Water Stains on Sand: Chemistry and Me

 

        My father, a physician by training and a professor of pharmacology in occupation, sparked my interest in chemistry.  In fact, his doctoral degree was a research degree in which he studied the chemistry and kinetics of blood clotting in the presence and absence of various anticoagulants.  He looked at the role of Calcium in particular.  So perhaps in some sense, he was a chemist himself.  As for myself, while the fascination with the chemistry set never quite expired, the attraction for experiments with electricity grew ever stronger as I entered high school as a teenager.  All my weekly allowances went into buying D-cells, at one point I had a 300-Volt power supply built entirely of flashlight cells!  By the time I got out of high school, I was firmly convinced that I would study electrochemistry some day.  Actually I had no clear conception of what electrochemistry involved except that it had something to do with my two favorite things.

 

        Academically I was never at the very top of any of my classes.  Nevertheless, my pursuit of a B.Sc. degree with honors in Chemistry and a M.Sc. degree in Inorganic Chemistry were rewarded with two prestigious national scholarships.  These were earned on the basis of nationwide competitive tests.  This was much to the envy of my peers who always left me behind in the routine examinations.

 

        I had an opportunity to work as a research scholar (Ph.D. graduate student) at the Indian Association for the Cultivation of Sciences (IACS) in Calcutta after getting my M.Sc. degree.  My research was to be supervised by Santi R. Palit, Distinguished Professor of Physical Chemistry.  In the Indian scene, IACS (where the only Indian Nobel laureate in science, Sir C. V. Raman, had earlier toiled to discover his celebrated effect) was a good place to work.  It had reasonably good supporting services (machine shop/glass shop/electronics shop) and most importantly, had a library equaled in few other Indian institutions.  Professor Palit was a singularly enigmatic and mercurial person.  In many ways, he molded the future course of my life.  Before joining IACS, he was a University professor and one of the best lecturers in the history of the graduate program in Chemistry.  His lectures were always lucid.  His ability to explain difficult concepts in a simple fashion is reflected in a number of textbooks he wrote, still immensely popular in India.  His early work was in Polymer Chemistry.  He contributed fundamentally to "end group analysis" and won many international accolades, including the Copeley Medal of the Royal Society.  His home was a mansion of impressive size.   However, half of it was made into a dormitory and he allowed needy students to live there, free of cost, meals and all.  At the time I joined the group, Palit's interest was primarily in electrochemistry:  he had serendipitously discovered that Faraday's Laws of electrolysis, the venerated laws that constitute the basis for the definition of the international unit of electric current, are likely only limiting laws . . .  true only under fairly limited conditions. 

 

He asked me to extend the experiments, which had thus far been confined to purely aqueous media, to mixed aqueous and nonaqueous solvents.  My experiments showed the same anomalies with the new solvents.  Six months into these experiments, I wrote a short report on my thoughts as to what is going on in these systems, and what may be the physical reasons behind these anomalies.  For the most part it was not well thought out and the author lacked the degree of mathematical finesse necessary to derive quantitative numbers from the crude model.  Still, the basic idea, the involvement of the solvated electron in these systems, was probably correct. The problem still remains largely unsolved and now it is much more widely recognized that a serious problem does exist with the Faraday constant.  Years of painstaking work by Harvey Diehl and the National Bureau of Standards (now NIST) have established beyond a doubt that there is a consistent and persistent difference between the physically and chemically determined values of the Faraday constant, even when the latter is measured under conditions that do not promote 'anomalous' behavior.  In any case, it took a good deal of courage for me to hand over my thoughts, written up, not asked for as it were, to the Professor.  Predictably, it came back to me, washed in a sea of red ink, both for linguistic and scientific reasons, with a note to see him.  I was mentally prepared to be rebuked for the basic mistakes in my account, but not at all ready for what he told me.  He said that the essence of my idea was good - I should really make an effort to study electrochemistry with all seriousness.  Uncharacteristically, he added that he does not have the background himself.  He suggested that I take the requisite tests and apply to U.S. Universities for graduate studies.  I did as I was asked and left IACS in less than six months hence, bound for Louisiana State University at Baton Rouge, ostensibly to work with Paul Delahay, a noted electrochemist.  My interactions with Professor Palit in my remaining days at IACS were greeted with characteristic Palitism, the terminal sentence generally being, "You are an intelligent young man, why are you so stupid?" Although I am no longer as young, the veracity of this profound question continues to haunt me in all too frequent a fashion (I must confess that I probably would like to substitute the word 'blind' for the last word in Palit's assertion, but a truly conscientious person probably will admit the truth and leave it as it is)!

I joined LSU(BR) in the spring of 1973 and was heartbroken to discover that Professor Delahay had recently left LSU for one of the upstate campuses of the State University of New York.  The department quickly assured me, however, that they were in the process of hiring another good electrochemist and so they did, in the spring of 1974.  I learned a good deal of electrochemistry, but mostly on my own and also made up my mind to get a sufficient number of courses in electrical engineering to get an official minor in my Ph.D. in that discipline.  By and large, the required course work (e.g., underwater acoustics, antenna theory) was so far removed from my desired expertise in utilitarian electronics, I decided also to take a home-study course in electronics technology offered by the Bell and Howell Schools.  Largely intended as a vocational training, I found this to be immensely useful in future years.  The relationship between my new supervisor and me was uncertain, however, and continued to deteriorate with time.  Although he had originally expressed interest in the problem I brought with me (anomalous nonfaradaic behavior), I never got to work on it.  I toyed in a minor fashion with electrogenerated chemiluminescence (A1) but mostly spent my time studying the electrochemical degradation of dinitroaniline herbicides by polarography.  In one year, I had run over 2,000 polarographic experiments.  Aside from the questionable utility of these studies, the repetitive nature of these investigations was singularly tedious and did little in terms of posing an intellectual challenge to an aspiring young scientist.  The net reward to me was authorship (third in the pecking order) of a paper (1), based entirely on my work that was to be my first scientific publication.  By mid-1975, I was in a state of utter depression.  I did not see either much meaning or purpose in the work I was doing and I found it increasingly difficult to get along with my mentor.  It finally got to a point that I decided to quit altogether and discussed the possibilities of finishing with a master’s degree with the departmental graduate advisor. 

 

Unbeknownst to me, the department was quite concerned with the eccentricities of my mentor (he headed the local section of a religious group quite far from the mainstream and later took to wearing swastika armbands on some semi-private occasions).  The graduate advisor suspected that my decision to quit might not have less to do with my inabilities than I thought.  (The university was eventually to make an early negative decision on my mentor's tenure.)  When he urged me to think about the possibility of continuing the pursuit of the Ph.D. degree working with some other member of the faculty, I told him that having been in the department for as long as I had and having worked for one member of the faculty for a considerable period, I found it singularly embarrassing to inquire with others if they would agree to have me in their group.  Professor Baddley, the graduate advisor, asked me, nevertheless, if there were anyone I would want to work for and asked me to postpone the decision about quitting.  I gave him only one name, that of Professor Philip W. West, knowing fully well that he had no particular interest in electrochemistry.  I like to think that my choice was dictated by the fact that West had devoted virtually his entire career on trace environmental analysis, an increasingly important area for mankind to continue an intelligent technological existence, rather than the fact that I had West for two graduate courses by then and was at the top of the class in both.  The latter may have influenced West’s decision, however, and Baddley obviously contacted him about my plight.  I received a note to see him and he told me in person that he would be happy to have me in his group.  I was entrusted with a challenging task:  the specific determination of airborne sulfates and aerosol sulfuric acid.  Considerable concern existed at the time in much of the United States on the aftermath of the mandatory installation of oxidative afterburners (referred to as "catalytic converters") in all new automobiles.   Specifically the issue was that the sulfur present in the fuel, and thus in the exhaust gas as sulfur dioxide, will be efficiently oxidized to sulfur trioxide.  Submicron size aerosol sulfuric acid is generated instantly as sulfur trioxide contacts ambient air and constituted (at least, what was then perceived to be) a major hazard.  Phil had considerable expertise and interest in the measurement of atmospheric sulfur compounds; the West-Gaeke method of measuring sulfur dioxide (this made a clever use of the Schiff reaction, among the oldest, best-known and much-used reactions in all of Chemistry) was by then, for nearly two decades, the most commonly used method worldwide for routine measurement of atmospheric sulfur dioxide.

 

Others before me had worked on the aerosol sulfuric acid/sulfates problem, developed techniques of varying degrees of complexity and success and left the group with their degrees.  My first contribution was to improve upon the synthesis of the analytical reagent being used (2) and to learn, in turn, a great deal of laboratory technique.  My solution to the sulfuric acid problem (5) was to be simpler than the previous ones (but in retrospect, still tedious) and probably more importantly, provided an extremely sensitive technique for measuring sulfate (3) even with simple inexpensive equipment (4).  The analytical problem was solved fairly early, by a color-producing reaction discovered in a moment of serendipity, but it took much longer to establish the structure of the colored product.  I had the good fortune of having advice and assistance of two excellent organic chemists and thanked my stars for the polarographic training, uninspiring as it was, which ultimately allowed me to solve the problem (7).

 

        If Palit had molded my thoughts, Phil put his own brand on it with no uncertainty.  While by no means an intellectual lightweight (one of the early distinguished professors at LSU and longtime editor of Analytica Chimica Acta), Phil taught me to be pragmatic and utilitarian in an indelible manner.  I wrote eventually in the preface of my dissertation of my interminable debt to him in Dylan's words:  "he said I'll give you a shelter from the storm."

A memorable exchange occurred one day as I was sitting at my lab desk.  I was trying to concentrate on the latest mass spectral results on "my" compound carried out by a friend at the University of Houston, as Phil happened to come by.  I felt enormously frustrated since it didn't make much sense (as it would turn out, the compound contained four nitro groups substituted in an aromatic ring system and as a result, merrily went.  . .  pfft . . .  in the heated probe of a mass spectrometer!). 

 

In reality, it was much more than that.  Phil was always concerned about the well-being of everyone in the group and he perceived that I was feeling blue.  The monologue went something like . . .  "Are you feeling blue Sandy? “(this nickname, originally bestowed upon me by my Venezuelan roommate, Carlos Brito, was firmly entrenched by now) – “Some analytical chemists feel blue because what they really want to be are                P-chemists.  Some P-chemists are depressed because they really feel like being physicists.  Some physicists are unhappy because they want to be mathematicians.  Some mathematicians are unhappy because what they really want to be are philosophers.  And all philosophers are unhappy " . . .  He said, "My boy, they'd all be happy to get jobs as garbage collectors!"

 

        Many problems in science turn out to be different from what one initially envisions, the sulfuric acid exhaust problem from automobiles turned out to be easily solved by putting the catalyst on a honeycombed matrix of basic alumina that captured the sulfur trioxide.  Also, ammonia in human airways was found to neutralize effectively much of the inhaled acid.  The determination problem would resurface again for ambient measurements as awareness increased about the general problem of atmospheric acidity and occurrence of acid aerosols (A2).  In the ambient atmosphere, sulfuric acid and other sulfates could presumably exist both as external and internal mixtures.  The method I had developed could specifically determine sulfuric acid aerosol only when it was not present as an internal mixture (A3). 

 

Later evidence showed unfortunately that my premise, namely that of sulfuric acid existing in an external mixture with other sulfates, was erroneous.  However, Phil was sufficiently impressed with the work I did to persuade the department that I would be the perfect person to hire following the formal award of my Ph.D. in 1977 for the position of an instructor then unoccupied.

        During my tenure in this position, I conducted some exploratory experiments with the aid of Pam Mitchell, an undergraduate, to understand how transition metal ions, originating from a variety of industrial emissions, affect the transformation of atmospheric sulfur dioxide to sulfuric acid (6).  The work disputed some established dogma and resulted in less than harmonious exchanges in the literature (A4).

 

At the same time, Phil was greatly interested in developing simple inexpensive devices that can be worn by workers to monitor noxious gases in the workplace.  I had an idea to measure chlorine by making eosin from a bromide-fluorescein mixture, and it turned out to be greatly successful.  I was a designated author of the resulting paper (8) even though I myself did not do any of the work.  The person who carried out the work is Jim Hardy, now a professor at the University of Akron and quite well known for his contributions to Internet based chemistry education (http://ull.chemistry.uakron.edu/analytical/http://ull.chemistry.uakron.edu/analytical/) .

 

Meanwhile the work (6) referred to earlier generated a good amount of debate and it also created an unexpected opportunity:  A large multidisciplinary research program at the University of California at Davis (UCD) was looking for a research chemist to undertake precisely the work I just started at LSU. I had no second thoughts about accepting the position at UCD.  It brought me the exposure to a new world:  the multidisciplinary study of inhalation toxicology --the generation and characterization of aerosols and an understanding of their properties, large-scale dilution systems and animal exposure investigations, respiratory physiology and smatterings of anatomy, biochemistry and immunology.  Some of the aerosol generation and chemical characterization work was highly challenging (9) and helped to unravel the pitfalls previous workers had suffered from (10). 

Most importantly, we performed the first definitive work on the respiratory toxicology of sulfite aerosols, which mimic the behavior of sulfur dioxide chemisorbed on particles (11).  I avidly read the literature on atmospheric sulfur compounds and took part in discussing the prevalent issues of the day in the appropriate forum (A5-A8).  By far the most satisfying work from my stay at UCD, however, was a new way to measure atmospheric sulfur dioxide.  By taking a long hard look at the West-Gaeke method, we developed a new technique that did not use toxic or corrosive reagents and permitted considerable leeway in the interval between sample collection and determination (12). 

Perhaps equally importantly, the mechanism of the Schiff reaction, by now more than a century old, was finally deciphered.  Ironically, this method would become the first real competitor to the West-Gaeke procedure and challenge its thirty-year reign.  At the present time, the method we developed has become the reference method in a number of countries.  The National Center for Atmospheric Research (NCAR) in Boulder automated it for routine determination of the dissolved sulfur dioxide content of atmospheric water.  The success of this chemistry in the hand of NCAR scientists would lead eventually to many visits to Boulder and much fruitful cooperative work.  In our work at UCD, we had chosen a very dilute solution of formaldehyde to collect and stabilize atmospheric sulfur dioxide; it provided excellent stabilization.  In fact, more detailed work in the future would reveal (46) that we originally underestimated the quantitative stability of the adduct by a large margin.  Nevertheless, it was soon obvious that this adduct formation must occur in atmospheric water (e.g. cloud droplets) as well (A10), in as much as formaldehyde and sulfur dioxide are both highly soluble and atmospherically ubiquitous gases.  This reaction (46) is a vital component of any present model of atmospheric reactions designed to explore the genesis of atmospheric acidity.  In any case, the new method for measuring sulfur dioxide turned out to be a lot of fun:  we found it suitable for routine measurements (13), were able to change the reactants to permit a more sensitive fluorometric, rather than colorimetric, assay (14) and were able to utilize it to measure traces of sulfite in a great variety of materials, including ashes from the Mount St. Helens eruption (15).  The last work also brought me to learn electron photomicrography with someone who was a true artist of her craft and this was an unforgettable experience.  Meanwhile my original work (6) that started the controversy was vindicated by a group of researchers from New York University (A9).  Although the debate would not end quite yet (A15), the facts by now were all but indisputable.  The success of the analytical work involving stabilization of sulfur dioxide with formaldehyde also made it worthwhile at this point to explore quantitatively the observed stability parameters for both this and a previous stabilizer system, from a fundamental standpoint of reaction kinetics (17). 

 

Meanwhile, an inhalation toxicologic study of ammonium persulfate (18) won the annual award of the Society of Toxicology and Applied Pharmacology as the outstanding toxicology study of the year.  It is fair to say, however, that the Lion's share of that credit belonged to Jerry Last, the first author of the paper.   Working with Jerry was never boring; he was always a riot, with a sense of humor I have not encountered since.

During my years at Davis, I began to develop a strong attachment to the environmental engineering program there and started teaching both aquatic and atmospheric chemistry (at both undergraduate and graduate levels) as an adjunct professor.  I developed some friendships strong enough to last a lifetime in the process but soon realized that if I wanted to pursue an independent academic career in chemistry, I should be in a chemistry department and not forever try to get engineering students to do chemical research. My interest in toxicologic research was also on the wane, I was esthetically unhappy with studies where whatever the conclusions were, they had to be defended largely on statistical grounds.  There were a number of openings in faculty positions in Analytical Chemistry as I started looking and I accepted an offer from Texas Tech University over that from another institution, a choice that I have never regretted since.

 

One of the legacies I brought over from my UCD days was my nearest and dearest friend, and someone who was, very newly, my wife.  A second legacy was an increasing affinity for ion chromatography, a technique invented by Small, Stevens and Bauman at the Dow Chemical Company in 1975.  The invention was commercialized by 1977 and in another 10 years would establish itself as one of the fastest growing methods of ionic analysis.  By 1979 I had one of the early models at UCD.  Of course, the first thing I used it for was analysis of sulfur compounds (16).  By the time I left in 1981, I was totally addicted to the possibilities of trace ionic analysis by this technique.

 

Except for a minor discussion (A14) relating to the atmospheric oxidation of sulfur dioxide, the 1983 publication year was barren, it took me some time to get adjusted to my new location and duties and do productive research after I joined TTU in mid-1981.  The calendar year 1983 was much better, however; after considerable initial frustration with granting agencies since coming to TTU, I got my first major independent extramural research support, from the U. S. Environmental Protection Agency.   Also, this year turned out to be particularly productive for my graduate students and me and is appropriately reflected in the publication year 1984.

One of the nagging problems in ion chromatography was that it used an ion exchange resin-packed column.  The resin changed from one ionic form to another during use and as such was totally expended in a few hours.  The column then had to be taken off-line and regenerated before it could be reused.  I had an idea that instead of using a resin-packed column, one should be able to use an ion-exchange membrane tube in a manner that will serve the same purpose.  Moreover, unlike the resin column, the membrane tube could be continuously rejuvenated by an appropriate solution flowing on its outside.  Although it is hard to say who had the idea first, some good people from Dow beat me to the punch.  However, in the end, the device I produced was better than theirs because making a continuously rejuvenated membrane-based ion exchanger was only half the story - the other half was to do this in the minimum volume possible.  My design involved putting in a very tightly fitting nylon fishing line inside the membrane tube and then coiling the assembly into a small diameter helix and thermosetting the coil by heat. The device proved to be far superior to the Dow product in all respects (20). More importantly, I provided a detailed theoretical treatment of the system such that in the future, performance could be assessed simply from the known parameters (19).  One of the best things that came out of this is an unreserved accolade that I got in the mail from Tim Stevens, one of the original inventors of ion chromatography at Dow.  Tim later came to Lubbock at my invitation to a regional meeting of the American Chemical Society, primarily to hear the research presented by our group (A17-A23).  This was the beginning of a wonderful friendship.  Over the years, Tim was of crucial importance in making me understand the nature of analytical problems faced by the chemical industry and in showing me the interesting and exciting aspects.  At the same time, he convinced Dow management that it is worthwhile to build formal ties with our research group.

 

In any case, the development of the helical membrane suppressor was an important event in my career.  To this day, we continue to use direct descendants of this device in our research.  The device was patented and licensed by TTU to Wescan Instruments in California (P1, P2).  The new suppressor led to an entirely new way of doing gradient ion chromatography, allowing separations not previously possible (22).

A paper completely unrelated to anything I had done until this point dealt with unusual and heretofore unknown effects of multipath light propagation in an optical cell (23) and this rounded out the fourth of my solo papers published in the most prestigious journal of the analytical sciences, Analytical Chemistry, in 1984.  Publication wise, this was a banner year for work done solely by myself in the laboratory.  I have never yet had again the time to myself to repeat that performance.  I had three more solo papers published that year in other journals, a total of seven.  One resulted from a search for an easily demonstrable example of a consecutive reaction for my undergraduate class.  I wanted something where I could readily show them how one colored product is first formed and then rapidly transformed into another (25).

 

One solo paper of 1984 was the result of the visit of a persistent salesman.  He wanted to sell me a general-purpose liquid chromatography detector that he claimed to detect density differences.  I thought his claims to be absurd and told him that I had no intentions of buying it but if he wanted to leave it with me for six months, I would tell him exactly how and why it does work, if it works at all.  At least, I said, he would not have to give nonsensical hand waving explanations to his future potential customers.  He turned out to be game for my offer.  Ultimately the device ended spending a year in my laboratory and I ended up building a syringe pump that provided pulseless delivery of liquids at pressures up to four and a half thousand pounds per square inch, because the commercially available reciprocating (back and forth moving) piston pumps produced far too much noise with this detector.  This project required the ultimate in terms of our machine shop and Jimmie Hall came through capitally.  Sadly, the visit of another salesman resulted in an untimely and premature (but temporary) demise of the pump.  While I intended to, I did not provide any safety shut off switches for the pump.  As I was working with it in the laboratory, a salesman came to visit and I was long enough gone from the laboratory to just come back and see the piston reach the end of its travel.  All the holding rods and plates were squealing and bending while the motor ran relentlessly and bent the one-inch diameter rigid steel worm screw like a spoon.  It was a deep shock.  I slowly acquired all the parts that were beyond repair and rebuilt the pump.  I should confess that I have rarely used it since; the density detector turned out to be, in fact, a viscosity detector (27).

 

The last of my solo papers from '84 may have been the most important one (24).  It involved a new non-invasive way of collecting atmospheric gases with a membrane device that does not collect the aerosols.  In many ways, we are still scratching the surface of an iceberg in terms of the potential of this technique and it has lent itself to achieve detection limits and reliabilities previously unattained by any atmospheric gas measurement instrument, particularly in terms of its very modest cost.  The concept of a tubular membrane based diffusive collector as represented by this initial work (24) was a key one.  Although it was to evolve greatly, aided by well over half a million dollars invested by the U. S. Environmental Protection Agency and the Electric Power Research Institute in subsequent years, the basic concept changed little.

 

My graduate students were productive as well; 1984 found Bill McDowell and Hoon Hwang with a nice paper apiece - Bill and I developed new chemistry using the Schiff reaction to improve further on the measurement of sulfur dioxide and recorded probably the first reliable measurements of sulfur dioxide in the Lubbock area (21).  Because Lubbock is quite distant from any major coal-fired utility and enjoys a great deal of sunshine permitting photochemical oxidation, sulfur dioxide levels here are very low, below the measurement capabilities of most instruments.  Hoon developed a very simple low-cost titrimetric method to measure sulfate, the signature of acid rain, in rain, snow and other natural waters (26).

The best and the most profound change that 1984 brought were far from Orwellian.  Sarah brought Michael into the world on an early sunny morning in November.  I became a father and life would never be the same again.

 

The work with ion exchange membranes continued at an excellent pace, with a bright young undergraduate named Quin Bligh, we collaborated with membrane manufacturers from New York to produce a detailed study of how permitted and forbidden (e.g., positive ions can move through a cation exchange membrane but negative ions are inhibited; the reverse is true for anion exchange membranes) ions move through ion exchange membranes and what are the governing parameters and equations for such transport (28).  With Quin and a new graduate student, Marita, we also improved the membrane suppressor in a fundamental manner.   Using a membrane tube within a membrane tube design gave us nearly twice the membrane area available for transport without increasing the internal volume.  This resulted also in an U.S. Patent, licensed in turn to Wescan (A25, 30).

 

At this point, I have not had very many graduate students at Texas Tech, partly because the department did not have an established reputation in analytical chemistry.  This made it hard to attract good students to this area.  But I was always fortunate to get some bright students and it has been a pleasure to work with them.  Hoon Hwang, from Korea, was my first Ph.D. student and I was indeed fortunate to have someone like Hoon as my first student.  Hoon and I co-authored five full-length papers in 1985.  The first of these established the thermodynamics of the hydrogen peroxide-water system and provided a method of generating standard peroxide gas at the sub parts per billion level (29). The physical constants measured in this work have been since verified by others and are used in virtually all atmospheric model calculations related to the production of atmospheric acidity.  Hydrogen peroxide is a key atmospheric oxidant responsible for the atmospheric oxidation of sulfur dioxide to sulfuric acid.  This reaction takes place in cloud droplets.  We were among the first researchers to become interested in the role that peroxide plays - in another two years, half of all atmospheric chemists will be trying to measure hydrogen peroxide at trace and ultratrace levels.  As the several peroxide-related papers were published (32,33,37), I was at first surprised why there were so many requests for these papers from medical schools, across the world.

 

I was to slowly learn that a whole host of physiologically important compounds (in blood and other matrices) are actually measured via measuring peroxide.  The strategy involves the reaction of oxygen with the substance of interest, e.g., glucose or cholesterol, in presence of the appropriate oxidase enzyme, e.g., glucose oxidase or cholesterol oxidase.  This results in an oxidation product and an exactly equivalent amount of hydrogen peroxide.  The peroxide is thus measured as a surrogate for glucose or cholesterol.  Even the routinely available Glucostix or Clinistix paper strips for measuring glucose in urine or blood use this principle.  One of the peroxide papers (32) also involved an interesting flow configuration - what we called a "nested loop," borrowing the term from computational algorithms.  We showed the utility of this concept by the ability to measure, in a single injection, two different analytes that otherwise react in the same fashion.  We also pointed out many other potential uses of this configuration, little did we know then that Hitachi Ltd. in Japan had applied for a patent on the same configuration 3 months before our paper came out.  While it eventually made Hitachi a good deal of money in the form of their flow analyzer, it was still gratifying for me to see the wide use of this concept.   Another work with Hoon that came out in 1985 is a spectrophotometric procedure for measuring sulfate (34) and we looked at the seasonal sulfate content of various playa lakes in the Lubbock area.

 

Meanwhile, my other graduate student at the time, Jae-Seong, also from Korea, and I presented a series of expositions on the merit of measuring the width, rather than the height, of a response peak in continuous flow analysis systems (35,36).  This technique was shown to provide a much greater applicable dynamic range than that available from the measurement of response intensities.  Earlier Jae-Seong had spent nearly a year on a very difficult project, making measurements at trace levels in a manner that required total absence of oxygen.  It was painstaking and difficult work and the results were gratifying:  we were able to reliably measure some of the physical constants relating to the dissolution and ionization of sulfur dioxide in water not previously possible (31).  By far the most important change in the orientation of my research that took place as 1985 rolled along is a firm belief in the merits of continuous flow analysis (31,33,35-37).

 

In the life of a university professor, there are few pleasures greater than doing enjoyable work with good enthusiastic people.  In visiting scholars, Shen Dong (from the People's Republic) and Vinay Gupta (from India), I found such souls.  In 1986, Shen Dong and I extended the porous membrane based ultratrace gas standard generation and thermodynamic characterization earlier developed by Hoon and me (29) to ammonia-water and formaldehyde-water equilibrium systems (38,44).  The porous membrane based Henry’s law equilibration technique has since become a benchmark in trace standard gas generation and for thermodynamic characterization.  Ammonia is the principal atmospheric gas that is basic and capable of neutralizing atmospheric acidity.  It is almost entirely biogenic.  Formaldehyde is produced both from natural sources via the photochemical oxidation of atmospheric methane gas (generated by bacteria) and man-made sources, e.g., the oxidation of certain unsaturated hydrocarbons present in automobile exhaust by ozone.  The same porous membrane based technique was used by Shen Dong to establish firmly the physical constants involved in the adduct formation between sulfur dioxide and formaldehyde in atmospheric water (46).  The physical constants derived in these studies have also become well-used property of all atmospheric modelers.  Vinay Gupta extended the technique of measurement of dissolved sulfur dioxide in atmospheric water by the Schiff reaction to automated continuous flow analysis and introduced, for the first time, membrane-based techniques to put reagents into a flowing stream without the need of a pump (43).  We used this technique to analyze samples for many researchers who wished to take advantage of our expertise. In particular, we measured the dissolved sulfur dioxide content of cloudwater samples collected over urban Los Angeles by a California-based research group on a number of occasions and these results helped unravel the chemistry of the cloudwater processes prevalent in urban southern California.

The membrane-based reagent introduction technique reached greater maturity in two other papers - the determination of hydrogen peroxide at low nanomolar (10^-9 M) levels and in amounts as small as a few picograms (10^-12 gram).  Hoon did a flawless job in this effort (42) and Huey Yang from Taiwan, completed his M.S. thesis with an equally commendable exposition of the membrane reactor reagent introduction systems (47).  Huey's work permitted a differential determination of various reduced sulfur compounds in a completely automated manner, with detectabilities not previously achieved.

       

Bill, Jae-Seong, and I characterized our first porous membrane diffusion scrubber.  With the fast response times and detectabilities this exhibited, it was clear by now that it was a winner.  I particularly wanted this technique to have European exposure at this point and published the paper in a British journal (39).

The membrane-based suppressor for ion chromatography reached another milestone in the form of the externally and internally resin packed dual-membrane device.  This was the result of excellent teamwork that involved a high school student over summer (Blakeley), a premed senior (Johnson) and orchestrated by Marita, a bright graduate student who, sadly, would get married and leave before completing a degree.  This device established new performance limits on the capabilities of a membrane suppressor (45).

       

In the department, thanks to my prolonged insistence, we finally had decent support in electronics in the person of Ellis Loree.  Together, we designed and built a pulse-width measuring device (40) that made the previous width-measurement based analytical systems (35,36) immensely more useful.

My continued preoccupation with the potentials inherent in the highly controlled and reproducible dispersion that can be generated in continuous flow analysis systems finally began to pay off.  After a year's worth of labor, Rathnapala Vithanage, a fresh Ph.D. from Virginia, was able to put together a highly sophisticated and extensively computerized analytical system that could generate the pH-dependent spectral characteristics of an indicator by a single injection.  Indeed, a single injection permitted the computation of the indicator dissociation constant of a pH indicator or the binding constants in a metal-ligand system for a metallochromic indicator.  These are parameters that previously required many painstaking repetitive measurements.  Of course, in science, it is the approval of one's peers that brings the final gratification; I was therefore particularly elated when Elo Hansen, the Danish inventor of flow injection analysis, called this paper (41) a 'tour de force', in his review.

 

The majority of my work demanded pushing the detection limits lower and lower.  As a measurement technique, we found ourselves increasingly using fluorescence (21, 31, 37, 42, 47), which is among the most sensitive of available methods.  Metals, by and large, are determined by atomic spectroscopy or colorimetric procedures; it seemed logical to design chemistry so that sensitive fluorescence techniques can be applied.  We carried out a particularly extensive study of how the fluorescence of metal complexes of a compound called sulfoxine is enhanced or diminished by surfactants.  Denise Phillips devoted her entire undergraduate research program to this end and she went on to carry out much more work after getting her degree.  Added to this were the efforts of Krystyna Soroka, a research associate from Poland and Dr. Vithanage.  Together, it resulted in a highly useful compilation of data for fluorescence detection of metal ions (48).

       

Dow Chemical decided at this time to fund our research program formally.  Interestingly, they did not ask me to solve any particular problem that was of immediate concern to Dow.  Rather, they suggested that I write a two-page proposal on a "blue sky" project - whatever I wanted to do.  If they liked it, they would provide the funds.  In due course the funds were indeed bestowed and since that time Dow has continued funding our research for one project or the other and even two at a time.  While I must ascribe it totally to happenstance, some of the best people I have had the pleasure to work with have spent their time in my laboratory as Dow Research Fellows.  Jamal Sweileh, a fresh Ph.D. from the University of Alberta, arrived with all the enthusiasm one could possibly expect, to work on my Dow "Blue Sky" project.  Unfortunately, the project was truly of the stargazing variety, we did not have half of the things necessary to do it and all the equipment had to be built from scratch.  As we waited for the various parts to arrive, I voluntarily asked Dow to give me a list of the then existing challenging analytical problems to see if I could have a crack at one.  They brought me back a humdinger:  simultaneous measurement of caustic alkali and phenol in concentrated process liquor, which is both highly caustic and toxic.  Our solution centered wholly on equipment designed and built in the laboratory: the phenol was measured by a color-forming reaction with the color intensity being measured by a light emitting diode-based photodetector.  The caustic alkali was measured by peak-width based conductometric titration with our pulse counter (40) providing an accurate width measurement.  While I probably would not rate it as my best contribution, the measurement technique was utilitarian and the time well spent (49).

Of course, the interval between the actual study and an account of it in print is significant; it typically ranges from one to five years.  Of the dozen publications that appeared in 1986, much of the work was carried out in the two preceding years.  The summer of '86 was the busiest one for conducting research, the diffusion scrubber based automated instruments reached their first (trans)portable prototype versions.  At the behest of the EPA and the California Air Resources Board, we spent months doing intercomparisons and extensive field studies.   Some of the results are have appeared in print (64,67), many followed later and others are still to be published.

I have never much enjoyed going to scientific meetings, especially large ones.  I was advised by well-intentioned peers as well as program officers from funding agencies that one should go to meetings - if for no other reason than just being more visible.  At the very least, one argued, you should not put two thousand dollars on the travel budget and then eight months into the year, ask that it be transferred to supplies, every year.  Well, sooner or later I had to comply.  The first time after being thus chastised I went to a major meeting, was the 1985 Pittsburgh Conference, held that year in New Orleans.  For the first time for myself, I made the important discovery that it is insane to rush continually from meeting room 29A to 53B, attracted by an intriguing title and a suggestive abstract.  I tried to find instead in the abstracts some of the people whose work I admired and made it a point to meet them.  Thus, I attended the plenary talk given by Richard Cassidy.  Cassidy had focused his attention for the last several years on one specific thing:  trace chromatographic determination of metals, especially transition metals, actinides and lanthanides, in a variety of challenging liquid matrices, of interest to the nuclear power industry.  In doing so, he reached a level of virtuoso performance on a routine basis, envied by many, but equaled by few.  Certainly, the early chromatograms from the Chalk River Nuclear Laboratories of the Atomic Energy of Canada that graced virtually all of the prestigious journals of analytical and separation sciences are in a class by themselves.  As I heard Cassidy talking about his forte, it was particularly elating to hear that he assessed high potential for the intriguing membrane reactors by "Dasgupta and his group."  I introduced myself afterward and we had a long chat and lasting friendship since:  I suppose I also can claim some minor role in his subsequent decision that he would be better off in a university setting.  Eventually he joined the Faculty of the University of Saskatchewan and spent the rest of his career there.

 

Metal ion chromatography, Cassidy style, is done in the following manner: the sample is injected on the top of a fine particle packed column while an eluent liquid is forced through the column at a pressure of a few thousand pounds per square inch. The composition of the eluent is changed in an intelligent manner throughout the course of a given separation.  Traces of different metal ions elute through the bottom of the column bed at different times.  They are visualized by mixing, with the column effluent, an indicator that reacts with the metal ions to form colored products.  This last, seemingly most trivial, part is actually the most difficult to fine-tune in practice and the attainable detection limits are strictly a function of the homogeneity of mixing on a fast time scale.  Mixing, in this case between two liquids, is far from a trivial matter; volumes have been written on it.  Mixing, especially ultrafast mixing, is of importance in the studies of many reactive fast processes of interest to chemists.  The traditional approach to doing this involves two high-speed ram-driven syringes to create two high-speed liquid streams and thereby generate extreme turbulence at a confluence point.  Some unbelievers wonder about whether such mixing, in and of itself, creates an event that affects the reaction process.  The task of mixing in chromatography had been carried out thus far by merging the two fluids involved at a tee.  The tee may be designed with various degrees of sophistication, the confluence point containing for example, fine-mesh screens to induce turbulence.  In the approach I took, the column effluent flows through a small segment of a narrow-bore microporous membrane tube while the second liquid is forced by pneumatic pressure from the outside of the membrane through the multitude of pores.  Because there are literally millions of pores and the membrane tube is of very small diameter (0.1 - 0.4 mm, the 0.4 mm dia. membranes were filled with 0.3 mm dia. Inert filament to reduce further the volume of the reactor to less than a microliter), the second liquid enters uniformly through the exterior and results in rapid and excellent radial mixing.  I sent several of our reactors to Cassidy and together we produced a paper that not only showed what good mixing means to such applications, but also examined, for the first time, the quantitative implications of the various noise sources present in such a system (50).  Such membrane reactors for chromatographic postcolumn reagent introduction are now commercially available.

 

The solution to a chemical analysis problem, I firmly believe, is most often solved best by chemistry.  The excitement that the fluorescence properties of metal complexes of sulfoxine started in our research group (48) led to a very detailed and definitive study:  we examined some 78 metallic species (counting oxidation state variations) for their reactivity toward sulfoxine and the fluorescence or quenching behavior of the resulting complexes.  This took a long-term dedicated effort by a number of people:  Denise, Krystyna, Vitha; even Brian Walker, our freshman laboratory coordinator, did a lot of the initial counting studies.  (Brian's love for chemistry has one common, essentially alchemical, theme with mine - we both love to mix this with that and watch for colors or fluorescence!).  Ultimately, these data were used to selectively determine various metals at unprecedented sensitivities using conventional, off-the-shelf equipment; many metals could be determined at subpicomole (10^-12 mol) levels, some at tens of femtomol (10^-15 mol) levels (51,62).  The work was recognized by the sponsor, the U. S. Department of Energy, as a "significant accomplishment," a designation given to less than 1% of the research sponsored by the agency.  In subsequent years, several researchers have used the fluorogenic sulfoxine chemistry to design capillary electrophoretic metal ion separation and detection schemes.

By 1987, a number of our efforts into chromatographic research began to bear fruit.  Certainly, much of this credit goes to Dr. Hideharu Shintani, of the National Institute of Hygienic Sciences from Tokyo, who was not only a conscientious, competent and extraordinarily hard working scientist but lightened the whole atmosphere in the laboratory with his unique rendition of the English language.  First, we explored the limits of gradient elution in ion chromatography and showed that detecting pH changes can be just as sensitive as conductivity measurements and in many cases, provide complementary information (53).  The second paper (60) described a different idea.  As ionic analytes eluted from the column in a background of essentially pure water, they were exchanged for highly fluorescent or optically absorbing ions with membrane-based ion exchange devices and detected the fluorescent or highly absorbing ions.  New horizons in the attainable limits of detection appeared.

 

With another valued colleague, Dr. Kazimierz Jurkiewicz from Poland, I showed that one widely held belief regarding the indirect detection mode in ion chromatography is erroneous.  This involves the assumption that in such a system one is obligatorily burdened with a high background and negative signals result from analyte elution.  We showed that it is possible, with judicious choice of a fluorescent eluent and operation in the self-quenched domain, to have a low background and positive signals resulting from analyte elution (56).

 

Although no concerted effort was ever spent on it, I have been interested in the multipath effects produced by reflective optics for some time (23).  A thought entered my mind as to the merits of performing absorbance measurements within a highly reflective enclosure, as within the 'etalon' of a laser cavity.  I performed some theoretical calculations.  To my considerable surprise, I discovered that the device should behave as a nonlinear absorbance amplifier.  This would, in principle, make possible detection of lower absorbance values, as well as extend the range of measurement to higher absorbances.  Somewhat skeptical about these results, I conferred with a friendly physicist.  He confirmed the validity of the calculations.  One of the reviewers of the resulting paper was Joel Harris, at the University of Utah.  Harris persuasively argued that since an etalon is an interferometer, I should approach the problem from that standpoint.  I had to first educate myself about interferometry and then recalculated the results.  Interestingly, either way they came out essentially the same.  In my own judgment, it is one of the more important contributions I have made (52).  At this point in time, no practical use of this concept has been made, either by me or by any other group.

       

This diversion firmed my interest in optics and thence generated the interest in the design of optical detectors.  Kaj Petersen, my new Dow-sponsored research fellow from Denmark, and I found a neat way to devise a tube "made of air."  Optical fibers can guide light with relatively low loss because they are coated with a lower refractive index (RI) exterior.  As light attempts to pass from a higher RI to a lower RI medium, it undergoes total internal reflection.  This is the same process that makes the air-water interface appear mirror-like to a fish, light from the higher RI medium (water) is reflected back by the lower RI medium (air).  In analytical chemistry, one is always trying to measure smaller and smaller changes in optical absorption.   One way to do this is to increase the pathlength of the absorption cell.  Unfortunately, beyond a relatively small length, the divergence of the incident light causes so much loss at the walls of the container, there is very little left to measure at the other end.  If the walls could be rendered highly reflective, this problem can be solved.  A simple mirror is not good enough for this purpose.  In order for the light to be reflected, it must first pass through the glass and then reach the reflective surface.  In multiple reflections, loss by passage through the glass is prohibitively high.  Chemicals put in the cell quickly tarnish a front-surface mirror.  If the cell itself could behave as a light guide, that would be perfect.  If the cell were a "tube made of air," this would be possible because light would be reflected totally by the air-water interface.  But how does one get a 'tube made of air'?

 

We found that while a free-falling column of liquid can indeed be maintained for short lengths, it is a better detector for vibrations (and what the next door radio is playing) than optical absorbance.  The solution we arrived at was a simpler one (57).  A porous membrane tube is used as the cell and it is mildly pressurized from the outside such that air forms a sheath around the liquid.  As far as the light beam probing the contained liquid is concerned, there is air around the liquid, i.e., it is a tube made of air!

 

Another nice device Ellis, our electronics man, and I fabricated was an on-tube photodetector.  The detector, mounted on rails, moved along a glass tube that served as the optical cell.  Such movement provided a very simple way to vary dispersion or residence time in continuous flow analysis systems and Jae-Seong found it to be a boon for peak-width measurement based analysis schemes (61).

 

The work on the diffusion scrubber based atmospheric gas measurement proceeded on its own momentum.  Denise Phillips worked out a way to measure nitric acid using anion exchange membrane tubes in a manner that eliminated interference from other nitrogen oxyacids and precursors (54).  The technique was not as sensitive as I would have liked, however.  More importantly, the first set of results from the automated instruments using continuous diffusion scrubbing and specific chemical reactions for fluorescence detection for the real-time measurement of gases came into print (59).  Formaldehyde was one of the gases we measured.  To do this we first needed a fast, sensitive, liquid phase method and Shen Dong did a splendid job on that (58) – we have since analyzed hundreds of samples for other research groups using this technique.

       

Of the papers that came out in 1987, probably the one piece of work that tickled my own fancy most, involved a continuous flow reactor containing enzymes.  The enzyme literature is literally awash with methods to immobilize enzymes.  We found that a simple physical entrapment procedure, in which the large enzyme molecules are contained within a tubular ultrafiltration membrane, can form the basis of a flow-through reactor.  This is performed easily, without carrying out any chemical immobilization.  We used it to contain both peroxidase and glucose oxidase and showed its applicability for the determination of hydrogen peroxide.  The direct determination of plasma glucose was the next obvious choice (55).  Predictably, the request for reprints from medical schools poured in.

       

Obviously, in a lot of the work that my students and I do, the activities of the research group, as it were, have by now acquired some momentum of its own.  It is with this perspective that I look at the final paper resulting from ten man-years of work:  "Continuous liquid phase fluorometry coupled to a diffusion scrubber for the determination of atmospheric hydrogen peroxide, formaldehyde and sulfur dioxide" (65).  These three gases play a more important role in the genesis and impact of atmospheric acidity than any other:  sulfur dioxide dissolves in cloud droplets and is oxidized by hydrogen peroxide to sulfuric acid.  But the transport patterns of gaseous sulfur dioxide and aerosol sulfuric acid once formed are quite different.  Given the same emission scenario for sulfur dioxide and some given set of meteorological conditions, the exact location of acid deposition is determined by the rate of conversion of sulfur dioxide to sulfuric acid.  At the present time, it is not certain if this process is rate-limited by the availability of hydrogen peroxide or by the availability of sulfur dioxide.  If it is limited by the availability of hydrogen peroxide, spending billions of dollars in sulfur dioxide emission control will not result in a proportionate reduction of acid deposition in regions near the source. 

 

Formaldehyde plays a uniquely duplicitous role in this saga.  Photochemically it is a precursor to hydrogen peroxide.  On the other hand, it dissolves in cloud water as well, and forms, as previously mentioned, a highly oxidation resistant adduct with sulfur dioxide.  However, the rate of formation of this adduct is much slower than the rate at which peroxide can oxidize sulfur dioxide.  Thus, if all three actors enter the stage at the same time (and a rather glistening wonderful stage a cloud droplet must be), formaldehyde just might as well not have been there.  Liken the situation to the maiden in distress (sulfur dioxide), the villain (peroxide) and the knight in (not quite so) white armor (formaldehyde).  If the knight is late, the damsel is done for.  To make intelligent informed decisions, we badly needed data on the levels of all the three analytes; both in the cloud droplets and the entrained air in near real-time.  The instrument described fills that measurement need (65).  Several European groups now use this type of an instrument to make trace gas measurements.  The behavior and reliability of our instruments are now well documented (64,67,71,72).

The diffusion scrubber (24) did indeed open a whole new way of looking at things.  Sulfur dioxide has remained a major preoccupation with me and I always try to take a fresh crack at it every time a new technique looks promising.  Sulfur dioxide can be translated into elemental mercury by a disproportionation reaction with a mercurous salt.  With an extraordinarily sensitive and specific solid state mercury detector and with the able assistance of Dr. Yoshiharu Hisamatsu from the Institute of Public Health in Tokyo and Dr. Ping Liu, a fresh Ph.D. from the University of Tokyo, we set about the business of measuring sulfur dioxide via mercury.  After we began the project, John McNerney, the president of the company that makes the mercury detectors called me up on the phone from Jerome, Arizona.  He offered to establish a "Jerome Postdoctoral Fellowship" to provide funding for the work.  This certainly did not dampen our enthusiasm. 

 

McNerney was a riotous bohemian; he is still much more at home prospecting for gold in the Nevada deserts than presiding over the boardroom.  He founded the company and invented essentially all of its products.  Looking for elevated concentrations of mercury is the high-tech way of prospecting for gold because the two often co-occur in elemental form and there is infinitely greater vapor pressure of mercury.  This is what caused him to invent his mousetrap in the first place.  The approach to measure sulfur dioxide in this manner proved successful although not extraordinarily so (84).  Later we attempted yet another translation reaction, translating sulfur dioxide to hydrogen peroxide via the enzyme sulfite oxidase, since we could determine the peroxide extraordinarily well.  Again, the gas-liquid collection interface was the diffusion scrubber.  Genfa Zhang, a visiting fellow from Shanghai (he came to visit but stayed on for ever), put his personal touches on this project and Hoon, who had accepted an assistant professor's position at a Korean University, came back briefly during the summer to assist us in this work (78).

 

By far the most interesting development that has taken place in our manifold games with the diffusion scrubber and measurement of sulfur dioxide is the direct coupling of the scrubber effluent to a chromatograph.  I asked for and received support from the Electric Power Research Institute to pursue the concept and was ably assisted in its execution by Per Lindgren.  Per was a Swedish student pursuing a Ph.D. at the University of Umeå in Sweden.  Through a special program, Per elected to do one year of his graduate research in my laboratories and the results have been good for both of us (72).  The Diffusion Scrubber Coupled Ion Chromatograph was repeatedly flown in the summer of 1988 aboard the NCAR Super King-Air research aircraft.  With a detection limit of 6 parts of sulfur dioxide per trillion parts of air, it was a breeze for the machine to put out flawless masses of useful data (97), albeit corrections must be made for low pressures encountered in high altitudes (83).  Of course, being a chromatographic technique, we could now look simultaneously at a number of gases; the possibilities became alluring.

By now, we made a sufficient impact on membrane suppressor design philosophy for manufacturers of tubular membranes to seek us out.  I was asked what characteristics of the membrane tubes are best suited for this application and attempts were made to produce the same.  Good material and good design is a hard combination to beat.  With new membrane tubes, especially extruded for us gratis, the membrane suppressor reached a new milestone --with an internal volume of less than 60 mL (about one drop of water) this device exchanges up to 16 milligrams of sodium per minute (63), a long way from the devices first made in 1981 that could exchange less than 0.23 milligrams sodium per minute and had an internal volume of over a thousand mL.  Dr. Syamasri Gupta, who put this new design into execution, came from India and she earned the love of everyone in the group by her sweet behavior.  Another postdoctoral fellow from India who came at the same time was Dr. Samir Roy.  Dr. Roy is a professor at my own undergraduate college, it was a treat to have him here.  Syamasri's suppressor paper, of which she was very proud, appeared in print (63).  Samir's work (66) addressed the question "if surfactants enhance the fluorescence of a metal ligand complex, what happens if we engineer the ligand molecule itself to be a surfactant?"

 

In the last few years, we also attracted other industrial sponsors for our work.  Shell Research and Development had wanted us to solve several highly specific and challenging problems.  The first of these solutions, measuring low parts per million levels of water in high purity nonpolar organic solvents continuously in a process stream took all our ingenuity and all the patience Jae-Seong Rhee could muster as a post-doctoral fellow.  It was solved largely due to new chemistry we invented (82).  The need for an automated syringe injector for introducing microaliquots of reagents were felt at this time and Jimmie Hall, our machinist, and I designed a very simple but effective one (85).

 

Another Shell project was carried out somewhat later by Wei Lei.  Wei and Ping came together as a husband wife team, both were from the People's Republic of China but both received their graduate education in Japan.  Wei had really wanted to be a musician when he was younger but his parents did not see this as a meaningful pursuit for a young man.  Ping was clearly in command in this family and Wei readily admitted that She was the better scientist of the two.  However, Wei always had a smile on his face and was a singularly pleasant fellow.  I figured out a way in which an analyte of interest can be preconcentrated on a sorbent bed from a sample matrix (e.g., mercaptans from gasoline on an anion exchange resin bed in OH^--form) and later eluted from the bed by an immiscible eluent (e.g., strong NaOH).  The sample matrix was an electrically nonconductive hydrocarbon.  The presence of the aqueous NaOH segment could be easily sensed by a pair of resistance-sensing wires inserted in the conduit.  Thus, a valve could switch and isolate the NaOH sample and redirect it to a purely aqueous system for performing further colorimetric chemistry specific for mercaptans in an intelligent manner (74).  Similar strategies for solvent extraction (organic sample, aqueous extractant) were later described in more detail and included parametric studies for such a system (90).

 

The detection and isolation of a nonconductive organic extractant introduced into an aqueous sample, a situation that is more typical of cases where solvent extraction is used, is a considerably more difficult task, however.  Several years would pass before we would solve this, thanks to the single-minded pursuit of the problem by Cherryleen Garcia, my lovely student from the Philippines (113).  Cherryleen was a special person to have in the laboratory and it did not take long for smart and handsome Mr. Lindgren, my student from Sweden, to figure out that she was good company outside the laboratory as well.  Before long, both were going about starry-eyed.  Cherryleen really moped and moaned when Per had to leave.  But on the other hand, she was in the lab night and day and before long her thesis was done, she was in Sweden and as the storybooks say, the church bells never pealed any sweeter.  Later, I got a picture of Cassandra Marie, in the full glory of her two-month old fat cheeks.  Who says there is no romance in chemistry laboratories?

 

Wei and I actually spent a lot of time playing with the determination of mercaptans and sulfide in various matrices of interest to Shell in different ways.  One on-line strategy of determining mercaptans in process gasoline streams was to use the technique of reverse flow injection analysis.  Here the reagent, dissolved in a suitable organic solvent, and a catalytic amount of base were simultaneously and directly pulsed into a flowing stream of gasoline.  The reagent requirement was extremely small and one could design such a system to operate for very long periods without attention (91).  Another application involved the determination of mercaptans and sulfide in strong caustic solutions that result when caustic alkali is used to sweeten sour gas; it is vital for efficient use of these scrubber liquids to keep a tab on the status of their exhaustion.  We used the well‑known methylene blue reaction; it produces a blue product (methylene blue) with sulfide and a red product with mercaptans.  Sequential detectors set at individual wavelengths monitored the signal to determine the individual components (89).

Few industrial sponsors have been as generous and as fun to work with as Dow Chemical.  However, the investment they made in us may not have been for naught.  The invention of the automated micro batch analyzer (AMBA, 68, 70) is the first step toward a true random access analyzer.  A random access analyzer is a device that can take any sample at random and perform any of a number of available determinations, in any order, on that sample. AMBA has the potential of being unusually intelligent.  AMBA can be easily adapted to perform digestions under various conditions.  Ping Liu and I demonstrated that by using •OH radicals generated by the Fenton reaction in an AMBA cavity, mercury present in various combined forms in a sample is readily converted into inorganic mercury.  A shot of sodium borohydride is next added and the solution is purged with N₂, the liberated elemental mercury is concentrated on a gold coil.  When the gold coil is flash heated, the mercury is released as a pulse and measured by our favorite Jerome mercury detector.  We first carried out manually and then later demonstrated completely automated analysis of environmental and urine samples containing ppb to sub-ppb levels of mercury, requiring less than ten minutes per sample, including digestion (79,92).  The AMBA concept was patented and licensed to Dow (P3).  Dow itself never licensed it to anyone but several years later, they went through corporate restructuring and gave most of the University patents back to the respective Universities.  A small start-up company, Bio-Array Solutions in NJ, promptly licensed the patent.  Michael Seul, the founder of Bio-Array claims that he knows of half a dozen or more companies engaged in combinatorial chemistry and PCR type work that currently violates the patent.  He says he would love to get his own company off and going and then get after all these people!  By far the biggest benefit of working with Dow was developing friendship with some extraordinary people, most notably Steve Gluck.  Steve is a singularly creative and intelligent chemist who is comfortable in thinking unorthodox thoughts and doing unorthodox things.  Generally his analytical strategies are also based on solid theoretical grounds.  I don’t know how much theoretical foundation this required, but Steve’s approach to chemistry is pretty well reflected in the style in which he acquired his house when he moved to Freeport.  He was being shown houses by a Real Estate agent in a particular neighborhood.  He did not like the house he was being shown so he told the real estate agent to wait a moment, walked up next door to the grand Victorian structure, knocked on the door and asked the owners if they had any plans to sell the house.  They did.  Our friendship really matured during the AMBA period and I will always cherish him, personally and professionally.

 

Without a shred of doubt, the biggest reward in doing what I do has been all the wonderful people I have had the good fortune to know and work with.  I don't know what the statistical probability is of picking such a wonderful group of people by chance alone.  If it is indeed merely my luck, I sure am thankful.  And of all of them, I always specially remember the very special warm people from cold Scandinavia.  From Elo Hansen's laboratory in Copenhagen came Kaj Petersen, whom I have already mentioned in connection with optical detectors based on porous membrane tubes.  Kaj is a wonderful human being and he looks at the whole world in that spirit.  Not surprisingly, the world finds him an equally kindred spirit.  Kaj came here for a year, ended up staying three.  He has been working for Perkin-Elmer since his return to Denmark (most recently in the US) and from all accounts, he was the best sales engineer the Nordic region ever has had, spreading, by his own account, ". . .  tales of FIDs with split and splitless tongues . . ." in lands near and afar!

 

The pulsed reagent introduction concept (69) matured in Kaj's hands.  This concept became integral to the air carrier continuous analysis system (ACCAS), a fast analyzer that could process 3,600 samples an hour.  The paper disclosing the ACCAS principle appeared in a special issue of the international journal Talanta honoring the pursuit of Analytical Chemistry in the U.S. (76).  Kaj also carried out an ambitious application of AMBA, measuring optical absorbance on a continuous basis while the aqueous sample is digested, under pressure, in the cavity in a medium of 50% chromic acid (93).  This is one of the most corrosive situations that can be imagined and the importance of this measurement is sufficient that the people at Dow built a prototype for their own evaluation and use.  A further equally ambitious AMBA application was carried out later by Shen Dong who was visiting the U.S. for a scientific conference and just about then the incident at Tianamen Square and all the other stuff broke loose in China.  He stayed for a while and we performed acid persulfate digestion in the AMBA cavity to convert all forms of phosphorus to phosphate.  The phosphate was then determined in-situ using the phosphomolybdate chemistry (102).

 

Obviously I am getting older; as I do so, different years merge into each other and I cannot keep track of the years and what happened when, as clearly demarcated as I could in the earlier years.  If it is a treat to work with pleasant people, it is probably even better to have someone in your laboratory who is not only a gentleman but who is, in all likelihood, a lot smarter than you are.  Tetsuo Okada was such a person.  It was a real privilege for me to have him in my laboratory for eight months.  During this time, he figured out the essence of what Dayong Qi (a visiting scientist from China) and I observed earlier in our stopped flow chronoamperometric experiments (81).  I have no shame in admitting that I did not.  The whole business originally came about partly because I love to tinker with electronics (despite the fact that I really do not have much talent either in the details of electronics design or in physically putting things together) and partly because of my parsimony.  Jim Tarter, at the time in North Texas State University, came to visit once and told me about a serendipitous finding that he had made.  When he put an electrochemical detector in a suppressed ion chromatography system after the suppressor (rather than before it as the manufacturer prescribes), he not only saw signals for every analyte ion whether they were electroactive or not (e.g., sulfate and phosphate), he claimed that the detectability was actually better relative to the standard conductivity detector.  I concluded that in the absence of a supporting electrolyte and the consequent high resistance of the cell, the system is essentially acting as a conductivity detector with an applied dc potential. 

 

Why not go ahead and build a dc conductivity detector?  So I went to Radio Shack, bought my $1.98 dual BIFET operational amplifier and built my $10 battery operated detector that indeed worked very well as long as the background conductance was low (80).  I had used, however, only half of the amplifier chip to build the detector. Being my parsimonious self (my wife would interject: read cheap), I had to use the other half of the IC.  So I built a second detector, both of them were contained within a cigarette pack size box.   Now there is only so much you can do with two identical conductivity detectors when you have only a single chromatographic effluent to monitor.  I put them in series and I put them in parallel, and they pretty much produced the same signal.  This was predictable (and boring).  I was also fooling around a great deal with fast solenoid valves at the time, evaluating their suitability for AMBA applications.  I had the bright idea of putting the two detectors in series, with one of these valves in between.  While looking at the second detector output as a peak emerged, I tripped the valve and flow in this detector stopped abruptly.  I noticed quite soon that if I trip the valve at the same current level but with different electrolytes present in the cell, the patterns of the rise and decay of the current (I called the chronoamperograms "conductograms" at the time) are quite different.  I could surmise that there must be some relationship with ionic mobility but I could not pin it down.  It was Tetsuo who figured out how the ionic mobility controlled the shapes of these conductograms (81).  Much later in 1991, another wonderful visitor from Japan, Professor Hisakuni Sato, spent a considerable amount of time attempting to improve the resolution and reproducibility through the utilization of longer migration conduits, higher applied voltages and multipoint measurements.  We improved our qualitative understanding some but basically succeeded only in establishing how complex these systems are and how much we still do not understand (136).

 

During his brief stay, Tetsuo carried out another project, although the idea in broad strokes were mine, the fine strokes and the finesse of experimentation were all his.  We were able to carry out gradient elution and generally applicable optical detection for weak acids in the ion exclusion mode, which had not heretofore been done, especially at the sensitivities we were able to attain (75).

 

Genfa Zhang came in 1988 from the People's Republic, originally at the recommendation of Shen Dong.  Within a short period, Genfa clearly established his value to the laboratory.  He is competent, responsible, and loyal and I can always rely on him.  I hope that it is possible for me to always keep him here.  As I write this a decade later, he is the most important person in the laboratory, responsible for virtually all the organizational aspects. 

 

The first major paper that Genfa and I published dealt with the measurement of ammonia/ammonium ion.  We showed that if sulfite is substituted for b-mercaptoethanol in the well‑known reaction involving the use of o-phthaldialdehyde for the determination of amino acids, ammonia is determined with great sensitivity and considerable selectivity over amino acids (73).  Later we coupled this chemistry to a diffusion scrubber and applied it for the measurement of ammonia with a detection limit of 50 parts per trillion with a few minutes time resolution (87).  Several European groups have now adopted precisely this technique for their needs.

 

From the beginning, Doug Strong was an unusual graduate student.  He was only two years younger to me, having been out and about for many years since getting his undergraduate degree.  He finally got tired of working as a technician, I suppose.  Anyhow, I developed a great liking for him immediately:  I have always been in love with poets -- and what he wrote in his younger days was very good.  Doug thought well and was good with his hands.  The first paper we published together was accepted in Analytical Chemistry (77) with rave reviews.  Membrane suppressors had brought a great degree of practicality to ion chromatography but they also brought one problem that was not present with packed column suppressors.  When an alkaline eluent flows on one side of a cation exchange membrane and an acid regenerant flows on the other side, the intended ion exchange of the eluent cation, typically Na⁺, for regenerant H⁺ does take place.  But at practical regenerant concentrations, some of the regenerant anion also penetrates into the eluent side.  This undesirable regenerant penetration increases background conductance, deteriorates detection limits (especially for the weaker acids) and degrades response linearity.  We had earlier shown that regenerant penetration could be minimized by using large and/or multiply charged regenerant counterions (28, 63).  However, I as well as several others had recognized that an electrical means of ion exchange should be preferable.  Some patents actually existed, ostensibly describing how to do this.  In reality these were valueless, largely constructed as a ruse to get around the Dow patent on membrane suppressors licensed to Dionex Corporation.  Indeed, these "electrical" suppressors only worked if you used them with acid regenerants and turned the electricity off!  It should be possible to work with just water as regenerant, with the negative electrode placed on the regenerant side and the positive electrode placed on the eluent side.  The electrical field provides the motive force for the eluent cations to be transported across the membrane, while H⁺ is produced at the positive electrode in the eluent.  Unfortunately, one also produces oxygen gas at the positive electrode in this process.  Gas bubbles and any flow through detector have a noted inharmonious relationship, they simply do not get along very well.  Doug and I solved the problem by isolating the positive electrode with a second membrane such that the eluent channel itself was in the electrical field but not in contact with either of the electrodes where gas is evolved.  We learned a lot more from this experience as will become apparent later.

 

In 1985 or so I finished a chapter on a book on ion chromatography that Jim Tarter edited.  It was nearly two hundred pages long (the whole thing was typed in my own one finger style that some people claim is inimitable, on an Apple II computer that boasted 48 K of RAM) and took me many months, not even including the time the disk crashed (fortunately, the disks held so little that this one chapter required three disks and I did not lose a great deal).  If someone told me that I will be writing another review on ion chromatography soon (Jim's book came out in late 1986), I would have attempted a style of scornful laughter that my college drama coach tried to get me to learn when I was doing the role of Pedro the pirate, but never succeeded (at least to his satisfaction).   But Jim is a miracle worker, he decided to edit a special issue of Journal of Chromatographic Science and he got me to do it again! The scope this time was limited (by my own choice) and concerned only post column reaction detection (by my own broad definition this included suppression).  Still, it required months of work and I really am not doing it ever again .  .  . (well, as long as Jim is not editing again . . .).  How good a review turns out is always a difficult thing for the author to judge but it was gratifying to hear from several people that it was very useful to them (86).

The final paper of 1989 I shall talk about is again a contribution with Kaj -- the idea was to see if we can inject compounds (the whole experiment is in the gas phase) on one side of a membrane and pick them up on the other side.  More importantly, we wanted to differentiate different compounds based on their different temporal response profiles (some will come through fast and loose, some slow and broad and so on).  The actual task turned out to be much more complex than we had originally anticipated because the response to mixtures was not a weighted linear combination of the individual components.  Probably the most significant thing that came out of it was not the experimental system but the method we developed to interpret the data.  We developed an algorithm that uses a linear model even when the overall behavior is nonlinear.  The model continually changes its choice of the calibration set (which is a small portion of the large set in its memory, the choice represents a limited space in the complete domain of possibility) based on the prediction on the sample it makes and repredicts it again (and again) until the results converge within specified limits.  Thus, we could obtain very acceptable predictions in highly nonlinear systems containing four sample components with fast converging linear models (88).  After all, everything is linear in a very small range.  A digital plotter draws circles routinely with small straight lines!

       

At the end of 1989, I reached my own human milestone, the age of forty.  Professionally, it was a remarkable year -- 20 papers were published (72-91), something I have not before or since equaled.  To be fair, I have not really made a conscious effort to exceed this; it is, at best, a record of dubious distinction.

 

The results of several intercomparison studies of atmospheric trace gas measurements using the diffusion scrubber technique were published in 1990 (94-95).  Dr. Huang Huiliang, originally from Xiamen University in China, came via the university of Göteborg in Sweden (where he got his Ph.D.) and a brief stay with Joseph Wang in New Mexico.  I had an idea earlier that the great water affinity of the perfluorosulfonate (PFSI) ionomer Nafion and its simultaneous increase in conductivity can be made into a useful water sensor.  During his stay, Per carried out a set of brief experiments that indeed produced encouraging results.  Dr. Huiliang picked up where Per left this project and pretty soon we had thin film microsensors responding to water from ppm to saturation levels in the gas phase (98).  Later we experimented with this technique further and learned how to make the sensors even more responsive, by incorporating phosphorus pentoxide in the PFSI (107).  Even later, further significant developments would be made to use the same principle to sense water in a variety of organic solvents (122) or to measure moisture levels in soil for agriculture (159).

 

The final paper of 1990 was co-authored with Osamu Nara; Osamu came to spend a year from his faculty position at the Tohoku College of Pharmacy in Japan and he really had a good time.  Considering that his musical passion was in Jazz music, Texas is a lot closer to Dixieland than Tokyo.  With Osamu's help, I showed that by making simple conductivity measurements of a salt and the corresponding acid, it is possible to measure the pK of an acid very conveniently.  While the technique can be practiced for multiprotic acids with the help of multiparametric fitting programs, it is particularly attractive for monoprotic acids where the alkali metal salt of the acid is readily available.  It requires a very small amount of sample and the sample conductivity can be measured as such and after passing through a small H⁺-form cation exchange resin packed microcolumn.  No titrations or calibrations of pH electrodes are required (96).   Professor Bartsch and his research group at TTU are experts in making macrocyclic ligands with attached acid groups; they immediately started using this approach.  Indeed, they could get so much better precision by this technique relative to painstaking titrimetric methods, a pattern in the pK values became apparent to them that was simply lost in the noise previously.

Some time back, I was very impressed by a paper in Analytical Chemistry authored by Vecera and Janak from the Czech Academy of Sciences.  I wrote to Dr. Janak (who happens to be one of the pioneers in gas chromatography) congratulating him on the nice work.  I also indicated that I would be happy to host someone from his laboratory who had worked in trace gas measurement (subject of that paper) and we can do something together.  Dr. Zbynek Vecera came to Lubbock as a result of that invitation.  Zbynek was a memorable character in many ways.  Back in Brno, he was a group leader; he did not do much with his own hands.  It took Zbynek a while to get used to cleaning up after himself, but once he got going, he went great guns, spending uncounted hours in the laboratory.  You could generally call the lab 1 AM and find Zbynek (and Huiliang) hanging out there.  When Zbynek first came to this country, Czechoslovakia was still very much an iron curtain country.  I even had visits from the FBI asking such questions as whether I knew if he hung around much with the post doctoral fellows from Poland in the Department, whether he went to Canon Air Force Base over the weekend to take pictures (this is a mere 200+ miles away and Zbynek didn't have a car) and so on.  On his part, Zbynek politely but scrupulously avoided entering discussions on political systems.  He made one very interesting comment early on.  He said, back in Brno, he liked to stand in his veranda and watch people go by.  He tried to imagine from their faces what kind of life they had and he thought he had a pretty good guessing record.  "But here," he said, "these people are smiling all the time, you cannot figure out a thing what is going on!"   As luck would have it, the whole political system in Czechoslovakia turned topsy-turvy while he was here. 

 

At first, he thought it was all American propaganda.  He took to listening to BBC.  I have a feeling that although he did call home on occasions, he did not yet feel that he was not being listened in on.  Zbynek and I concentrated on nitrous and nitric acids.  The first task was to use a porous membrane diffusion scrubber and couple it to a chromatograph in much the same manner as Per did earlier, except that here we chose to use UV detection for greater sensitivity.  We also provided for the first time a reliable contamination free calibration source for nitrous acid (99).  In carrying out ambient measurements, we were puzzled by one fact.  According to the pundits, the daytime concentrations of HONO should be very close to zero because it is easily photolyzed to OH • and NO.  Although we observed marked diurnal cycles, the daytime concentrations were persistently 0.2-0.4 ppbv, substantially above instrumental noise levels.  Later we went on to develop a wet denuder (this required nearly alchemical techniques to produce a layer of highly wettable porous glass inside a glass tube) in which we continually wetted a glass tube, by flowing water down it while air was sampled countercurrent through it.  The collected nitrate and nitrite, representing HNO₃ and HONO, were preconcentrated on a 6 mm long ion exchange column and then separated on a 10 mm long column.  Following separation, the effluent was passed through a cadmium reductor to reduce nitrate to nitrite, the well-known Griess-Saltzman reagent was added on-line to form an intensely colored azo dye and the product was measured colorimetrically.  Detection limits meaningful for ambient applications were possible even with very inexpensive light emitting diode based detectors.  Once again, we observed the anomalous daytime persistence of HONO (111).  To date, I am not certain what is going on here -- experiments with potential interferents, oxides of nitrogen, do not indicate that there is enough of an artifact to account for the observation.  Are alkyl nitrites responsible?  Very recently, the Swiss have observed the conversion of NO₂ into HONO on soot, perhaps the future would reveal more about this.

 

When I visited Elo Hansen's laboratory in Copenhagen, Kaj was already back in Denmark.  Another of his students, Kim Sonne, was interested in spending some time in Texas.  Elo's secretary told me that I have the good fortune of enticing two of the nicest fellows they have ever had around.  It was indeed a tall order for someone to come in and fill Kaj's shoes but Kim proved to be a wonderfully likable individual in his own rights.  Kim was very independent and required little supervision.  Kim's fondness for ice cream at our dormitory cafeteria lunches (I frequently ate lunch with them) became a legend in the group.  Kim's manners were the same: sweet and soft-spoken.  We completed the Shell project of measuring several sulfur anions by colorimetric flow injection analysis methods (100) and built a field deployable instrument for them.  Kim went out with the Shell crew to measure various things.  It worked out so well that Don Olson, the project supervisor from Shell, really tried to get him to spend some time at Shell.  Kim was determined, however, to get back to Denmark after his year's stint was over and he did so.

       

The original work that Doug and I started in electrical suppressors took on an important new twist, with encouragement, collaboration and funds from the people at Dionex Corporation.  During the original work on the electrical suppressor, it did not escape our attention that the sodium that we electrically transport across the membrane forms sodium hydroxide.  More importantly, the purity of this sodium hydroxide that we generated was limited essentially by the quality of water we fed in.  This is in contrast with any NaOH that can be bought or otherwise made; these always contain significant quantities of various amounts of impurities.  As a result, when using NaOH as an eluent, the conductivity after suppression is substantially larger than for pure water.  This compromises both ultimately attainable LODs and response linearity.  Overmore, if we are not necessarily operating the device as a suppressor but as an electrodialytic NaOH generator, we do not need to transfer all the Na⁺ from one side to another.  Rather, we transfer as much Na⁺ as we want by controlling the current.  Thus, varying the current could readily vary the eluent concentrations.  This concept was first demonstrated using a single membrane device in which the generated gas in the product channel was removed by a second porous membrane device (101).  In a later paper, Doug and I showed how it is possible to produce gas-free NaOH directly using two membranes and perforated plates (104).  Albeit this method has limitations in how much NaOH can be generated and it is not as current efficient as the single membrane technique, it does not need means for gas removal.  The combination of an electrodialytic suppressor and an electrodialytic generator is truly complementary; this marriage allows both techniques to attain their full potential and was described in a symposium paper (105).  The electrodialytic eluent generator was also patented (P4).  A slightly differently engineered version of this, that holds a static cation donor source was marketed by Dionex, won the silver medal at the Pittsburgh Conference and Exposition and has enjoyed considerable commercial popularity since.  Dionex coined a catchy slogan, Just Add Water!

 

Our interest in measuring hydrogen peroxide had continued over the years.  In field applications, we found it a nuisance that the enzyme solutions had to be refrigerated, not even counting the substantial cost of the enzyme.  In one study, following the lead of an earlier Japanese paper, Genfa and I found that the UV irradiation of the H₂O₂‑substrate mixture causes the same reaction that the enzyme catalyzes (103). However, the blank is higher and the product yield is poorer and so the search continued.  In looking at various substances similar in structure to horseradish peroxidase, Genfa discovered that hematin, a common inexpensive bovine blood product, is very effective and per unit peroxidatic activity, it is 500 times less expensive than the enzyme preparation we were using.  With p-cresol as the substrate, a very sensitive and inexpensive fluorometric method for peroxide results (114).  This method was subsequently coupled to a diffusion scrubber and with vapor phase introduction of        p-cresol through a membrane, permitted a very simple single flow channel system capable of detecting 5 parts per trillion gaseous H₂O₂ (116).

 

I have had students and associates from many countries but Poruthoor Simon was my first graduate student from India.  Simon had a good sense of humor.  However, initially it was most often lost on people because they found it hard to understand him.  As this wore off, he was always the guy for exchanging jokes.  Simon and Zbynek, both working independently on wet denuders (I had tried to get them together but both these souls were so adamantly independent, past a certain point I gave up) were always trying to outdo each other.  Although in this account I have mentioned Zbynek's wet denuder first (111), Simon's was actually first realized.  He used thin layer chromatographic grade silica gel bonded to glass by tetraethylorthosilicate.  His system used preconcentration columns to concentrate the wet denuder effluent.  We found out serendipitously that it is not necessary to maintain an exact liquid inflow outflow balance; small amounts of air injected into the ion chromatograph are easily dissolved under pressure and do not reappear to cause problems.  Simon achieved single digit parts per trillion detection limits for SO₂ with a time resolution of 7 min with this system (106).

 

We had done some work in the previous years to better design the membrane based diffusion scrubbers, both with regard to improving collection efficiency and minimizing particle deposition.  The latter aspect of the studies was conducted at the Inhalation Toxicology Research Institute in Albuquerque with the help of Yung-Sung Cheng.  This study showed that straight inlet scrubbers have the minimum deposition.  But it also showed that at least with dry aerosols, deposition loss is increased by PTFE surfaces because they are particularly prone to acquiring a significant amount of static charge.  A hydrophilic membrane such as Nafion proved to be superior for collecting H₂O₂ and HCHO and better detection limits were attainable relative to previous work.  Genfa chose to measure the formaldehyde levels in his home, which contained plenty of particleboards, and other wood faced paneling.  Although the panels were at least 20 years old, they were still exuding formaldehyde.  The formaldehyde levels were directly related to the air exchange rates -- the concentrations always took a nosedive whenever the air conditioner was turned on.  One important and originally unexpected finding reported in this same paper (108) was that cation exchange membranes might not necessarily be the best surfaces to collect cationogenic gases like ammonia.  The collection efficiency is excellent but the resulting ammonium cations are removed from the membrane only with difficulty, requiring high concentrations of displacing ions.  This generally results in deterioration of detection limits and poor response times.

 

Also this year I wrote a chapter for Lenny Newman's book "Measurement Challenges in Atmospheric Chemistry," a volume in the American Chemical Society (ACS) Advances in Chemistry series.  Lenny got many of us together in Boston at a symposium within an ACS national meeting, fed us a 23 course Chinese dinner at some back alley gourmet Chinese restaurant (he is known as a real connoisseur) and then in our weak moments got us to promise that we would deliver these goods for his book.  I was apologetic that I was late in keeping the promised date but then he revealed that I was still the first one to come through!  The reviews on this review were particularly gratifying; I hope that it would turn out to be of some use to others in the field (125).

 

I had the honor of being the "Fakultesopponent" (a formidable title for the external thesis examiner, the actual role does not involve an adversarial attitude!) for the Ph.D. dissertation of Knut Irgum in 1986 at the University of Umeå in Sweden.  Later we became close friends and collaborators.  Through such collaboration came the report on the enhancement of ion transport through ion exchange membranes (110).  It is also through the Irgum connection (Per's advertising did not hurt either), came my second Swede, Ingemar Berglund, the gentle giant.  (For quite a while my secretary Tonia, who is a bit of a movie buff, insisted on referring to him as Ingemar Bergmann!).  Ingemar was deeply into martial arts (Aikido) and he quickly became Sarah's special favorite; Sarah had by this time plunged herself heavily into similar stuff, her affliction being with Tae-Kwon-Do.  Ingemar and I demonstrated a two dimensional conductometric detection approach to ion chromatography (109).  Following a conventional suppressed chromatography system, the eluite (in the form of HX) is passed sequentially through a cation exchanger and an anion exchanger membrane device, both bathed externally with dilute NaOH.  The first device converts HX to NaX and the second converts NaX to NaOH, and a second detector detects this.  While the extent of conversion is dependent on the degree of ionization of HX, very weak acids produce very weak signals in suppressed ion chromatography.  Thus, borate, sulfide etc., can be much more sensitively detected after conversion.  The response linearity also improves markedly for most carboxylic acid analytes upon conversion.  The most important aspect about this technique, however, is that it enables one to utilize the signals of the two individual detectors simultaneously to produce unique information, e.g., as to the purity of the peak, pK of the eluite, etc.  Approximate quantitation without reference to standards is also possible from the converted signal.  We pursued this theme for a good bit.  The next approach in this line was to improve one weakness – weak acids are weakly ionized and therefore the HX®NaX conversion does not proceed well, especially as some conversion to NaX results in a buffer that permits the existence of a relatively low level of free H⁺.  To circumvent this, we carried out simultaneous cation and anion exchange with a dual tubular cation/anion exchange membrane device and converted HX into a salt (123).  The real breakthrough came later, though.  I realized that what I want to do is best done by introducing NaOH in micro amounts after the detector in a conventional ion chromatography system and putting in a second detector.  Osamu Nara, who came from the Tohoku College of Pharmacy in Japan, began the initial work.  Osamu was not super smart, but he was extremely thorough.  The other interesting thing about him is although his English vocabulary was quite limited, his English pronunciation was absolutely superb.  He was an accomplished Jazz pianist, it is amazing how well and without accent he sang some of the popular tunes from the 50’s.  Ingemar, however, completed the work.  A whole new micro electrodialytic NaOH generator was developed to introduce very small amounts of NaOH in a pulseless manner (129).  Anna Sjögren, who also came from the same Swedish University as Ingemar, would do the final touch on this even later.  Anna had barely finished her undergraduate degree in Sweden when she came here but she matured with great rapidity and turned out to be a smart, methodical and very thorough student.  Anna’s work will be the last in this series, at least with conventional scale systems (155).  It is really unfortunate that we did not patent this.  A further improvement came later when Anna and I demonstrated that superior results could be obtained with a more robust planar device (204).

 

Without a doubt, the greatest gift to my research in 1990 was that the department finally convinced the Dean that we needed a full time machinist and especially that Kavin Morris filled that job.  Kavin is the best machinist I have had the opportunity to work with and he is truly inventive.  In making optical detectors smaller and smaller, we were continuously facing hurdles in making tubing connections.  Kavin figured out a trivially simple way to do this - thread the tubes themselves!  In this way we have made not only connections to our detectors but also unions, tees . . .  (115).

 

Pioneered by Jim Jorgenson, capillary electrophoresis began to become a major tool in Analytical chemistry research.  This technique permits incredible abilities to separate charged substances in tubes that have bores substantially smaller than human hair.  Total sample necessary is less than 1/1000^th that of a raindrop!  Thus far, people had only analyzed liquid solutions by this technique; we showed how to analyze soluble gases (117).   This was also my first paper with Liyuan Bao, an incredibly nice guy who came from China, subsequently made a major invention with me in being able to apply the technique of suppressed conductometry, widely used in ion chromatography, to capillary electrophoresis (127). This invention was a race to the finish with scientists at Dionex but I cannot tell the details of that story publicly.  Nevertheless, it was patented and Texas Tech licensed that invention to Dow, who licensed and eventually sold it to Dionex.  The University made close to $500,000 already from that invention.  Three separate patents actually resulted from this invention (P5- P7).  One of them related to the use of the timing of the water dip time to measure electroosmotic flow and thus correct the complete electropherogram.  I never actually wrote this up in the general scientific literature.  Bao later went to work for Dionex and still works there.  Bao loved Michael very much.  Every Saturday, I would take Michael to the lab and Bao was his very best friend.  Bao was and is a truly generous person, by heart and by nature, there are few in this world like him. 

 

Originally from Korea, via the laboratory of Jim Ingle at Oregon State for his Ph.D. came Hyung-Keun Chung.  Chung, as he preferred to be called, turned out to be a really sweet, competent and very hard working individual.  With my erstwhile colleagues in the Department of Civil Engineering at the University of California at Davis, we had been selected by the American Water Works Association Research Foundation to develop a monitor for measuring residual ozone.  The first person I appointed to work as a postdoctoral fellow on this project left after a few months, leaving me in a lurch.  Chung came in and picked up the project in midstream and did so admirably.  He saw the project through the field-testing stages at Davis and also at the Los Angeles aqueduct water processing facility.  One approach proved to be interesting but we did not have the opportunity to pursue it fully within the scope of the project involved.  This utilized aqueous chemiluminescence reactions of ozone (119).  The principal approach used was different, however, and involved the measurement of the decolorization of indigo by ozone.  A time-differentiated measurement was made to compensate for any interference from permanganate.  This was compared with the nebulization and gas phase ozone measurement system developed by the group at University of California at Davis.  However, writing this up between the two groups took forever and the paper finally came out in 1995 (150). 

 

The University increasingly recognized my research efforts.  In 1991, the Texas Tech Mom’s and Dad’s Association bestowed their Rushing Faculty Distinguished research award (to my wife’s delight, that contained a substantial check as well!) upon me.  A statistical magazine, Science Watch later published a paper, ranking various Universities in the world in the different disciplines of chemistry for the seven-year period 1984-1990, by the impact index (number of times the papers were cited elsewhere divided by the number of total papers published).  In the combined area (they did not subdivide further) of inorganic, analytical and nuclear chemistry, Texas Tech was ranked second in the world, only trailing Yale.  Since we did not have much of a inorganic program at the time and no nuclear program while Yale had no analytical program, one could logically assume that we were the first in the world in analytical chemistry, as judged by the impact of publications.  It was a wonderful feeling because I knew I authored 70% of the papers that they counted for us.

 

Professor Vlastimil Kuban came as a Fulbright Fellow from Masaryk (now Mendel) University in Czechoslovakia.  We had earlier met at the Royal Institute of Technology at Stockholm where he was in residence as a visiting scientist and I came to present a lecture.  Dr. Kuban was (and is) an enigma.  Anyway, there is little doubt that he is a good scientist and we were able to do some really useful work together on the determination of traces of sulfide in wastewater using a silicone membrane interface at the front end.  The really important thing that happened through this work, however, was that we did a thorough study of the methylene blue and nitroprusside reactions, both used for measuring sulfide for more than a century.  With the help of John Marx, we were able to do away with a considerable amount of mystique about these reactions in regard to reaction mechanisms and the exact products involved (112).  Vlastic’s stay here was at least partly supported by Shell Development.  They had a strange problem of corrosion of steel vessels that was finally traced to trace levels of HCN.  But the presence of overwhelming amounts of hydrogen sulfide made it very difficult to determine.  We developed a means of determining HCN in such a challenging system (118).  This had substantial practical importance.  Our Shell contact, Don Olson, got sufficiently interested in this such that he took it upon himself to apply it to process situations (140) and after he retired, he formed a company, that among other things, sold such devices.

 

By now, some thought of me as some sort of an Ion Chromatography Guru, rightly or wrongly.  Analytical Chemistry published a collection of 50 or so landmark papers, published in the journal over the past 40 years, which changed the world of analytical chemistry, as it were.  Included in it was the first paper on ion chromatography, by Small Stevens and Bauman.  They asked me to write a brief introduction to this paper.  Later, I wrote a review on the state of the art of ion chromatography for this journal for the magazine pages (120). 

 

Capillary electrophoresis relies on electroosmotic flow, a phenomenon discovered more than a hundred years ago.  We conceived an idea to use electroosmotic flow as an instantly reversible pump and use it in flow injection analysis, using incredibly low flow rates.  This was my first paper with Shaorong.  As I write this in 1997, Shaorong was the best of my students thus far.  He came from China with the highest of recommendations and it is ironic that initially I did not even want to accept him; I thought I had too many students at the time.  We have a system of cumulative examinations at Texas Tech as the qualifying barrier for the PhD.  To wit, we have these on one Saturday every month and a student must pass so many of such exams within the first two years.  One particular Saturday, Shaorong took the analytical examination and having finished rather quickly with the Analytical Cume, he asked if he could have a copy of the Organic test.  He aced both tests.  This has never happened in the history of this Department and it is not likely that it is going to happen again.  In mathematical ability, Shaorong was (and is) substantially better than I am and he had good experimental skills to match that.  This would be the first of many papers (121) that I shall publish with Shaorong.  Shaorong has gone on to Molecular Dynamics, developing DNA separations on glass chips, after a brief stint at Barry Karger’s.  We remain close friends.  Shaorong’s ambition remains to have his own business.  I do not know whether his nighttime acquisition of an MBA is an essential ingredient but I know that he will eventually succeed!

Some things are always difficult to write.  But they are important events in one’s life and I cannot avoid writing about them.  At the time I started out as an assistant professor, I used to tell Sarah that this continuous preoccupation with my work would surely cease after I am promoted and tenured.  By this time, many years have passed by since I was promoted and tenured.  But by now Sarah clearly saw that my life will always continue this way, the Chemistress will always have a greater hold on me than her or my family.  Also, even though I was professionally successful and we were not poor, I was still a University professor and far from rich.  Sarah was brought up in a rather well to do family.  In 1992, she decided this is not the life she wanted and filed for divorce.  I asked her to reconsider but she had apparently made up her mind for some time.  It is ironic that later in the year, Texas Tech bestowed its highest honor, the Paul Whitfield Horn Distinguished Professorship, so named for the first president of the University, upon me.  At the time this came about, I was the youngest person to be so honored.

 

As I moved alone into a life of my own, living in an apartment, I probably would not have survived without the continuous care and attention paid to me by my students.  Ingemar stayed with me in my apartment for a week to make sure I would survive and was my guardian angel.  Life has tendency of moving on, all on its own.  

 

I have always believed in simple and inexpensive instrumentation.  Light emitting diodes (LEDs) are inexpensive virtually monochromatic light sources that span the entire visible-near infrared range.  We have been working with LEDs for many years; based on this experience I authored a paper with many of my students and associates as to how to build LED-based photometric detectors (124).  Reprints were gone in a record time! 

 

When I first came to the United States to Louisiana State University, I was amazed and pleased to find Santasib, who was one year ahead of my class at Bankura Zilla School, was there, working for his Ph.D. in Civil Engineering.  Santa and I were very close but there was an element of insanity and instability in him that no one could understand.  When he was good he was a wonderful and most generous human being.  When he was cross, he would not give me the time of day even if my life depended on it.  Santa got his degree in 1975, long before me and got a job in Midland, Texas.  After little more than a year there, he decided to go back to India and not to return. He was married to Kajori while in India.  However, he came back by himself a month after he was married - I was still in Louisiana and he stayed with me until he found another position.  A few months later Kajori (She was called by her nickname, Moinu, by most of us) joined him and I met her for the first time - a very sweet and very young girl with an enchanting smile.  Santa and I were intermittently in contact, he was living in Houston, Texas and by this time I had moved to Davis, California.  When Sarah and I were married in 1981, I asked him to give Sarah away in the Hindu wedding ceremony we had.  Moinu and other Bengali friends dressed Sarah in the wedding sari and adornments.  Santa’s moodiness kept on only increasing with time, however and I found it difficult to maintain any reasonable relationship with him over the next few years.  In 1983, Santa made up his mind again that he was going to be living in India and left, Moinu then pregnant with their first and only child, Shuva.

 

        Santa could not keep the same resolve for long; he did not last very long in India.  He and the family were back, living in Houston, by 1984.  I recall getting up early one morning (Michael was still a very young baby and kept us up most of the night) by a phone call from Moinu, telling me that they were back.  I was disappointed that he could not call me.  After this, I heard about them now and then through common friends, but there was no direct contact.

 

In the late summer of 1991, one such common friend called me to give me the sad news that Santa had died, he had committed suicide.  I talked to Moinu shortly thereafter; I told her that She should never blame herself for this.  Aside from her, his wife, I said, I was the one person who knew him very very well.  He was a ticking time bomb who could really go at any time and it was my judgment that he lived as long as he did only because she was there to care for him and appease him.  We both cried.  Slowly it would come out that the intervening years had only seen a deterioration in Santa’s mental well being, he was abusive to most around him, especially to his wife.  I had intended to go over to see her or to have her and Shuva come visit us.  Neither worked out since my own marriage was in the process of disintegration at this time.

I had talked to Moinu off and on and both appreciated what it means to be alone.  She and Shuva flew down to from Houston to visit me in my little apartment in Lubbock.  This was the first time I met Shuva and I was immediately struck as to how much the eight year old resembled his father.  I visited them some months later in Houston.  Then one late night in Lubbock, I got a call from Moinu, in great panic, she had called an ambulance; Shuva was throwing up blood.  It took several days to make a final diagnosis – he had a tumor in his right lung and good part of his right lung had to be removed.  I went to Houston to help her through his surgery.  These times made both of us realize, how much we appreciated and needed each other.  I told Moinu perhaps she could move to Lubbock – She could find a similar job there and if anything were to develop between us, it would not do so if we were separated by 600 miles.  In the summer of 1993, she moved to Lubbock, had her apartment at the other end of the city.  Through the next several months, we spent more time together than apart.  Early in 1994, with my long time friend Ken Reiszner, who flew in from Louisiana, as the witness, Moinu and I were married, without any fanfare, in the front of a judge.  I gained a wife and a son; I had a real family again.

 

Optical detectors and optical systems continued to constitute an important theme in my work.  Jarda Ruzicka, the charismatic inventor of flow injection analysis, showed a sandwich type optical fiber based cell that is particularly useful for utilizing the technique of gas diffusion in Flow Injection Analysis.  We improved on the cell considerably and showed that actually this geometry greatly reduces refractive index change induced artifacts as well (126).  The only meaningful improvement to this and the previous paper (124) was constituted by a sequel some six years hence, through the development of a commercial version (sold by Global FIA) of a fully optical fiber coupled detector that used reflective optics, had readily changeable LEDs and was quite immune to trapping air bubbles (214).

 

Simon and I made a really useful invention in making wet denuders, especially of the parallel plate type (128).  Many research groups in the world are using such devices today for collecting gases and analyzing them on-line.  I am often told that we should have applied for a patent.  I have come to understand something about patents that I did not understand before - no one wants to commercialize an invention given to them freely when it is not protected by a patent because they feel that they have no protection from copycats!

As his last contribution before he left for Czechoslovakia, Vlastic looked at the relative merits of conductivity based and indicator/photometry based methods for determining carbonate/bicarbonate/dissolved CO₂ by gas diffusion and concluded that conductometry was superior (130).

 

After the ozone study, Chung had some extra time.  I asked him to look at a means to measuring my favorite analyte, H₂O₂, by a colorimetric method that will produce a blue product so that we could use simple sensitive LED based detectors (131).  Meanwhile my friends in Switzerland, Andreas Sigg and Albrecht Neftel, miniaturized our fluorometric H₂O₂ measurement system and put it on a motorized glider to make airborne measurements (135).  More work continued with the LED based detectors.  Hanghui was the first person to come to the group with a reasonable background in electronics and computers; in time the whole group benefited from this.  For now, I asked him to apply the switched integrator chip, then recently developed by Burr-Brown for making sensitive X-ray detectors for clinical Catscan machines.  It did indeed produce excellent results, yielding noise levels lower than any available commercial detector (132).  Later, Hanghui continued with LED-based detectors, this time with two colors, pulsing them alternately or sometimes using wavelength selective detectors.  In flow injection extraction experiments, one normally carries puts in alternating segments of a sample and an organic solvent and analytes of interest are extracted from the sample into the solvent.  The two phases are then separated prior to the measurement of the analyte in the organic solvent.  Using the two LED scheme, we used one color to monitor the analyte and one color to sense which of the solvents was in the detection volume (137).  Later on, this scheme was extended to a conductometry/photometry approach where conductometric sensing was used to judge what solvent was present in the detector and the LED based photometry was used to determine the analyte (162).

 

The high point of dual LED detection came with a dual cell dual wavelength detector arrangement.  Hanghui and I showed that it is possible to selectively determine absorbance in the presence of turbidity and refractive index artifacts.  We could selectively determine the amount of an indicator dye, Bromthymol Blue present in a mixture of milk and alcohol (139)!

 

Shaorong and Hanghui often worked in the same laboratory and had a considerable degree of mutual admiration for each other.  One is fortunate to have one student of this caliber.  To have both of them at the same time was simply euphoric.  Shaorong and I continued on the theme of electroosmotically pumped flow analysis in the capillary scale, introducing valve-based injection for such systems (133) and means to increase the optical path length (134) and finally described in detail the real power of this technique that goes well beyond simple flow injection analysis (141).  We used the electroosmotic pump as the heart of an ultramicroscale sequential injection analyzer (147) and used a membrane interface with this to measure ammonia (152).  When we applied auxiliary electroosmotic pumping to augment or inhibit flow in a CE system, we observed that under the right condition, a surprising increase in separation efficiency was possible (145).  The concepts of electroosmotically pumped flow analysis and an auxiliary electrooosmotic pump in capillary electrophoresis were both patented (P8, P9).  Satyajit (Bill) Kar and I later added some through studies on how auxiliary electrooosmotic pumping is superior to other alternatives in improving resolution and efficiency through bulk flow control (211, 213).

 

Another important milestone in my personal life arrived.  Rivu Nalok Dasgupta arrived in his full glory towards the end of 1994.  Ever since the day he was born, people have been telling me that he is the spitting image of his father.  I hope he has as good a life as I have had!

 

Since the original experiments carried out by Dayong Qi and the explanations by Tetsuo Okada on the stopped-flow chronoamperometric experiments (81), I always wanted to study this further.  When Professor Hisakuni Sato came on a sabbatical leave from Yokohama National University in Japan, we started back on this.  It was established beyond doubt that the chronoamperometric profiles are due to electromigration and are essentially dependent on the ionic mobility.  Thus, ion identification is possible (136).  On a similar sabbatical leave came Professor Qijia Fan from the Beijing Institute of Microchemistry.  Professor Fan worked on the gas phase formaldehyde instrument we earlier developed (65) but instead of the acetylacetone used before (58) we changed to cyclohexanedione and a straight inlet Nafion membrane diffusion scrubber (108).  The Tennessee valley Authority largely sponsored the work – these guys needed a helicopter-borne instrument to measure formaldehyde during the southern oxidant study.  Altogether, this research led to a reliable gas phase formaldehyde instrument that could measure ambient formaldehyde at the parts per trillion level (138).  Subsequently, this instrument was field tested in an intercomparison study sponsored by the National Center of Atmospheric Research; it produced results identical to their $200,000 tunable diode laser but ours had better time resolution (187)!.  They were impressed enough that they built one of their own.  Paul Shepson at Purdue and his student Ann Louise have also built one, taken it to the arctic and found wonderful things to publish in Nature (the formaldehyde in the arctic comes mostly from the snowpacks)!

 

Formaldehyde is a carbonyl compound.  One is also interested in measuring the occurrence of other carbonyl compounds in various industrial products because these often polymerize and produce discoloration, etc.  Shell came to us with this problem.  Of course, in industrial situations the levels one needs to measure are much higher than those necessary in ambient air experiments.  We chose a well-known chemistry, adapted to a LED based colorimetric flow injection analysis method; the catch was that different carbonyl compounds should produce the same response.  Through the variation of reaction time and temperature this too was accomplished.  The reaction was age old, but many mechanistic details were unknown.  John Marx came through capitally, with well thought out experiments to decipher what was going on.  The net result was an excellent paper in Analytical Chemistry (142).  A commercial analyzer that incorporates this chemistry has since been marketed by Global FIA.

 

As ion chromatographers, we have been using conductivity detectors a whole lot.  The principles of the bipolar pulse conductance detection system were already well established in the literature.  Hanghui wrote the program and built the electronic interface to a PC.  Satyajit (Bill) Kar originally came to my laboratory as a post-doctoral fellow from Mark Arnold’s laboratory in Iowa, to fill Bao’s place, a big task.  I thought of how to make reproducible conductivity sensors for capillary systems using bifilar conductors where each one of the wires is substantially smaller than the finest human hair.  Kar and I put all of it together.  My very first graduate student, Hoon Hwang, by now an associate professor at Chuncheon University in Korea, came back for a brief stay and helped in this quest as well (143).  Satyajit was methodical, careful and most organized in his work (wish I could say the same of myself!)

 

In our continuing work on atmospheric measurement instrumentation, as his last act before returning to now free Czechoslovakia (shortly to be divided into Czech and Slovak Republics), Zbynek applied the wet denuder to measure nitrous and nitric acid emissions from domestic open flame sources such as stoves and heaters.  These put out sufficiently large amounts of nitrous acid to be of significant concern in indoor environments (144).  In the same atmospheric instrumentation theme, Simon and I developed an automated approach to determine the chemical composition of the soluble fraction of atmospheric fine particulate matter, in near real time.  This was in many ways a very important development (151).  Atmospheric particles in the 0.1-10 micron size range are the ones of primary interest to the atmospheric chemist.  These are the ones that affect visibility, are responsible for various health effects and so on.  It is very difficult to collect submicron size particles without using a filter or such like because very small particles tend to follow the air streamlines.  After trying to do the same in a futile manner, Simon hit upon the idea of condensing steam on the particles to make them grow, just as in a cloud chamber.  When this was drawn through a cooled maze, impaction and thermophoretic effects both caused the droplets to be collected.  The mixed liquid/air stream was then phase separated and the liquid, already containing the soluble fraction was sent to an ion chromatographic analysis system.  Detection limits for particulate sulfate, nitrate, etc. were in the low nanogram per cubic meter level, with a time resolution under 10 minutes.  The system is equipped with a parallel plate wetted denuder as the front end to remove and measure all the soluble gases.  A subsequent study reported on the near simultaneous measurement of nitric acid gas and particulate nitrate and nitrous acid gas and particulate nitrite in ambient air (153).  Some time had elapsed between the development of these instruments and the papers coming out.  In the meantime, we helped our Swiss friends to build a copy.  The first paper that came out was actually work carried out in Switzerland, describing the measurement of gaseous ammonia and nitric acid particulate ammonium and nitrate to make deposition measurements (148).  An account of this instrument appeared later in the magazine format (177) and later we also demonstrated a new improved version, without a maze, for potential use in Plutonium aerosol monitoring, using Cerium (III) as a surrogate (197).  If sufficient steam is introduced it is also possible to collect the particles through a bead packed column (210) although this isn’t quite as efficient and the other approaches remain preferred.  This technology was used later to solve several other problems.  In the wake of energy conservation measures, the same air is recycled through a building many times and some times this leads to an unhealthy atmosphere in the building.  The reasons can be biological or chemical; the occurrence of a so-called “sick building syndrome” has been well documented.  It is difficult in such cases to distinguish whether the problem is of biological or chemical origin.  We hypothesized that if the problem were biological in origin, then there would be more proteinaceous aerosol in the air of such a building than otherwise.  A system was developed to measure that and the whole concept was validated when excess protein was indeed found in the air of a building known to have fungal contamination (193).  Another application involved the determination of strong acidity in the aerosol.  The analytical system was composed of two halves: in the first half, the cations (basically ammonium) are concentrated; in the second half, the anions are concentrated.  Basically total cations and ammonium are each determined by elution and suppressed conductometric detection.  Acidity is determined by the difference between the total anion equivalents and the total ammonium equivalents (198).  This is a bittersweet story.  The instrument went through several years of development, and two different EPA grants supported it in succession.  Chung was the first to work on it, followed by Dr. Ito from Hiroshima University Japan, and finally Clay Chasteen.  This constituted Clay’s Master’s thesis.  The instrument went through several field tests.  The first was at Ranchos Los Amigos hospital in Los Angeles and we determined to what extent the sulfuric acid aerosol introduced in the test chamber was neutralized by the expired ammonia from the volunteer subjects (all of it!).  The second study was a collaborative study at Ithaca on known and unknown aerosol acid contents.  After the presentation of the performance of the final version of the instrument at a meeting in Park City, Utah, the EPA people encouraged us to begin some collaborative testing with health effects people, notably a group at New York University.  I had to leave the country at the time and instructed Clay that he should drive it to New York.  He decided to pack it up and ship it instead, spending three days to crate it properly.  Worse, he shipped it by airfreight without insurance. 

The shipment never reached New York.  Several others (most notably Lung Chi Chen and his colleagues at NYU) and I spent months (two years, really) but the instrument seemingly vanished from the shipper’s hub in Columbus, OH.  I offered a substantial personal reward from my pocket, to no avail.  I gave the final performance report to the EPA in Research Triangle Park orally, and in the end, a grown man, I couldn’t help breaking into tears.  That instrument represented 10 man-years of effort.  I half-heartedly wrote a proposal to rebuild it.  It was never funded and in a way it was good - I never really wanted to rebuild it again. 

 

Julie Zheng also came from China.  I always worried about her health because she not only looked and acted frail; she even fainted at least once in the laboratory.  She was among one of the very few of my students who came in to get a MS degree and stopped there instead of going on to a PhD.  Regardless, the work she did for her MS thesis was quite good and novel.  Everybody has been doing capillary electrophoresis in aqueous or at best, highly polar water-miscible solvents.  I wanted to know what will happen if I dissolve lipophilic salts such as tetrabutylammonium perchlorate in water immiscible solvents such as chloroform and then inject a slug of this in the conventional aqueous electrophoretic system.  What we found was a delightful concentration effect – a suitable analyte dissolved in the aqueous phase can be concentrated at the organic solvent interface.  The process was probably more akin to flow injection based solvent extraction that is electroosmotically pumped than capillary electrophoresis (146).

 

In Korea, my ex-student Jae-Seong Rhee, now a senior scientist at the Korea Institute of Science and Technology, and his coworkers published a paper on the determination of dissociation constants of organic acids by flow injection analysis.  I discussed this idea sometime ago with Jae-Seong and I was really surprised when he sent me this paper in Korean, with my name as an author (149)!  Now I could honestly be accused of not even having read (incapable of reading!) a paper I supposedly authored.

In 1995, Shaorong and I published the first of our many papers on drops (and films), an area of affliction that will always continue with me.  When I was in high school, I thought that rain water must be very pure, like distilled water - water is evaporated and recondensed - only later, terms like acid rain came into vogue and all of us learned that rain is not so pure after all.  The impurities in rain are acquired both at the cloud stage and during the passage of the raindrops through the atmospheric column as they acquire various gases.  One day, when my son Shuva asked me how can you tell if the rain is acid, I told him that we don’t often have acid rain in arid dusty West Texas because the calcareous soil is alkaline and neutralizes any acid.  However, I added, if it rained really acid somewhere and your taste buds were fine-tuned, you might be able to detect acid rain by its slight sourness.  Next I saw him in our rare West Texas rain, he had his tongue struck out big time!  As I laughed looking at him, what occurred to me was that if raindrops involuntarily pick up gases, why couldn’t we create a pristine drop and expose it to gases and then analyze the drop for information on the nature of the atmosphere it was exposed to?  Thus we adapted our electroosmotically pumped sequential injection analyzer so that a drop of liquid was created at the tip of a capillary and then gases of interest such as ammonia or sulfur dioxide at trace concentrations were sampled around this microdroplet and then we withdrew the droplet back into the analyzer to determine how much gas was collected.  The analyzer itself was impressive: it required only 1 microliter of sample.  Thus, from a droplet 18 microliter in size (for reference, a typical small drop from a medicine dropper is about 50 microliters) we could make some 17 sequential analyses and thereby probe the spatial composition of the drop after the analyte was sampled.  Shaorong did a wonderful job of pinning down the underlying mathematics: all in all, it was an excellent paper (154)!

 

Gary Tarver was the first Texas Tech graduate who came to do graduate studies with me.  The Tarver family is well established in the Lubbock area and Gary’s father and others were very much in the electrical business.  I applied and got a grant from the State of Texas to construct a laboratory in a motorhome so as we can go make measurements as to how much hydrogen sulfide and other reduced sulfur gases are emitted to the atmosphere in oilfield operations.  At this time, Gary had already received his BS degree for several years and has been working as an electrician.  Although he grew up in an environment where higher education was not only not encouraged but also rather scoffed at, he never was happy with a hardhat job.  He always wanted to get a graduate degree.  As providence would have it, he was visiting Jerry Mills, my colleague in the Department to inquire about possibilities of graduate studies.  Jerry sent him to me and I recruited him on the spot.  When he cane back to school in this fashion, he was only a few years younger to me.  Together we bought a $40,000, 29-ft long Southwind motorhome and completely gutted its inside and made it into a laboratory with all the amenities of a standard laboratory.  Then he designed a measurement system for hydrogen sulfide (originally we thought that we can use a commercial gas chromatograph equipped with a flame photometric detector but this turned out to have insufficient sensitivity); an account of this was published in Atmospheric Environment  (156).  Most people have one or two idols – for a long time my idol has been the eccentric English scientist Jim Lovelock.  As it turned out, he reviewed this paper – with favorable comments.  It took several years and many field measurements and much modeling efforts to carry out the intended study, however.  In the end, we concluded that much larger amounts of sulfur gases, primarily hydrogen sulfide, are emitted from oil field operations relative to what the operators say, the total emission contributes up to 1% of global sulfur emissions (188).  Over the years Gary has been much more than an ex-student, he has been like a brother to me that I never had, one of my best friends.

 

When Harvey Bellamy came to graduate school at Texas Tech, he came from having worked for many years at the Pantex nuclear armaments reprocessing plant as a senior technician.  Thinking that this experience would be very valuable to my laboratory, I wished that he would join our group.  He elected to work with my colleague Kasem Nithipatikom.  Kasem, however, left after one year; he wanted to be back in Wisconsin, where he was educated.  So it was in Harvey’s fate, after all, to be my student – he joined my group after Kasem left.  Harvey has not had an easy life.  Between this and the economic needs, which prompted him to go back to Pantex, 140 miles away from the University so that he could come and work only on Friday and Saturdays, greatly delayed his progress.  Although he has been a co-author of several papers by this time (119,124,126), constructing a spectrometer based on a liquid crystal shutter array was Harvey’s first major solo work (157). 

 

1995-1997 were three years where I really concentrated on drop and film based analytical systems.  As I write in 1997, Hanghui maintains two different web pages that describe some of our drop work: http://pegasus.acs.ttu.edu/~vephl/drops.html/ and

http://lynx.neu.edu/home/httpd/h/hliu/.  While Shaorong did the first and the seminal paper on drop based analysis systems (156), the work from this point on was continued by others.  From the Araraquara campus of the University of São Pãolo in Brazil came Professor Arnaldo Cardoso; his wife Elisabeth (Beth) Pereira, also a chemist, joined him later.  Brazil is the only country in the world where the principal automotive fuel is not based on petroleum.  Rather, it is alcohol produced from the fermentation of sugar cane juice.  There is a massive amount of sugar cane waste that is produced as a consequence.  When this is burned off seasonally, there is a large emission of nitrogen dioxide and formaldehyde.  Inexpensive but sensitive and reliable monitoring methods that can be deployed in the field are needed for these species.  Arnaldo and I created a drop/film of Griess-Saltzman reagent on a wire support, with optical fibers leading to an LED and a photodiode.  As nitrogen dioxide reacts with the liquid drop, a purple color develops that is measured by the LED-photodiode based absorbance detector.  Low parts per billion levels of nitrogen dioxide are detectable in a few minutes (158).  Beth’s task of developing a formaldehyde sensor was more challenging because the reaction she chose to utilize required that the air sample be contacted with one liquid contained in the drop.  Then a second reagent was to be added to the drop; it was not permissible to have that reagent present in the drop initially.  Beth accomplished her task as well, reaching similar low parts per billion limits of detection in a few minutes (183). Arnaldo also developed a fluorometric drop based sensor for measuring hydrogen sulfide; there are a lot of advantages in exciting a drop from within!  Again, we were able to achieve low parts per billion limits of detection within a short time resolution.  A few finishing touches were necessary and Hanghui carried these out when Arnaldo left (182).  We did patent the concept of determining various things using a liquid drop or film as the sampler (P10). 

 

Films have even better surface area to volume ratios than drops so if someone has an analysis system that can work with very small samples, films would be ideal.  I proposed to the EPA that we make essentially a micro version of a child’s soap bubble wand using a capillary tube as the stick.  A platinum wire loop is formed at the end and the film thereon is exposed to a gas sample.  After a minute or two of actual sampling, part or all of the film contents is transferred to the capillary and separation commenced by electrophoretic means.  Bill Kar and I showed that with suppressed conductometric detection, even a one-minute sample is sufficient to measure several gases simultaneously with detection limits in the parts per billion level (160).  Subsequently, we showed that with UV detection, UV absorbing substances could be similarly sensitively detected (170,184).  Later, Dr. Kaz Surowiec came from Marie Curie University in Poland and we used the same basic technique with indirect optical detection.  This detection technique is not very sensitive but we made up for that by introducing the analytes exhaustively from the loop using electromigration, using the loop itself as the high voltage electrode (174).  This was patented and licensed to Dionex (P11).

 

Most of the drop work was done through the remainder of Hanghui’s graduate career at Texas Tech.  A small person with an extraordinarily big imagination and a great heart and with unequaled talents as an experimenter, Hanghui and I shared the love for experimental science and had a truly good time with drops.  I gave many an invited talk in many countries on drops, generally always opening with Tagore’s unforgettable lines:

 

I have been across the continents to see the highest of the high mountain peaks

I have spent all my fortune to go across the seven seas

Alas, Never have I taken the time

To see two steps from my door

One dewdrop resting on a blade of grass…            (Rabindranath Tagore, 1867)

Hanghui and I first ventured into dynamic drops (drops that are continuously forming and falling) rather than the static drops which we have worked with where we formed the drops and then carried out sampling.  We used such a system to monitor gaseous chlorine, a project initially funded by Dow.  With a drop containing tetramethylbenzidine (which forms a yellow product upon reacting with chlorine) forming and falling every minute or so and a blue LED/photodiode combination, we could detect chlorine at the low part per billion level.  Each drop exhibited incredible reproducibility and at a constant pumping rate, the time between drops was a measure of the evaporation and thence the relative humidity of the sample air.  While we submitted this paper in the regular fashion, Royce Murray, the editor of Analytical Chemistry took it upon himself to publish this as an accelerated article (161).  The next publication with Hanghui was also an accelerated article in Analytical Chemistry and possibly the more important of the two.  We showed the solvent extraction of an analyte, as an ion pair, from a flowing aqueous outer layer to a stationary interior drop of an organic solvent of microliter volume and the in-situ optical monitoring of the extracted analyte.  This got rave reviews as well (166).  A drop has many virtues.  In yet other papers, we exploited the utility of a drop as a volumetric container, a reactor without walls and a windowless optical cell (168), used the act of the falling of a drop as an injection event for a capillary electrophoresis system (178) and authored two separate invited reviews (173,189), aside from one for lay readers (194).  Hanghui’s Ph.D. dissertation was titled “Analytical Chemistry in a Drop” and received the Song Prize, the best annual dissertation award in Chemistry and Biochemistry.  At this time the Song Prize had only been 5 years old and including Hanghui, my students have garnered that prize four out of five years (Simon, Gary, Shaorong, Hanghui…).

 

My interest in peroxides continues to this date.  In 1995, I published a letter with my Swiss friends on the pitfalls in sampling peroxides by cryogenic means (163).  Huiliang, Genfa and I described a simple electrochemical sensor coupled to a diffusion scrubber for measuring hydrogen peroxide (169) and in subsequent work eliminated the diffusion scrubber using essentially a falling film, to measure both hydrogen peroxide and industrially important organic hydroperoxides (180).  Dr. Huiliang has a much better formal education in electrochemistry than I do; together we authored a review on electrochemical sensing of gases (185).

 

I think the greatest good fortune I have had in my life is that I have had the pleasure of working with so many wonderful people.  Statistically, it is extraordinarily unlikely that I can be blessed this way simply by chance.  So I am particularly thankful.  In this account, I have only mentioned those that I had published papers with.  Actually, there have been several people who have not been specifically mentioned and several others who came and stayed for a short period of time – we did some interesting studies together but did not come to a point of finishing it.  I have had a particularly good number of students from Sweden and postdoctoral research fellows from Japan.  Dr. Mitsunori Murayama came from the same institute as Dr. Shintani.  We developed a means of measuring polynuclear nitro compounds in environmental samples in an unambiguous fashion by using high performance liquid chromatography with tandem electrochemical and fluorescence detectors.  The electrochemical detector reduced the nitro compounds to the corresponding amino-/hydroxylamino- moieties; these then produced intense fluorescence.  Both detectors had to respond for a peak to be classified as a nitrated polynuclear aromatic hydrocarbon (165).

 

I have mentioned Dr. Kazimierz Surowiec (everyone called him Kaz) before.  He had actually spent some time in Japan but he is hardly Japanese.  He was Polish in the best of the traditions.  As of this time, the two Polish gentlemen that came to my laboratory were both from Lublin and both were named Kazimierz, after the famous Polish king.  Quite a coincidence!  The first Kaz (Dr. Jurkiewicz - Shintani overlapped his stay, started calling him Dr. J., and it stuck) had a difficult time here.  A number of close relatives people died back home while he was here and it was a difficult time for him.  It was a different story for the younger Kaz (Dr. Surowiec).  He came here to spend a year, ended up spending nearly three and we had a wonderful time together.  His wife Barbara and beautiful daughter Kasha joined him for a year.  Kaz was another of those good souls; if I never knew him, my journey through this life would doubtless have been incomplete.  Experimentally he had great eyes; with his bare hands, he could make loops at the end of a capillary tip with ultra-fine platinum wires.  Some of these loops were so small I could not actually see them without a magnifying glass.  Kaz and I introduced the quantitative introduction of the contents of such a micro-loop, holding as little as 12 nL of a liquid (1/4000^th of a raindrop!), by pneumatic means (164). In a similar experiment but with somewhat larger loops, we applied exhaustive electromigration to introduce large amounts of analyte without a concomitant volumetric introduction of liquid.  Backed by extensive theoretical modeling and days and days of computer time spent running numerical models, this was published in Analytical Chemistry (174).  Coupled with the previously mentioned gas analysis technique (184), we published detailed description of the simple but powerful homebuilt instrument that allows one to do all these experiments and more (192).  Even with indirect photometric detection, we were able to detect low ppb levels of gaseous carboxylic acids with a loop film - CE instrument coupled to exhaustive electromigration (196). 

 

I have talked about Anna Sjögren, my Swedish student from Umeå, before.  Using our homemade capillary electrophoretic apparatus, Anna made the serendipitous discovery that if high voltage is applied with an electrolyte filled capillary when one side of the capillary is still hovering over an electrolyte filled vial, current flows and the resulting electromigration can have a very beneficial effect on the stacking efficiency.  It took some time to fully elucidate this phenomenon but it was a worthwhile effort in the end (167).

 

Before he left to accept a postdoctoral stint at Barry Karger’s lab in Boston (Hanghui followed him there to the same place in about a year but by then Shaorong has already left for Berkeley), I appointed Shaorong as a postdoctoral fellow in my own laboratory. A unique concern of relevance to aerosol composition measurements relates to the monitoring of hazardous nuclear material that may be released in particulate form.  Since the end of the cold war, large quantities of used nuclear materials were being brought to the US for destruction and reprocessing.  The principal storage site for Plutonium in the US was the PANTEX facility, so named for its location in the Texas panhandle.  Projections were that it could be three decades before a viable approach is found towards the ultimate disposition of this material.  In the interim period, one had to have stringent and redundant monitoring methods in place to insure public health and safety.  In the case of a major disaster, the threat of nuclear contamination to the population at large is largely from the atmospheric transport of the dispersed particulate material, amply proved in the Chernobyl incident.  Sensitive atmospheric measurement instrumentation is also expected to be of great help in identifying clandestine nuclear activities in which fingerprint radionuclides are typically released to the atmosphere.  Shaorong developed two micro collection interfaces for aerosol collection – with direct coupling to ion chromatography.  Both of these operated on electrostatic collection of the aerosol, the aerosol was charged in the first case with corona discharge (171) and in the second case by field charging (172) – the second approach avoided the production of a lot of unwanted nitrogen oxyanions.

 

One publication that I will always remember is that with my cousin Anjan Dasgupta (175).  Anjan is four years younger to me; we grew up together in our very early childhood in a joint family.  He still lives in our ancestral home in Bagbazar, in Calcutta and commutes daily to Kalyani University where he is a Professor of Biophysics.  Anjan is very good with abstract mathematical concepts and so on.  He applied the difficult to grasp concepts of nonlinear scaling to a set of copper ore leaching data from a real mine in India.  When he wrote it up, however, it was rather indigestible for most people.  He asked me to see if I could help and I did the best I could to make it more understandable.  He insisted that I have contributed to it sufficiently that I should now be an author – this was a rare opportunity for the two of us to be on the same paper as co-authors, how could I refuse?

 

Ion chromatography, when first introduced in 1975 by Small et al., brought about a revolution in how ionic analysis, especially trace anionic analysis, is conducted. By late 1980’s, with the advent of capillary electrophoresis and its demonstrated application to similar analyses as done by ion chromatography led many people to believe that ion chromatography will now finally be displaced.  The techniques are complementary, however, and as I write this in the late 1990’s, there is no indication that ion chromatography will be displaced.  It is possible in principle to couple one separation effluent to another.  Bill Kar and I described such a two dimensional separation system where the effluent from an ion chromatograph is automatically analyzed in a second dimension by a suppressed conductometric capillary electrophoresis system (176).  The mismatch between the scales of the two systems (we were using a conventional scale ion chromatograph with a flow rate of 1 mL/min where as the electrophoresis system used a flow rate 1/1000^th of that) became rather obvious. 

 

I began exploring the merit of open tubular ion chromatography even before I tried packed column work.  Ordinarily, getting good results in the open tubular format requires very small tube diameters, to reduce the diffusion distance from the flow channel to the tube wall where the active sorption sites are.  This requires very low dispersion conditions and places great demands on the injector and the detector.  At least initially, we decided to take an easier way out, use small analyte ions that have relatively large diffusion coefficients and use an elevated temperature to further increase diffusivity.  This allowed the use of 50 mm i.d. capillaries and commercial injectors and detectors.  We were capable of making very small conductivity detectors but at an increased temperature, the thermal instability caused too much noise to use them.  So we stayed with on-capillary optical absorbance detectors, already available for use in capillary electrophoresis.  Virtual anion exchange sites were put on fused silica capillaries by using a cationic surfactant, cetyl trimethyl ammonium acetate, as the eluent.  This worked, but only up to certain temperatures.  The extent of adsorption of the surfactant on the silica surface decreased dramatically with increasing temperature.  Thus, although efficiency increased with increasing temperature, retention decreased rapidly and practical separations became impossible to attain at higher temperatures.  My friend Doug Gjerde told me about Leon Yengoyan, a professor at San Diego State who had developed a method of rendering the capillary wall cationic by passing a suspension of cationic latex through the capillary.  We had previously bought such a capillary from Doug’s company Sarasep and tried to use it for suppressed conductometric capillary electrophoresis.  In that case, the combination of electrophoretic and ion exchange separation mechanisms did not lead to results that were overall favorable.  But in this case, it will be ion exchange chromatography alone and the results should be better.  Dongjin Pyo, a young faculty member from Korea (from Hoon’s department, as a matter of fact) began this attempt, with a capillary directly procured from Professor Yengoyan.  We pretty quickly destroyed it and so learned to make our own.  The latex supplied by Dionex was superior in quality and led actually to better results.  I wanted to expand on the very encouraging initial results but it became important for Pyo’s career to publicize this.  He was invited to give a plenary lecture at ASIANALYSIS in Japan and the account was published in the special issue of Analytical Sciences devoted to this (190).  Ultimately, the real problem with this approach is that the quaternary ammonium group decomposes at high temperatures, especially under alkaline conditions.  As a result, suppressed conductometric ion chromatography will be very difficult by such a technique.  The surfactants should really be investigated further.  Ingemar did do a lot of preliminary work etching the inside of capillaries and creating a porous silica layer so as to increase the surface area of the capillaries, ultimately to increase adsorption on the capillary wall and thus to increase retention.  With new thermally stable, easily sorbed surfactants, I would really like to get back to this some day.

 

Among other things, and the fact that I developed a close friendship with Jim Alexander at Rohm and Haas, a wonderfully self taught capillary chromatographer, led me to seriously consider doing suppressed ion chromatography in the capillary scale.  I talked with Dionex and they consented to give Jim some of their AS-11 packing and share with him how it is actually packed.  Jim went down to Sunnyvale to do this and returned with the news that the columns he packed out there were actually more efficient than what Dionex quoted for their macrosized columns with the same packing.  After Jim sent us some of these capillary columns, first we used them by splitting the flow from a Dionex DX-500 microbore pump.  We already had developed suppressors and detectors that could be used with capillary scale systems.  However, prior to this time I have repeatedly tried to draw attention in various talks to the fact that liquid chromatographic pumps are grossly over designed in terms of the power they consume.  In the capillary scale where flow rates are 1 or 2 microliters a minute against pressures of a few thousand pounds per square inch, the actual power spent is only a milliwatt or less.  I bought a little stepper motor driven glass syringe pump, and this was sufficient to do the necessary pumping.  Anna and Brad Boring (Brad came from S. F. Austin State University in East Texas where Brad’s father is a professor (and the head) of Chemistry) put a microscale electrodialytic eluent generator together and a wonderful capillary IC system was born (179).  Brad worked very hard later to put this whole thing into a portable briefcase like package and we exhibited this at both the 1997 Wintergreen capillary chromatography meeting and the International Ion Chromatography Symposium in Santa Clara, California.  I chaired and organized the latter meeting.  Anna was awarded a travel grant by our panel of judges (Chuck Lucy and John Lamb) to come in from Sweden to present her poster on this instrument.  All participants at the meeting (nearly 300) were asked to vote about what the best poster was.  Anna won (A69) and the paper was published in the symposium volume of the Journal of Chromatography (195).  Subsequently, the instrument was used for both IC and conventional LC (199).  Also, Simon made a miniature parallel plate denuder, Brad improved on it and we coupled it to the Capillary ion chromatograph to devise a miniature IC based gas analyzer (203).  The effort then went to gradient LC and designing simpler pumping systems capable of operating at a very high pressure.  My new student Scott was brought up amidst stock car racing and his penchant for fooling with all things mechanical led to an inexpensive capillary scale gradient LC with pressure capabilities well beyond what is commercially available (206).

 

One thing about capillary liquid chromatography systems is that it is relatively simple to miniaturize most components except for the detector.  When we built our field portable ion chromatography system, it was relatively easy to do so because the conductivity detector is an exception to the rule – it is relatively easy to miniaturize.  The optical absorption detector, however, is the most widely used detector in liquid chromatography and something needed to be done about making a smaller affordable optical detector.  The switched integrator based circuit that Hanghui built earlier (132) is ideal for making an optical absorbance detector for capillary systems.  I assigned this task to Brad and the results were gratifying (181). 

 

I had always thought that if water has extraordinary properties as a liquid, ice has even more unusual properties as a solid.  In particular, I had been thinking for a long time if one could conduct normal phase liquid chromatography (e.g., with things like hexane as the eluting solvent) with particles of ice as a stationary phase.  My student Youwen Mo demonstrated for the first time that this could be done in a realistic basis (186).

 

Shortly after Simon first made the parallel plate denuder, we thought of putting multiple parallel plates together in a single enclosure so as to either allow very high flow rates or to put different liquids on different plates and send the same to different analysis systems.  Lizhen Ni was assigned this task when she arrived from China.  She ingeniously made the plates out of two sheets of polymeric net material such that a pocket of water is formed in between.  This was the key to a small (only 30 cm long) wetted denuder with eleven parallel plates, capable of sampling with reasonable efficiency even at a flow rate of 100 L/min (191)! 

 

Water.  Hydrogen Peroxide.  Sulfur Dioxide.  Nitrogen Dioxide.  Chlorine.  Carbon Dioxide.  Hydrogen Cyanide.  Ammonia.  All small molecules but together they have consumed a large part of my life.  Ammonia in particular has been a recurring theme (38, 73, 87).  It is interesting for atmospheric chemists because it is the only atmospheric base.  It is interesting to me as an analytical chemist because there are sensitive and selective techniques for its determination, both by colorimetry and fluorometry.  You can easily play analytical tricks with it like gas diffusion, preconcentration, etc.  Yet, because of omnipresent contamination issues, it is a challenge to determine at trace levels.  There is real practical importance in the determination of low levels of ammonia.  Semiconductor wafer fabrication facilities (FABs) are acutely concerned about the residual levels of ammonia.  Over the background ocean, the levels of ammonia are still difficult to measure with certainty and thus the direction of nitrogen exchange (land Û ocean) still remains unclear.  The key is to measure the gaseous ammonia concentration over the ocean and the potential upward flux from the water.  Our previous attempts at developing a gas phase ammonia instrument led to several people making better and more improved copies.  In particular, Roy Harrison’s group In the UK and Lise-Lotte Sørensen in Denmark both made significant improvements, Sørensen’s group pushed the limit of detection down to 10 parts per trillion (we were stopped at the 50 part per trillion level).  A chance encounter with Lise’s beau, who just happened to be the program officer of atmospheric research programs at the Office of Naval Research (ONR), led to a project to look into the second part of the problem, the measurement of the potential flux.  Basically, we pumped ocean water on a flat surface without turbulence and collected the ammonia evolved from it on a pure water surface flowing on the other side.  This was then preconcentrated and measured.  We needed help to get this project to a meaningful point of completion.  Wilfried Winiwarter (I read his papers in the literature originally - he was a student of my friend Hans Puxbaum in Austria - and when he came to give a seminar in Texas, Kim immediately dubbed him as little boy genius) agreed to help us numerically simulate what goes on this device and Tony Clarke and his student Tomoe Uehara at the University of Hawaii collaborated in setting up a fieldable version of this instrument and doing field experiments at their laboratory facility at Coconut Island in Hawaii.  Overall this resulted in a really great useful paper (200), my 200^th, and Tomoe got her MS degree on this project. 

 

On the ONR project we also had another productive collaborative effort, with Bill Hoppel and his associate Glenn Frick at the Naval Research Laboratory, we flew our hydrogen peroxide instrument aboard a blimp along the California-Oregon coast, following ships plumes (209).  It was not unusual to find that [H₂O₂] decreases while [SO₂] increases in the center of a ships plume but the more interesting finding was that [H₂O₂] increases sharply again at the plumes edge, reaching concentrations significantly higher than the background.  This occurs presumably because of ongoing photochemistry with reactive hydrocarbon emissions.

People in other professions imagine scientists as thinkers and experimenters but not many realize that an equal part of a scientist’s life goes into being an author.  I cannot speak for others but in my own case, the process of discovery is very much alive during the process of writing a paper.  It is often then that a need for conducting yet another set of last (such an elastic word, one of my students said) experiments emerge.  Indeed, in one recent instance, 90% of the data with which I started to write a manuscript got dumped in favor of the results of altogether new and different experiments commissioned during the writing!  Anyway, authors obligatorily deal with editors.  In my career, the bulk of my papers have been in journals devoted to analytical chemistry, notably, Analytical Chemistry, Analytica Chimica Acta, and Talanta collectively has published probably 80% of our papers.  Good editors are indispensable in shaping, nurturing and cultivating good authors.  I am old enough now to realize that the goodness of an editor does not lie in accepting all the papers I send.  I also realize that good science alone does not make a good paper, good authorship is still required.  Analytical Chemistry is a very large operation and I have dealt with many editors over the years.  Probably the funniest incident relates to the time when Georges Guiochon, then associate editor for Analytical Chemistry wrote a private note to a reviewer about a manuscript of mine he was being sent to review.  In due course, the editorial staff (then in D.C.) sent me the reviews, along with his original notes and all.  Adopting a completely innocent attitude, I sent it directly to Georges, saying that perhaps he should take out some of the things in that package.  I cannot divulge the contents of his note to the reviewer, suffice it to say that Georges was acutely embarrassed.  This is trivial, however.  As a scientist, I matured most during the reign of Royce Murray as the Chief Editor of Analytical Chemistry and without a doubt Royce has had great influence on me as an author, not just by accepting my manuscripts (on one occasion, promoting one to an accelerated article status (161)) but making constructive criticism and returning them when he simply did not think that they cut it.  To my credit, I never took a paper rejected from one journal and simply reformatted it and sent it to another one.  At Analytica Chimica Acta, Harry Pardue was a compassionate editor who is also a gentleman and a scholar.  Harry has made many executive decisions, accepting many a paper with only one review in, he was also the one editor who prevented me from doing a foolish thing that I would have later regretted.  Analytica Chimica Acta published a paper by Chinese authors, which neither acknowledged our prior work but also provided a significant amount of misleading information.  I wrote a caustic letter to the editor.  Harry suggested to me that I should forget about it but if I insist, he will forward it to the authors to respond.  I insisted.  The author’s response came back, even more caustic.  Meanwhile, I sought a second sober opinion from my good friend, Charlie Patton.  Charlie said to me, “Sandy, you are going to come out looking like the heavy on this one”, almost paraphrasing what Harry said to me originally, “It is not going to benefit you ultimately, Sandy”.  I took Harry’s advice and withdrew the letter.  Hopefully, never again will I waste so much time on something of so little consequence.  But above all, thank you Harry, for telling me like it is. 

Finally, Gary Christian at Talanta reshaped that journal altogether since he took it over and my association with that journal is virtually all “post-Christian”.  I have known Gary in so many different capacities; it is difficult for me to merely think of him only as the Editor of Talanta.  He has shaped my professional life in too many indelible ways that it is impossible to make an adequate acknowledgment.  I have had my opportunity to be an Editor of a scientific journal, but decided against it.  I am still having a great deal of fun being immersed in my own work.  Perhaps, another time will come.

 

Editors cannot save you from all unseen horrors, however.  There are two publications, which I wish never existed.  I wrote a critique of ambient gaseous HF measurements in California (A71).  Although I still believe that there is something wrong with the reported data, my chronology on the accurate measurement of fluoride by ion chromatography was incorrect.  The second one is more consequential and more difficult to retrieve.  And ironically, there was absolutely nothing wrong with what was written in the paper but I wish that it were never published.  In 1997, Professor Dipankar Chakraborti from the School of Environmental Sciences in Jadavpur University first drew my attention to the unfolding saga of arsenic contamination of drinking water in Bangladesh and the adjoining parts of Eastern India.  As I write this at the close of the century, with nearly a million people already showing symptoms of arsenic poisoning, this is likely to be the greatest environmental disaster in recorded human history.  Of necessity, I will keep this account short and therefore there will probably be errors of oversimplification.  The irony is that it started with no malicious industry but good intentions.  Some thirty years ago, when waterborne diseases like Cholera and Typhoid were endemic in the region and made their appearance with every rainy season when floods appeared.  UN organizations actively supported the erection of tube wells in the area.  Initially people wouldn’t drink the tube well water, referring to underground water as Devil’s water.  As it became apparent that clean water was so easy to obtain in this fashion, this attitude did not last long.  Bangladesh currently boasts of more tube wells per head than any other country and was the first in the region to declare the eradication of Cholera.  Indeed, as diesel generators and pumps became plentiful, people found it expedient to use tube well water for irrigation.  These are regions where there is hardly a lack of water, there is plenty of surface water and floods are annual events.  Still it was easier to pump the water from underground than to have it pumped from a lake some distance away.  With the use of pumped groundwater in irrigation, the rate of consumption of groundwater increased to a value 4 to 5 orders of magnitude greater than what would have been the consumption solely due to drinking and cooking.  However, nothing would still be particularly wrong with this except for one fact.  The Ganges, some eons ago, washed down a particularly mineral rich region of the Himalayas, and this mineral rich layer underlies a large part of Bangladesh and West Bengal, both sitting on the Gangetic Delta.  This layer is not that far below the surface in most of the affected region and the fact that the region sits on a tectonically active plate ensures that the layer does not subside.

        As groundwater is pumped out, fresh, oxygen-bearing surface water takes its place.  It has taken some time but it is clear that arsenic is leaching from these originally undisturbed sediments and coming out in unprecedented quantities.  The normal recommendation of the World Health Organization (WHO) is that the arsenic content in drinking water should not exceed 10 parts per billion.  For these regions, recognizing that such a limit would be simply unworkable, a special limit of 50 parts per billion has been recommended.  In Bangladesh, in more than half the districts, very much more than half the wells are above this limit.  Indeed in both Bangladesh and neighboring West Bengal, there are still wells in use with containing not part per billion but part per million level arsenic.  Dipankar came to Texas Tech in 1997 and gave a seminar.  The arsenic problem in the region, its full discovery and monitoring, evaluating technological solutions, its social, moral and economic implications, has become Dipankar’s sole focus in life.  Indeed he is, in his own fashion, an evangelist.  I cannot recall being in a “scientific” seminar after which there were so many moist eyes.  Simple field-portable measurement techniques for As was still elusive at the time and Dipankar recruited me easily to do something about this.  (It wasn’t a hard sell, after all, Kajori, my wife, attended his seminar!).  We settled on a stripping voltammetric technique and everything worked OK.  I presented an account of the technique at an International Conference on Arsenic.  More extensive laboratory work was done and everything still seemed fine.  I wanted to get an account of the technique out in the literature quickly and it was published shortly (202).  Only during the construction of a commercial version and the testing of some landfill leachate samples did it come out that there are unknown interferences with a number of the actual samples.  We have gone back to the drawing board, taken an entirely new approach and are confident that we will have a new and different solution in the near future.  I am able to caution people about the shortcomings of the technique who write and request for reprints.  But for every one of those, there may be many who may end up wasting their time.  I will never forget this episode. 

 

Daniel Wayne (Dan) Armstrong, has made lasting contributions to Analytical Chemistry.  Regardless of what further he does, he will always be remembered for his invention of micellar chromatography, the invention of a variety of techniques related to chiral separations (harnessing cyclodextrins and various chiral antibiotics for this purpose) and many other things too numerous to mention here.  Before he moved to the University of Missouri at Rolla, Dan and I were colleagues at Texas Tech, where as Dan likes to reminisce, we terrorized the students together.  Don’t anybody tell Dan, but this is wholly untrue!  Despite all the outward appearances of being stern (which he works very hard on), Dan is really a pussycat.  Anybody that figures that out, can take advantage of him.  (For that matter, I have figured out that contrary to Dan’s statement, it actually requires some degree of meanness to actually terrorize anyone.  I am working on it.  For starters, I put up a sign on my door under my name that proclaims BAD AND MEAN.)  The trouble is, as we age (Dan and I are only a few months apart), we are both mellowing, perhaps a bit too much for our own good!  When my first PhD student Hoon Hwang, who is now himself a professor (and chair of his Department) came back to visit briefly on a sabbatical, he spent most of his time telling my students that they should really appreciate how lucky they really are:  they should have worked for me 15 or 20 years ago.

        So, Dan sent this gentleman from Novus International in Missouri, to talk to me about a problem they had.  Dr, Bill Shermer, the director of Research at Novus, not only had an analytical problem but he was also convinced as to how it should be solved.  Edible oils and fats autooxidize first leading to hydroperoxides which then decompose to a variety of secondary products (most of which contribute to rancidity).  The onset of rancidity is accompanied by the loss of nutritional value and beyond that, hydroperoxides are considered by many as one of the main contributors to arteriosclerosis.  The quality of edibles oils and fats are judged by their hydroperoxide content; this is traditionally determined by iodometric titration.  Bill extensively researched the determination of hydroperoxides by direct spectroscopic means and I had to muster all available resources to convince him that it was not such a good idea, especially when half of the exact hydroperoxides were not characterized and standards were not available.  Novus gave us funds originally to work for a year but pretty soon it dragged into two.  Tian, who worked on this project, is one of the nicest guys anyone would ever meet.  To give an example, Tian and I went to a major process analysis meeting in San Antonio where each of us would make a presentation.  San Antonio being about 450 miles from Lubbock, we drove in my car.  Tian was extremely nervous, this being his first presentation (A77) and I listened to him about 4 times the night before.  The next morning he did very well in his presentation.  Relieved, that evening I went out to dinner with my friends from Dow, who were there en masse.  That night after I came back, I decided to look at my overheads for my talk the next morning and discovered that much to my chagrin, I had picked up the wrong folder.  As I was debating what to do (I already called my session chair, but he was out, partying late, I suppose!), Tian offered in all seriousness to drive to Lubbock and then drive back with the correct set of overheads.  (Pretty soon it occurred to me that since my whole life is in my laptop and since that travels with me wherever I go, the overheads must be in there in some form or other.  So at that point the scope of the adventure got considerably reduced - we merely scoped out the city of San Antonio and finally located a Kinko’s that was open through the night and presto, I had my overheads.)

        Tian was indefatigable.  He knocked off the peroxide measurement problem quickly.  We devised a very simple non aqueous flow injection system based on the ferric thiocyanate chemistry (ferrous ion being oxidized to ferric by the hydroperoxide), that was far more sensitive than needed and could analyze 100 samples an hour easily (208).  We now devoted our energies to the more important question, how do you get to a better index of the potential oxidizability?  Consider the fact that one has two oil samples, both of which have an equivalent, acceptable amount of hydroperoxides.  How do you tell which one is going to oxidize faster over a period of storage?  The accepted method is to take a multitude of bubblers, fill them with a fixed volume of the oil sample and bubble air through them at a prescribed rate while holding them at 98°C.  Every two hours, a bubbler is removed from the system and the hydroperoxide content is measured by iodometric titration.  This is called the Active Oxygen Method or AOM.  The results show that for a considerable period, the hydroperoxide content increases only slowly but then it begins to rise sharply, attaining a plateau, characteristic of an autocatalytic reaction, proceeding through radical intermediates.  In appearance, this is quite similar to a traditional pH titration curve of an acid being titrated with a base except that in an AOM curve, the hydroperoxide content may actually decrease past attaining a maximum.  The important thing is that the time from the initiation of the experiment to the onset of the initial sharp rise of the hydroperoxide content, called the induction period, is a measure of the oxidative stability of the sample.  An AOM test requires anywhere between 20-40 hours to run, with considerable manual intervention.  An abbreviated AOM test simply measures the increase in the hydroperoxide content after 4 hours of bubbling air, but this is known to have limited predictive abilities.  A more automated attempt to simulate the AOM passes the exit air from a hot bubbler containing the oil through deionized water and the conductivity of the latter is measured.  As hydroperoxides are formed and then decompose further (most commonly to formic and acetic acids), the acids are purged out of the oil and render the water more conductive.  A plot of conductivity vs. time thus results in a plot very similar to that obtained in the AOM experiment, except that it is further delayed.  This instrument thus measures the Oxidative Stability Index, commonly referred to as the OSI.  Its virtue is automation, its fault is that it is one step further removed from hydroperoxides and there is no certain knowledge as to how the different hydroperoxides from the different samples decompose. 

        More importantly, the AOM and OSI both fail in one important aspect.  They have an accurate predictive ability only if one is solely interested in stabilities at the temperature at which the experiment was conducted (typ. 98°C).  Most of the time, one may be interested in the stability of the sample at the storage temperature (typically ambient) or at the use temperature (which can be, for example, frying temperature: How long McDonalds can keep their oil before the French Fries?).  Direct experimentation at these temperatures are difficult.  In the first case, it will take a very long time to determine stability at the storage temperature and in the second case, it will be formidably difficult to conduct the experiment. 

We decided to take one step back at and look at the more fundamental step of oxygen consumption, which must take place before any oxidation can occur.  We devised a system, largely a gas phase flow injection system with an oxygen sensor as a detector, with the sample loop being a temperature programmable two-phase reactor. A low concentration of oxygen, typically of the order of 0.1%, balance being nitrogen, flows at a slow flow rate through the system.  A fixed amount of the sample is loaded into the reactor and the carrier oxygen flushes the headspace above it.  Next, the valve is switched and the reactor is isolated and heated rapidly to a preprogrammed temperature and held at that temperature for some fixed period of time.  During this time, oxygen is consumed.  When the valve is switched again and the carrier gas sweeps the headspace, the decreased oxygen content of the reactor is registered by the detector as a negative peak.  What we found is that these systems always obey Arrhenius behavior.  That is, the logarithm of the oxygen concentration loss is linearly related to the reciprocal of the absolute temperature.  The instrument was fully automated; it made replicate measurements at one temperature then increased the temperature to some other preprogrammed value and repeated the measurement.  Because of the low oxygen concentration used, the sample is altered little and the same sample goes through a complete multi-temperature run and generates a complete multi-temperature Arrhenius curve in 2-3 hours with no operator attention.  The resulting data can then be used to predict stability at any temperature (207).  The technique was since extended to solid samples (218).  Novus applied for patents on both the peroximetry and the Arrhenius Oxidative Stability measurement techniques. 

 

Without a doubt, the biggest scientific excitement in my laboratory in the last two years of the decade had to do with the advent of a new material Teflon AF, its extraordinary properties and what you could do with it.  Our previous effort with making light guides from porous Teflon tubes (tubes made of air!) (57) may be recalled.  Efforts like this or filling Teflon tubes with ethanol (ethanol has a significantly higher RI than most types of Teflon) may represent interesting curiosities but they do not really represent practical solutions towards making reliable and robust long path optical absorption cells.  The introduction of Teflon AF (AF stands for amorphous fluoropolymer) from Dupont in the late 90’s became a major milestone in this regard.  Teflon AF is a dioxole - tetrafluoroethylene copolymer that is transparent in the 200-2000 nm wavelength range and throughout that range it has a RI less than that of water.  Teflon AF is readily solution processable and tubular varieties of Teflon AF, including fused silica tubes coated with Teflon AF also became available, although at $20,000 per lb. or so, it was the most expensive commercially available polymer.  Several publications and patents quickly appeared on long path absorption cells and also on launching an excitation beam along the axis of the liquid-filled tube and looking at the backscattered fluorescence or Raman signal through an appropriate filter.  All this stuff was already on the table, as it were, at the time we got our first sample of a Teflon AF tube.  What interested me as to what makes this particular variety of Teflon have such a low RI.  I studied the literature on the characterization of this material by traditional polymer chemists; this showed that it has an extraordinarily high free volume, as it is called in the jargon of polymer chemistry.  What this basically means is that there is a large amount of interstitial space in the molecular network.  I immediately reasoned that this must make Teflon AF far more permeable than most polymers.  A further study of the literature revealed that this is indeed already known, the permeability of gases like H₂, O₂, N₂, etc., had already been measured and these permeabilities were three orders of magnitude higher than the corresponding values through common Teflon. 

        This gave me the idea of making a simple versatile gas sensor based on selective chromogenic reactions and a Teflon AF tube to act as a light guide.  Basically there is a reagent-filled Teflon AF tube with tees at each end through which acrylic optical fibers communicate through the liquid core waveguide.  The whole tube is then jacketed on the outside so a gas or air sample can be aspirated through this jacket and thus come in contact with the exterior of the AF tube.  As the gas permeates through the AF tube it reacts with the stationary reagent inside forming a colored product.  An LED of suitable color is connected to one of the optical fibers to act as the light source and the optical fiber at the other end acts as the conduit to a photodetector, measuring the absorbance of the long path waveguide gas-sensing cell.  The reaction being irreversible and the reagent being stationary, the absorbance signal is cumulative with time and if the concentration were constant then the absorbance will increase linearly with time.  Thus, the absorbance at any point in time gives an account of the cumulative exposure while the derivative of the absorbance signal with time is a measure of the instantaneous concentration.  When the detector reaches some upper practical range of absorbance it can measure, a simple solenoid valve opens and fresh reagent refills the tube by gravity.  This is a fully automated yet extraordinarily powerful and inexpensive device that can be configured to measure a different gas merely by changing the reagent and the color of the LED.  This was published as an accelerated article in Analytical Chemistry (201) and for whatever reason, caught the imagination of others and received quite a bit of press.

        I stumbled upon the second unique utility of liquid core waveguides pretty much accidentally.  For some time now, I have been fascinated by the so-called scintillating fibers, available from novelty suppliers like Edmund Scientific.  These are traditional acrylic optical fibers but with the core region being heavily doped with a strongly fluorescent dye like, fluorescein, rhodamine, or coumarin, etc.  Ambient light falling on the fiber results in fluorescence of the dye molecules and a significant fraction of the emitted fluorescent light proceeds down the fiber such that the two ends of the fiber glows quite brightly.  Originally, I thought that it would be pretty neat to use this as a (nearly) monochromatic light source; one could use one end as the source and the other end as the reference.  It would be nearly impossible to find a better-matched source and reference and the light source wouldn’t consume any power!  I am sorry to say I have had this idea for some time but have done nothing about it so far (LEDs appear so much brighter!)  One day, I was playing with the fluorescein doped fiber and just seemed to notice that while the glow from the ends did become brighter or dimmer as I went from the fluorescent lighting in my office to a tungsten desk lamp to afternoon sunlight streaming out of the window at the end of my corridor, the color of the emission that I perceived did not change, at least qualitatively.  It bothered me in the way that it was something that I did not really understand but I felt that I should.  However, I could not finger precisely what it was that bothered me.  A few days later, I mentioned it to my good friend and colleague, Dennis Shelly.  Dennis looked at me, puzzled and said, But Sandy, Fluorescein always does glow green!  And of course, he was one hundred percent right.  Later that afternoon, I realized exactly what it was that was bothering me.  The perceived color of the fluorescence was unaffected by the nature of the excitation light.  This means that I was looking at the fluorescent emission with very little interference from the excitation light.  I quickly looked at the fiber optic literature and learned that when an unjacketed optical fiber is illuminated from the side, only one in a million or 10 million of the incident photons actually propagate down the lumen.  It was then obvious that applying this principle to an optical fiber that is based on a liquid core waveguide could lead to a very simple fluorescence detector that does not require a monochromator.  If a liquid core waveguide (LCW) is interfaced through a tee to a high numerical aperture optical fiber that goes to a sensitive photodetector and liquid flows through the tee and through the LCW and out while the LCW is illuminated from its side, we have a fluorescence detector with no optics and no monochromator.  When only a background liquid flows through the LCW the light received by the photodetector is very little but as soon as a fluorescent or otherwise scattering analyte enters the detector, the photodetector receives this luminance that is guided through the tube.  A very extensive account of this was published (205).  Here, I cannot but say something about the frailty of the peer review system.  Even given my obvious own bias in the matter, I knew what we have here has a great deal of practical consequences.  At the capillary chromatography and electrophoresis meeting in Park City, UT when I described the applications of this technique for capillary detection systems, Jim Jorgenson, one of the pioneers in analytical chemistry, came up to me and said, Sandy, that was a great talk.  As far as I am concerned, that was the best talk in this meeting.  While I thanked him, I forgot to tell him that the two reviewers who reviewed this manuscript for Analytical Chemistry, said the paper was passé, wasn’t state-of-the-art and should not be published.  (Obviously, the Editor had more sense than these reviewers and the paper was published!)  There are some obvious consequences of this technique for detection on micro scale systems; only the future will be able to reveal how useful the technique will be.  Meanwhile, in collaboration with Mark Holtz at our Physics Department (Mark has a great Raman setup), I showed the technique is equally well applicable to Raman spectrometry, demonstrating, for the first time, successful Raman spectroscopy with a single monochromator instrument without a “notch” filter to block the exciting laser radiation (212).  We showed that the LCW cell is also the ideal conduit to carryout chemiluminescence (CL) experiments, because very fast reactions can be followed and sensitive detection does not require large area phototubes (215).  We demonstrated in this work another unrelated invention, the generation of unstable electrolytic products with the isolation of anodic and cathodic products by the nature of the flow geometry (215).  We demonstrated the detection of nanomolar levels of ammonia by the sulfonatoisoindole reaction (73) with a LCW cell and a photodiode detector (216). 

 

The first paper of a new decade, I hope, is some kind of an omen.  The way to accurately determine acid and base concentrations has been titrimetry for more than a century now.  Together with good friends Chris Pohl (Chris is a phenomenal scientist - he has an innate grasp of what chromatography is all about that is second to none) and Kannan Srinivasan at Dionex, several of my students and I demonstrated how acidity can be rapidly and accurately determined by a simple ion chromatographic technique (217).  To be fully useful for determining very low levels of acidity, we still must develop stationary phases that have very little residual carboxylic acid content.

Undoubtedly, it is too early to make a meaningful retrospective evaluation of what I have done and what it means to me, much less to anyone else.  But memory is fickle; indeed, one presently favored theory by Bolles claims that human memory is simply an act of imagination.  I write this down now for myself.  Perhaps twenty years hence, if my mind is still in working order, I will reread it and try to find out if I remember these things the same way.  I used to think that for a scientist, an important achievement is a contribution that is a lasting one.  I was responsible for replacing the West-Gaeke procedure for measuring sulfur dioxide; it did not seem to bother Phil the least!  The West-Gaeke Procedure reigned for thirty years.  The replacement, I am convinced, will not be useful nearly half that length of time.  Does this mean that the longevity of equivalent achievements have decreased, or did I simply replace it with something that is a lesser achievement for its time than the West-Gaeke reaction was for its era?  These are questions I try not to ponder about; I am certainly not the best person to answer them.  If all that I have done and all that I do turns out to be nothing more than water stains on sand, as it were, I pray that twenty years hence I can still say I am having too much fun to care.  It is surprising, as someone once remarked, that people would actually pay you for having fun!  And chemistry (chemistress, some maintain) and me, we have had lots of fun.  At least at this point in time, it seems to be no less colorful than when I first mixed copper sulfate and ammonium hydroxide in my chemistry set.