected to the vacuum system using a twelve stopcock adapter, frozen with liquid nitrogen and evactated. Upon thawing, the ampoules were sealed under vacuum after the bubbling due to dissolved gas ceased. The ampoules with octane as a solvent were stored until used in the refrigerator, and the ampoules with cumene as a solvent were stored in liquid nitrogen until used, especially the ampoules containing cresol.  Kinetic Runs The kinetic runs were done in a 195° silicone oil bath with the ampoules being quenched in a bath of tapwater, at approximately room temperature, the time recorded being the time of quenching. The time of immersion in the hot oil bath, until the time the f i r s t sample was taken, 0.0 minutes, was 10 to 15 minutes. This interval allowed for temperature equilibration in the ampoules. The concentration determined for the 0.0 minute sample is the stated i n i t i a l concentration of the run. The quenched samples were numbered and immediately analyzed on the Perkin-Elmer 900 chromatograph, (K-20M, 80°/7 min  3 2  °/  m i n  ,200°) and the  glc traces were integrated for the internal standard, DHA, and in some cases adamantane by cutting and weighing from three Xerox copies of the traces and averaging the values with respect to the internal standard. The reproduceability of the integration of the peaks was tested by comparing the weight of peak ratio of twenty copies each of three sets of peaks, consisting of two peaks about 10 cm. high, two peaks, one 10 cm. high and one 3 cm. high and two peaks, one 10 cm. high and one 2 cm. high. The standard deviations expressed in percent were 1.5%, 2.3% and  -49-  3.8%. A reasonable figure for the integration error as a whole in the kinetic runs would be thus 4%, since a large portion of the DHA peaks, especially on the longer runs, were less than 3 or 4 cm. The order of the decomposition was determined by plotting the data according to the integrated rate expressions for an order of h, 1, 3/2, and 2, and observing the best straight line visually.  OctaneSolution Product Studies One preparation of DHA made at -50° as given earlier was dissolved in 20 ml.  n-octane previously degassed under reduced pressure by bubbling  dry nitrogen through, and was degassed and sealed in a Carius tube under vacuum. The Carius tube was then heated at 185° for five days in the silicone oil bath, after which the tube was removed from the bath. A white coating quickly formed on the inside of the tube as the tube was cooling. After .cooling to room temperature, the solution in the tube was frozen with liquid nitrogen. After the solution melted, the tube was opened and a white precipitate was filtered out, collected and dried in air for one week. The remaining solution was vacuum d i s t i l l e d to leave a yellow oil and a white precipitate with the same glc retention time as the white precipitate filtered from the solution. The oil was centrifuged overnight to settle the white precipitate, and the about 0.2 ml. of oil was decanted off. The slightly yellow oil was passed through a h" x 2' alumina column (Alupharm Woelm neutral) in hexane and reconcentrated to give a colourless oil containing four main components, glc retention times 4.75 minutes, 5.25 minutes, 5.75 minutes and the dissolved white solid at about 40 minutes (K-20M, 200°).  -50-  The o i l was separated into about 25 mg. amounts of the three components, all liquid, by preparative glc at 135°, using K-20M, using very many small shots due to the very close retention times. The estimated purities were  >99% for the 4.75 and 5.75 minute peaks and  >95% for the 5.25 minute peak. The 4.75 minute peak gave 100 MHz nmr: 0.896(m), 6H; 1.266(m), 8H; 1.526 and 1.656 (m), 12H; 1.93fi(v. broad s), 3H, CDC1 solution. The 3  5.75 minute peak gave 100 MHz nmr: 0.85fi(m), 7H; 1.286(m), 7H; 1.506 and 1.666 (m), 12H; 1.96fi(v. broad s), 3H, CDC1 solution. This nmr data is 3  consistent with monosubstituted adamantanes with 12H at about 1.56 and 1.656 and 3H at about 1.956. The very small amounts of liquid obtained prevented further study. The white solid was identified as l,l'-biadamantane, XXI, m.p. 272280° (corr) sealed tube, l i t  3 0  288-290°, the depressed and broadened  melting point being due to impurities in the solid. Identification was made by comparison to a genuine sample, retention time 40 minutes, available in this lab and prepared by other means , m.p. 285-290° (corr) 31  sealed tube. 60 MHz nmr:. White solid-  1.606(m), 4H; 1.966(v. broad s),lH in CDC1 Biadamantane- 1.606(m), 4H; 1.956(v. broad s), IH in CDC1 3  3  White solid-  1.536(m), and 1.706(m), 4H; 1.976(v. broad s) IH in benzene Biadamantane- 1.5l6(m) and 1.686(m), 4H; 1.945(v. broad s) IH in benzene  IR in CS  solution in cm" : White solid- 2910(vs), 2740(vw), 2320(vw), 1350(w), 1338(w), 1300(w), 1100(w), 1039(vw), 970(vw), 815(vw). Biadamantane- 2905(vs), 2840(vw), 2310(vw), 1354(w), 1340(w), 1309(w), 1103(w), 1037(vw), 966(vw), 815(vw). 1  0  -51-  IR in a KBr pellet, in c m : White s o l i d - 2900(vs), 2650(vw), 1445(m), 1340(m), 1310(w), 1110(w), 1040(w), 970(w), 820(w), 805(vw), 775(vw), 705(m). -1  Mass Spectrometry: High resolution of the white solid gave a parent peak of 12.9% at m/e 270.2340, calculated for biadamantane C QH , 2  3Q  270.2347, base peak at m/e 135.116, calculated for the adamantyl ion C^QH^+ ,135.1173. The low resolution spectrograms of the white solid and biadamantane agreed well, giving strong peaks at about 270-271, 136, 135, 134, 93, 79, 67, with the base peak at 135. Analysis: Calculated for C  2 Q  H , C 88.82%, H 11.18%, ratio of 3 0  C%/H% 7.94, found C 87.37%, H 10.86%, ratio 8.05. Due to the incorporation of oxygen in the DHA used by the unavoidable exposure to air when f i l l i n g the Carius tube, the C, H values are expected to be a bit low, but with a ratio of 7.94. The Carius tube was broken open and, after washing the white coating on the inside of the tube repeatedly with CS , CHCl^, and CCl^, the 2  solid coating was scraped off and collected. The white deposite was identified as containing higher polymers of DHA, including the tetramer and pentamer, by high resolution mass spectrometry. The extreme insolubility of this solid prevented the taking of nmr and solution IR spectra, but an IR spectrum was taken in a KBr pellet. The melting point was >325°, the material only browning in a i r , sealed tube. Mass Spectrometry: In low resolution, peaks were seen at about m/e 404 and 540, exact counting being impossible due to the lack of substantial fragmentation seen in the spectra of adamantanes in general. Base peak was m/e 135. High resolution values were obtained for peaks at m/e 538.4538 and 672.5561. Calculated for C ^ H ^ , the DHA tetramer,  -52-  538.4538, and for C  5rj  H , the DHA pentamer, 672.5633. The r e l i a b i l i t y of 72  the mass value of the pentamer was, according to the technician who determined the values, "questionable", due to the very small size of the peak. IR in a KBr pellet, in cm- : 3400(w), 2910(vs), 2670(vw), 1440(m), 1340(s), 1305(w), 1105(w), 1040(w), 968(w), 820(w), 808(vw), 765(m), 740(w). 1  Cumene Solution Product Studies One preparation of DHA made at -50° as given previously was dissolved in 20 ml. of cumene that had been deoxygenated under reduced pressure by bubbling dry nitrogen through, and was degassed and sealed off in a Carius tube under vacuum. The Carius tube was then heated at 185° for five days in the silicone oil bath, after which the tube was removed from the bath and allowed to cool to room temperature. No precipitate was seen. The solution was then frozen with liquid nitrogen, and upon thawing, the tube was immediately opened and a white precipitate was filtered out, collected and dried in air for one week. The remaining solution was vacuum d i s t i l l e d to leave a yellow o i l , which partially solidified upon further drying at about 30° for one week in air. This solution gave adamantane plus four major components by glc, retention times(K-20M, 200°) of 9 minutes, 13.5 minutes, 25 minutes and 39.5 minutes. The f i r s t three peaks were isolated by preparative glc (K-20M, 200°) but the 39.5 minute peak was omitted being the white precipitate. A control sample of cumene in five one ml. ampoules was heated at 185° for one week, using degassed cumene. An extremely small amount of the 13.5 minute peak was seen by glc.  -53-  The white precipitate, glc retention time 39.5 minutes, m.p. 269275° (corr) sealed tube, was identified as l,l'-biadamantane,  XXI.  Identification was made by comparison to a genuine sample , glc retention 31  time 40 minutes. 60 MHz nmr: White precipitate- 1.60s(m), 4H; 1.956(v. broad s), IH in CDC1 Biadamantane- 1.606(m), 4H; 1.956(v. broad s), IH in CDC1 3  3  Analysis: Calculated for C H 2 Q  3 0  C 88.82%, H 11.18%, ratio  C%/H% 7.94, found C 88.33%, H 11.15%, ratio 7.92. The 13.5 minute peak was identified by IR, nmr, melting point, mass spectrometry and analysis as 2,3-dimethyl-2,3-diphenylbutane, XXII, called bicumyl, m.p. 114.5-116.5° (corr) sealed tube, l i t 1 1 5 ° , 3 3  lit *118°. 3t  IR in CS solution, in cm" : 2295(s), 1950(vw), 1890(vw), 1808(vw), 1372(m); 1150(m), 1095(w), 1060(w), 1033(s), 782(s), 703(vs). IR in CHCU solution, in cm" : 2980(s), 1970(vw), 1890(vw), 1810(vw), 1740(vw), 1610(m), 1500(m), 1480(m), 1440(m), 1394(w), 1375(s), 1148(m), 1090(w), 1067(w), 1031(m). IR in KBr pellet, in cm" : 3000(s), 1960(vw), 1900(vw), 1830(vw), 1600(w), 1500(w), 1475(m), 1440(m), 1375(m), 1215(m), 1145(m), 1090(m), 1067(w), 1030(m), 1000(vw), 920(m), 776(vs), 733(m), 704(vs). 1  ?  1  1  60 MHz nmr: 1.306(s), 6H; 7.136(s), 5H in CDCK 1.276(s), 6H; 7.006(s), 5H in CS J  0  Mass Spectrometry: Parent peak 238, expected 238, base peak 119, major peaks at 91, 79, 77. Analysis: Calculated for  C  H 1 8  22'  C  9 0  -  7 0 %  »  H  9  -  3 0 %  »  f  O  u n d  C 90.80%, H 9.40%. The 25 minute peak was identified as 2-(l-adamantyl)-2-phenylpropane, XXIII, called cumyladamantane. The white hairlike crystals, m.p. 70.572.5° (corr) sealed tube, were identified by IR, nmr, mass spectrometry,  -54-  and elemental analysis. IR in CS solution in c m : 2860(vs), 2640(w), 1930(w), 1850(vw), 1790(w), 1730(vw), 1370(w), 1350(m), 1310(w), 1298(m), 1260(w), 1220(m), 1175(m), 1090(m), 1066(m), 1052(m), 1027(s), 970(w), 914(w), 816(w), 779(m), 776(s), 749(m), 699(vs), 673(vs). IR in KBr pellet, in cm" : 3450(m), 2920(vs), 1610(w), 1500(vw), 1480(w), 1440(m), 1390(m), 1360(m), 1345(m), 1320(m), 1240(m), 1190(w), 1110(w), 1080(m), 1065(m), 1040(s), 975(w), 925(m), 825(m), 788(m), 772(s), 745(m), 701(vs). -1  9  1  v  60 MHz nmr: 1.306(s), 6H; 1.56fi(m), 12H; 1.926(v. broad s), 3H; 7.286(s), 5H in CDC1 3  1.246(s), 6H; 1.51<s(m), 12H; 1.846(v. broad s), 3H; 7.1ls(s), 5H in CS . 2  1.256(s), 6H; 1.526(m), 12H; 1.886(v. broad s), 3H; in benzene, aromatic protons obscured. Mass Spectrometry: Parent peak 254, expected 254, base peak 135, major peaks at 119, 93, 91, 79, 67, and 55. Analysis: Calculated for C H i g  2 6 >  C 89.70%, H 10.30%, found  C 89.68%, H 10.15%. The only other products seen in any quantity were adamantane, identified by its gel retention time, and the glc peak in the cumene product studies which was not identified due to the very small amount collected, retention time 9 minutes. A 100 MHz nmr spectrum was obtained giving peaks at about 1.36, 4 or 5H; 1.54s(s), 4 or 5H; 1.736, 6 or 8H; 2.146, 3H; 2.386, 4 or 5H, in CDC1  3  No further information was obtained due to subsequent contamination of the sample.  Preparation of Solvents The dry ether used was the anhydrous one pound size (Mal1inckrodt  0848) and was, upon opening, immediately stoppered with a wired down rubber septum, and was deoxygenated by blowing dry nitrogen through for one minute using syringe needles. The interior of the can was afterwards kept at a positive pressure with added dry nitrogen at all times. The hexamethylphosphoramide used (Fisher H-343) was stored over molecular sieves type 5A. A sample of HMPA dried and d i s t i l l e d over sodium was found to cause a yellowing of the DHA produced for undetermined reasons. The HMPA used was thus only dried over molecular sieves. All DHA made with the d i s t i l l e d HMPA was discarded. The n-octane used was 99% minimum grade(PhiHips).  One l i t r e of  n-octane was washed twice with AR grade H^SO^O), twice with d i s t i l l e d water and dried overnight over anhydrous Na^SO^. The n-octane was then refluxed over sodium pellets for 20 hours and then d i s t i l l e d from the sodium with a constant boiling point of 122° (uncorr). The n-octane was then stored in the cleaned original bottle until use. Before use the bottle contents were deoxygenated by bubbling dry nitrogen through for about five minutes under about 0.5 atmosphere pressure. The cumene used (MCB CX2105) was freshly d i s t i l l e d over sodium under nitrogen less than one half hour before use and subsequently kept under nitrogen. Extreme caution was used to prevent the cumene from being exposed to oxygen after distillation due to the autoxidation of cumene producing cumene hydroperoxide , an unfortunately good source of 37  free radicals at elevated temperatures . Sufficient cumene was washed 38  twice with 20 ml. AR grade HpS0*(l), three times with d i s t i l l e d water,  -56-  and dried over anhydrous ^ S O ^ . The filtered cumene was placed in a combination reflux-distillation apparatus with sodium pellets, and the apparatus was evacuated and flushed with nitrogen twice. The cumene was refluxed over sodium under nitrogen for a minimum of overnight. The cumene was d i s t i l l e d as needed and kept under nitrogen until used.  -57-  REFERENCES  1.  D. Ginsburg, Accts. Chem. Res., 5_, 249 (1972)  2.  M.D. Newton and J.M. Schulman, J . Amer. Chem. S o c , 94, 773 (1972)  3.  P.E. Eaton and K. Nyi, J . Amer. Chem. S o c , 93_, 2785 (1971)  4.  P. Warner and R. LaRose, Tetrahedron Lett., 2141 (1972)  5.  P. Warner, R. LaRose and T. Schleis, Tetrahedron Lett., 1409 (1974)  6.  K.B. Wiberg, et a l . , J . Amer. Chem. S o c , 94, 7402 (1972)  7.  K.B. Wiberg and G.J. Burgmaier, Tetrahedron Lett., 317 (1969)  8.  P.G. Gassman, A. Topp and J.W. Keller, Tetrahedron Lett., 1093 (1969)  9.  K.B. Wiberg and G.J. Burgmaier, J . Amer. Chem. S o c , 94, 7396 (1972)  10.  K.B. Wiberg, E.C. Lupton and G.J. Burgmaier, J . Amer. Chem. S o c , 91_, 3372 (1969)  11.  K.B. Wiberg', J.E. Hiatt and G.J. Burgmaier, Tetrahedron Lett., 5855 (1968)  12.  D.H. Aue and R.N. Reynolds, J . Org. Chem., 39, 2315 (1974)  13.  W.D. Stohrer and R. Hoffmann, J . Amer. Chem. S o c , 94, 779 (1972)  14.  P.E. Eaton and G.H. Temme, III,  15.  K.B. Wiberg, G.A. Epling and M. Jason, J . Amer. Chem. S o c , 96_, 912 (1974)  16.  J . J . Dannenberg, T.M. Prociv and C. Hutt, J . Amer. Chem. S o c , 96_,  J . Amer. Chem. S o c , 95_, 7508 (1973)  913 (1974) 17.  M.D. Newton and J.M. Schulman, J . Amer. Chem. S o c , 94, 4391 (1972)  18. 19.  J . J . Dannenberg and T.M. Prociv, Chem. Commun., 1973, 291 R.E. Pincock and E.J. Torupka, J . Amer. Chem. S o c , 91_, 4593 (1969)  20.  R.E. Pincock, E.J. Torupka and W.B. Scott, Canadian Patent 903199  21.  R.E. Pincock, J . Schmidt, W.B. Scott and E.J. Torupka, Can. J . Chem., 50, 3958 (1972)  22.  W.B. Scott and R.E. Pincock, J . Amer. Chem. S o c , 95, 2040 (1973)  -58-  23.  C.S. Gibbons and J . Trotter, Can. J . Chem., 51, 87 (1973)  24.  J.O. Schmidt, Masters Thesis, University of British Columbia, 1972  25.  R.R. Perkins and R.E. Pincock, Tetrahedron Lett., 943 (1975)  26.  V.N. Leibzon, et a l . , Chem. Abstr., 81_, 488 (1974), abstract 119564z  27.  Prepared in this lab by Bruce Scott, 1970-1971.  28.  R.H. Bauer and G.M. Coppinger, Tetrahedron, 19, 1201 (1963)  29.  N.P. Neureiter, J . Org. Chem., 28, 3486 (1963)  30.  H.F. Reinhardt, J . Org. Chem., 27, 3258 (1962)  31.  Prepared by a summer student and purified by R. Perkins, m.p. 285290 degrees.  32.  T . J . Broxton, et a l . , Applied Spectroscopy, 25, 600 (1971)  33.  K.M. Johnston and G.H. Williams, J . Chem. S o c , 1960, 1168  34.  D.H. Hey, B.W. Pengilly and G.H. Williams, J . Chem. S o c , 1956, 1463-  35.  P. von R. Schleyer, et a l . , J . Amer. Chem. S o c , 86, 4195 (1964)  36.  G.A. Russel, J . Chem. Educ., 36, 111 (1959)  37.  G.A. Russel and R.C. Williamson, J r . , J . Amer. Chem. S o c , 86_, 2357 (1964)  38.  J.W.L. Fordham and H.L. Williams, Can. J . Res., 27B, 943 (1949)  39.  H.C. Bailey and G.W. Godin, Trans. Faraday S o c , 52, 68 (1956)  40.  W.R. Foster and G.H. Williams, J . Chem. S o c , 1962, 2862  41.  "Free Radicals", page 161.  W.A. Pryor, McGraw-Hill, Inc., New York, 1966,  -59-  APPENDIX  The half order plots seen in this Appendix are plots of the data to  2[DHA]o - 2[DHA]^  log[DHA] of  0  2[DHA]  [DHA]  -1  versus time. The f i r s t order plots are plots of  - log[DHA] 2  - 2[DHA]  versus time. The three halves order plots are plots 2 0  versus time. The second order plots are plots of  - [DHAJo versus time. 1  FIRST. THREE HALVES AND SECOND ORDER PLOTS FOR RUN 2  FIRST. THREE HALVES AND SECOND ORDER PLOTS FOR RUN 3  FIRST, THREE HALVES AND SECOND ORDER PLOTS FOR RUN U  OJ  FIRST. THREE HALVES AND SECOND ORDER PLOTS FOR RUN 8  480  HALF. THREE HALVES AND FIRST ORDER PLOTS FOR RUN 10  HALF. THREE HALVES AND FIRST ORDER PLOTS FOR RUN 11  


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items