00:00:01 Good afternoon. I'd like to welcome everybody to the afternoon session on Pioneers and LC-MS.

00:00:12 Hopefully this session, this two-hour session, will be a little different than your standard technical session

00:00:19 in that the emphasis that will be presented is not really focused on technology,

00:00:30 but rather, paraphrasing John Fenn, those who fail to learn from the past are doomed to repeat it.

00:00:40 We've invited six speakers this afternoon to discuss the field of LC-MS from a historical perspective.

00:00:52 The presenters, even if I wanted to, I probably couldn't tell them what to say,

00:01:00 but I asked them to consider certain questions.

00:01:03 One of the questions I asked them to address was some of the barriers that they were faced with 15 or 20 years ago

00:01:13 in overcoming the, at least perceived, fundamental incompatibility between LC-MS and mass spectrometry at that time.

00:01:26 I also asked them to add a little more personal components to their lectures

00:01:37 and describe some of the serendipity and unexpected results that may have led to developments in LC-MS.

00:01:46 I asked them if there were any critical technologies that enabled or facilitated the growth of LC-MS.

00:01:54 Many of the people in the audience today are graduate students and weren't around 20 years ago,

00:02:01 and a lot's changed in the last 20 years.

00:02:04 It's not just the interfaces that have made LC-MS a multi-hundred million dollar industry, instrument industry,

00:02:14 and a tremendous analytical tool for many analytical chemists.

00:02:20 So, without further ado, I'd like to introduce the first speaker, Bruce Thompson,

00:02:28 and he will be discussing atmospheric pressure ionization and LC-MS together at last.

00:02:36 I'd also like to acknowledge his award for his contributions in analytical chemistry in Canada.

00:02:46 He was just awarded this past weekend the prestigious McBride Medal, and if you get a chance, you should congratulate him on that.

00:02:55 Bruce.

00:02:58 Thank you.

00:03:07 Thank you, Ross.

00:03:09 Thank you for inviting me to this session and including me amongst these other icons of mass spectrometry.

00:03:14 The inspiration for the title of this talk came partly from this cartoon that Patrick Arpino used to show at LC-MS talks back in the early 80s.

00:03:33 Patrick was another pioneer of LC-MS in the DLI method, some of you may know.

00:03:37 And he used to show this cartoon that represented to him the conundrum of getting the liquid-phase LC compounds and the gas-phase mass spectrometer together.

00:03:48 And you have the liquid-phase compounds looking longingly at the gas-phase mass spectrometry, wanting to get together, and how can we best put them together.

00:03:56 The other inspiration came from the fact that at the same time I was giving lectures or talks at ASMS and other places on LC-MS, and promoting the idea that atmospheric pressure was really the best way to do it.

00:04:08 And so I can finally say that I think LC and MS are together at last, and they fish fly, if you like.

00:04:17 Fish now fly. That may not be the best solution to it, but it's a solution that works today.

00:04:21 I really stretch a metaphor, I could say that whales fly now, that we have protein ions flying from electrospray.

00:04:29 Before I go any further, I want to recognize an even earlier pioneer of atmospheric pressure, LC-MS, and that was Evan Horning at the Baylor College of Medicine, who was really the first to show LC-MS by atmospheric pressure ionization.

00:04:41 He used a corona-discharge source and a heated block into which the LC-LU inflowed, and he showed this demonstration in a paper in 1974 of LC-MS using atmospheric pressure corona-discharge ionization.

00:04:53 It wasn't picked up on at that time, in fact it wasn't picked up on for a lot longer than that, but he was really the first to do it.

00:05:00 So I want to recognize his contribution, his earlier contribution.

00:05:04 My hook into this area came from being a graduate student with Julio Eberni at the Department of Physics at the University of Toronto in the mid-70s.

00:05:14 We were working on the ways in which charged droplets lost charge, cloud droplets lost charge, evaporated and lost charge.

00:05:23 We came up with this conceptual model of the fact that as the droplets shrunk and evaporated, that individual ions would pop out into the gas phase.

00:05:29 That was a mechanism for their losing charge.

00:05:33 The conceptual model is represented here by the ions surrounded by some water molecules.

00:05:37 We were working with water always at the time.

00:05:40 It would come out into the gas phase surrounded with a few clusters of water and then they could eventually be evaporated away and you'd be left with the individual ion.

00:05:46 We used ion mobility and mass spectrometry to study these ions.

00:05:51 From that we developed the model of ion evaporation, which is probably the best accepted model now for the process that goes on, at least reluctantly accepted in many cases.

00:06:05 From a conceptual point of view, which was really what we intended to do with this model, it still explains best the process that goes on in the spray electrified methods of LC-MS and spray evaporation.

00:06:18 Where you create charged droplets and the ions in the droplet, which are initially solvated and so sit in the bottom of a large potential well,

00:06:26 then their potential is raised as the droplet evaporates and the electric field due to the net charge in the droplet eventually kicks them out into the gas phase.

00:06:34 We actually put a lot of numbers and detail around this model, but probably wasn't meant to go beyond the general concept of the fact that individual ions would come out probably clustered with some water molecules.

00:06:45 Sometimes it's over-analyzed today in terms of the details of that model, but I think from a conceptual point of view it still holds water, if you like.

00:06:52 There's obviously lots of things still to be understood about it, particularly in the case of proteins, where this simple model may not be the one that works.

00:07:00 This was the instrument that we put together at the University of Toronto. Just focus on the top part of it now.

00:07:11 It shows a mass spectrometer, a quadrupole mass spectrometer, some ion optics and a sprayer out in the gas phase, where we were just spraying water with some various inorganic ions in it and they would come off and we'd sample them into the mass spectrometer.

00:07:24 We used this orthogonal sprayer, if you like, with the spray directed through an elbow, so it was directed across the front of the orifice.

00:07:32 That's an idea that seems to be coming back into popularity today with another name.

00:07:39 This is a picture of the instrument that we had, this tank of a mass spectrometer, but a tank of a vacuum chamber with the mass spectrometer in it and the quadrupole, or the ion source, would be in this region.

00:07:50 It's not shown here, but just sitting in the air, spraying in this region in front of the front flange there.

00:07:57 This was the sprayer that we used. These were the days before color film, obviously.

00:08:07 This was a cross-flow nebulizer, the liquid coming out here and the air coming out here to spray in this direction.

00:08:15 This was a grounded sprayer, which is also often used in electrospray today.

00:08:20 The high voltage was applied by means of what we called an induction electrode, or it could be called a counter electrode here.

00:08:25 This rod was supported at a high voltage and it charged the droplets by what we called induction, basically by applying a high field at the tip of the sprayer.

00:08:33 The net result was a lot of charged water droplets.

00:08:36 They were driven through here by the airflow and then in front of the mass spectrometer through this high-tech copper elbow.

00:08:44 Now, although we were looking at inorganic ions and processes and so on,

00:08:47 Ira Barney was enough of a chemist to know there may be an analytical application for this.

00:08:52 In fact, he was a physical chemist from the old days.

00:08:55 So he ran across the street to some people in pharmacy at the University of Toronto and said,

00:08:58 what's a good application, organic ion, for us to look at?

00:09:02 And they gave us an example of these catecholamines.

00:09:05 Unfortunately, he didn't go into biochemistry and have them suggest a peptide or a protein.

00:09:10 So we had to look at these as our model compounds initially.

00:09:13 And after I graduated and I moved to SCIEX and started working there,

00:09:16 and we were working on an atmospheric pressure mass spectrometer,

00:09:19 then we cooperated, Ira Barney and I, on just demonstrating some of the applications of ion evaporation

00:09:24 with what was in those days the TAGA mass spectrometer.

00:09:27 And these are just some spectra of these three simple catecholamines done by this, what we called, ion evaporation method.

00:09:35 And then we decided to couple an LC to it, and I found this slide.

00:09:41 This was a figure that was prepared for the Asilomar meeting in 1981.

00:09:45 And Ira Barney had done some LC-MS at SCIEX, and he put these,

00:09:48 this was not a computer-controlled instrument,

00:09:50 put these graphs together and glued them on a piece of paper and pinned this note on it.

00:09:54 You can see the date on it, August 7th, 1981,

00:09:56 showing left side, three separate runs at corresponding masses for the three catecholamines here.

00:10:02 And on the right side, sensitivity, injections.

00:10:04 Now this is online LC-MS at a milliminute,

00:10:07 probably with injections of something like half a mil or a mil,

00:10:09 because we didn't really know what we were doing in terms of LC.

00:10:14 And you can see the lowest amount here, 10 to the minus 7th molar of dopamine,

00:10:17 with a little bit of a signal there.

00:10:18 So this is about, I think, about 20 nanograms injected on column.

00:10:24 So we were proving principle those days that you could do LC-MS by ion evaporation.

00:10:27 Here's just a separation of some amino acids and creatine, not outstanding sensitivity.

00:10:34 One of the things that was happening was that we were running a milliminute of mostly aqueous water,

00:10:39 sometimes with some organic,

00:10:40 spraying that without the application of any heat at all.

00:10:43 And any of you who do electrospray today

00:10:45 knows how much sensitivity you're going to lose under that situation.

00:10:48 We weren't smart enough to think of lowering the flow.

00:10:52 So one of the things that limited us, I guess,

00:10:54 was that low-flow LC wasn't something that people used commonly in those days.

00:11:01 And all I thought about was a milliminute for LC.

00:11:07 This is a really impressive slide.

00:11:11 The title is off the top.

00:11:12 This was actually on column, two nanograms of antipyrene injected on column.

00:11:16 Sometimes the sensitivity was amazingly good.

00:11:18 This was full scan LC-MS,

00:11:20 where we did have a computer system now attached, a PDP-8 attached to the system.

00:11:24 And you can see extracted ion current profile for the parent ion there

00:11:27 and the mass spectrum down below.

00:11:31 And then we, in the early days of developing the TAGA6000 triple-quad system,

00:11:35 this would be in about 1982,

00:11:37 did some SRM or MRM, 20 nanogram, 200 nanogram of arginine,

00:11:42 which was another model compound

00:11:43 because it was considered challenging, involatile, labile, difficult to do.

00:11:47 This was an MRM. There should be an arrow here.

00:11:53 And one of the interesting things was, I think, in 1981,

00:11:55 ASMS, we showed this spectrum of a multiply-charged ion from ion evaporation.

00:11:58 This was, again, the title is off the top,

00:12:00 adenosine triphosphate, the disodium salt.

00:12:03 So what we saw when we ran that was the doubly-charged anion.

00:12:06 And that actually elicited a fair bit of interest at the time

00:12:09 because it was reported, or reputedly,

00:12:11 the first negative doubly-charged anion that had been observed in the gas phase at that time.

00:12:17 And you can see the MSMS spectrum down below,

00:12:19 if you can see below the floor,

00:12:20 of how it fragmented to the two halves of the molecule very nicely.

00:12:25 Now it kind of languished because

00:12:28 another of the impediments to the development of this technique

00:12:30 was nobody else really had atmospheric pressure mass spectrometers at the time

00:12:33 to even try this or play with it.

00:12:35 So it sat there and languished.

00:12:37 Although there were a couple of visionaries,

00:12:40 one of them was Steve Unger at Squibb

00:12:42 who had a compound that he thought might work by this method.

00:12:45 This was about 1983. It's TRENAM.

00:12:47 And we ran that.

00:12:48 This was a compound that didn't work at all well by FAB,

00:12:50 didn't work at all well by thermospray,

00:12:52 but we got this beautiful spectrum of just the parent ion here from it.

00:12:56 And he was very excited about this.

00:12:58 Unfortunately, he wasn't excited enough to be able to convince his management

00:13:01 to buy a system based on it,

00:13:03 but he recognized the power of this technique in any event.

00:13:10 Now, from there, at the same time,

00:13:12 really simultaneous with all this,

00:13:14 we were working on a heated nebulizer interface

00:13:17 for our corona discharge ion source.

00:13:19 And the genesis of this was from a little bit of work

00:13:22 that we did with Jack Kenyon and Peter Dawson.

00:13:25 And Jack really stimulated this.

00:13:27 He's going to tell you the story of that early work initially.

00:13:30 But that got us going on developing the heated nebulizer interface at SIACS.

00:13:34 And this was the conceptual idea of the heated nebulizer,

00:13:37 a sprayer where the liquid would be sprayed through a heated tube to vaporize it

00:13:41 and then ionized by corona discharge in the ion source.

00:13:45 And this was the ion source that we had with the Tega at this time,

00:13:48 corona discharge ion source.

00:13:50 It turned out that it was a bit more complicated than that.

00:13:55 Droplets didn't just evaporate in the gas phase.

00:13:57 We finally had to nebulize it to a very fine mist

00:14:00 and try to ensure that the droplets actually impacted on the walls of this heated tube

00:14:04 in order to vaporize and without decomposing these fairly fragile molecules.

00:14:10 So we did a lot of work on optimizing the sprayer

00:14:13 and the gas flows around here and the mixing

00:14:15 to make sure these were fully vaporized

00:14:17 and what we called flash vaporized on the walls of the tube.

00:14:21 We ran this heater at about 450 degrees centigrade

00:14:25 in order to maintain the walls hot enough.

00:14:27 Of course, the sample was never exposed to that kind of temperature.

00:14:30 But you needed that to get where this liquid was impinging

00:14:33 the temperature enough to flash vaporize the liquid at that sort of flow rate.

00:14:36 But it scared a lot of people at the time to see a temperature controller

00:14:39 reading 450 degrees for an LC interface

00:14:41 and thinking that their sample was actually exposed to that temperature.

00:14:48 We used dexamethasone as our thermometer for this compound.

00:14:51 This is the parent ion here at mass 333

00:14:54 and this was a decomposition product at mass 333

00:14:57 and we could tell when the temperature of the probe was too low

00:15:00 because you got this more fragmentation and tailing on the fragment ion.

00:15:04 So this was a matter of turning up the heat

00:15:07 in order to get better chromatographic integrity and less decomposition.

00:15:13 This was the heated nebulizer probe as we commercialized it

00:15:16 in about 1983 or 1984

00:15:19 as an inlet on the Tega 6000 triple quad

00:15:22 which was mostly an environmental application instrument at that time.

00:15:27 But we had this inlet probe designed

00:15:29 and again, not many people took up that torch

00:15:32 except for Jack Hennion and his group at Cornell

00:15:35 who really picked up the torch on this and carried it forward

00:15:38 and developed a lot of really beautiful applications

00:15:41 for LC-MS using atmospheric pressure ionization.

00:15:44 This is a picture of the probe on the Tega instrument in their lab

00:15:48 in about 1984, the Tega 6000.

00:15:51 And they did some beautiful applications of that, of course,

00:15:54 simultaneous with working on taking electrospray and turning it into ion spray

00:15:58 and making it a very practical LC-MS instrument.

00:16:01 That was eventually where ion evaporation ended up

00:16:04 was with ion spray and electrospray as very practical implementations

00:16:07 of this spray electrification process.

00:16:13 And this is really a quintessential, I think a beautiful application

00:16:16 that was done at Cornell in those years, 1985-1986

00:16:19 of very rapid quantitation.

00:16:21 This is a pharmacokinetic study, basically,

00:16:24 doing with the heated nebulizer

00:16:27 from administered phenylbutazone to a horse

00:16:30 extracts of plasma and urine interjected with standards

00:16:33 at the rate of one sample a minute using an autosampler

00:16:36 doing LC-MS-MS and measuring

00:16:39 quantitating on the residual

00:16:42 parent drug plus the metabolites, the number of metabolites.

00:16:45 And this was done at a milliminute with a very short

00:16:48 small particle column, a 3 micron

00:16:51 by 3 centimeter long column.

00:16:54 And separations actually in less than a minute

00:16:57 for each of these. So these were 60 samples injected

00:17:00 in an hour's time, one a minute.

00:17:03 So that's 10 years old and it stands amongst

00:17:06 the best of the LC-MS-MS today

00:17:09 and the applications it's being put to in these high throughput

00:17:12 quantitation labs.

00:17:15 And you can see that even in that

00:17:18 doing this MRM study, they were separating in the course of

00:17:21 less than a minute metabolites, oxyphenylbutazone

00:17:24 and hydroxyphenylbutazone from the parent drug

00:17:27 in that time. So that was really cutting edge LC-MS.

00:17:30 This is a parent ion scan now for the same

00:17:33 sample showing separation of

00:17:36 3 or actually 4 metabolites plus the parent drug by

00:17:39 scanning, doing a parent ion scan and monitoring mass 93

00:17:42 which is common to all of these fragments. And that was published in

00:17:45 1986. So

00:17:48 the final thing I want to recognize here in terms of pioneers

00:17:51 are some of the visionaries who really eventually

00:17:54 took these methods and turned them

00:17:57 into practical instruments and started to use them

00:18:00 in their laboratories and show the rest of the world what could be done

00:18:03 with these atmospheric pressure techniques. People like

00:18:06 David Wang-Iverson of the Sanfuda and the pharmaceutical

00:18:09 labs, Jerry DiDonato. They were some of the early visionaries

00:18:12 who took the chance on this new dedicated

00:18:15 LC-MS method, an instrument that later became

00:18:18 the API-3 and started to

00:18:21 make it the popular method that it is today.

00:18:24 Thank you very much.

00:18:35 Are there any questions for Bruce?

00:18:46 I have one. The original

00:18:49 API work used steel wool.

00:18:52 Did you ever do that?

00:18:55 That's a story that Jack will tell you in the next talk. Yes, it did.

00:18:58 No, we didn't follow up on that.

00:19:04 Thanks, Bruce.

00:19:10 The next speaker

00:19:13 is Jack Hanyon

00:19:16 and his topic is

00:19:19 From DLI to Ion Spray and LC-MS

00:19:22 MS and

00:19:25 LC-MS Journey.

00:19:28 Thank you, Ross. I have mixed feelings about giving a lecture like this.

00:19:31 It's an indication of age when you're talking about history.

00:19:34 I'm a long ways from being ready to stop this stuff.

00:19:37 I'd like to start

00:19:40 with LC-MS as I have seen it.

00:19:44 The first slide. There we go.

00:19:47 If we could get that focused a bit and take you through

00:19:50 what I would see in my

00:19:53 world, the steps that were important to me

00:19:56 in leading us to where we are in our laboratory now.

00:19:59 I'm going to talk about direct liquid induction. This goes back

00:20:02 a few years before Bruce just entered into the

00:20:05 picture he talked about. I'm going to talk about atmospheric pressure

00:20:08 ionization for LC-MS. It will pick up on what he described

00:20:11 in ion spray or pneumatically assisted electrospray.

00:20:14 That's where we're going on this journey.

00:20:17 I hesitate to put my in here. These are not milestones in LC-MS.

00:20:20 These are milestones in my world of LC-MS.

00:20:23 There are certainly other milestones that we all know about.

00:20:26 I purchased a Finnegan system

00:20:29 in 1974 and

00:20:32 first started taking a look at direct liquid induction

00:20:35 in March of 75.

00:20:38 In 77 was when we moved to micro-LC-MS

00:20:41 to get better sensitivity.

00:20:44 In 81 where I saw what Bruce just described

00:20:47 at Asilomar. That was a pivotal point to me at Asilomar

00:20:50 to see those results he just showed you.

00:20:53 They were extremely exciting to me. I distinctly remember that day

00:20:56 and Neil Reed gave a talk that showed their instrument

00:20:59 which I had never even heard about SIACS of course then.

00:21:02 It had a box. All these things leading to their instrument.

00:21:05 It was an empty box and I said I think I can help you fill in that box.

00:21:08 We've been down a long road since that time.

00:21:11 And then ion spray as you just heard it described

00:21:14 a little bit. I want to tell you what led us to that.

00:21:17 Barriers to development.

00:21:20 There's a lot of negative people in all areas

00:21:23 and there certainly was an incompatibility mindset.

00:21:26 Mass spectroscopists including myself didn't want to think about

00:21:29 putting liquid into the mass spectrometer. We take that for granted now

00:21:32 although I remind you that we don't put it into the mass spectrometer.

00:21:35 We spray at it and we put ions in it. That's what we want after all in the first place.

00:21:38 There was clearly a mindset in those days that we didn't want to

00:21:41 dirty these instruments and burn the filament out.

00:21:44 That was a barrier for sure. There was little precedence

00:21:47 other than Horning's work that you just heard.

00:21:50 I too acknowledge his pioneering efforts there

00:21:53 as well as the demonstration of the feasibility of DLI

00:21:56 by Mike Baldwin and Professor McLafferty

00:21:59 in the early 70s.

00:22:02 The other comment was why not just collect the peak and put it in the probe.

00:22:05 You can still think about that. We still see that with

00:22:08 Maldi right now. I want to see online Maldi, true online Maldi

00:22:11 LCMS. I once made the mistake of saying

00:22:14 you'd never see online fab and I've learned not to

00:22:17 say that anymore so I'm not going to say we're not going to see online

00:22:20 and of course there have been some feasibility demonstrated but I don't think

00:22:23 any of us want to buy it yet.

00:22:26 That was a mindset and an immature LC equipment.

00:22:29 I made a lot of enemies. I still seem to be able to do that

00:22:32 by criticizing equipment. The Waters equipment was a joke

00:22:35 in terms of competing with a tank of helium.

00:22:38 In 1974 and 5 and 6

00:22:41 those pumps went in and out. They went in and out

00:22:44 but I mean their flow rate went in and out and everything was all over the place

00:22:47 and it couldn't deliver a steady stream of solvent and we depended

00:22:50 upon a steady stream of that effluent. So those were barriers

00:22:53 to make this work. Why bother to? Every

00:22:56 technique you should answer this question, is this worth doing?

00:22:59 Well GCMS was certainly showing practicality

00:23:02 and this was 1974 now so you've got to have a mindset

00:23:05 some of you may not have even been around at that time

00:23:08 but GCMS was in its infancy and clearly starting to be

00:23:11 sold and work and it was clear to me that HPLC

00:23:14 was going to be a technique that was going to be

00:23:17 important and we needed to know what those peaks were

00:23:20 and that's why I do LCMS. And separation science

00:23:23 and mass spectrometry are just a natural combination

00:23:26 not natural that it's easy to put them together but natural that we need them together

00:23:29 that if you're going to have a chromatographic peak and a separated mixture

00:23:32 you want to have a mass spectrum of those peaks. So that was why bother

00:23:35 to me. Why we embarked on this

00:23:38 Why won't my advancement go?

00:23:41 Parallel technologies at that time, micro HPLC

00:23:44 techniques were just beginning very very much in their infancy

00:23:47 and I soon learned that if we were going to get more sensitivity we had to

00:23:50 reduce the flow. API mass spectrometry

00:23:53 you just heard about that from Bruce that was on the horizon

00:23:56 it was there and only a few of us saw it. There was a lot of people

00:23:59 that just didn't buy into it. Tandem mass spectrometry

00:24:02 that was an exciting area. I remember taking the two or three day course

00:24:05 at Purdue with Graham and colleagues to learn

00:24:08 and become familiar with that. And then this ion evaporation interface

00:24:11 that Bruce just described. That was terribly exciting. That was pivotal

00:24:14 for me to see those results that he just showed at El Silamar in 1981

00:24:17 I distinctly remember that. Now let me take you on these

00:24:20 benchmarks in my laboratory and my experience

00:24:23 with this thing where I bought that Finnegan twin stack

00:24:26 in 1974 and in March of that year

00:24:29 started doing what we now call infusion or even

00:24:32 FIA, trying to get ion current. And it was not until

00:24:35 a year later that we actually did it online. And this was

00:24:38 off hours in the night time

00:24:41 when the service laboratory could be used for interesting things.

00:24:44 The interface, if you will, was a glass

00:24:47 tube. It was quarter inch OD

00:24:50 and I can't remember the inside diameter, 100 microns

00:24:53 or so. And when I see all this nano spray it just makes me

00:24:56 shiver to think of all the headaches I used to have with a 5 micron hole here

00:24:59 plugging this thing. And it was just dumb determination

00:25:02 that kept me going. I could not and I never did

00:25:05 promote that everybody should do this. This was headache city.

00:25:08 This was enough to drive you to beating your kids and drinking alcohol

00:25:11 and everything. You literally had

00:25:14 success by playing the odds. And I could tell you many

00:25:17 stories that I won't. But this was ground down

00:25:20 to fit inside the ion source of the Finnegan 3300

00:25:23 and I did it on a belt sander in the machine shop. I had to take a

00:25:26 course to get the key at night. To graduate from the course

00:25:29 you got the key at night and I'd go in and put it on a belt sander

00:25:32 and I'd get these broken down. I'd take a spare source to make sure it fit

00:25:35 I'd get impatient and push too hard and it would break it right off.

00:25:38 Right there. So you make lots of these things. And it was real fun

00:25:41 to demonstrate that they were open by putting high pressure helium

00:25:44 on here and shooting it across the laboratory at 300 PSI.

00:25:47 That's a real shocking experience.

00:25:50 And of course it bounced off the walls and I never told

00:25:53 the department too much about that. I did always

00:25:56 make sure I stayed away from it. I need some help with balancing. There we go.

00:25:59 There's that glass tube. Literally you could buy this

00:26:02 in a meter length and cut it off. I spent a lot of time making it.

00:26:05 There's the split. A new probe valve. This was a split

00:26:08 at 1 mL per minute. About 1% would go into the mass spectrometer.

00:26:11 There's the inlet of the 3300. And this was clearly the Achilles heel.

00:26:14 It was a much worse Achilles heel than we have now.

00:26:17 But keeping that open was a pain. I had a T there with a septum in it

00:26:20 where I could inject if I didn't

00:26:23 get squirted in the eye because this was under high pressure.

00:26:26 There wasn't a reodine injector or a Velco injector at that time.

00:26:29 And I learned soon the importance of eyeglasses

00:26:32 in the laboratory because I would inject through the septum

00:26:35 and either the plunger would go to the ceiling or the liquid squirt out of my face.

00:26:38 So there was no chromatography there. It was just getting it to work.

00:26:41 That's why I say it was a year later before we actually got online results.

00:26:44 But that was the early DLI. This probably

00:26:47 was one of the first HPLC systems on wheels, on a cart

00:26:50 that we all now do. This belonged to Jerry Meinwald

00:26:53 in the Department of Chemistry at Cornell and I knew he had it

00:26:56 and I said, if I put it on a cart and build this thing

00:26:59 and give it to you, can I use your HPLC after hours?

00:27:02 And he was on the fifth floor and I was on the second and I would literally go up

00:27:05 and wait for his students to be done with it and come down and work all night long

00:27:08 off hours to make this work and couple it with the Finnegan 3300

00:27:11 and did quite a lot of that.

00:27:14 This is just one example. I used to experiment with colors and slides

00:27:17 and this was one of the failures, but these are corticosteroids

00:27:20 and this was online DLI LC-MS.

00:27:23 We had fragmentation not from ion chemistry

00:27:26 but from heat, the hot ion source, and we did a lot of things

00:27:29 with that, but that was a hard way of doing it.

00:27:32 This was an improvement, believe it or not.

00:27:35 This is a photograph that Bob Boykster made

00:27:38 in, I don't remember when, 79 or 80.

00:27:41 This is the direct probe, the DLI probe that Hewlett-Packard

00:27:44 marketed and Extrel at the time also co-marketed this device

00:27:47 and had a diaphragm pinhole here

00:27:50 and we sprayed that in. This replaced the glass probe

00:27:53 and believe it or not, it was a big improvement, although it still had that little hole.

00:27:56 1% of the total effluent went through.

00:27:59 This was HP's drawing at the time. The diaphragm, a laser drilled

00:28:02 5 micron hole right there and effluent

00:28:05 at 1 mL per minute from the LC would come by, go past

00:28:08 that hole, squirt into the ion source, the mass spectrometer.

00:28:11 99% would go to the bottle on the floor. We'd use water,

00:28:14 cooling water, in here to cool the tip because it was in contact

00:28:17 with the ion source. It was usually 250 C or greater.

00:28:20 That was an advance for DLI. The biggest advance

00:28:23 is you didn't blast anything against the wall when it slipped off the pressure

00:28:26 device. You just replaced the diaphragm pinhole and

00:28:29 sonicated it and played the odds. I like this one

00:28:32 because it made the mass spectrometer look like a tractor or a steam engine

00:28:35 because one way to increase the pumping

00:28:38 in a simple way is to cryopump and we now know

00:28:41 that. HP had a big thing that

00:28:44 you could pump nitrogen in and the thing would erupt like this every once

00:28:47 in a while and after a while I got used to it but in early days

00:28:50 it would scare me and anybody around me but this was the Hewlett-Packard

00:28:53 5985B that did an awful lot of this work.

00:28:56 These are compounds trichloromethazide

00:28:59 and reserpine. Heckling drug testing days and I'm taking

00:29:02 the days when I moved across campus out of the chemistry department to the

00:29:05 racehorse drug testing program and I defy you even today

00:29:08 to get these through without druidization through GCMS.

00:29:11 Trichloromethazide, diuretic, reserpine. If you all wonder why

00:29:14 everybody's using reserpine, it's from horses. It's a great tranquilizer.

00:29:17 I wouldn't want to show in this mixed audience what

00:29:20 reserpine does to a horse. Ask me in private and I'll tell you.

00:29:23 But it's tough to do it at 10 pg per ml.

00:29:26 In fact, this was 1981 or 2

00:29:29 and only recently did we have a paper in Jazz Mass

00:29:32 where we showed 10 pg per ml by

00:29:35 LCMS, API LCMS of that compound

00:29:38 in horse urine. It's very low levels.

00:29:41 And this was 3 micrograms, I think, negative ion

00:29:44 of each of these compounds and that was done by the

00:29:47 HP DLI probe. Showed it could be done. Here's the mass spectrum

00:29:50 positive ion. Many of us know that very well. Negative ion

00:29:53 detection. So that was a breakthrough. This extended what you could do

00:29:56 by GCMS. Now we move to micro LCMS because

00:29:59 it's in this racehorse drug testing business and trace analysis is

00:30:02 real important in that area and we needed to get more

00:30:05 sensitivity and my colleague on the bench

00:30:08 right down there, Bill McFadden, brought to my attention this.

00:30:11 1976, Bill. I visited Finnegan and you showed me a photocopy

00:30:14 a bad photocopy, I might add. I couldn't read the address

00:30:17 of the Jasko Familic micro LC system.

00:30:20 And I purchased that thing in 1976 or

00:30:23 7 and we still have it. We don't use it for micro LC

00:30:26 but we use it for infusion, if you will. Delivering it

00:30:29 instead of the other little pump. It's a true syringe pump

00:30:32 and it actually had a syringe. Glass syringe which would break.

00:30:35 The column was there. It was Teflon tubing and this

00:30:38 operated at 5 to 20 microliters per minute and really did

00:30:41 give us an increase in sensitivity. I converted

00:30:44 or modified the DLI probe

00:30:47 from HP to a micro LC DLI probe.

00:30:50 So this is a homemade one with the same diaphragm in there.

00:30:53 So the total effluent. The key was that micro LC, the total effluent

00:30:56 and no more split and their peak concentration was higher

00:30:59 and we had better sensitivity. And we pushed back the limits

00:31:02 there. I used to jokingly say that this technique is real easy. I always

00:31:05 admitted that it wasn't. But you could show that a young child

00:31:08 could do this, which was my son at the time. Now he's

00:31:11 three times that high. But the probe is inside

00:31:14 the mass spectrometer. The micro bore column

00:31:17 this is usually 25 centimeter long.

00:31:20 One millimeter ID was part of the probe and the injector was right there.

00:31:23 And it really did improve sensitivity. Well then we used

00:31:26 this micro pump from ABI at the time or actually Bob Brownlee

00:31:29 the late Bob Brownlee generously provided that. We still had that very

00:31:32 instrument working in our laboratory.

00:31:35 And these are results that were representative of what we do. Tom Covey

00:31:38 personally did these. This is diethylstibestrol. The key is tissue

00:31:41 and urine analysis. You need good sensitivity. Here's a UV chromatogram

00:31:44 and this my friends, manufacturers and users are what

00:31:47 chromatograms from mass spectrometers are supposed to look like

00:31:50 without smoothing. Not ratty peaks and crummy stuff. We want

00:31:53 accurate areas of those things. And when you do it right, when the moon is right

00:31:56 and you've got a good person like Tom Covey, this is what chromatograms can look like.

00:31:59 This is 50 nanograms.

00:32:02 I think it was selected ion monitoring at the time. Easily detecting that sort of

00:32:05 level. So that's what microLCMS did to us

00:32:08 for us by DLI. But I knew at that time

00:32:11 that that was a hard way to make a living.

00:32:14 And it was tough. And it was only blind, dumb determination that pushed us.

00:32:17 And when I saw the capability of atmospheric pressure ionization

00:32:20 that Bruce just talked about, I knew I had to get involved

00:32:23 in that. Even though I had seen Elvin Horning's work

00:32:26 in the early 70's, I didn't

00:32:29 catch on until I could see that there was a commercial instrument that I could

00:32:32 somehow get access to. I used to show this slide

00:32:35 and people would, either it didn't sink in. You ever try to get a duck wet?

00:32:38 They don't get wet. It doesn't sink in. You've got to

00:32:41 wash them with soap before they really get wet. And people were like ducks.

00:32:44 They couldn't conceive that you could make ions at atmospheric pressure.

00:32:47 What a difference 10-15 years makes.

00:32:50 So we don't have to show that slide anymore.

00:32:53 This is a TAGA 6000. It was before it was called

00:32:56 an E. They extended it from 450 mass range

00:32:59 to the massive mass range of 1200.

00:33:02 And that was what the early TAGA

00:33:05 looked like. TAGA, I might add, was trace atmospheric

00:33:08 gas analyzer. The thing was made to put in a van and drive around a dump

00:33:11 site and sniff the air. No LCMS.

00:33:14 It was people like us that suggested maybe we could do LCMS.

00:33:17 Well there is the HP DLI probe. This was Peter Dawson's

00:33:20 instrument at NRC in Ottawa. And I saw

00:33:23 and talked with the Sykes folks in

00:33:26 Minneapolis in 1981. 3? 1983

00:33:29 I think. And made a date. This was

00:33:32 June 83 and by

00:33:35 July 1st, just before our July 4th

00:33:38 I think that's when it was. And early, very early July, less than a month later

00:33:41 I was in their laboratory. Bruce Thompson met with me

00:33:44 in Peter Dawson's lab. And I had solvents and real samples.

00:33:47 And this was a physics laboratory and they didn't know. We had an audience

00:33:50 of physics people. We heard comments earlier about physics folks

00:33:53 and how we ought to work with them. But they were in awe that we were doing

00:33:56 real chemistry. And so that probe was converted.

00:33:59 I went there with the understanding that Bruce was going to somehow

00:34:02 help me make this work. And it took urine extracts and so forth.

00:34:05 And that was our first LCMS results. This is the origin of the

00:34:08 pneumatic nebulizer. This messy

00:34:11 looking thing that got real hot and black. We worked one week intensely

00:34:14 from early in the morning to late at night. And accomplished it.

00:34:17 There's a chronodischarge needle. And that became the heated pneumatic

00:34:20 nebulizer. This is the drawing we had in the 82 paper

00:34:23 or 80 whatever year it was. Analytical chemistry.

00:34:26 And Ross asked about stainless steel or

00:34:29 mesh, wire mesh. The way we finally made this thing work. This is the

00:34:32 DLI probe with a pinhole diaphragm. I spent all week trying to

00:34:35 keep that open. And did so. This is the nebulization

00:34:38 coaxially. And this is a, I think it was a

00:34:41 platinum mesh that Peter somehow

00:34:44 had laying around. Until we put that in there, it didn't work very well.

00:34:47 We had to break up those droplets. But that probe allowed us to get

00:34:50 these results. A UV chromatogram and extract ion

00:34:53 chroma profiles. MSMS for urine extracts of

00:34:56 sulfa drugs. I see today there's a poster on sulfa drugs.

00:34:59 Sulfa drugs work very well with this technique. And I believe

00:35:02 this was the first real world application of online LC-MSMS

00:35:05 in this paper. I want to show you with interest

00:35:08 this. You can't see this very well. These mass spectra

00:35:11 are from a HP data system. The science data system

00:35:14 in 1980, whatever it was, was a dog.

00:35:17 And I took numbers and entered it into the

00:35:20 HP data system. And that's the way I plotted the spectra. I never omitted that in the paper.

00:35:23 But their data system didn't make the mass spectra plots like I

00:35:26 wanted to. I still have some concerns here and there. But

00:35:29 we'll talk about that. This is the instrument

00:35:32 that Bruce showed you. I spent almost

00:35:35 two years drumming up 525,000 U.S. dollars.

00:35:38 At that time, in 1984, that was a lot of money.

00:35:41 And EPA finally funded our work. And CyEx, to their credit,

00:35:44 let us earn off the last chunk of it. About a third

00:35:47 of it. And Tom Covey was a new student with me. And this is in that

00:35:50 blank laboratory right after we got it. And we got it just days before

00:35:53 ASMS in 1984. So we could say we really did

00:35:56 have a triple quadruple for ASMS. Of course, nobody asked or cared.

00:35:59 At that time, this was our business.

00:36:02 This was the working end of our laboratory animal.

00:36:05 This was actually not a real one.

00:36:08 I have a story about Tom Covey hijacking this horse

00:36:11 with a truck in the wintertime. But in any case, there were times

00:36:14 with this early instrument that I wanted to get a real horse, lift the tail, and let it

00:36:17 really come out. But this was the origin

00:36:20 of our, the exit, I guess, of our DLI interface.

00:36:23 We did away with that with that pressure tubing, if you can see it.

00:36:26 But we did a lot of equine drug testing. Where is that sensor up back there?

00:36:29 And this is the heat and pneumatic nebulizer that you heard about.

00:36:32 And Bruce described that. And with that, we did

00:36:35 this work that he showed. This was one minute window

00:36:38 from there to there. That's the story he described.

00:36:41 And this was paradigm scanning. And that was a real exciting thing for us.

00:36:44 I might give you one little story that the autosampler injecting

00:36:47 every minute needed a graduate student to put, whose name

00:36:50 was Ed Lee at the time, to put the sent vials in the

00:36:53 same hole. Because the autosampler spent too much time reaching over and getting

00:36:56 to the, and we still have that problem to these days. Autosamplers don't work

00:36:59 fast enough. My time is up. Oh, iron spray.

00:37:02 How could I miss that? Is my time really up? Yeah.

00:37:05 Alright. Iron spray. These are the key people.

00:37:08 This is the father of iron spray, Andres Bruins, who

00:37:11 came and wanted to do electric spray. I said, go for it. He spent three

00:37:14 weeks of headaches and said, there's got to be a better way.

00:37:17 Tom Covey, who was there at the time, and Ed Lee, and that threesome was a

00:37:20 hell of a team in that time to make this work.

00:37:23 And this is this ion evaporation interface that Bruce described.

00:37:26 And it was from that that we got the pneumatic nebulization idea for iron spray.

00:37:29 It was not a novel idea. And we could get spectra like these that we know

00:37:32 now we can regularly get. That was our original drawing

00:37:35 in 86 in Cincinnati. I'm almost done, Ross.

00:37:38 That's the system as we designed it and

00:37:41 built it. And it looked like this. And Sykes copied that and

00:37:44 made that plastic front end, as you're all familiar with. These are students.

00:37:47 Some people are surprised I have so few students, but it's quality, not quantity,

00:37:50 that helps. And these folks have done a great job.

00:37:53 Postdocs, a number in this area. And these

00:37:56 three are here with me at the present time at Cornell.

00:37:59 And these are visiting scientists that have come and stayed in my laboratory.

00:38:02 Andres Bruins at the top of the list, chronologically.

00:38:05 And key people, Timothy Wachs, who is still with me,

00:38:08 keeps things going when I travel and do other things. And these other people,

00:38:11 and they kept veterinary college. And these are the people that make

00:38:14 things possible by giving money. And almost all these people

00:38:17 are still currently providing the research.

00:38:20 I'll close with just one group shot of several people

00:38:23 of these representative people. So I thank you, Ross, for the opportunity to

00:38:26 present this. And sorry it was so fast, but thank you very much.

00:38:29 Applause

00:38:32 Applause

00:38:35 Applause

00:38:38 We're very proud today to have Bill McFadden

00:38:41 come out of retirement and honor us

00:38:44 with some of his stories

00:38:47 about the early days in LC-MS.

00:38:50 Bill's title is

00:38:53 LC-MS, How Did We Ever Do It?

00:39:02 What am I doing here?

00:39:06 When Ross Willoughby called me

00:39:09 and asked if I'd like to give a review paper,

00:39:12 I pointed out to him that I have been

00:39:15 completely inactive in mass spectrometry

00:39:18 for eight years. Well, you know,

00:39:21 he sort of rubbed my ego a little bit and said,

00:39:24 oh, that's all right, Bill. We don't need to hear anything new.

00:39:27 All we want to do is hear about the wonderful things you used to do.

00:39:30 And so I thought, that's good.

00:39:33 But what I didn't realize at the time

00:39:36 was when I started to bring out my material,

00:39:39 I'd forgotten everything I used to do.

00:39:42 So in any event,

00:39:45 I feel a very appropriate title is

00:39:48 How Did We Ever Do This?

00:39:51 And on the first slide, am I supposed to do something here?

00:39:54 Do I get these?

00:39:57 Nobody instructed me and I

00:40:00 I hate to waste a whole minute and a half

00:40:03 just getting a slide. It says

00:40:06 forward, hit forward.

00:40:09 This first slide

00:40:12 really says it all.

00:40:15 I have here the, do you hear me

00:40:18 when I talk this way, away from the mic?

00:40:21 Good. I have here a very old slide

00:40:24 which shows a comparison

00:40:27 between the conditions for gas chromatography

00:40:30 and liquid chromatography.

00:40:33 And if we look down here, carrier, chemical nature,

00:40:36 gas additives, stability of samples and so on,

00:40:39 everything is beautiful. And I used to tell people

00:40:42 in the sixties, gas chromatography is easy,

00:40:45 just get into it and don't worry about it.

00:40:48 But when we come over here to liquid chromatography,

00:40:51 we find everything is

00:40:54 unacceptable. We have high flows

00:40:57 of a carrier gas, somewhere in the range of about

00:41:00 a hundred times more than we could possibly

00:41:03 want to consider. We're going to add horrible

00:41:06 buffers to this thing.

00:41:09 And we're going to want to work with samples

00:41:12 that we can't evaporate into a mass spectrometer.

00:41:15 So it really is a very big problem.

00:41:18 And looking at it in the

00:41:21 1960s and 1970s,

00:41:24 we know why we wanted to do it. GCMS

00:41:27 was so successful. But

00:41:30 we also had the emergence

00:41:33 of high pressure liquid chromatography,

00:41:36 a very important thing that sort of catalyzed

00:41:39 together and people said we've got to have liquid chromatography

00:41:42 with mass spectrometry. Now

00:41:45 this slide shows the first four people

00:41:48 that really published in this subject. And I think

00:41:51 it's very interesting to point out the first of these

00:41:54 was done in Russia by Tallrose in 1968.

00:41:57 Shortly I'll show the slide

00:42:00 and we'll see that his results were not wonderful,

00:42:03 but he must be credited as giving the introduction

00:42:06 of this game. From there,

00:42:09 and I'm wanting to emphasize how everybody

00:42:12 really builds on something that happened before them.

00:42:15 Fred McLaperty, very cleverly,

00:42:18 said the trouble with Tallrose is he

00:42:21 took a very small amount of his sample into the mass spec

00:42:24 and he wanted to do electron impact.

00:42:27 We'll take a larger amount and we'll

00:42:30 do chemical ionization. And that was

00:42:33 what Fred came out with, building as I say

00:42:36 on what Tallrose had done with the original DLI.

00:42:39 At about the same time

00:42:42 we had Ray Scott, who was a chromatographer

00:42:45 but well known

00:42:48 around a lot of mass spectrometry

00:42:51 in the 1960 period.

00:42:54 And he came along with a moving wire.

00:42:57 This moving wire, a picture of which I'll show you later,

00:43:00 was available commercially

00:43:03 and it had been used to take sample off a liquid

00:43:06 chromatograph, evaporate the solvent

00:43:09 and take it into a flame ionization detector.

00:43:12 So that's where the

00:43:15 moving wire system came from. And then of course

00:43:18 we've already heard a great deal of the

00:43:21 comments about Evan Horning's work with the

00:43:24 atmospheric pressure ionization. Names that should

00:43:27 be on here are Dave

00:43:30 Carroll and Zydig, Ismael Zydig.

00:43:33 And they were the people in the

00:43:36 laboratory who were actually doing the work.

00:43:39 And where did the idea come from for this API?

00:43:42 Dave Carroll had worked with a company that made

00:43:45 this drift mass spectrometer.

00:43:48 Kind of a poor instrument actually because the resolution was

00:43:51 terrible. It wasn't a mass spectrometer at all and

00:43:54 operated at high pressure and just let ions drift down

00:43:57 a tube. But they used an atmospheric

00:44:00 pressure ionization source, so that's what he

00:44:03 was dealing with. Now, on the next slide

00:44:06 this is the work of Talrose.

00:44:09 And we look very quickly, we see, my gosh

00:44:12 the chromatography is slow, very wide

00:44:15 peaks. He was doing electron impact

00:44:18 on compounds, methylene and benzene

00:44:21 and naphthalene that weren't very important for

00:44:24 electron impact anyway. And his sensitivity

00:44:27 was incredibly low because of the fact that

00:44:30 he was taking such a small amount into the

00:44:33 mass spectrometer. Of course, as I say then

00:44:36 Fred McLaperty said, let's take

00:44:39 a larger amount and we'll use

00:44:42 the solvent in the carrier as

00:44:45 the chemical ionization gas. Very brilliant concept

00:44:48 at the time because chemical ionization had been fairly

00:44:51 well established, but we also had very well controlled

00:44:54 gas, methane and butane and helium

00:44:57 and it depended what you wanted to do with it.

00:45:00 Now again, here was the chromatogram from

00:45:03 a UV detector and you'll notice

00:45:06 kind of bumpy peaks he was getting off of these

00:45:09 steroids, but one must always remember

00:45:12 when you're looking at the first results that somebody

00:45:15 comes out with, they don't always

00:45:18 look so good, but they're the pioneers and the ones that we've

00:45:21 got to respect. And the next slide

00:45:24 was, oh, I'm sorry, I'm a little out of

00:45:27 line, it doesn't matter. This slide then is the work that

00:45:30 Horning did. Horning and Dave Carroll

00:45:33 and Ziddig was the API source.

00:45:36 And you might say, how in the world

00:45:39 did we see in 1974

00:45:42 this kind of work going on and we didn't

00:45:45 seem to pay any attention to it? Well, it's

00:45:48 not as if we didn't pay attention to it. I actually

00:45:51 spent two times going down to Horning's lab, spent

00:45:54 two or three very delightful weeks with them,

00:45:57 but what was driving us

00:46:00 at the time was electron impact.

00:46:03 Everybody wants electron impact

00:46:06 so they can compare with the standard

00:46:09 spectra. And when you said, well, we have this

00:46:12 nice API source and it gives pretty

00:46:15 good results, as you can see, a selected ion

00:46:18 monitor here, derivative of glycine,

00:46:21 people weren't interested in that.

00:46:24 And, you know, in case you think, well, did we all

00:46:27 just forget about it? Somewhere in the middle

00:46:30 70s, at a meeting in

00:46:33 St. Louis, we had a whole symposium on API

00:46:36 sources. So it was in front

00:46:39 of us. But, you know, it's hard to say.

00:46:42 Should we be driven by our

00:46:45 customers? I think the answer is yes

00:46:48 because the customer is ultimately the one that's

00:46:51 going to buy an instrument. But on the other hand,

00:46:54 you might feel that the scientist

00:46:57 should just go in there and do his work and do what he

00:47:00 thinks is right. The problem with that is you've

00:47:03 always got to get a sponsor. So we're driven

00:47:06 by a lot of things, not just what we want to do.

00:47:09 And then I promised, as I promised,

00:47:12 the result of Ray Scott. And it was

00:47:15 really quite good. This is the result of a

00:47:18 fermentation extract. This is the

00:47:21 ultraviolet. No, this is the ultraviolet.

00:47:24 This was a total ion monitor of it.

00:47:27 And this was a selected ion monitor of it. And as

00:47:30 you can see on the selected ion monitor, quite

00:47:33 good separations and very nice looking spectra.

00:47:36 Well, what comes next

00:47:39 now? Oh, I thought I should show you this

00:47:42 little slide. Here are these wires, these

00:47:45 tubes of wires that come in through a vacuum lock,

00:47:48 in through the mass spectrometer and out again.

00:47:51 And the problem with that, of course, was you could only hold a

00:47:54 very small amount of sample on this wire.

00:47:57 So we at Finnegan thought

00:48:00 about how we can improve in this because it had the advantage of

00:48:03 electron impact. And we felt, well, we'll go

00:48:06 from that and we'll put a broad belt

00:48:09 on it where we can take more sample.

00:48:12 But in addition, there were other things going on. There

00:48:15 were attempts, for example, at membranes. They have the

00:48:18 disadvantage that a membrane, of course, has

00:48:21 your solvent and solute

00:48:24 are going to have fairly close polarities.

00:48:27 Since they've got such close polarities or personalities,

00:48:30 if you like, it's very difficult to get a separation

00:48:33 on them. Look, here is, got Bruce

00:48:36 Thompson's name here and he was starting to do the ion

00:48:39 evaporation that we've just heard of at that time.

00:48:42 1976. We had,

00:48:45 as I mentioned, we expanded the belt to

00:48:48 take a larger amount of sample.

00:48:51 And then, of course, there was direct liquid introduction.

00:48:54 That was an extension of McLafferty's and it was primarily

00:48:57 pioneered in the commercial world

00:49:00 by Hewlett-Packard with Malera, one of the

00:49:03 important, Malera and Chris

00:49:06 Kenyon being the important developers in

00:49:09 that. And then, of course, about that same time

00:49:12 in this late 70s, we had

00:49:15 Marvin Vestal, Gene Fertel, and Takauchi

00:49:18 in Japan doing nebulization

00:49:21 and which, in a related way, with

00:49:24 the DLI work. Now this, again,

00:49:27 you start to think of this and you see this all together and you say,

00:49:30 gee, there's the DLI. Now we're starting to get

00:49:33 to nebulize it. And if you think of all the things that are

00:49:36 commonly being used today, you see there's a

00:49:39 whole pattern of the whole picture and it's actually

00:49:42 come together. And I only

00:49:45 regret that I haven't been able to be a part of it.

00:49:48 This machine, you've already seen it from a picture

00:49:51 from Jack Henion. And I've

00:49:54 often wondered why in the world chromatographers

00:49:57 cannot be browbeaten into going to

00:50:00 liquid chromatography. This is the Jasko

00:50:03 Familic, a commercial instrument that was

00:50:06 available in the middle 70s. I used to

00:50:09 implore people, if you want to do

00:50:12 LC-MS, go to micro-LC

00:50:15 because then you've reduced your problems by a

00:50:18 factor of 100. That's a big reduction in a

00:50:21 problem, you know. How many people could I convince?

00:50:24 A few Brits. They're very brave souls.

00:50:27 Jack Henion, who I

00:50:30 was delightfully grateful that he acknowledges

00:50:33 my contribution there. But the real

00:50:36 thing is, even today

00:50:39 the chromatographer is still

00:50:42 reluctant to go

00:50:45 into the small quantity. Why? He says,

00:50:48 Well, we can't get as good as separations,

00:50:51 we can't control our pumps as well, and

00:50:54 we've got all our methods worked out. Well, of course you have.

00:50:57 But if you don't try the other, you're never going to get it.

00:51:00 And if you look on your program today, it's rather

00:51:03 interesting that right on the top of our today's

00:51:06 program, on the next column,

00:51:09 there's a micro-LC

00:51:12 paper being given, or was given in one of the other

00:51:15 sessions. Now, I thought you might

00:51:18 be interested in seeing the kind of results that you could get

00:51:21 with a belt using

00:51:24 small columns and low flows. This was done

00:51:27 on aflatoxins. We have

00:51:30 this is the UV, we have the total ion monitor,

00:51:33 and we have the four samples here

00:51:36 showing very nice separation, very nice

00:51:39 results. But how

00:51:42 interested are people? These are electron impact, by the way.

00:51:45 How interested are people in doing that? If you've got to

00:51:48 use a liquid, sorry, a micro-liquid

00:51:51 chromatograph, we don't want it.

00:51:54 Now, I want to show a couple of results

00:51:57 that were available

00:52:00 in this period.

00:52:03 This was done by

00:52:06 Thruston and McGuire, EPA people,

00:52:09 in 1980 on a tannery extract.

00:52:12 Quite good results, nothing extremely

00:52:15 hard, dihydroxy, whatever that is,

00:52:18 I can't, phenylpropanol, not the hardest

00:52:21 compounds in the world, but not a bad piece of LC-MS

00:52:24 work. At the same time, we also

00:52:27 had a great deal of work being

00:52:30 done by Hewlett-Packard, and this is the kind of results

00:52:33 they were showing. The slide doesn't show up quite so well,

00:52:36 but this is Hewlett-Packard work

00:52:39 done by Chris Kenyon on these sugar derivatives.

00:52:42 Incidentally, how many people can see this fuzzy

00:52:45 little thing down here? Can you see that at all?

00:52:48 Or is this, it's hidden out, is it? Well, it doesn't matter.

00:52:51 It's not important at all. These are old slides,

00:52:54 you know, and since we're no longer allowed to put

00:52:57 the names of the

00:53:00 sponsor or

00:53:03 manufacturer on a slide, I had to take and block

00:53:06 that fuzzy little thing out. That's the F-word that you might

00:53:09 guess.

00:53:12 Now,

00:53:15 going on, where are we going from here?

00:53:18 We've got DLI, and we've got

00:53:21 the belt. Both of them are showing plenty of

00:53:24 results, and there's a lot of customers. A hundred

00:53:27 people or so using these things around the world in the

00:53:30 late 70s, but it wasn't enough.

00:53:33 We knew we had to get more, and look at what

00:53:36 we're doing. We had thermospray.

00:53:39 Marvin Vestal will be talking about that, and that was

00:53:42 discovered in the late 70s, a serendipitous

00:53:45 result. And in this time,

00:53:48 1980 or so, 84, 12

00:53:51 papers. That's a rather interesting historical

00:53:54 fact that we'll talk a little bit more about.

00:53:57 Suchiya was doing liquid surface ionization,

00:54:00 and more work was coming out of Thomson's lab.

00:54:03 And also, perhaps one of the biggest things we've

00:54:06 seen to date, Dr. Finn, Whitehouse,

00:54:09 and Yamashita published on the

00:54:12 electrospray. Again,

00:54:15 coming back to the idea that we all

00:54:18 stand on someone else's shoulders,

00:54:21 the first electrospray was actually done by

00:54:24 Malcolm Dole in about 1960 and published

00:54:27 in 1968. And if

00:54:30 Malcolm had been around a lot of good mass spectrometers,

00:54:33 it's just possible that the work of Finn

00:54:36 would not have been 1984. It might have been 1974.

00:54:40 But again, it took people to sort of pick up on these

00:54:43 things and see just what

00:54:46 all has been done.

00:54:49 Another important point, of course, at this stage is

00:54:53 Dr. Browner and Ross Willoughby were coming out with

00:54:56 aerosol generation. And their reason

00:54:59 was exactly the reason why

00:55:02 I worked with

00:55:05 the Bell. They wanted to get electron impact systems.

00:55:08 And this has turned out to be the more popular

00:55:11 electron impact system because

00:55:14 the Bell had two kinds of people.

00:55:17 Those who could make it work and those who couldn't make it work.

00:55:20 And unfortunately, I think there were more

00:55:23 of the latter. And as a consequence,

00:55:26 they were saying, I don't want this darn belt rumbling

00:55:29 through my ion source. I once heard it

00:55:32 sort of compared to a gardener having an

00:55:35 elephant sort of walking around through his garden.

00:55:38 And they felt this was not a nice thing

00:55:41 to do to a mass spectrometer.

00:55:44 And in addition, at that time, we had a lot of interest

00:55:47 in supercritical fluid chromatography.

00:55:50 Somehow or another, that seems to have gone on the wayside.

00:55:53 But it was back in that

00:55:56 middle 80 period, there were some very important

00:55:59 developments coming along there.

00:56:02 And so just to continue on,

00:56:05 other important trends that were coming along were

00:56:08 thermospray in this period was the most prominent.

00:56:11 I think there's no question that from about 1980

00:56:14 up until the latter part to about

00:56:17 1990, thermospray was clearly the most

00:56:20 important LC-MS system. And it was

00:56:23 providing good sensitivity

00:56:26 and ease of operation. And now

00:56:29 what is not so here is it could be

00:56:32 coupled up with another new thing coming

00:56:35 in mass spectrometry, what is today very common,

00:56:38 MS-MS in its various many forms

00:56:41 today. Of course, at the beginning,

00:56:44 in my particular environment, we used to think of that as a

00:56:47 triple quad. But there's other things too.

00:56:50 And continuous flow fab

00:56:53 was starting to come along.

00:56:56 And also, we were just on the edge of starting to see

00:56:59 zone electrophoresis. Now,

00:57:02 I believe that's my last slide.

00:57:05 Oh no, this is slide two of three, but I'm not showing the third one.

00:57:09 Other things, micro-LC

00:57:12 was starting to be used a little more

00:57:15 and they're showing that there are a variety of techniques

00:57:18 you can start to get electron impact from if you use

00:57:21 micro-LC. These were

00:57:24 the Stenhagen group in Sweden and

00:57:27 Takauchi and Ishii. Ishii was very, very

00:57:30 he was the Japanese that would really push micro-LC.

00:57:33 And also, aerosol

00:57:36 beam generation was coming along and that

00:57:39 was perhaps the main basic development.

00:57:42 Now, in 1989,

00:57:45 we had five new laboratories reporting

00:57:48 on LC-MS. That almost

00:57:51 sounds silly when you think of it, but

00:57:54 look at this slide.

00:57:57 In the 80s, we had less than 20 papers at these meetings

00:58:00 on LC-MS. Of course, it came along

00:58:03 quickly, so that by the time we're out to the 90s,

00:58:06 we're somewhere in the range of 150 papers.

00:58:09 Now, I'm not going to bother to take into account

00:58:12 how many we've got here today. I leave that as an

00:58:15 exercise to the student.

00:58:18 Anyway, my last slide.

00:58:21 The important thing today is our work with high molecular

00:58:24 weight. Why do I show a crummy

00:58:27 cartoon like this? And the important

00:58:30 thing about this cartoon isn't even on it.

00:58:33 Not mass, but we mentioned, gee, I shouldn't

00:58:36 have tried mass 2500. This

00:58:39 cartoon was, came out, we

00:58:42 came out with this or saw this about

00:58:45 1978. And in

00:58:48 1978, it was a big deal to

00:58:51 think of mass 2500. In fact, we used to

00:58:54 at that time, we were trying to improve it and expand

00:58:57 our range into 1000. And so

00:59:00 you can see how far we've come in the last

00:59:03 decade or two. I

00:59:06 really wish I had been a bigger part of it, but I wasn't.

00:59:09 I want to thank Ross very much for

00:59:12 having me here. And I want to say that it's

00:59:15 been a pleasure. And with a little bit of luck,

00:59:18 maybe 10 years from now, you'll have another paper back

00:59:21 here, or sorry, another symposium back here.

00:59:24 And I'll be invited to give a paper again.

00:59:36 I meant to warn you it'll be the same paper.

00:59:44 That's pretty good for

00:59:47 being a little rusty, Bill. That was a great lecture.

00:59:50 The next

00:59:53 speaker is Marvin Vestal,

00:59:56 20 years lost in the interface between

00:59:59 LC and Mass Spec.

01:00:02 Ross, it's certainly a pleasure to be here.

01:00:05 It's nice to be on a program where I'm not the oldest

01:00:08 speaker.

01:00:14 These are hard acts

01:00:17 to follow, but fortunately I get to speak before John

01:00:20 Vestal, so I know we'll get the last word.

01:00:23 I just wanted to say one personal thing about Bill.

01:00:26 Bill and I both graduated from the University of Utah.

01:00:29 I think he was about 20 years ahead of me.

01:00:32 But we both

01:00:35 benefited from being mentored by Austin

01:00:38 Baroftig, who's one of the people in this business that

01:00:41 I think not many people know as well as they should.

01:00:44 But he's certainly been a great help to me at the

01:00:47 University of Utah.

01:00:53 That's my title, and I asked my colleagues

01:00:56 at Perceptive to come up with a suitable cartoon to illustrate this,

01:00:59 but they were all too busy getting ready for their own presentation,

01:01:02 so you'll have to imagine what you like.

01:01:05 I was thinking of some bald-headed old scientist

01:01:08 sort of bewildered between a dripping faucet

01:01:11 and an instrument, but you'll have to imagine

01:01:14 for yourself.

01:01:17 Bill did a lot of this already,

01:01:20 so I won't repeat it all, but when I was thinking about this,

01:01:23 I thought, you know, what were the things that

01:01:26 had gone before this in mass spectrometry, which kind of led us

01:01:29 to where we were when we started out.

01:01:32 Going back to the

01:01:35 is this the pointer?

01:01:38 Going back in the 30s, there was some work by Loeb

01:01:41 and his students on spray ionization,

01:01:44 and in particular, a paper by Chapman in 1938

01:01:47 where he observed ions in a mobility apparatus,

01:01:50 which clearly were the right mobility to be solvated

01:01:53 in molecular ions. He didn't do any mass spectrometry,

01:01:56 so of course, none of us took it seriously.

01:01:59 I must confess, I wasn't even aware of this paper until sometime

01:02:02 in the late 70s when we started observing ions and we were trying

01:02:05 to explain where they came from. But he certainly observed

01:02:08 ions from spray ionization back in the 30s.

01:02:11 And of course, Dole's work has already been alluded to also,

01:02:14 where he was making ions by spray techniques, but wasn't able

01:02:17 to do mass spectrometry on them, so we never took it seriously

01:02:20 until John showed us the error of our ways.

01:02:23 Also in this period was chemical ionization,

01:02:26 which was really kind of the first new ionization technique

01:02:29 which allowed people to get molecular ions on things

01:02:32 that were otherwise difficult, so it gave us an alternative

01:02:35 to electron impact. And I think field desorption

01:02:38 was kind of the first technique that really

01:02:41 started dealing with non-volatile molecules.

01:02:44 The first laser desorption work was in 1972,

01:02:47 not with LC separation, but with the

01:02:50 showing that you could start to get less volatile materials

01:02:53 in this way, and the other work that's been referred to

01:02:56 in terms of the pioneers of LC-MS.

01:02:59 And the last one on here, which you may not be able to see,

01:03:02 but it's behind the table,

01:03:05 is the work that Friedman and his coworkers did at Brookhaven

01:03:08 where they showed that fast heating from relatively

01:03:11 inert surfaces allowed you to get quite non-volatile

01:03:14 molecules into the gas phase as ions.

01:03:17 They did some small peptides and so forth,

01:03:20 and it was really this work that led us to thinking about

01:03:23 how we might do LC-MS on relatively non-volatile molecules.

01:03:33 Is there a focus on that?

01:03:36 This is basically the same table as Bill showed, at least my version

01:03:39 of his table, and so I won't belabor it as what the problem was,

01:03:42 but I think the major problem with LC-MS

01:03:45 is that, as he said, people wanted

01:03:48 to do one milliliter per minute of aqueous

01:03:51 effluent. If you couldn't do that, they really

01:03:54 didn't think it was a serious LC-MS technique.

01:03:57 And they also wanted to do electron impact, if possible,

01:04:00 because they wanted to be able to compare with the library.

01:04:03 So that was kind of the boundary conditions

01:04:06 that we had going into the game.

01:04:09 So Calvin Blakely and I, who had

01:04:12 built a cross-beam apparatus for doing ion-molecule

01:04:15 reactions, in which we studied complicated systems

01:04:18 like H3 plus colliding with D2, et cetera,

01:04:21 and got data ad nauseum on the

01:04:24 reaction cross-sections, et cetera,

01:04:27 proposed to NIH in, I think, 1975

01:04:30 to build what we called a cross-beam LC-MS.

01:04:33 And this is a schematic diagram of that apparatus,

01:04:36 which I think may even be from the proposal,

01:04:39 I'm not sure, in which we used

01:04:42 six diffusion pumps, varying from a small

01:04:45 4,200 liter per second here, which is a fat 10-inch

01:04:48 pump, a 4-inch pump, and

01:04:51 several smaller pumps, a huge mechanical pump pumping

01:04:54 on this vaporization region. There were six

01:04:57 mechanical pumps, I believe, and five diffusion pumps,

01:05:00 and a CO2 laser. Because what we

01:05:03 proposed to do was take a liquid jet out of the

01:05:06 end of a tube here, which we knew would come up,

01:05:09 and in a reduced pressure region

01:05:12 hit that jet with a CO2 laser.

01:05:15 So this is where the cross-beam idea came from. We were going to

01:05:18 vaporize the sample in free space without it touching

01:05:21 any surfaces, because we knew if it touched surfaces, we would get

01:05:24 pyrolysis. And we wanted to be able to

01:05:27 vaporize intact, non-volatile molecules, and then ionize them

01:05:30 by either electron impact or by chemical ionization.

01:05:33 So the elaborate pumping was to get the pressure low enough to do

01:05:36 electron impact. The electron impact source was

01:05:39 let's see, where was that? That was down here, I believe.

01:05:42 We produced chemical ionization up here,

01:05:45 and anyway, a very

01:05:48 elaborate apparatus, and Calvin dug out of his file,

01:05:51 being very well organized, he dug out an old picture of this

01:05:54 apparatus, which you may or may not be able to see.

01:05:57 It looks a little dark in this.

01:06:00 We really did have an LC on it. Here it is. It's a Perkin-Elmer

01:06:03 syringe pump LC, which was, I think, the world's

01:06:06 largest LC, and it fit very well

01:06:09 with the rest of the apparatus, as you can see.

01:06:12 The CO2 laser is up on top here. The beam went down

01:06:15 through here. There's a 12-inch aluminum cube here, which is

01:06:18 the center of the world, with pumps coming out in all directions.

01:06:21 It looks rather neater than it was because the mechanical pumps

01:06:24 were over here in the other room. There was a small hole in the wall here

01:06:27 where the lines went through. Well, this system

01:06:30 actually worked, and it worked well enough for us to get our NIH

01:06:33 grant renewed so we could go on and do some more work,

01:06:36 and I think that's the main purpose that this apparatus

01:06:39 served, was to get us started.

01:06:42 We did do both EI and CI spectra.

01:06:45 Mary Jane McAdams, alias Finney,

01:06:48 got her PhD thesis running a lot of

01:06:51 samples on this instrument. This is

01:06:54 CI and EI spectra on this

01:06:57 molecule here. I don't think you can see the EI spectrum,

01:07:00 but it's there.

01:07:03 So we were able to get both kinds of spectra, and this

01:07:06 worked, actually, up to about a half a milliliter per minute.

01:07:09 We were trying to do one milliliter per minute, but we really couldn't do EI

01:07:12 spectra above about a half a milliliter a minute.

01:07:15 This shows the driving force behind

01:07:18 this was to be able to do matching

01:07:21 with the library. This is adenosine, which we considered a rather

01:07:24 non-volatile molecule in those days. It didn't give a

01:07:27 real strong molecular ion, but you can see in the library it's there,

01:07:30 and it's also there in our spectrum. So we were able to match

01:07:33 the libraries of a number of compounds

01:07:36 with the spectra that we were getting

01:07:39 from this apparatus. This shows the sort of

01:07:42 kind of sensitivity that we were able to get. This is 80 nanograms injected

01:07:45 of hexachlorobenzene

01:07:48 in the negative ion mode, which admittedly was a fairly favorable case

01:07:51 because it's a strong electron capturing. And we have

01:07:54 a paranoid here blown up by a factor of 10. And of course,

01:07:57 this is chlorine and Cl2 minus, which are not very interesting,

01:08:00 but were there as a result of this sample.

01:08:03 And we really did do LC-MS on this. I believe this is

01:08:06 from this period and not later. Some of these slides

01:08:09 kind of run together after a while, but I think this is from

01:08:12 Mary Jane's thesis. These are some polynuclear aromatic

01:08:15 compounds. In fact, I recognize these. These are anthracene

01:08:18 and phenanthrene, that doublet there, which is resolved.

01:08:21 Respectable LC-MS under the most favorable conditions,

01:08:24 although when I think about how hard we had to work

01:08:27 to get some of these data, it looks easier now than it did then.

01:08:31 Well, we went from that apparatus. We realized after a while

01:08:34 that heating the, well, actually I forgot

01:08:37 the most important part of it. It didn't really work as advertised.

01:08:40 We weren't able to vaporize the liquid jet by

01:08:43 hitting it with a CO2 laser. Because when you try to

01:08:46 heat liquid droplets with a laser in free space, they

01:08:49 start evaporating. They evaporate on one side and they go off

01:08:52 the other direction. And so that never really

01:08:55 worked. In fact, we tried it. What we found is we made

01:08:58 beautiful ice sculptures inside of our vacuum

01:09:01 system. And so we decided we had the

01:09:04 world's most expensive ice machine.

01:09:07 But if we moved the laser over a little bit so we heated

01:09:10 the end of the capillary tube, we discovered we in fact could

01:09:13 vaporize the liquid and we could get good spectra. And

01:09:16 somewhat to our surprise, even relatively labile molecules

01:09:19 like adenosine worked perfectly okay this way.

01:09:22 So, well, we thought maybe we don't have to heat over such a small

01:09:25 region after all. So if we're going to heat only the end

01:09:28 of a tube, a laser seems like kind of the hard way to do it.

01:09:31 It's a little moist. So our next idea is thinking we

01:09:34 still needed to have a very high power density

01:09:37 in order to vaporize over a very small region. Otherwise

01:09:40 we would get pyrolysis. So the next

01:09:43 most dense source we could think of was an oxy-hydrogen

01:09:46 torch. Which, from a practical

01:09:49 point of view, there may be some problems with this. But in fact

01:09:53 we did make this work. This was actually a ray of four torches

01:09:56 around the capillary tube. Now this was all atmospheric

01:09:59 pressure now, so we vaporized, shot into the

01:10:02 vacuum system with a skimmer.

01:10:05 Any particles or droplets hit this

01:10:08 heated probe here, were vaporized, and we were only doing

01:10:11 chemical ionization with this. We gave up on the electron

01:10:14 impact fairly early, realizing that the amount of pumping

01:10:17 required to do this was probably not worth the trouble and that

01:10:20 we would really accept it as a useful technique.

01:10:23 So we produced ions, deflected them into a quadrupole

01:10:26 analyzer, and the rest is fairly

01:10:29 well known. This instrument actually was moved to

01:10:32 Jim McCloskey's laboratory, who was one of our collaborators

01:10:35 in this early work. And in fact it was through his efforts that

01:10:38 we were able to get the NIH grant in the first place, because he was well

01:10:41 known in biological mass spectrometry.

01:10:44 The rest of us were physical chemists, and at that time

01:10:47 it was very suspicious of physical chemists coming along trying to get money

01:10:50 out of the granting system.

01:10:53 Klaus Biemann was the chairman of the

01:10:56 site visiting committee that gave us the money, and the fact that Jim

01:10:59 had been one of his students, I'm sure helped immensely

01:11:02 in his review of our proposal.

01:11:05 So, this instrument

01:11:08 was actually moved to Jim's lab. It was

01:11:11 in continuous service, more or less, until about a year ago

01:11:14 when it was finally taken out of service,

01:11:17 and he published an enormous amount of excellent work, particularly

01:11:20 on identifying previously unknown nucleosides using this instrument.

01:11:23 Now we did, as time went along, replace this with different

01:11:26 versions of the vaporizer, which I will mention a little bit

01:11:29 later.

01:11:32 I should wait to show this one.

01:11:35 While we were evaluating this instrument, and trying to tune it up and figure

01:11:38 out how to make it work better, John Carmond and I,

01:11:41 who was one of my students at the University of Houston,

01:11:44 we were doing a series of experiments in which we were

01:11:47 injecting adenosine to see how the system responded,

01:11:50 and at one point

01:11:53 we noticed that the response had changed rather markedly.

01:11:56 We were getting a lot more molecular ion, and things were rather

01:11:59 different, but we obviously had found the magic tuning

01:12:02 condition, so it worked. At that point, Calvin Blakely

01:12:05 walked in and looked at the instrument, and looked at some of the dials

01:12:08 with red currents and so forth, and he said,

01:12:11 and watched this peak come out, and he said,

01:12:14 did you inject that sample before or after the filament burned out?

01:12:17 And we said, well, we must have injected it before,

01:12:20 because clearly the filament had burned out because the emission gauge was

01:12:23 reading zero, but just for the fun of it, we injected another sample,

01:12:26 and sure enough, we still got a signal.

01:12:29 I don't know whether this is the actual

01:12:32 original data, but it was taken within a few days

01:12:35 of that observation. This is adenosine

01:12:38 injecting one microgram, just a flow injection experiment.

01:12:41 This is with the filament on, and you can see we're getting a small

01:12:44 molecular ion here. We have a peak here

01:12:47 which has a little tail on it, which is mainly the 136 fragment

01:12:50 ion, and lots of background in our total ion current.

01:12:53 At this point, we turned

01:12:56 the filament off.

01:12:59 In fact, this is the 136 trace, I'm sure. There was a little bit of tailing

01:13:02 on the 136. This is the total ion current.

01:13:05 We turned the filament off intentionally this time,

01:13:08 and did a couple more injections, and you can see we got a very strong

01:13:11 molecular ion signal, a very weak fragment ion signal,

01:13:14 and almost no background.

01:13:17 Well, we scratched our head about this a little bit,

01:13:20 and we had kind of looked a few times for ions without having the filament on,

01:13:23 not really expecting to see anything.

01:13:26 And I have to confess, at this point, even though they had published

01:13:29 a paper a few months earlier than this using mass spectrometry

01:13:32 for ion evaporation experiments, at the time we made this

01:13:35 observation, I was totally unaware of Bruce Thompson

01:13:38 and Ehrbarn's work, and was somewhat

01:13:41 embarrassed when I went to the Journal of Chemical Physics,

01:13:44 and I think it was an issue of the journal which also

01:13:47 had one of our papers on a different subject in it,

01:13:50 but I had not read his paper, of course, because I wasn't really

01:13:53 interested in what he was doing, or at least I didn't know that I was interested in it

01:13:56 at the time, and discovered that they'd already explained a lot of these phenomena

01:13:59 that we later realized were responsible

01:14:02 for ion evaporation, or what we

01:14:05 called at the time direct thermospray ionization, but clearly

01:14:08 it's very closely related to what he was observing

01:14:11 and what electrospray has made even more

01:14:14 prominent lately.

01:14:17 Well, obviously even an

01:14:20 oxyhydrogen torch is not a very good way to heat the end of a tube,

01:14:23 and we started increasing the length

01:14:26 of the heated region. Our next version used a

01:14:29 copper block about an inch long in which we put some cartridge

01:14:32 heaters, and vaporized directly into the chemical ionization

01:14:35 source, extracted the ions

01:14:38 out to the mass spectrometer, and this was really kind of the

01:14:41 first commercial version of the thermospray system.

01:14:44 It worked quite well, and we found that

01:14:47 even though we heated over a longer length, it didn't

01:14:50 make any difference. We still got about the same kind of results. In fact, we got

01:14:53 rather better results than we were able to get with the oxyhydrogen

01:14:56 torch, because we could control this much more accurately.

01:14:59 About this time, Gordon Ferguson, who's in the audience,

01:15:02 joined us at the university

01:15:05 and said, gee, that's, you know, if we

01:15:08 heated the whole, just heated the capillary tube directly by conduction,

01:15:11 we could have really good control over it, and we could put thermocouples

01:15:14 along it and really monitor what's going on.

01:15:17 And so he developed the electronics and

01:15:20 really was the one who did that work of

01:15:23 going to a probe that had a

01:15:26 heated capillary, and this is

01:15:29 an example of a spectrum from that system

01:15:32 which we, the first time we saw really large non-volatile

01:15:35 systems, we could see glucagon, and

01:15:38 somewhat to our surprise, as we got up to larger molecules,

01:15:41 we started seeing higher charge states. This is the 3 plus

01:15:44 and 4 plus states of glucagon.

01:15:47 We probably had the 1 and 2 plus too, but our quadrupole mass spectrometer

01:15:50 was really struggling by the time it got to mass 1200, so we

01:15:53 really couldn't look any higher than that. And

01:15:56 several of my students and associates applied this to things like

01:15:59 peptide sequencing, and I intended to have a slide

01:16:02 of that, but it got lost in the shuffle somewhere.

01:16:05 But quite a lot of work was done sequencing peptides

01:16:08 by enzyme digestion

01:16:11 using immobilized enzymes on columns.

01:16:14 But we were at the nanomole level to be able

01:16:17 to do this. Typically 1 to 10 nanomoles of sample

01:16:20 was required, and that was not really acceptable

01:16:23 to the protein chemists. This is the

01:16:26 version that really became the standard for thermospray,

01:16:29 the heated capillary, and we found we could

01:16:32 heat over a very long length and still get good results.

01:16:35 And the reason for that is shown on this slide

01:16:38 where we've measured, in fact Gordon, I believe,

01:16:41 did these experiments initially. We put thermocouples all along the tube

01:16:44 so we could measure the temperature along the tube, and

01:16:47 the interpretation of this profile is that you're getting,

01:16:50 you start to get vaporization at some point in the tube, it continues

01:16:53 and if you put too much heat in, you get complete vaporization

01:16:56 before you get out of the tube, and then you do get pyrolysis

01:16:59 and all sorts of bad things happening. But if you control

01:17:02 the heat input so that this interface

01:17:05 moves out to the end here, then you can get nebulization

01:17:08 and only partial vaporization

01:17:11 and it works beautifully, and that's the way that thermospray

01:17:14 was done, and I guess is still done in some places

01:17:17 these days. So we coined

01:17:20 this word thermospray to define what we were doing.

01:17:23 It was contrary to some

01:17:26 people's suggestion, and I think Bill may have been responsible for this

01:17:29 suggestion first, that it really meant

01:17:32 equal applications of heat and prayer was what was required

01:17:35 to make it work. That was not really the

01:17:38 actual coining of the word was

01:17:41 here. I guess I'm out of time.

01:17:44 I would like to just put this into a little bit

01:17:47 of context of the other developments in LCMS

01:17:50 and ionization techniques. We were developing

01:17:53 thermospray in the 80s here, and Bill

01:17:56 alluded to some of this before,

01:17:59 but I think what happened in 1988

01:18:02 has changed the world as far as mass spectrometry

01:18:05 and LCMS is concerned. Of course, that's electrospray and MALDI,

01:18:08 and these are the techniques which we are

01:18:11 working on presently, and I think everybody else is in

01:18:14 one form or another who are interested in large non-volatile molecules.

01:18:17 I will skip our other

01:18:20 contributions to LCMS, which are more recent and

01:18:23 probably not so exciting, and conclude with the acknowledgments

01:18:26 of the people who really did all of this work. My colleagues

01:18:29 at the University of Utah helped get this started, including

01:18:32 Charlie Edmonds, who worked in Jim McCloskey's lab and did a lot of the

01:18:35 beautiful early applications. Calvin Blakely you will see all the

01:18:38 way across here because he's been a vital part

01:18:41 of my efforts in this area from the beginning until recently.

01:18:44 We are continuing to do some LCMS work at

01:18:47 Perceptive, but I'm not talking about that now. These students

01:18:50 in particular are the ones who

01:18:53 did most of the work that I'm talking about now.

01:18:56 I'd like to just finish with a photograph.

01:18:59 Hank Fales is responsible for this.

01:19:02 I don't know whether he's in the audience or not, but that's Hank there

01:19:05 if you don't recognize him from about 20 years ago.

01:19:08 He came down and spent a sabbatical few months with us in

01:19:11 Houston, and while he was there he bought

01:19:14 t-shirts for the group. His says, I guess that's supposed

01:19:17 to be honorary or ornery. I'm not sure which vestal is virgin.

01:19:20 And this is the head virgin down here.

01:19:23 Mary Jane McAdams was the one who did

01:19:26 the early work with the LCMS. This is Calvin Blakely

01:19:29 if you don't know him. Dwayne Harden did some work with laser

01:19:32 desorption for moving belts, and Joe Hiller worked

01:19:35 on photo dissociation. Julie Robeson

01:19:38 went off and got a degree and became an MD

01:19:41 so she's really successful. Thank you very much.

01:19:44 There's a change in the program today.

01:19:47 Professor Briner had the flu

01:19:50 and couldn't make it

01:19:53 to the meeting this year.

01:19:56 Fortunately, the organizer of this session was a graduate

01:19:59 student at Georgia Tech

01:20:02 about 15 or 20 years ago

01:20:05 when particle beam was developed.

01:20:08 So I am honored to be here today

01:20:11 True to form

01:20:14 with Professor Briner

01:20:17 I call him Rick now

01:20:20 because it's a few years after I graduated.

01:20:23 The time I was at Georgia Tech, it was Professor Briner.

01:20:26 He would always

01:20:29 drop things off at me

01:20:32 and say, you know,

01:20:35 I'm not going to do this anymore.

01:20:38 He would always drop things

01:20:41 on you at the last minute.

01:20:44 Friday afternoon at 5 o'clock he gave me a call

01:20:47 and said that he wasn't coming to the session today.

01:20:50 I was faced with the responsibility of preparing

01:20:53 a lecture. So I do not have polished

01:20:56 slides, but I do have some insights

01:20:59 about the early days and hopefully

01:21:02 from a graduate student's perspective

01:21:05 it may be of interest to you.

01:21:21 Since I didn't have a title slide, I was asked

01:21:24 to give a plenary lecture at the Aerosol Conference

01:21:27 a year or so ago. The title was Aerosols

01:21:30 and Analytical Chemistry. Ignorance is not always bliss.

01:21:33 I thought that might be appropriate for this lecture

01:21:36 because when I went to Georgia Tech

01:21:39 in 1977

01:21:42 I was a first year graduate student.

01:21:45 Professor Briner was a

01:21:48 first year assistant professor.

01:21:51 I was sort of the first group of graduate students.

01:21:54 Professor Briner had worked in

01:21:57 Weinfurtner's group as a post-doc

01:22:00 and really had distinguished himself

01:22:03 more in the field of atomic spectroscopy

01:22:06 than either liquid chromatography or mass

01:22:09 spectrometry. But he had an idea that

01:22:12 possibly there could be

01:22:15 some advancement in the field of LC-MS

01:22:18 if someone took a different perspective

01:22:21 and really looked at the aerosol process

01:22:24 more thoroughly rather than

01:22:27 the traditional mass spec approach looking at ions

01:22:30 or the liquid chromatography approach looking at condensed

01:22:33 phase.

01:22:36 So I thought that was fairly interesting.

01:22:39 I picked a graduate

01:22:42 advisor a year or so later

01:22:45 and was excited about

01:22:48 starting a career in LC-MS.

01:22:51 I read the reviews by Bill McFadden in the late 70s.

01:22:54 There was an excellent review by Fred McLafferty

01:22:57 and Patrick Carpino and George Guisson.

01:23:00 I didn't have a basis of

01:23:03 30 years of input to develop

01:23:06 or to attack this problem.

01:23:09 I really was a young graduate student

01:23:12 with not a technical basis learning how to do

01:23:15 research in a laboratory that

01:23:18 had no mass spectrometers

01:23:21 nor did we have a liquid chromatograph.

01:23:24 We didn't even have an LC pump.

01:23:27 But we did have a few instruments to measure particles

01:23:30 and some insights in aerosols.

01:23:33 So that's the perspective

01:23:36 that I began my career at Georgia Tech.

01:23:39 Just another comment.

01:23:42 This is

01:23:45 data, real data

01:23:48 from LC-MS.

01:23:51 Particle beam LC-MS in the early 80s

01:23:54 showing

01:23:57 the state of the pumping

01:24:00 world back in those days.

01:24:03 You see the reciprocating pumps

01:24:06 fluctuations in the baseline.

01:24:09 The world has come a long way

01:24:12 in the last 15 or 20 years

01:24:15 in terms of pump technology, fittings, and so on.

01:24:18 Another point of view, you see all my data

01:24:21 was collected on a strip chart recorder.

01:24:24 I didn't have a mass spectrometer until

01:24:27 four years into my thesis.

01:24:30 Basically, when the aerospace department

01:24:33 at Georgia Tech bought a new

01:24:36 CI Hewlett Packard GC-MS

01:24:39 they gave me their old

01:24:42 EI only instrument.

01:24:45 So I couldn't do any other ionization mode but EI.

01:24:48 So I didn't have a choice but to develop an LC-MS

01:24:51 technique that could do EI.

01:24:54 I only had a singly pumped system.

01:24:57 And the only way to detect signal on this instrument was a strip chart recorder.

01:25:00 The way I quantitated was by cutting and weighing.

01:25:03 So for those of you

01:25:06 who have PCs with 200 MHz

01:25:09 processors, only a few short

01:25:12 years ago people were still cutting and weighing.

01:25:15 You didn't have automated data reduction.

01:25:18 But you could still think about problems and solve problems

01:25:21 and look at processes.

01:25:24 So as I said,

01:25:27 I didn't have an LC

01:25:30 in this process.

01:25:33 This is some work

01:25:36 that I performed over about

01:25:39 a three and a half or four year period.

01:25:42 This work wasn't funded either.

01:25:45 More insight to graduate students.

01:25:48 So I used to have to get hangers and try to steal

01:25:51 stainless steel out of the stock room

01:25:54 through the cracks

01:25:57 in the mesh

01:26:00 and use the student shop.

01:26:03 I made just about every part that you'll see today

01:26:06 myself in a machine shop.

01:26:09 So maybe this is a lesson.

01:26:12 You don't have to be funded to solve problems either.

01:26:15 But it took me five and a half years to get my degree.

01:26:18 So if you're funded, you could probably do it faster.

01:26:21 So as I said,

01:26:24 I had a syringe pump

01:26:27 to deliver liquid to the system.

01:26:30 I machined all the parts.

01:26:33 This is a glass chamber.

01:26:36 I borrowed a roughing pump from the

01:26:39 physical chemistry group across the hall.

01:26:42 I did have a flange mount

01:26:45 with a conflat flange.

01:26:49 I could attach this device

01:26:52 to borrowed instruments in other labs.

01:26:55 So our approach was

01:26:58 to study the aerosol process

01:27:01 and hopefully

01:27:04 get rid of the solvent.

01:27:07 Bill McFadden pointed out in his review

01:27:10 very clearly that you needed a 10 to the 4th enrichment

01:27:13 of this liquid in an interface

01:27:16 to do EI.

01:27:19 So it was pretty straightforward for me.

01:27:22 I didn't have to read the old literature.

01:27:25 I just had to figure out a way to get rid of 99.99%

01:27:28 of all the stuff going into the system

01:27:31 and make sure that 0.01% was all solute.

01:27:34 That was pretty straightforward.

01:27:37 It was hard to do, but the problem was easy to define.

01:27:40 So the approach was to generate

01:27:43 particles. I used a Berglund-Lew design

01:27:46 that was published by an aerosol group

01:27:49 at Minnesota. It makes small uniform

01:27:52 particles. The problem was that this

01:27:55 picture is an idealized picture.

01:27:58 In reality, you don't get

01:28:01 an aerosol coming out

01:28:04 of an aerosol generator to go straight down the tube

01:28:07 and into the mass spectrometer. Usually it went off at a

01:28:10 dog-leg angle and stuck in the side of the wall.

01:28:13 You had droplets dripping, and that certainly

01:28:16 wasn't a very efficient or effective

01:28:19 LC-MS.

01:28:22 We deserted this design for

01:28:25 a more sophisticated approach.

01:28:28 We thought that we could control the direction of the

01:28:31 particles by using induction.

01:28:34 This was a standard technique.

01:28:37 Bruce Thompson alluded to using induction techniques to

01:28:40 charge his spray. Well, if you applied

01:28:43 several hundred volts to this

01:28:46 cylindrical electrode, you could charge the

01:28:49 uniform particles very

01:28:52 uniformly. Then we had

01:28:55 deflection plates, and we tried to

01:28:58 pulse the particles.

01:29:01 Again, the picture looks better than the performance

01:29:04 Most of the time, the liquid impacted with the walls,

01:29:07 and it was better on the drawing board

01:29:10 than in reality.

01:29:13 Another important point is, as we all know today,

01:29:16 charged particles don't follow straight lines.

01:29:19 They repel each other, right?

01:29:22 We know that this is a very divergent aerosol.

01:29:25 Again, the tubes coated with the aerosol

01:29:28 and nothing ever made it into the mass spectrometer.

01:29:31 How did we heat this system?

01:29:34 We learned something from that, and I had to steal some more

01:29:37 stock metals and build another system.

01:29:40 Actually, I used to have to buy

01:29:43 the glass blowers

01:29:46 cases of beer

01:29:49 to make all these

01:29:52 chambers were made out of glass.

01:29:55 I had to beg these glass blowers

01:29:58 to make all the parts. Each one of these systems

01:30:01 took probably three to six months to have made

01:30:04 because it was a slow, tedious

01:30:07 get-in-line process. But we won the intramural

01:30:10 softball championship at Georgia Tech, so

01:30:13 there were some side benefits.

01:30:16 This was just a more sophisticated version

01:30:19 of the last induction charge device

01:30:22 with more electrodes trying to control

01:30:25 the direction of charged particles in a vacuum

01:30:28 more clearly. What you see by this is

01:30:31 you have uniform droplets, you have uniform

01:30:34 charge on droplets, you have sophisticated optics

01:30:37 trying to control the direction. We tried to pump

01:30:40 more and more, and as

01:30:43 the more sophisticated we got, the more complex

01:30:46 we got, the result was really that you

01:30:49 collected liquid further downstream.

01:30:52 You weren't really

01:30:55 doing LC-MS.

01:30:58 Just another version.

01:31:01 These are all in my thesis, if you want to

01:31:04 I put

01:31:07 failures as well as successes in my

01:31:10 thesis. Again, this was a mesh design

01:31:13 longer. In fact, I had

01:31:16 a design that was over a meter long at one

01:31:19 time, and I actually quantitated the liquid

01:31:22 from one end to the other and found that I lost

01:31:25 virtually none of the liquid into the vapor phase.

01:31:31 So I decided that the meat machine shop

01:31:34 isn't really how I was going to solve this problem.

01:31:37 I think that maybe one of the lessons

01:31:40 of this is that

01:31:43 a PhD is supposed to be a research degree, a learning degree.

01:31:46 In this process, I was learning how to do research,

01:31:49 and building apparatus is probably not the

01:31:52 first place to start. It's certainly

01:31:55 time-consuming. So I went to the

01:31:58 library

01:32:01 and started to learn about

01:32:04 aerosol physics and about thermodynamics and about

01:32:07 the physical processes involved in creating

01:32:10 aerosols. This is just another version

01:32:13 of the vacuum system. Longer, more sophisticated,

01:32:16 another six months, but no

01:32:19 LC-MS.

01:32:22 Finally, I read a few

01:32:25 papers, and ha-ha, we weren't

01:32:28 evaporating any of the solvent. I wasn't getting the

01:32:31 10 to the 4th enrichment.

01:32:34 I got this huge infrared lamp, and I figured

01:32:37 Marvin could afford a CO2 laser.

01:32:41 I went to the hardware store and bought the biggest

01:32:44 infrared lamp I could.

01:32:47 I had been to the library. I had the calculations.

01:32:50 I actually figured out the radiant flux

01:32:53 after I built it, of course.

01:32:56 I was a slow learner.

01:32:59 It wasn't enough energy to evaporate the particle

01:33:02 going 10 to the 3rd centimeters

01:33:05 per second with that amount of energy.

01:33:08 I needed a laser if I wanted to do something like that.

01:33:11 After 15 years, Marvin said that the particles move

01:33:14 when you do that anyway, so I'm glad I didn't have

01:33:17 a laser.

01:33:20 I went on, and this is actually the first

01:33:23 particle beam,

01:33:26 or we called it magic interface.

01:33:29 The real revelation

01:33:32 or insight after all those previous

01:33:35 designs was that you don't spray into a

01:33:38 vacuum. You spray into atmospheric pressure.

01:33:41 Because you have high heat transport, you have

01:33:44 orders of magnitude more heat transport at atmospheric

01:33:47 pressure, and you could evaporate all the solvent

01:33:50 at atmospheric pressure, change the phase

01:33:53 or change the state of the mobile phase

01:33:56 components into the gas phase.

01:33:59 The less volatile solutes or analytes would stay as

01:34:02 microparticles.

01:34:05 Friedlander

01:34:08 had published a paper in the late

01:34:11 60s, I believe, describing

01:34:14 aerosol beams. It was very different than

01:34:17 molecular beams, but it was clear that you could have

01:34:20 highly collimated particle beams

01:34:23 with uniform particle

01:34:26 size, and you could shoot beams

01:34:29 over very long distances and remove the

01:34:32 gases radially. So, after those

01:34:35 insights were gained,

01:34:38 it was simple to develop an interface where you

01:34:41 simply desolvated at atmospheric

01:34:44 pressure, used the natural aerodynamic

01:34:47 processes and viscous forces to

01:34:50 drive the aerosol through a nozzle

01:34:53 and get your 10 to the

01:34:56 4th enrichment of solute on the axis

01:34:59 of a particle beam and remove all the solvent vapor

01:35:02 due to radial expansion.

01:35:05 And I have a few...

01:35:08 So, that's sort of the progression.

01:35:11 And I have a few slides of the early

01:35:14 devices.

01:35:17 Thanks, Jack.

01:35:27 This is just what it was all about.

01:35:30 Electron impact was not only the goal

01:35:33 at that time, but also the only alternative

01:35:36 I had since I had an EI-only instrument with one

01:35:39 small pump. I had to make it work

01:35:42 with a low gas load.

01:35:45 This is a collection of a particle beam

01:35:48 just to show you that, in fact, the solutes

01:35:51 did make it across a vacuum. These are dry

01:35:54 solute particles. This is the first

01:35:57 chromatograph, and I

01:36:00 found it in my slides. And actually, I

01:36:03 wasn't responsible for collecting it. Paul Winkler was

01:36:06 after I left. So, although I

01:36:09 probably get credit for this, Paul was actually the person

01:36:12 who did the first particle beam LC-MS

01:36:15 at Georgia Tech after

01:36:18 I left.

01:36:22 This is the first interface. All glass.

01:36:25 This is the desolvation chamber. This is the monodisperse

01:36:28 aerosol generator. This is a particle beam interface.

01:36:31 These are all the parts that I made by hand

01:36:34 in the shop. So, if I

01:36:37 can't make a living in mass spectrometry in the future,

01:36:40 I can always get a job in a machine shop,

01:36:43 I think, after that experience.

01:36:46 Anyway, this is a Hewlett-Packard

01:36:50 5930A,

01:36:53 if you recall.

01:36:56 And there's the strip chart recorder that I used to

01:36:59 collect data. And so, maybe the lesson is

01:37:02 go to the library,

01:37:05 think a little bit before you spend a lot of time building complex

01:37:08 apparatus. And I think, you know,

01:37:11 I've developed a philosophy over the years that you

01:37:14 really should look at fundamentals before you jump into

01:37:17 costly development projects.

01:37:20 And I learned the hard way. Ultimately, we had

01:37:23 an interface that had the same ion source geometry

01:37:26 as a traditional GCMS. And that was the goal.

01:37:29 You could essentially introduce

01:37:32 solute into a mass

01:37:35 spectrometer with an EI source and

01:37:38 tune and calibrate and operate

01:37:41 the instrument in the classical manner.

01:37:44 This is just to reinforce

01:37:47 that you basically look at fundamentals

01:37:50 before you waste a lot of time building

01:37:53 systems. And that's it.

01:37:56 Thank you.

01:37:59 This afternoon

01:38:02 is familiar to us all.

01:38:05 John Fenn, and his

01:38:08 title is Electrospray Ionization Mass Spectrometry.

01:38:11 The MS tail that wagged the LC dog.

01:38:20 Bill McFadden said that

01:38:23 he found when he sat down and started to think about what

01:38:26 he was going to talk about that happened some years ago, he'd forgotten

01:38:29 everything. Well, let me

01:38:32 tell him that he's got something to look forward to because I'm a bit

01:38:35 older than he is. And if

01:38:38 he waited till he was my age to sit down, he wouldn't remember

01:38:41 anything.

01:38:44 In fact,

01:38:47 I'm at the age now where

01:38:50 the famous Supreme Court Judge Oliver

01:38:53 Wendell Holmes is my role model.

01:38:56 He was active at the bench into his

01:38:59 nineties, if you remember. And there are a number of stories

01:39:02 about him. One of my favorite ones was

01:39:05 one afternoon when he was sitting on the veranda

01:39:08 of a country club with an octogenarian

01:39:11 friend of his as they were having a drink before going into

01:39:14 dinner. This barefoot young girl

01:39:17 came running across the lawn in front of him and behind her was a young

01:39:20 man chasing her. And as they passed

01:39:23 Holmes turned to his friend and says, isn't that a lovely

01:39:26 sight? That barefoot lass

01:39:29 tripping across the greensward and behind her in hot

01:39:32 blooded pursuit, that eager young man.

01:39:35 And his friend looked

01:39:38 and thought him on it and said, well, yes, Holmes, he said, but

01:39:41 I've forgotten why.

01:39:50 Actually, it was

01:39:53 a liquid chromatographer that got me into

01:39:56 electrospray, although he had

01:39:59 been tipped off by a

01:40:02 mass spectrometrist. And so

01:40:05 I became

01:40:08 fascinated with the

01:40:11 way electrospray works,

01:40:14 how ions are formed or not formed. And I've done

01:40:17 really very little in liquid chromatography.

01:40:20 And so I have forgotten why

01:40:23 the electrospray

01:40:26 is so important in LC.

01:40:29 We've heard ample reasons why it should be.

01:40:32 So you'll forgive me if I don't

01:40:35 do honors, do justification to the

01:40:38 chromatographic half of this

01:40:41 title.

01:40:44 To my knowledge, well, I'll just tell you, it was

01:40:47 John Woods who was working for VG up at the time who

01:40:50 discovered Malcolm Dole's paper in the Journal of

01:40:53 Chemophysics, and he brought it down to Yale. That was

01:40:56 the year after I got to Yale from Princeton

01:40:59 and showed it to Sandy Lipsky.

01:41:02 And Sandy Lipsky, who was a combination mass spectrometrist

01:41:05 and chromatographer, had a young man,

01:41:08 Chaba Horvath, working with him. Chaba Horvath was also

01:41:11 spending part of his time in the chemical engineering department.

01:41:14 And Sandy showed this paper to Chaba.

01:41:17 This was Malcolm Dole's first paper on electrospray.

01:41:20 And Chaba noticed that my name,

01:41:23 Dole had been kind enough to reference

01:41:26 some of our early papers on free jet

01:41:29 expansions and acceleration of heavy molecules by light ones.

01:41:32 And so Chaba said,

01:41:35 Hey, John Fenn has just arrived at Yale. You ought to come over and talk to him.

01:41:38 So Sandy Lipsky came over and showed me Malcolm Dole's

01:41:41 paper. And I got

01:41:44 kind of excited about it because to me at that time it was another

01:41:47 application of these supersonic free jets, these big leaks

01:41:50 into vacuum systems, which I'd been having a love affair with for

01:41:53 then many years. And in fact,

01:41:56 I had a new young graduate student who had just come in.

01:41:59 This was 73 or so.

01:42:02 And we had bigger vacuum pumps than Dole did

01:42:05 and understood a little bit more about free jet expansions

01:42:08 than he did. And so I asked this young man,

01:42:11 Mike Lebowski, whether he'd be willing to repeat Dole's

01:42:14 experiments. And so he did. He built a system

01:42:17 and we repeated Dole's experiments.

01:42:20 But then we quit and he quit pretty much for the

01:42:23 same reasons. Dole was convinced, and so were

01:42:26 we, that if we got the big ions that he thought he was getting,

01:42:29 that

01:42:32 electron multipliers, these big ions

01:42:35 would not produce secondary electrons when they struck a dyno

01:42:38 unless they were accelerated up to half a million volts.

01:42:41 This is what Lou,

01:42:44 what's his name, and Bob Buehler

01:42:47 and Lou Friedman down at

01:42:50 the University of New York at

01:42:53 Brookhaven had discovered

01:42:56 with their big water clusters. And we didn't want to,

01:42:59 he was in a hotbed of accelerator builders, so

01:43:02 half a million volts was nothing to him, but we didn't want it around our lab.

01:43:05 And so we gave up because

01:43:08 if you've ever used a vibrating reed electrometer, you know

01:43:11 it's a terrible machine to have to humor.

01:43:14 So we forgot about it, and then some years later

01:43:17 when a young Japanese chap came to my lab,

01:43:20 Mashimichi Yamashita, and wondered

01:43:23 what he ought to do some research on. He was going to visit from

01:43:26 Japan for a year. And I said, let's go back and take a look

01:43:29 at electrospray again. And instead

01:43:32 of using great big molecules like Dole was

01:43:35 trying to do, we'll use little ones. Because we had a little

01:43:38 quadrupole mass spectrometer, and we said we can analyze these little

01:43:41 ions if we get an interquadrupole mass spectrometer. And so

01:43:44 Gatto, as we called him, put the system together that we

01:43:47 started all our work on, and so I've been having a love affair

01:43:50 with these charged droplets ever since.

01:43:53 To get back to where

01:43:56 LC and ES get together,

01:43:59 the first reference to this combination

01:44:02 that I know of was in fact by Dole himself.

01:44:05 How do I get the first slide?

01:44:17 Of course, there's the man that started it all. This is Malcolm Dole.

01:44:20 I almost went to graduate school

01:44:23 at Northwestern just after he'd arrived there

01:44:26 as an assistant professor, but for a number of reasons

01:44:29 which I won't tell you with now, I went to Yale instead. I'd gotten an offer

01:44:32 from both places. Mainly the thing that

01:44:35 made my final decision for me was

01:44:38 that the treasurer of Berea

01:44:41 College where I'd gone to school

01:44:44 was taking his son up to be a freshman at Yale that fall

01:44:47 and he offered me a ride from the middle of Kentucky

01:44:50 up to New Haven with all my luggage

01:44:53 and that was the straw that

01:44:56 delayed my meeting Malcolm Dole for 50 years.

01:44:59 I didn't meet him at all until an ASMS meeting

01:45:02 in San Diego in 1985

01:45:05 and we had a delightful time, an encounter

01:45:08 with each other and kept up some correspondence until

01:45:11 he died. And he was

01:45:14 also the first guy that put LC

01:45:17 and electrospray together.

01:45:20 Not in quite the same way that we've been

01:45:23 speaking about because he used

01:45:26 I shouldn't say LC, I should say chromatography

01:45:29 he used chromatography to analyze

01:45:32 his electrospray ions instead of using electrospray

01:45:35 ions to analyze his chromatographic

01:45:38 peaks. You may remember that

01:45:41 after giving up on the mass spectrometry

01:45:44 mass analysis part of his

01:45:47 ion work, he used what he called, what was then known

01:45:50 as plasma chromatography in which today we call ion

01:45:53 mobility. So if I may be permitted

01:45:56 a pun, Dole was not only first with electrospray

01:45:59 but he was first with electrospray and chromatography.

01:46:02 But

01:46:05 here is the device

01:46:08 which he used

01:46:11 I guess, is this thing dead?

01:46:21 Oh, it's on here.

01:46:24 It's okay now?

01:46:35 Is that alright?

01:46:40 It was a vertical machine

01:46:43 here's his electrospray

01:46:46 rising up, being lifted by the

01:46:49 special gradient along here. This was a chamber

01:46:52 which he could sweep his nitrogen through

01:46:55 and in all his work, as those of you who have looked at his papers

01:46:58 know, he always had his

01:47:01 bath gas. He was the one, the thing that Dole really

01:47:04 recognized was he needed the enthalpy of

01:47:07 a bath gas to provide for

01:47:10 evaporating the solvent from the droplets.

01:47:13 People had been trying to inject electrically charged droplets

01:47:16 into vacuum systems for a long time in what is known as

01:47:19 electro-hydrodynamic ionization

01:47:22 and they always got into trouble because

01:47:25 there was nothing in the vacuum to evaporate the solvent

01:47:28 and so it didn't evaporate. And Dole recognized the importance

01:47:31 of needing this bath gas. So he always used bath gas

01:47:34 and it was atmospheric pressure. But he always had it concurrently

01:47:37 because he didn't

01:47:40 understand and appreciate, he never fooled around with

01:47:43 free jets at all, that the adiabatic expansion that takes place

01:47:46 in this free jet cools things down and brings about re-solvation

01:47:49 if you don't have all the solvent vapor out. And there was no provision

01:47:52 in his way, his system here, to remove any solvent

01:47:55 vapor. So all his ions were

01:47:58 pretty much solvated. They went on

01:48:01 up into this aperture here, into his drift tube

01:48:04 and from there, that's what he was using to

01:48:07 analyze these ions, beginning about 1973

01:48:10 after he moved to Baylor from Northwestern, having retired

01:48:13 from Northwestern.

01:48:16 And it was at that ASMS

01:48:19 meeting in San Diego that I was telling you

01:48:22 about that I recognized Dole

01:48:25 pointed him out in the audience and asked him to stand up

01:48:28 and take a bow. And this

01:48:31 apparently pleased him greatly because he later sent me a copy of his

01:48:34 autobiography and

01:48:37 a very nice little book which I commend to you called

01:48:40 My Life in the Golden Age of America.

01:48:43 And the fact that he was

01:48:46 recognized at that

01:48:49 meeting really gave him a big thing. He labeled it as a very

01:48:52 remarkable event when he was down there and that happened.

01:48:55 And he also referred to the fact that after that session

01:48:58 was over, what he called the eminent mass

01:49:01 spectrometrist, Professor Paul Cabali

01:49:04 came up to him and said,

01:49:07 Professor Dole, isn't it nice that you're still alive

01:49:10 while electrospray that you started is beginning to

01:49:13 get some attention. Fortunately, Dole

01:49:16 lived long enough to go to the first electrospray workshop

01:49:19 which was run by ASMS up in Chicago

01:49:22 and by that time there was a real

01:49:25 good sized crowd of people working in the field and so he had

01:49:28 the satisfaction of realizing that

01:49:31 this work, which was really a very minor portion of his

01:49:34 research effort that he didn't count as all that

01:49:37 important at the time, had

01:49:40 such far-reaching effects. I don't know what he'd think

01:49:43 if he knew today

01:49:46 I think there were some close to 800 papers

01:49:49 published involving electrospray one way or another in

01:49:52 1996.

01:49:55 The next slide shows some of his

01:49:58 plasma chromatographs.

01:50:01 This is the pure solvent. This is what he obtained

01:50:04 with a mixture of polyethylene glycols

01:50:07 from a mixture of 51,000

01:50:10 and 250,000 and the thing that bothered

01:50:13 him was, here's the 51,000, here's the 200,000

01:50:16 and the drift

01:50:19 time for the 51,000 is longer than for the 200,000

01:50:22 and it just

01:50:25 didn't make any sense to him and it was

01:50:28 if you realize the difference between

01:50:31 this spectrum here

01:50:34 and this spectrum here, this is 25 degrees. When he

01:50:37 raised the temperature of his bath gas to 75 degrees

01:50:40 this is what he got. When he went to 125 degrees this is what he

01:50:43 got. Practically everything he was seeing in his peaks

01:50:46 were solvents

01:50:49 droplets

01:50:52 probably

01:50:55 recondensing around the ions

01:50:58 during the free jet expansion.

01:51:01 Well, in the right way around

01:51:04 the first record that I found

01:51:07 in which an LC was hooked

01:51:10 to an electrospray apparatus is in this patent

01:51:13 by Wade Fite in 1980

01:51:16 and it's a fairly complete patent.

01:51:19 Here's his design. I don't know, is this in the way so you can't see it?

01:51:22 I hope not. Here is his design

01:51:25 built on Dole's system

01:51:28 and here is his electrospray needle

01:51:31 spraying in an aperture into the vacuum system

01:51:34 and a quadrupole mass spectrometer

01:51:37 and

01:51:40 Ross tells me, he did build this equipment

01:51:43 and he did get some ions with it but it was kind of doomed to failure for the same reason

01:51:46 that Dole had his problems. There's no provision

01:51:49 for providing enough heat

01:51:52 to vaporize things and a means of heating enough

01:51:55 so that you get the temperature high enough so that recondensation

01:51:58 can occur or for getting rid of the solvent vapor.

01:52:01 And so it was doomed to failure

01:52:04 and nothing much ever

01:52:07 did come out of it. It was sometime

01:52:10 after that

01:52:13 you can see this is essentially the thing

01:52:16 that was in the first slide and then he was showing the possibility of adding

01:52:19 stage pumping here but otherwise things are much the same

01:52:22 and he also had a variety of

01:52:25 arrangements for his electrospray needle

01:52:28 and he took the trouble, this is Wade Fite

01:52:31 showing various kinds of shapes that you could make the injection

01:52:34 needle at its end and

01:52:37 with some implication that perhaps one of them would work better than the other

01:52:40 although there was no evidence or no data in the patent

01:52:43 to indicate that he'd ever done any

01:52:46 real experiments.

01:52:49 This paper has

01:52:52 liquid chromatographs and electrospray

01:52:55 in the title and

01:52:58 there's a lot more chromatography

01:53:01 in the title than there is in the paper.

01:53:04 Because the last paragraph

01:53:07 in that paper pointed out

01:53:10 that we realized because we had

01:53:13 discovered by then that you had to have pretty low flow rates in order to get

01:53:16 a good electrostatic dispersion of the liquid

01:53:19 and that meant micro-bore

01:53:22 chromatography and those things weren't very common now but we

01:53:25 borrowed what was called a micro-bore column from somebody

01:53:28 it was fairly thin, I don't remember what it was, and hooked it up with

01:53:31 an injection valve

01:53:34 which we could squirt in a sample

01:53:37 of 20, I think it was

01:53:40 20 or 30 microliters

01:53:43 and the column was such that the elution time

01:53:46 for the one peak that we had, we only put one column in

01:53:49 was 52 minutes

01:53:52 so we forgot about that

01:53:55 but this paper was the first one in which

01:53:58 what's now become familiar to all of you, our version of the

01:54:01 electrospray ion source with the counter current gas flow

01:54:04 to help dry things out before the solvent vapor

01:54:07 gets out of the gas

01:54:10 that entrains the ions and comes out here to the free jet expansion

01:54:13 and on into the mass spectrometer

01:54:16 that was

01:54:19 we did do a few experiments later and this is the general

01:54:22 scheme of things

01:54:25 and, oh dear, this is upside down but

01:54:28 that's alright, we did actually get some peaks

01:54:31 these were the ones that were shown for the UV detector

01:54:34 but these are the peaks that came out and these are only a few minutes wide

01:54:37 but our chromatography was just absolutely ridiculous

01:54:40 so I will hasten on

01:54:43 and here, these are the guys that, as you already heard

01:54:46 were the first ones to make electrospray

01:54:49 and chromatography really work and do bonafide chromatography

01:54:52 Jack Hennion and his colleagues

01:54:55 along with Bill Budde from the EPA

01:54:58 and

01:55:01 they got bonafide

01:55:04 peaks

01:55:07 you've already seen this

01:55:10 they were using

01:55:13 what they call a pneumatic assist

01:55:16 a blast of gas to help atomize the liquids

01:55:19 because he knew too, by that time he had found out

01:55:22 that electrostatic dispersion of liquid flow rates

01:55:25 of the one milliliter a minute

01:55:28 flow rate that everybody was using then in chromatographers

01:55:31 just plain wouldn't work

01:55:34 now the story I heard, and Jack will have to avouch for its truth

01:55:37 do I really have to stop now?

01:55:40 everybody can leave if they want to leave

01:55:43 I'm the last one, let me finish the sentence

01:55:46 you don't have to ask any questions

01:55:50 Jack, as he's already mentioned

01:55:53 got this tagger machine from Cyax

01:55:56 and

01:55:59 we had published a couple of papers on electrospray by that time

01:56:02 and Andres Bruins

01:56:05 and Tom Cully

01:56:08 had read them

01:56:11 and they were having some trouble with some dyes

01:56:14 and so they wanted to try electrospray

01:56:17 and it was very easy to do because they had this atmospheric pressure

01:56:20 now this is a story I heard Jack, you can deny it or not

01:56:23 so one week when he was away, they bootlegged it and put it in anyway

01:56:26 and when he came back, he was very impressed with the results that we got

01:56:29 they got

01:56:32 what? another one?

01:56:35 well, I don't know

01:56:38 but anyway, here it is with the offset

01:56:41 from the orifice leading into the vacuum system

01:56:44 and these are the bona fide

01:56:47 respectable chromatograms that they've got

01:56:50 here is ion spray LCMS for example

01:56:53 two monosulfated azo dyes

01:56:56 extracted from fortified wastewater

01:56:59 I don't know what it was waste from

01:57:02 but selected ion monitoring of

01:57:05 the negative ions

01:57:08 and I noticed that the flow rate

01:57:11 was 40 microliters a minute

01:57:14 now the big

01:57:17 advantage claimed by

01:57:20 ion spray as they call it

01:57:23 pneumatic assist was

01:57:26 one to use large flow rates

01:57:29 but practically all of the good

01:57:32 spectra that I've ever seen have always been taken

01:57:35 at flow rates that were down very close to

01:57:38 with pure electrospray

01:57:41 I think this was good

01:57:44 bona fide advertising

01:57:52 and actually they could get even at these low flow rates

01:57:55 which of course worked better because the charge to mass ratio was better

01:57:58 or more stable, it was very difficult to get stable

01:58:01 sprays at any flow rates above

01:58:04 5 or 10 microliters a minute

01:58:07 so as you know today everybody including you at Packard

01:58:10 can now offer electrospray

01:58:13 with flow rates up to 1 milliliter a minute

01:58:16 and the bandwagon is really rumbling down the street

01:58:19 I got off it a long time ago and maybe it's about time

01:58:22 for me to get back on. Thank you very much for your attention

01:58:25 and now you're all dismissed

01:58:38 I'd like to thank the

01:58:41 speakers today for a very informative

01:58:44 and entertaining session

01:58:47 thank the audience