00:00:00 It's a great privilege for me to have the opportunity to participate in this tribute

00:00:07 to Bob Woodward, whose work I believe was an inspiration to every organic chemist and

00:00:14 truly to every scientist.

00:00:19 Today I'm going to tell you some of the very recent adventures that students in our research

00:00:28 group have been having primarily in the area of organic synthesis of molecules of a very

00:00:37 peculiar and interesting kind.

00:00:41 And in order to concentrate on this more recent material, I'm going to have to ask you to

00:00:48 accept a certain number of statements that I make with less than the customary scholarly

00:00:57 documentation.

00:00:59 In other words, I'm going to make some statements that you may want to argue with.

00:01:03 We think that these statements are correct, but I won't have time to justify every one

00:01:10 of them.

00:01:11 And I'll try to point those out as we go along.

00:01:14 There will be a certain amount of hard fact in today's lecture, however, and about these

00:01:20 I make no excuses.

00:01:30 This slide shows a group of molecules that we've been working on for a number of years

00:01:37 now and about which we think we have reached some understanding, some level of understanding.

00:01:44 There are four of them.

00:01:46 There are two bi-radicals here.

00:01:49 One of them is a triplet substance with paired spins of the electrons.

00:01:54 Another is a singlet molecule.

00:01:58 And we believe that both of these species are true local minima.

00:02:04 That is to say, these are not transition states, but they do have finite lifetimes.

00:02:09 And I'll have more to say about that in a moment.

00:02:12 The two molecules to which these bi-radicals are connected are the ones shown at the upper

00:02:19 right of the slide.

00:02:20 This is a bicyclo-310-hex-1-ene with a bridgehead double bond here.

00:02:26 And at the lower left, an alkylidene bicyclo-210-pentane.

00:02:33 Now these two hydrocarbon systems are extremely strained.

00:02:38 We don't know the exact strain energy, but we guess that in the molecule at the upper

00:02:43 right, the bicyclohexene, the strain energy may be 75 kilocalories per mole.

00:02:51 There's another very peculiar thing about this group of molecules, and that is of the

00:02:56 four of them, the most stable species is the triplet bi-radical.

00:03:03 And in these hydrocarbons, the bicyclic hydrocarbons, we have reached, we believe, the border of

00:03:10 covalency because these compounds are less stable with this bond intact, or that bond

00:03:18 intact, than they are with it broken.

00:03:21 And that means that the global minimum here is a non-Kekulé substance.

00:03:26 So this is a, if you will, a violation of Kekulé's rules of valence.

00:03:33 Now the facts and line of argument that lead us to make these statements, which you may

00:03:42 think to be outrageous, are given, at least in preliminary form in the literature, and

00:03:49 I won't have time to go into them today, but I do want to show you the energetic relationships

00:03:56 that we believe to prevail among these molecules.

00:04:00 The numbers here have been derived by successive approximations over a period of years.

00:04:08 They are subject to change, as is any datum, but they are the best that we have available

00:04:13 now.

00:04:15 In some cases, we don't have exact numbers, but merely inequalities.

00:04:20 The zero of energy I'm just arbitrarily putting here at this alkylidene bicyclopentane, the

00:04:28 310-hexonene system is less stable than that by an amount that we don't know, so we just

00:04:33 put it higher, it's higher than zero.

00:04:37 The singlet biradical is approximately 13 kilocalories per mole higher in energy than

00:04:42 the bicyclopentane.

00:04:46 And the triplet biradical is lower than this one by some amount that, whose exact value

00:04:53 we don't know, but we suspect may be at least one kilocalorie.

00:04:57 So these three substances are very close to each other in energy, probably.

00:05:02 This one is considerably higher in energy.

00:05:04 Incidentally, this energy separation between the singlet and triplet biradicals is considerably

00:05:11 higher than we at first thought it was, and it is now reasonably close to the values that

00:05:20 are being calculated by the most sophisticated quantum chemical means.

00:05:27 But the value of this energy is an experimental number, I want to emphasize.

00:05:32 We don't do calculations of that degree of sophistication, but there is some gratification

00:05:38 in the correspondence of these two approaches.

00:05:43 Now separating the bicyclo-210-pentane from the singlet biradical is a transition state

00:05:53 whose energy is approximately two kilocalories higher than that of the singlet biradical

00:05:59 itself.

00:06:02 At a higher level, how much higher we don't know, but higher than this transition state

00:06:08 is a transition state leading from the singlet biradical to the other bicyclic compound.

00:06:15 This 2.3 kilocalories per mole energy gap is something that we can measure.

00:06:20 It is, in fact, one way we can measure it roughly is by the difference in two activation

00:06:30 energies measured kinetically.

00:06:32 One of them, the energy for opening this bond in a reaction leading to the singlet biradical,

00:06:37 and another, the activation energy for going from this compound to the singlet biradical

00:06:43 to the triplet molecule.

00:06:45 The difference in those two activation energies can be shown to be, to a pretty good approximation,

00:06:50 the value of this barrier.

00:06:53 The barrier can be detected, although not directly measured in magnitude, by trapping

00:06:58 reactions that we will show you in a moment, but I want to emphasize that although this

00:07:05 barrier is extremely small, it is nevertheless finite.

00:07:10 And it may be, in fact, although one has to be very careful about priority claims,

00:07:17 we think there are certainly very few examples in which a finite barrier to the closure of

00:07:25 a singlet hydrocarbon biradical has been demonstrated.

00:07:29 We don't take the position that this barrier is typical of such barriers.

00:07:34 In fact, we think it is rather large for such a species.

00:07:39 And in particular, the barrier in the simple trimethylene biradical itself, we believe,

00:07:45 must therefore be lower than 2.3, probably quite a bit lower, very close to zero.

00:07:52 Now that barrier has a practical consequence in that it is the thing that has allowed us

00:07:58 to carry out the chemical transformations that identify the presence of this species.

00:08:05 Were that barrier much lower, it would be very difficult, if not impossible, to intercept

00:08:10 this thing.

00:08:12 The value of this barrier predicts a lifetime for this molecule.

00:08:21 The disappearance, the unimolecular disappearance of this molecule in the absence of a trapping

00:08:25 agent consists of cyclization to this one and intersystem crossing to the triplet.

00:08:30 The sum of those two rate constants, the reciprocal of that is the average lifetime of this species.

00:08:35 And from the numbers that we have available for those transformations, we can predict

00:08:40 roughly a lifetime of 100 picoseconds, or thereabouts, for this species.

00:08:47 Now that is, with modern fast kinetic techniques, a spectroscopically detectable species in

00:08:54 principle.

00:08:56 And I wish that I could report to you the outcome of an experiment that is currently

00:09:01 underway in the laboratory of Dr. Peter Rencepis at Bell Laboratories, which constitutes an

00:09:09 attempt to detect this species and measure its absorption spectrum, which incidentally

00:09:16 has been predicted in advance in the literature by Goddard and Davis.

00:09:21 Unfortunately, that experiment is not quite finished, although I cannot forbear to say

00:09:27 that the results are extremely encouraging at this stage.

00:09:40 Now the chemistry of these species can be entered in the triplet case by following a

00:09:49 device that was first worked out in the case of the trimethylene methane bioradical itself

00:09:56 by Paul Dowd in the late 1960s.

00:09:59 We've merely adapted his procedure to this bicyclic diazine.

00:10:06 And when that material is irradiated in a glassy matrix, one can generate a strong signal

00:10:12 that we have ascribed to the presence of this triplet molecule.

00:10:16 The EPR spectrum intensity follows the Curie law, which is shown here, and all that tells

00:10:22 you is that there is no singlet state that is populatable thermally through the temperature

00:10:29 range that is available for investigation.

00:10:35 When this species is either photolyzed or heated in fluid medium, it gives rise to a

00:10:41 series of dimers derived from this triplet molecule.

00:10:45 And those same dimers can be obtained when this low-temperature photolytic preparation

00:10:53 is allowed to melt.

00:10:55 There's no question that this photolysis produces massive amounts of triplets.

00:11:00 In fact, this reaction is essentially quantitative at the low temperature.

00:11:09 Now, some of you will recall an experiment in the 1950s by Criguet using a diazine similar

00:11:18 to this but without the exocyclic alkylidene group.

00:11:21 And in that photolysis, the nitrogen was eliminated and a new bond was formed here.

00:11:27 That is to say, bicyclo-210-pentane itself was the product of that reaction.

00:11:34 And one might ask whether that same trick might not be used for the synthesis of 5-alkylidene

00:11:42 bicyclo-210-pentanes.

00:11:45 When we made these observations on the triplet, it was an unfortunate consequence that this

00:11:55 very exciting EPR observation muddled our thinking about the possibility of closing

00:12:06 this bond.

00:12:08 Because the low-temperature photolysis ought to be the best way, one would think, of making

00:12:15 a very unstable molecule.

00:12:17 And we expected that bicyclic hydrocarbon or the other one from ring closure here to

00:12:21 be very unstable molecules.

00:12:24 And so the logic of this idea was that if one was making the triplet quantitatively

00:12:32 at 77K, there was no hope of closing this ring at a higher temperature.

00:12:38 But as is usually the case, the excuses that one makes up for not trying an experiment

00:12:44 don't hold water.

00:12:46 That was a favorite expression of Professor Woodward, many times repeated to those of

00:12:52 us working in his laboratory.

00:12:54 And so it proved in this case.

00:12:57 Because the photolysis of this diazine in fluid medium at a much higher temperature

00:13:04 than 77K, in fact, does give rise quantitatively to this 5-isopropylidene bicyclo-210-pentane.

00:13:15 At minus 78 degrees, this is a stable molecule, despite the fact that at minus 196 degrees,

00:13:23 the triplet is formed.

00:13:25 Now the reason for that is this very simple fact that I told you earlier, namely the thermal

00:13:30 barrier to the closure of this biradical.

00:13:33 The thermal barrier for intersystem crossing is probably close to zero.

00:13:37 The barrier for ring closure is finite, about 2 kilocalories per mole.

00:13:41 So the product composition is going to be temperature dependent.

00:13:46 And in fact, it is.

00:13:47 And it changes from almost completely this at minus 78 to almost completely this at minus

00:13:53 196.

00:13:54 That's a tricky thing that we did not discover until much later than we should be proud to

00:14:02 confess.

00:14:05 But it does provide us with a way, then, of generating these compounds.

00:14:11 Now as I mentioned, this compound is thermally unstable at higher temperatures above about

00:14:16 minus 40 degrees.

00:14:18 It breaks this bond, goes back to this biradical, undergoes intersystem crossing to the triplet,

00:14:23 and the triplet then dimerizes.

00:14:25 And one can study the kinetics of that process and of the dimerization of the related methoxy

00:14:34 compound shown here at the top of this slide.

00:14:39 The products of these reactions quantitatively are the dimers of the trimethylene methanes.

00:14:44 The activation energies here are 13.2 for the dimethyl compound and 16.9 for this methoxy

00:14:50 compound.

00:14:51 We don't know the cause of that small difference, but we believe it's real.

00:14:57 The interesting thing are the pre-exponential terms, which are approximately four or five

00:15:06 orders of magnitude smaller than they should be for a thermal unimolecular reaction.

00:15:11 These reactions are clean first-order processes.

00:15:13 The rate-determining step, then, involves the formation of some reactive intermediate

00:15:18 in the unimolecular step.

00:15:20 And if that were a spin-allowed reaction, we would expect a pre-exponential term here

00:15:25 of about 10 to the 14th.

00:15:29 We interpret these low pre-exponential terms as being associated with a necessity to form

00:15:37 the triplet bi-radical, which is the actual precursor of the dimer of the trimethylene

00:15:43 methane.

00:15:44 Now, this behavior is strikingly different to the kind of thing that one sees with one

00:15:52 additional methylene group in this larger ring.

00:15:55 This is work of Gajewski and of Roth.

00:15:58 There are many other examples in the literature of methylene cyclopropane rearrangements

00:16:02 and stereo mutations.

00:16:03 I've just picked these out to indicate that you can roast these molecules at high temperatures

00:16:08 for a long time.

00:16:09 You never get any trimethylene methane dimers from them.

00:16:12 So the inter-system crossing in this system is very slow compared to all the other processes

00:16:19 that can go on.

00:16:20 And the pre-exponential terms here are about what one would expect for a thermal unimolecular

00:16:26 spin conservative reaction.

00:16:33 Now the chemistry of these alkylidene bicyclopentanes can be outlined here.

00:16:44 First let me point out that one can make cycloadducts of the structure shown at the bottom of the

00:16:51 slide here by carrying out the thermal decomposition of these diazines in the presence of an olefinic

00:16:56 trapping agent.

00:16:57 Here is the case with the crylonitrile.

00:17:00 Above 50 degrees, nitrogen is eliminated and the intermediate bi-radical is trapped.

00:17:06 Or this reaction can be carried out by photolysis of the azo compound in fluid medium at minus

00:17:11 78 in the presence of the trapping agent.

00:17:13 In both instances one obtains these cycloadducts to position isomers of the cyano group.

00:17:25 The intermediate in the formation of these cycloadducts is not the bicyclopentane.

00:17:31 That's fairly simple to demonstrate because it does not react with the crylonitrile at

00:17:36 minus 78 degrees.

00:17:39 However it does react with the crylonitrile at minus 55 degrees and it gives exactly the

00:17:44 same products.

00:17:46 So at higher temperature this species is capable of going to something else that can give that.

00:17:53 Now what is that other thing?

00:17:59 We'll come back to that question in a moment but let me point out first that this is not

00:18:06 the first instance by any means of the cycloaddition reaction of a strained sigma bond with a pi

00:18:14 bond.

00:18:15 I believe the first work along that line was actually carried out by Paul Gassman and his

00:18:21 co-workers in the 1960s.

00:18:26 This is maybe a little too big to get on the slide but there is a date down here, 1968.

00:18:32 Thank you.

00:18:34 Now at the top of the slide I show the reaction that Gassman and Mansfield studied.

00:18:41 That is a reaction between propiolic ester, the head of which has been removed by the

00:18:49 edge of the screen, and bicyclo-210-pentane itself.

00:18:54 The kinds of products that they observed are shown here.

00:18:57 Two of them are formal Ene products and another one is a cycloadduct.

00:19:03 Now note that the kinetics of this cycloaddition reaction are, as reported by Gassman and Mansfield,

00:19:11 second order.

00:19:14 First order each in the hydrocarbon and the olefin.

00:19:19 And there is no sharply responsive solvent effect.

00:19:29 That is to say in polar solvents the rate is about the same as it is in nonpolar solvents.

00:19:40 The mechanism that they proposed, and a similar mechanism was proposed by Cairncross and Blanchard

00:19:46 who studied the cycloaddition reactions of the next lower homologue, namely bicyclo-110-butane,

00:19:54 involved a bimolecular interaction between the strained hydrocarbon and the olefin, giving

00:20:02 a biradical intermediate, which then could lose stereochemistry, which it did when the

00:20:08 olefin was stereochemically labeled, and it could then cyclize or transfer hydrogen to

00:20:13 give the products observed.

00:20:15 Now the crucial point here is that this mechanism is a bimolecular, then unimolecular process.

00:20:23 Not the alternative that one might have anticipated, namely a unimolecular process breaking this

00:20:29 bond and giving a biradical, which then was captured by the olefin.

00:20:34 That process was specifically ruled out or rejected by these workers.

00:20:43 Now I certainly have no reason to quarrel with the conclusions drawn here.

00:20:51 They certainly are consistent with the information given.

00:20:57 But this is not the only mechanism by which strained hydrocarbons react with alkenes,

00:21:05 as we'll see.

00:21:07 Now on this slide I outline two kinetic possibilities for the reaction of the alkylidene bicyclopentane

00:21:16 with acrylonitrile.

00:21:18 One of them is the same mechanism, would have the same overall kinetics as the mechanism

00:21:25 proposed by Gassman and by Cairncross.

00:21:30 And that is a bimolecular process, mechanism A, in which a collision between these two

00:21:37 is required in the rate determining step.

00:21:40 What may happen after that, whether biradical intermediates are involved, one doesn't know.

00:21:44 The brutal overall kinetics would be second order.

00:21:49 Now an alternative mechanism is the one involving a bond cleavage as the first step to form

00:21:57 the biradical intermediate.

00:21:59 This species, we believe, must be a singlet for a substantial number of reasons, which

00:22:04 I don't have time to go into.

00:22:08 The chemistry of this entity as reflected in the cycloaddition reactions we're talking

00:22:15 about is very different from that of the triplet biradical.

00:22:19 This singlet biradical can either return or be captured by an olefin.

00:22:25 Now that is a reaction which does not have clean second order kinetics, at least not

00:22:30 under all conditions.

00:22:32 And in fact, this mechanism B can have a kinetic pattern that depends upon

00:22:53 the individual circumstances, as shown here.

00:22:59 M A, as I mentioned a moment ago, would predict the second order kinetic pattern.

00:23:04 If one carried out that cycloaddition under pseudo first order conditions in the presence

00:23:10 of a large excess of the trapping agent, then the observed pseudo first order rate constant

00:23:14 would be directly proportional to the concentration of the trapping agent M with some second order

00:23:21 rate constant K 3.

00:23:23 And in a double reciprocal plot, that would have this form.

00:23:26 It would be linear, but the intercept would be 0.

00:23:30 Now in contrast, the two step mechanism involving the preliminary formation of this intermediate,

00:23:42 mechanism B, would have kinetics shown here.

00:23:49 In the double reciprocal plot, again, this would be linear with the reciprocal of the

00:23:57 trapping agent concentration, but there would be a finite intercept.

00:24:00 And that finite intercept is, in fact, the reciprocal of the rate constant for the ring

00:24:05 opening step.

00:24:07 Now depending upon which step is rate determining, this reaction can either take on second order

00:24:13 or first order form.

00:24:15 And if neither step is rate determining, that is to say if there is no well-defined

00:24:19 rate determining step, the kinetics can be intermediate.

00:24:23 One should expect to see first order behavior when the rate determining step is the ring

00:24:30 opening.

00:24:31 And that would be the case when the trapping agent is sufficiently reactive to capture

00:24:38 the bi-radical every time it's formed.

00:24:42 On the next slide, I show the results of a double reciprocal plot of this pseudo first

00:24:47 order rate constant against the trapping agent reciprocal, trapping agent concentration,

00:24:54 for two different trapping agents.

00:24:56 The upper one is acrylonitrile, and the lower one is maleic anhydride, which is a far more

00:25:02 reactive trapping agent.

00:25:05 These data fit very well for the exclusive operation of mechanism B, the two step mechanism.

00:25:11 Because the intercept is finite, it is essentially the same intercept for the two trapping agents,

00:25:17 and the more reactive trapping agent leads to essentially first order behavior, that

00:25:21 is to say a zero slope.

00:25:24 So the chemistry of alkylidene bicyclo-210 pentanes, at least as I've outlined it to

00:25:36 you so far, seems to be well described by a mechanism shown here, in which there is

00:25:44 a reversible unimolecular ring opening to give this bi-radical intermediate, ring closure

00:25:50 of which is opposed by a small but finite barrier, which permits the interception of

00:25:56 this species by a trapping agent.

00:25:58 That trapping agent has to react at nearly the diffusion control limit in order to be

00:26:03 compatible with other information that we have.

00:26:07 And if it is not trapped, it undergoes intersystem crossing by an electron flip to give the more

00:26:13 stable triplet bi-radical, which also can be trapped by the trapping agent to give cycloadducts.

00:26:20 But if no trapping agent is present, this just dimerizes at these two positions.

00:26:26 So that's alkylidene bicyclo-210 pentanes, at least as far as we knew them until very

00:26:31 recently.

00:26:32 And we're going to return to some special molecules in this class in just a few moments.

00:26:40 I want to discuss with you now some attempts that have been made recently in our group

00:26:47 to come to grips with the problem of this much more elusive, even more elusive molecule,

00:26:56 bicyclo-310 hexonene.

00:27:00 I don't know if I have control over the focus here.

00:27:13 No, I have no control over the focus.

00:27:18 But perhaps you...

00:27:19 Can you focus that?

00:27:23 Thank you.

00:27:24 That's good.

00:27:26 This molecule was first generated by Kerberich and Heinemann as a reactive intermediate by

00:27:34 an ingenious cyclization, which I show here in modified form.

00:27:40 They generated an alkylidene, a vinylidene really, in a molecule which contained in a

00:27:48 remote position a double bond onto which cyclization could occur.

00:27:53 And that alkylidene was generated by alpha elimination.

00:27:57 They actually used the terminal vinyl chloride.

00:28:00 We have modified this slightly to use the dibromide.

00:28:08 Now these dibromides, I point out, are prepared by a kind of Wittig procedure from the corresponding

00:28:14 ketone.

00:28:17 This is following a recipe of Posner.

00:28:20 And this works quite well in many instances and gives us access to structures of this

00:28:25 type.

00:28:26 And when these are treated at low temperature with butyllithium, one can generate these

00:28:30 species.

00:28:31 Now Kerberich, and subsequently our own work that confirmed that, showed that when you

00:28:38 generate that particular compound with two methyl groups here and another methyl group

00:28:42 on this double bond, at low temperatures, at minus 78 degrees for example, the bicyclo-310-hex-1-ene

00:28:52 system holds together because the dimer that you obtain, the principal dimer that you obtain,

00:28:58 has this pi plus pi structure.

00:29:00 That's an orbital symmetry forbidden reaction taking place at minus 78 degrees.

00:29:05 And you see why I'm really unqualified to talk about the topic that Martin Saunders

00:29:10 originally announced.

00:29:13 But it just goes to show how fantastically reactive this double bond is.

00:29:19 We've also found another isomer, incidentally there's a methyl group missing here at that

00:29:24 point, which is a product of the dimerization of this in which the sigma bond of one molecule

00:29:31 adds to the pi bond of the other so that we have one open system and one closed system.

00:29:38 But these are the products at low temperature.

00:29:39 Now Kerberich noted, and we also have found, that if this reaction is carried out at higher

00:29:45 temperature, namely 0 degrees, one gets dimers of the trimethylene methane that is formed

00:29:51 by cleavage of this bond.

00:29:54 Now it is this set of observations that has started us on a search for these compounds.

00:30:05 That is, what one would like to do is to actually have these in a bottle and find out what their

00:30:09 properties are.

00:30:17 If you generate this substance at a temperature intermediate between 0 degrees and minus 78

00:30:23 degrees, namely at a temperature at which the bicyclo-210 pentane, alkylidene bicyclo-210

00:30:33 pentane shown here, is reasonably stable, or stable for a few minutes, it is possible

00:30:41 to observe the NMR spectrum of that species and the way this experiment is carried out

00:30:46 is to carry out the Kerberich cyclization at minus 30 degrees.

00:30:51 That generates this species, this bond breaks, a new bond forms here, in other words this

00:30:56 is a methylene cyclopropane rearrangement, and this rearranged compound is sufficiently

00:31:03 stable for a few minutes at minus 30 degrees, so it can then be cooled back down to minus

00:31:08 80 degrees, where it's indefinitely stable, and one can see the NMR spectrum of this,

00:31:13 among other things, in the probe.

00:31:17 Now the other things that you get are the dimers that are formed from the corresponding

00:31:24 trimethylene methane.

00:31:26 If one adds to that reaction mixture, having kept it cold, one adds a trapping agent and

00:31:32 then allows it to warm to room temperature, one can isolate the cycloaddots of the trimethylene

00:31:37 methane.

00:31:38 So one of the kinds of chemistry, unimolecular chemistry, that bicyclo-310 hexenes are capable

00:31:46 of doing is undergoing this methylene cyclopropane rearrangement to the more stable alkylidene

00:31:53 bicyclopentanes, and this is one of the pieces of evidence that leads us to the statement

00:31:58 that the energy of this is lower than the energy of that, because this rearrangement

00:32:02 is apparently unidirectional.

00:32:07 Well now, one of the things that people do in attempting to cut down the reactivity of

00:32:15 extremely reactive molecules is to put tertiary butyl groups on, and we have had some experiences

00:32:24 along this line.

00:32:27 This tertiary butyl ketone would be a desirable precursor of this tertiary butyl substituted

00:32:35 by cyclo-310 hexonene.

00:32:38 This should be a compound with a diminished reactivity of the double bond, because that

00:32:42 bulky substituent is there.

00:32:44 Perhaps we could isolate it.

00:32:48 Well, difficulty immediately presents itself, because the synthesis of these dibromoalkene

00:32:54 precursors needed to apply the Kerberich-style ring closure doesn't work with a tertiary

00:33:03 butyl group next to the carbonyl.

00:33:06 This Wittig reaction fails completely, and this line of approach seemed to be shut off

00:33:13 to us until quite recently.

00:33:17 And what has made this all possible is the extremely interesting observation by Professor

00:33:23 Jack Gilbert and his co-workers at the University of Texas, that one can generate the equivalent

00:33:29 of a vinylidene by a kind of Wittig reaction between a ketone and dimethyl diazo-phosphonate.

00:33:41 In the presence of a strong base, these undergo what is essentially a Wittig reaction to generate

00:33:47 something that can't be isolated, at least has not yet been isolated, namely a diazoethene.

00:33:52 As far as I know, compounds of this have never been isolated, but they are obviously of extreme

00:34:00 importance in this connection, because they lose nitrogen under the reaction conditions

00:34:05 generating this type of species.

00:34:07 And if the reaction is carried out in the presence of an olefin, one can obtain reasonably

00:34:12 good yields of these alkylidene cyclopropanes, even when the starting ketone has a tertiary

00:34:23 butyl group at that position.

00:34:25 This methyl tertiary butyl ketone, for example, about 50% yield of this cycloaduct was realized.

00:34:32 So that became available in 1979, and we immediately plunged into a study of the application

00:34:41 of this kind of chemistry to three systems in which tertiary butyl substitution was gradually

00:34:50 increased in the precursor ketone.

00:34:53 Here is this phosphonate reagent, and the target molecules now are this monotertiary

00:34:58 butyl, tertiary butyl dimethyl, and the ditertiary butyl bicyclo-310-hex-1-ene.

00:35:13 When this reaction was applied to the monotertiary butyl case, shown here, the result was the

00:35:22 formation of the sigma plus pi dimer in high yield.

00:35:28 In other words, even the tertiary butyl group is insufficiently bulky to diminish the reactivity

00:35:36 of this double bond to the point where one could isolate the monomeric compound.

00:35:44 The presence of that group has changed the mode selectivity.

00:35:48 That is to say, there is none of the pi plus pi dimer that we found when a methyl group

00:35:54 was at this position.

00:35:56 But nevertheless, dimerization is virtually the only result observed.

00:36:02 So that was a disappointment.

00:36:06 But we reasoned that perhaps with two methyl groups at this site, that would add sufficient

00:36:16 additional bulk so that one might be able to see monomeric product.

00:36:22 And we treated this ketone with a tertiary butyl group here, gem dimethyl substitution

00:36:29 at the terminal double bond with a diazephosphonate reagent and butyl lithium, and we obtained

00:36:36 a preparation.

00:36:38 This reaction is carried out at low temperature.

00:36:40 This preparation, when treated with aqueous ammonium chloride at 25 degrees, gave two

00:36:46 products.

00:36:47 Now, you can see how badly we wanted this, because even when I drew this slide, I couldn't

00:36:52 help but draw it.

00:36:53 That's not the compound we obtained.

00:36:55 It's the one with the double bond down here.

00:36:57 So I apologize for that error.

00:37:01 These are the two products formed at high yield from this cyclization.

00:37:05 But the striking thing is that they are monomers, and that's the first time that monomeric

00:37:11 products have been found in the Kerberic cyclization.

00:37:13 So we were very excited and wondered what is in this reaction mixture that is surviving

00:37:19 long enough to be treated with ammonium chloride at room temperature.

00:37:24 And so the next obvious thing to do was to try to isolate this compound by more conventional

00:37:28 means, namely distillation.

00:37:30 And one can distill a compound from that which has all of the correct properties, volatility,

00:37:39 and weight to be a monomer.

00:37:43 But again, it is not the bicyclo-310-hex-1-ene.

00:37:46 Surprisingly, what it is, or perhaps not surprisingly in retrospect, is this alkylidene bicyclo-210-pentane

00:37:55 with a tertiary butyl group at this position.

00:37:57 Now, this is the first persistent alkylidene bicyclo-210-pentane that we've obtained.

00:38:04 It's stable at room temperature indefinitely.

00:38:06 It's exquisitely sensitive to acid.

00:38:10 And if it is heated to 75 degrees centigrade, it slowly and reluctantly goes away.

00:38:20 What it goes to is this open chain diene, the same compound obtained over here.

00:38:26 We don't know that this is a thermal reaction.

00:38:28 It's conceivable that may be caused by adventitious acid catalysis in the glass.

00:38:34 But in any case, this is a lower limit for the threshold temperature for the thermal

00:38:38 decomposition of this compound.

00:38:39 It's really fantastically stable for a substance of that structure.

00:38:47 Now just to give you an idea, what we believe is happening is that the Kerberic ring closure

00:39:08 does take place to give this compound, which at some temperature below 25 degrees centigrade

00:39:16 undergoes the alkylidene cyclopropane rearrangement to give the bicyclo-210 structure.

00:39:25 And again, the equilibrium constant here is apparently very much larger than 1, which

00:39:29 accounts for the fact that the product that we see has only that form.

00:39:36 Now to compare this compound, the tertiary butyl substituted derivative, with the other

00:39:44 members of the series that we've prepared, I show in the right-hand column some threshold

00:39:50 temperatures.

00:39:51 The unsubstituted one is unstable above about minus 60 centigrade.

00:39:55 The dimethyl one we've talked about, about minus 40.

00:39:58 The methoxy one is stable up to plus 5.

00:40:02 And this one is stable up to at least 75 degrees.

00:40:06 In terms of activation energy for the thermal cleavage of this bond, we can see that the

00:40:12 presence of this tertiary butyl group has strengthened this carbon-carbon bond by about

00:40:16 11 kilocalories per mole.

00:40:21 Now the chemical origin of that 11 kilocalories is not difficult to discern, because the reaction

00:40:32 that we are looking for is the formation of the trimethylene methane dimers.

00:40:38 That requires the formation of the triplet bi-radical as a precursor.

00:40:42 The triplet bi-radical, in order to be stable, must be planar.

00:40:46 And if you cleave that bond when R is tertiary butyl, then you have a terrible allylic strain

00:40:51 between the tertiary butyl group and the methyl group.

00:40:54 And that is apparently worth at least 10 or 11 kilocalories per mole.

00:40:59 And that's what protects that compound.

00:41:05 Now to return for a moment to the bicyclo-210 pentane chemistry, we have studied very recently

00:41:19 some cycloaddition reactions of this molecule.

00:41:23 And it seems to behave differently from either the Gassman-Cairncross kind of chemistry that

00:41:34 we saw with just a bicyclo-210 pentane itself, the saturated hydrocarbon, or from the chemistry

00:41:41 that we had observed with the alkylidene bicyclopentane, where there was no tertiary butyl group here.

00:41:49 This seems to be still a third kind of cycloaddition chemistry.

00:41:53 We don't understand it fully, but I can tell you what the preliminary results are, at least

00:41:59 in a qualitative sense.

00:42:00 The reaction is definitely not a second-order reaction.

00:42:04 And with the acetylene dicarboxylate, for example, at 25 degrees, this reaction goes

00:42:10 quite smoothly.

00:42:12 The products that one obtains, two of them shown at the lower right of the slide, are

00:42:15 analogous to those that Gassman and Cairncross saw in their cycloaddition chemistry, namely

00:42:22 ene products and cycloadducts.

00:42:25 But the third product, which is formed in substantial amount, cannot be accounted for

00:42:31 on that basis, we believe.

00:42:33 And in fact, we believe the only reasonable intermediate to generate this structure is

00:42:37 the one shown here, that is, its zwitterion.

00:42:40 I've arbitrarily shown that as the common precursor of these two other products, although

00:42:46 that's not a necessary feature.

00:42:48 Those could be formed by other mechanisms.

00:42:50 It is very difficult to see how a biradical intermediate, for example, could form this

00:42:57 rearranged product, which no longer contains a tertiary butyl group.

00:43:03 And you see what's happened here is the tertiary butyl group has been broken up, a methyl group

00:43:07 has moved to that position, and now cyclization has occurred there to give this bicyclic compound.

00:43:14 So exactly how one gets from these precursors to that zwitterion is not yet known, and we

00:43:20 hope to find out eventually how that happens.

00:43:25 But that is still a different chemistry.

00:43:29 Well now, we've failed twice to make a bicyclo-310-hex-1-ene.

00:43:40 There's another chance at it with two tertiary butyl groups in these positions, and the idea

00:43:45 here is that we thought that we had outsmarted this molecule now because what it wants to

00:43:54 do is to rearrange to a bicyclo-210-pentane.

00:43:59 But we said, ha ha, we got you now because if you do that, you will put two tertiary

00:44:05 butyl groups cheek-by-jowl in an eclipsed conformation, an unavoidable clash of very

00:44:15 large groups.

00:44:17 And that kind of steric interference is worth, from data in the literature, probably about

00:44:24 10 kilocalories per mole, perhaps more.

00:44:28 And so one asks the question, is it conceivable that with this substitution pattern, the equilibrium

00:44:33 constant can be inverted, or at least modified, and the reaction driven back to this side

00:44:41 of the equilibrium?

00:44:44 Well, the answer, of course, is that these molecules are really much smarter than we

00:44:50 are.

00:44:52 And we had a moment's thrill when this reaction was carried out.

00:44:59 This ketone, two tertiary butyl groups here, terminal methylene, this Gilbert-style reagent

00:45:08 was used, and there was isolated from the reaction mixture in good yield a single monomeric

00:45:14 product.

00:45:17 But it was quickly evident that it was not the bicyclo-310-hexonene, but instead was

00:45:23 this di-tertiary butyl methyl cyclopentadiene, obtained by a hydrogen shift from the ring

00:45:30 position into that point.

00:45:34 That is a reaction that takes place at some temperature below 25 degrees centigrade.

00:45:40 We believe it's intramolecular and thermal.

00:45:44 The exact mechanism, perhaps, is speculative.

00:45:48 But neither of these two compounds is formed, and an escape hatch that we had not anticipated

00:45:55 was cleverly found by this molecule.

00:46:02 So we have been close but no cigar.

00:46:06 Now one final approach that we are investigating I just want to tell you about briefly.

00:46:16 This work began with the study of the high temperature chemistry of these bi-radicals

00:46:26 generated in flash vacuum pyrolysis with a residence time of about 20 milliseconds, 600

00:46:34 degrees centigrade.

00:46:35 This diazine loses nitrogen very rapidly and gives the products shown here.

00:46:41 Now this triene, cross-conjugated triene, of course results just from cleavage of the

00:46:46 carbon-carbon bond at the back of this bridge.

00:46:49 This is a hydrogen shift product, which is easy to imagine obtaining from the bi-radical

00:46:54 that would be generated by deacetation.

00:46:57 This is a secondary product from hydrogen shift here.

00:47:01 And there's a very interesting product, X, whose nature we did not know for a long time,

00:47:09 but which we discovered serendipitously in the study of another reaction of the bi-radical

00:47:17 carried out under conditions about as diametrically opposed to the ones shown here as one could

00:47:21 imagine.

00:47:22 This is the photolysis at the minus 196 degrees centigrade of the species generated when that

00:47:30 diazine is photolyzed, namely this triplet bi-radical.

00:47:34 When that preparation is photolyzed at 310 nanometers, the same products I showed you

00:47:39 on the previous slide, namely the open-chain triene and its transformation products and

00:47:44 that cyclic diene are formed.

00:47:47 But in addition to that, this olefinic alkyne, terminal alkyne, is formed.

00:47:57 This product is almost the exclusive product when the reaction is carried out in ether

00:48:02 glass, and it's formed at about 40% of the product mixture in hydrocarbon glasses.

00:48:09 At first glance, this is really an astonishing transformation, but when you think about it

00:48:13 for a moment, you see that this could be derived if this bi-radical were to close between these

00:48:21 two positions to give this bicyclo-310-hex-1-ene, which then can undergo a retro-carbene reaction

00:48:31 by cleavage of these two bonds to give this intermediate, which is a vinylidene.

00:48:38 Now of course we've seen this type of vinylidene cyclize in this position, but that simply

00:48:44 means this reaction may be reversible.

00:48:47 But what has never been the case before in the work that we have done is the presence

00:48:53 of a hydrogen at the end of this double bond.

00:48:56 Hydrogen being an easily movable group in contrast to alkyl, this alkylidene has an

00:49:04 escape route that is not available to the alkylidenes that we had generated previously.

00:49:10 And it does that by migration to give this terminal alkyne.

00:49:15 Now I should mention that the exact mechanism of these transformations, which of these are

00:49:23 excited state reactions and which are ground state reactions, is still under investigation.

00:49:28 I should point out that there seems to be little question that this material must be

00:49:36 formed in some sort of excited state, because the—well, I shouldn't say that there seems

00:49:45 to be no question.

00:49:47 It is probable that this is formed in an excited state.

00:49:50 The reason that I say that is that there is precedent in the literature, in the work

00:49:55 of Brinton and also of Kendi and of Gilbert, that shows that methylene cyclopropanes undergo

00:50:04 this retrocarbene cleavage photochemically.

00:50:07 There is no thermal precedent for this reaction.

00:50:10 So it may be that this transformation can occur thermally if the strain energy is high

00:50:20 enough that remains to be established.

00:50:25 Now one might think that it should be possible to get back into this energy surface starting

00:50:35 with this terminal alkyne, because in principle all one has to do is to just reverse these

00:50:40 steps to shift the hydrogen back to here, cyclize to the bicyclo-310-hex-1-ene, and

00:50:49 then open that bond, and that will take one back into this manifold and the products derived

00:50:55 from it.

00:50:56 And on the next slide I show the result of this experiment, which has recently been carried

00:51:02 out.

00:51:03 This is pyrolysis at 700 degrees, residence time of about 20 milliseconds.

00:51:09 This terminal alkyne does in fact give these same products that we saw when the biradical

00:51:16 was generated from the diazine, namely this cross-conjugated triene and this monocyclic

00:51:25 diene.

00:51:26 And we believe that is taking place by the thermal reversal of that photochemical reaction

00:51:32 that we saw earlier, which must then pass through a bicyclo-310-hexene intermediate.

00:51:41 And what this observation tells us now is that we can do the Kerberich-style ring closure

00:51:46 in the gas phase.

00:51:48 And that leads to the possibility, we hope, that it may be possible to trap this on a

00:51:54 cold window before it has a chance to reactivate, become reactivated, and go on to these products.

00:52:02 That may be a vain hope.

00:52:03 There are other possibilities that we are investigating.

00:52:05 But at least it is going to be examined.

00:52:11 On this slide I show the names of the persons who have done this work.

00:52:16 Richard Salinaro did all the work on the tertiary butyl substituted systems that I have talked

00:52:21 about.

00:52:22 Mark Mazur did the kinetics of the thermal reactions of alkylidene bicyclo-210 pentanes.

00:52:29 And he and Alan Pinhas and Matthew Platts were responsible for the photochemical study.

00:52:35 Mark Ruhl and Mike Lazaro were the ones who solved the synthetic problem of making alkylidene

00:52:42 bicyclo-210 pentanes.

00:52:43 There is one name left off here, and I apologize to Stephen Potter, who was the one who carried

00:52:49 out the original work on the high-temperature chemistry of the diazine.

00:52:54 These granting agencies have kindly lent us their support.

00:52:58 I thank all of these people and agencies, and I thank you for your kind attention.