Subtitles section Play video Print subtitles >> I want to move on and start talking about 2-D NMR spectroscopy and what we're going to do we'll be using this as a tool very, very useful for structure solving. There's a whole sort of alphabet soup of different techniques but rather than just unleashing a torrent, I mean people do research in this area just like they do research in organic chemistry and so big thing is invent a new technique to solve specialized problems, but rather than trying to sort of talk broadly about everything we're going to focus on getting a few tools in our toolbox and see how to use these techniques to address different problems. We'll start out with 2 tools in the toolbox that will be HMQC and COSY techniques and then we'll add some more tools and I'll try to put them into some sort of context. There are 2 additional lectures that aren't specifically on 2-D that will come in either possibly next time or the time after that so we'll be talking specifically about the Nuclear Overhauser effective, which applies to 1-D NMR as well and we'll be talking about dynamic NMR and dynamic effects in NMR spectroscopy, but we're going to start. Our next homework set will start to bring in 2-D and I'd like to get you familiar with the tools. All right theory I'm going to start really simple minded and I think this is actually a good way to think about things. So, in 1-D, we said the basic idea was your pulse and then you observe, that's your 90-degree pulse. The observe is your FID. Have you now seen your FID on the spectrometers? Have you seen the little wiggly, squiggly cosine wave with a die off [phonetic]. So this is your FID and, of course, what you've got here is an amplitude domain and then over here you have time. This is literally your signal dying off with time and the cosine wave that corresponds to the periodicity of the various nuclei. So the whole idea in 1-D Fourier transform is this time domain on the X axis ends up getting transformed to a frequency domain and that's your parts per million and so your spectrum still has amplitude on the vertical axis and it has frequency in the units of PPM on the horizontal dimension and the reason we call this 1 dimensional NMR spectroscopy is not because this is a 1-D graph, it's not, you'd say this is a 2-D graph. It's because you have 1 time dimension and that gets transformed to a frequency dimension. Now, in 2-D NMR, you get 2 time domains, 2 time dimensions in the FID and they get transformed into 2 frequency domains. So I'm going to give you just as I have given you my simplified version of an NMR spectrometer, an IR spectrophotometer and a mass spectrometer and so forth. I'll give you my simplified version of a 2-D pulse sequence. A 2-D pulsate sequence is going to be pulse weight pulse observe and so what you do when you do this is you get 2 time dimensions because the weight is you're waiting for some time, you're going to vary the weight and then you observe. So this first weight becomes time 1 and we'll call that t1 and the second weight becomes t2. Now these are not to be confused with the capital Ts we talked about for relaxation. Remember we talked about Capital T1 is vertical is spin relaxation where the magnetization returns to the Z axis and Capital T2 is spin lattice relaxation where the magnetization spreads out in the X, Y plane. These are lower case ts and they in turn transform when you do a 2-D ft they transform to 2 frequency domains and so you get a spectrum that might look like this where you have 1 domain here and this is called your f2 domain and then another domain here and that's called your f1 domain. Now what does this mean? As you're varying, well, you understand here, of course, in t2, you're collecting a signal and it's dying off with time. So you understand that basic transform that if the periodicity of this signal is 1 cycle per second, we get a line at 1 hertz and if the periodicity of this line is 2 cycles per second, you get a line at 2 hertz and if it's a composite of 1 cycle per second and 2 cycles per second and others you get a spectrum consisting of many lines. Now similarly as you vary this t1 let's say starting with hypothetically a millisecond in the first experiment, then the next experiment 2 milliseconds, the next experiment 3 milliseconds, the next 4. Another periodicity occurs. In other words, your FID what you observe in this time also shows variation that occurs in time. Variation, amplitude, a periodic variation. Those variations transform to the second frequency domain and so you get a spectrum now that consists of 2 frequency domains. It is, of course, plotted 2 dimensionally but it is really just as this is actually a 2-D graph this is 3-D graph if you will and typically these days the way we express it is as a topological map so you'll typically see a series of contours that's just like if you've ever seen a topographical map of the mountains each contour represents a certain height. So a very tall peak has many contours and a short peak has fewer contours. So it's 3 dimensions being represented being projected in two, but again the reason we call this 2-D NMR is not because there are 2 dimensions in the graph but rather because there are 2 time dimensions. All right that's what I want to say about sort of the basic mechanics of the experiment. There are 2 general types of 2-D NMR experiments. One of these experiments is one of these families the one that we'll be talking mostly about, is correlation experiments. Correlation means connectivity. It means literally what's connected to what. Another way of thinking of this is coupling. It can be proton-proton coupling, it can be proton-carbon coupling, that's what correlation experiments give you information on. You've already been using this type of information from coupling patterns and coupling constants. When you see a triplet here, you say, oh, that's a methyl group and then it integrates the 3 hydrogens you say, oh, that's a methyl group that's next to a CH2 group. When you see a quartet here and it integrates to 2 hydrogens, you say, oh, that's a methyl group that's next to 3 hydrogens. Maybe it's next to a methyl group and correlations give that same type of information. When you see a 17 hertz coupling in a trans alkene, you say, oh, that 17 hertz coupling must have a partner somewhere. Ah, here is its partner that also has a 17 hertz coupling. So you're already using connectivity information in helping to deduce your structures. Two-D experiments provide that information in a more general term. The other type of 2-D experiment that we'll be talking about are Overhauser effect experiments. We'll be talking more about the Nuclear Overhauser Effect in a couple of lectures. Those give rise to information on spatial proximity. [ Writing on board ] These can be very useful for information about stereochemistry and conformation. All right my philosophy on teaching 2-D NMR spectroscopy as I said before there's a whole alphabet soup of techniques out there. My philosophy is not to bombard us but to give us a small box, a small tool box of what I'll call core techniques. In other words, techniques that if we are good at we can use to solve a variety of problems and then if you're good with those techniques you'll be able to say oh here's a whole in my tools where I have a very specialized problem that isn't being solved by these tools and you can go to Phil [phonetic] or go to the NMR manual and say, oh, I'm encountering this particular problem with a COSY and A Toxi [phonetic] isn't helping me out but I remember him saying something that there was some type of technique called a relay COSY and saying I can add that to my toolbox. So, okay, the first 2 tools that we'll be talking about are COSY, which was really the first main 2-D technique. It stands for correlation spectroscopy. So this is typically proton-proton or let's just say homo-nuclear coupling and then the second technique that we're going to add to the toolbox is HMQC and this is heteronuclear correlation. Well, I should say something. So we're learning about the modern versions of the experiments. HMQC uses something that's inverse detection. That means on the f2 dimension you're detecting proton and on the f1 dimension you're detecting carbon. The older, less sophisticated version of this experiment was called het core [phonetic]. I'm going to put it in parentheses but that's not really, it's not the same thing. Het core was heteronuclear correlation spectroscopy and now that's what you'd call HMQC. Het core was an experiment where you would collect carbon data on the f2 dimension and proton data on the f1 dimension and it was a slower, less-efficient experiment. So we're going to start with these 2 techniques as our initial starting point for building our toolbox and we'll see that they're extremely powerful. We're then going to add in Toxi [phonetic]. Toxi is what stands for total correlation and I'll put that in quotes. It's like a super COSY that gives cross peaks with all other nuclei in the spin system. I'll show it to you today but you won't have the, you won't yet have the experience to see where it's useful. We'll bring in some problems later on, but I don't want to bombard you with too much and HMBC is sort of a long range het core. In fact, that's the version of the experiment that it used to be. It is basically these two experiments are conceptually more complicated because initially you're going to say what do I need them for and it gives you a ton of data but when you start to encounter specific problems of overlap in the case of the former and in the case of the latter fragments that you can't put together they'll be very helpful. So all of these are correlation techniques and then we will also throw into the mix of core techniques NOSY and ROSY. These are both Overhauser effect experiments. [ Writing on board ] They both give rise to information on proximity. NOSY is good for molecules that are small and molecules that are very large, but there's a whole right in the middle of medium-sized molecules that don't work well in it and ROSY ends up working well with medium-sized molecules. [ Pause ] All right. Let's start with COSY and HMQC and let me just show you the general gist of the 2 experiments. So let's start with COSY. Imagine for a moment that you have propanol and so if you think of your H1 NMR spectrum of propanol you'll probably think of something that looks like this. You'll see a triplet with a 1 to 2 to 1 triplet for the CH2 that's next to the oxygen. You'll see a singlet for the OH typically unless you're very free of acid or very free of water and the singlet is going to correspond to the OH that's going to be exchanging rapidly and not coupling unless you, as I said, are very acid free. You'll see something that looks kind of sort of like a sextet in a 1 to 5 to 10 to 10 to 1 ratio. I guess that's not the prettiest of sextets. Let me make my outer peaks a little smaller. Then you'll see something that looks like a triplet in a 1 to 2 to 1 ratio. As I said, you already know correlation. You know that when I see this triplet here downfield it's telling us that I have 2 hydrogens, it's telling us that I have a CH2 next to a CH2 and when I see this triplet up field I see, I know that I'm having a methyl group and I need to go off 3 hydrogens. I'm having a methyl group and it's next to a CH2. When I have this sextet here, you know that I'm having one methyl group is it's 2 hydrogens and by being a sextet I know it's coupling with equal coupling constants to find different hydrogens. So for this simple problem you're very good at reading this.