Opportunities in Chemistry: Today and Tomorrow (1987)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

CHAPTER IV Intellectual Frontiers in Chemistry A remarkable bounty of benefits has been shown to flow from chemistry. This chapter will provide abundant evidence that these benefits will increase greatly in the years to come. The basis for this optimistic expectation is that this is a time of special opportunity for intellectual ad- vances in chemistry. The opportunity comes from our developing ability to in- vestigate the elemental steps of chemical change and the ability to deal with ex- treme molecular complexity. His

The Time It Takes to Wag a Tail When your pet dog sniffs a bone, instantly his tail begins to wag. But it must take some tune for the northernmost canine extremity to send the news all the way south where enthusiasm can be registered! How long does it take for that delicious aroma to lead to the happy response at the other end? Chemists are now asking questions much like this about their pet molecules! If one end of a molecule is excited, how long does it take for the other end to share in the excitement? That time may determine whether the excitation will result in a chemical reaction in the part of the molecule where the energy was Projected, somewhere else, or nowhere at all. For the canine expenment, we need a hungry dog, a quick hand with the bone, and a quick eye to read the stopwatch. For molecules, it's much harder. Only within the last few years has it been possible to measure the rate of energy movement within a molecule. But chemists now have pulsed lasers gong bursts of light with durations as short as a millionth Of a millionth of a second (a "picosecond''). Comparing a chemical change that takes place one picosecond to a one-second tail-wag delay involves the same speed-up as a l~second instant replay of all historical events since the pyramids were built. The aLa~yl benzenes provide an example. Each of these molecules has a nod benzene Iing at one end and a flexible alkyl group at the other. At room temperature, this flexible "tail" vibrates and bends under thermal excitation. But to act like our hungry dog, the molecules must be cooled to cryogenic temperatures, while avoiding condensation. Supersonic jet expansion makes this possible. When a gas mixture cows through a jet nozzle into a high vacuum, the molecules can be cooled almost to absolute zero. An alkyl C' ~ :~/C~1 1 benzene molecule camed along in such a stream loses all its vibrational energy, thus relaxing the molecular tail. Then, the cold molecules intersect a brief pulse of light with color that is absorbed by the benzene ring. With careful "color-tun~ng,~' extra vibrational energy can be placed in the head without any v~b~onal excitation In the tail. Then we must watch the molecule to see how long it takes for the tail to wag. Fluorescence lets us do this. When a molecule In a vacuum absorbs light, the only way it can get rid of the energy is to reelect light; Such fluorescence can be recorded with a fast-response detection system to give a spectrum that carries a tell-tale pattern showing where We extra energy was at the instant the light was emitted. Those molecules that happen to emit right away after excitation show the molecule head vibrating and the tail still cold. Those Mat emit later have an emission spectrum Mat shows that the tail is wagging. In this way, we have learned that the time it takes for the aLky! benzene ~ to begin to wag depends on how long the tad] is. Su~pnsingly, the longer the aLkyl, the faster the movement out of the nng. The result shows what detentes energy flow within molecules (the "density of sagest. Such information might one day charm combustion and help us make fine chemicals out of coal. ~6 r

IV-A. CONTROL OF CHEMICAL REACTIONS IV-A. Control of Chemical Reactions Ultimately, success in responding to society's needs depends upon the ability to control chemical change, a control made possible by our understanding of chemical reactivity. Today, this understanding is being broadened and deepened at an astonishing pace because of an array of powerful new instrumental techniques. These instruments permit us to pose and answer fundamental questions about how reactions take place, questions that were beyond reach only a decade ago. They account for the recent acceleration of progress in the most basic aspects of chemical change. MOLECULAR DYNAMICS Chemistry is the science concerned with the changes that occur around us when one set of chemicals turns into another set of chemicals. Such a change, a chemical reaction, is understood at the atomic level in terms of one set of molecules rearranging into another set of molecules. The study of these rearrangements is called molecular dynamics and it encompasses: · molecular structure, the stable geometries of the reactant and product mole- cules; · chemical thermodynamics, the energy effects that accompany the change; and · chemical kinetics, the time it takes for the reaction to occur. The theory behind all chemical behavior rests in quantum mechanics. Quantum mechanics is the mathematical description of atoms and molecules devised by Erwin Schroedinger in 1926. It is based upon a wave-picture of the atom that has the potential for explaining all of the chemistry of that atom. Though this has been known for over 50 years, most of the predictive power of quantum mechanics has been out of reach because the mathematics has been too difficult to solve. In contrast, experimental progress on stable molecules has been extremely rapid. This is evident in the fact that chemists have prepared more than ~ million compounds, 95 percent of them since 1965. On the other hand, our understanding of the speed aspects of chemical change has been limited by reaction steps too fast to be observed. Now a new era has begun. Chemical theory, supported by the power of modern computers, has emerged from empirical modeling. At the same time, we have expenmental techniques that open the way to understanding the time dimension of chemical change. Over the next three decades we will see advances in our understandings of chemical kinetics that will match the advances in molecular structures over the last three decacles. Fast Chemical Processes A chemical reaction begins with mixing reactants and ends with formation of final products. In between, there may be a succession of steps, some extremely rapid. To understand the reaction completely, we must cIanfy all the steps between beginning and end, including identification of all of the intermediate molecules that are involved in the steps. ~7

118 INTELLECTUAL FRONTIERS IN CHEMISTRY Fifteen years ago, we could track intermediate molecules only if they hung around at least as long as a millionth of a second. The many interesting studies on this time scale only increased the chemist's curiosities because it became clear that a whole world of processes took place too rapidly to be detected at that limit. Nowhere was that more apparent than in the centuries-old desire to understand combustion, perhaps the most important type of reaction known. Laser light sources have spectacularly expanded these experimental horizons over the last decade. One of their unique capabilities is to provide short-duration light pulses with which to investigate chemical processes that occur in less than a millionth of a second all the way down to a millionth of a millionth of a second (i.e., down to a picosecond, 10- ~2 see). At the state of the art, physicists are learning how to shorten these pulses even more; pulses as short as 0.01 picoseconds (10 femtoseconds) have been measured, and kinetic studies are beginning in the 0.1-picosecond range. At one-tenth of a picosecond frequency accuracy is limited ~ ~ . ~ . . . . · , · · . .. .. . · , ~ · · . to about 50 cm-l by a fundamental physical principle the Uncertainty Principle (See Section V-A). These developments imply that chemists can now investigate a reacting mixture on a time scale that is short compared with the lifetime of any intermediate molecular species involved. The exploitation of this remarkable capability has only just begun. The absorption of visible or ultraviolet light by a molecule adds enough energy to redistribute the bonding electrons, to weaken chemical bonds, and to produce new molecular geometries. The outcome might be a high-energy molecular structure difficult to reach by chemical reactions stimulated by heat. So the excited electronic states reached by absorption of light furnish a new chemical world that we have only begun to understand and put to practical use. When a molecule absorbs light, it gains energy. One of the ways it can dispose of the energy is to reemit light, generally of a different color than the absorbed light. If this emission occurs quickly, it is called fluorescence. "Quickly" can mean anywhere from within a microsecond to a picosecond. The blue light emitted by a Bunsen burner flame and the spectacular display of the Northern Lights are examples of fluorescence. If the light emission occurs more slowly, it is called phosphorescence. "Slowly" can mean anywhere from a millisecond to several seconds or even minutes. Some clock dials that glow in the dark and the blue glow of evening ocean tides are examples of phosphorescence. We have some basic understandings about the differences that cause these two behaviors. When two electrons are shared in a chemical bond, they must have opposite magnetic spins (as expressed in the Pauli Principle). But if absorption of light adds enough energy to move one of these electrons to another part of the molecule, the Pauli Principle no longer limits the electron spins. Then they can be oriented opposed to each other, like two magnets whose fields cancel each other, to give a "singlet" state. But they can also be oriented parallel so that the two magnetic fields add together. This is called a "triplet" state. We have learned to associate fluorescence with light emission processes that begin and end in singlet states. Phosphorescence, however, requires moving from a triplet to a singlet state (or the reverse). Apparently, the need to change the electron spin makes the emission much more difficult, so it occurs more slowly.

lV-A. CONTROL OF CHEMICAL REACTIONS There has been a spectacular increase in our ability to clarify what is going on in these excited states since lasers have come into the chemistry laboratory. We can now excite particular states (by control of the laser color, or wavelength), and we can measure the time it takes for reemission to occur (by use of laser pulses of very short duration). Even for the fastest fluorescent processes, we can measure the radiative lifetimes, and by measuring the wavelength of the light emitted (spectral analysis) we can see how rapidly energy moves within the molecule and where it goes. Thus, we are beginning to map and understand the high-energy electronic states of molecules so that they can be used to open new reaction pathways. Benzophenone is a substance that demonstrates how lasers are being used to probe these high energy states. When benzophenone in ethanol solution absorbs ultraviolet light at a 316-nm wavelength, it reemits light at two different colors, at wavelengths of 410 and 450 nm. If the exciting light (316 nm) is delivered in a laser pulse of 10 picoseconds duration, "prompt" emission is seen at 410 nm, with intensity that decreases with a 50-picosecond half-life. This fluorescence is fol- lowed, however, by weaker emission, still at 410 nm, but with a longer half-life (a microsecond). This slower fluorescence disappears at lowered temperatures and is replaced by longer-wavelength phosphorescence at 450 nm with an even longer lifetime (a millisecond). Photochemists have been able to interpret these clues about the excited states of ben- zophenone. Absorption at 316 nm reaches a singlet state (S~) but with extra energy placed in the vibrational motions of the benzophenone. This vibra- tional excitation is lost so quickly in liquids (warming the solvent) that even the "prompt" fluorescence back to the ground state (S0) occurs at longer wavelengths (410 nary). On the other hand, the low-temperature behavior shows that benzophenone also has an excited triplet state (T~),that can be reached via So. Once occupied, To emits phosphorescent light with the characteristic long lifetime of a triplet-singlet transition (T~ > Ski. The temperature dependence of the delayed fluorescence shows that To is lower in energy than So and by how much. The set of processes clarified here have lifetimes that range from 50 picoseconds to a millisecond, a difference of 20 million. The observations reveal the excited states of benzophenone and the rates of movement between them. These under- standings are of extreme significance because they can all be applied to natural photosynthesis, a process scientists would like very much to master. There are many other types of laser-based, real-time studies of rapid chemical reactions now o FLUORESCENCE (3 _ _ ~ - 'A _ _ S 1 S I _ ~ T I 'PROMPT' 'DELAYED' he he' he he/' 50 psec 1 User _ _ _ _ _ ~ _ _ so sow _ 50-10-12 sec 1. 10-6 sac PHOSPHORESCEN CE ~3 _ ~ . ~ T1 ~ LOW T he ~ he" _ / I msec _ ~ _ _ Hi= so- 1~10-3 sec EXCITED BENZOPHENONE EM ITS LIGHT WITH TWO COLORS AND THREE CLOCKS 119

120 HOT RING COLD TAIL _ COLD RING COLD TAIL INTELLECTUAL FRONTIERS IN CHEMISTRY being made, including chemical isomerizations, proton transfers, and photodisso- ciations. Some of the phenomena to follow also depend upon use of short-puIse laser excitation instrumentation. Energy Transfer and Movement In all chemical changes, the pathways for energy movement are determining factors. Competition among these pathways determir~es the product yields, the product state distributions, and the rate at which reaction proceeds. This competition is highly important in stable flame fronts (as in Bunsen burners, jet engines, and rocket engines) and in explosions, shock waves, and photochemical processes. When two gas phase molecules collide, vibrational energy can be transferred from one molecule to another. Thus, a vibrationally "cold" molecule might be heater! up and caused to react or a vibrationally "hot'' molecule might be cooled off so it cannot react. These transfers of vibrational energy between and within molecules as a result of collisions between them have long been recognized as central to determining reaction behavior ire flames. But progress has been slow because the processes have been too fast to measure. Now a variety of tech- niques almost all based on laser methods—has opened the way to providing critical data related to the pathways and rates of energy flow. These data, in turn, furnish a basis for the develop- ment of useful theory. As much has been learned about vibrational energy movement in the last 15 years as was learned in the preceding half- century. As tuned lasers became available they were used to ex- cite particular vibrations in a molecule. Then, experiments were devised to permit us to watch this carefully placed en- ergy move into other parts of the molecule or into another molecule if collisions occurred. Fluorescence provides one way to follow this energy movement. The light reemitted during fluorescence carries a spectral signature that shows what part of the molecule is vibrating at the moment of emission. A clear-cut example is provided by recent studies of the alkyd benzenes, C6Hs-(CH2)nCH3 with n from 1 to 6. This molecule has a structure like that of a tadpole, where n determines the length of its tail. Tuned-laser excitation allows - FLUORESCEN CE REVEALS INTRAMOLECULAR VIBRATIONAL REDISTRIBUTION _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ~ ~ a ~ c= l- ~~' COLD RING HOT TAIL

IV-A. CONTROL OF CHEMICAL REACTIONS us to deposit prescribed amounts of vibrational energy in the benzene end of the cold molecule (in the head of the tadpole). When this energy is reradiated, its spectral signature displays its vibrational excitation at the instant of radiation. Since this light emission is a time-dependent process, we can monitor the movement of energy from the original location of excitation into the rest of the molecule. This movement in absence of collisions is called Intramolecular Vibrational Redistribution (IVR). Light emitted in the first picoseconds shows that the energy has not yet left the benzene unit where it was absorbed. The time scale for appearance of vibrational excitation in the alky! tail depends upon the tad] length. For n - 4, vibrational energy moves out into the tail in 2 to 100 picoseconds. In contrast, for n = 1 (ethy~benzene), it is a thousand times slower, it takes 100 nanoseconds or more. Thus, we have direct evidence about the . . .. . . .. · . factors that determine IVR energy movement in an isolated molecule. State-to-State Chemistry When two gaseous reactants A and B are mixed and react to form products C and D, the outcome is determined by statistical probabilities. The different encounters that may happen between A and B include all the possible energy contents, specific different types of excitation, and all the ways molecules may be oriented in space at the moment of collision. Not ad of these collisions are favorable for reaction most collisions have too little energy, or the energy is in the wrong place, or the collisions are at an awkward geometry. If we are to understand finely the factors that permit chemical reactions to occur, we should control the energy content of each reactant, i.e., control the "state" of each reactant. Then we could systematically vary the amount and type of energy available for reaction. Finally, we would like to see how the available energy is lodged in the products. Such an experiment is ceded a "state-to- state" study of reaction dynamics, and 20 years ago it was beyond all reach. Now, with modern instrumentation, chemists are realizing this goal. The earliest efforts, based upon chemiluminescence, revealed a part of the picture: the energy distribution among the products. For example, when a gaseous hydrogen atom and a chlorine molecule react they form hydrogen chloride and a chlorine atom. These reaction products emit infrared light. Analysis of the spectra from that light shows that the energy released in the reaction is not randomly distributed between the final products. Instead, a large fraction of it (39 percent) is initially located in the vibration of the hydrogen chloride product. Discoveries like this won John Polanyi (University of Toronto) a share of the 1986 Nobel Prize in Chemistry. This measurement led directly to the demonstration of the first chemical laser a laser that derived its energy from the hydrogen/chlorine explosion. Chemical lasers differ from conventional lasers in that the energy to produce their light comes from a chemical reaction instead of an electrical source. These beginnings led to the discovery of dozens of chemical lasers, including two sufficiently powerful to be considered for possible initiation of nuclear fusion (the iodine laser) and for possible military use in the "Star Wars" program (the hydrogen fluoride laser). "Molecular beams" move even closer toward "state-to-state" investigations. A molecular beam is a stream of molecules produced by a suitably hot oven. A 121

122 Cot c Cat INTELLECTUAL FRONTIERS lN CHEMISTRY CHEMICAL LASERS REVEAL THE PRODUCT ENERGY DISTRIBUT ION F+ H2 ~~k ~ ko: ~ = 0 HF~ H RERCTIQn COORDInRTE substance is placed in this oven, and when it melts and vaporizes the vapor is directed out a tiny hole to form a unidirectional beam of molecules. Out- side the oven the pressure is kept extremely low- so low that no molecular collisions occur. The molecular beam can then be directed toward reac- tants. In such experiments, the reactants collide at such low pressures—10-~° atmospheres that each reactant molecule has at most one collisional opportunity to react, and the products have none. These sophisticated instruments depend upon ul- tra-high vacuum equipment, high-intensity super- sonic beam sources, sensitive mass spectrometers for detectors, and electronic timing circuitry for time-of-flight measurements. With this incredible control it has become possible to predetermine the energy state of each reactant molecule and then to measure both the probability of a certain reaction and the energy distribution in the products. For bringing such elegant experi- ments into chemistry, Yuan-Tseh Lee (University of California, Berkeley) and DudIey Herschbach (Harvard University) shared in the 1986 Nobel Prize in Chemistry. For example, a current study has explained a key reaction in the combustion of ethylene. These molecular beam experiments show that the initial reaction of oxygen atoms with ethylene produces the unexpected short-lived molecule CRECHE. With this starting point, calculations have confirmed that a hydrogen atom can be knocked out of an ethylene molecule by a reacting oxygen atom more easily than that atom can be moved about within the molecule. This combustion example illustrates the intimate detail with which we can now hope to understand chemical reactions. Multiphoton and Multiple Photon Excitation Photochemistry has traditionally been concerned with what happens when a single photon is absorbed by an atom or a molecule. This productive field accounts for the energy storage in photosynthesis, the ultimate source of all life on this planet. Photochemistry also provides us with new ways to synthesize organic compounds and, through photodissociation, to produce a variety of short-lived molecules that play critical roles in flames and as intermediates in reactions. Now lasers give us optical powers lO,OOO times higher at a given frequency than even the largest flashiamps ever built. Clearly, these devices do not simply extend the boundaries of conventional light sources, they open doors to new processes as molecules interact with such intense photon fields. For example, at normal light intensities, the simultaneous absorption of two photons by a single molecule takes place so rarely that it cannot be detected. However, the probability of this happening increases with the square of the light intensity. Thus, if a laser increases

lV-A. CONTROL OF CHEMICAL REACTIONS light intensity by a factor of 10,000, then the chance of two-photon absorption increases by four orders of magnitude over the chance of one-photon absorption. This lets us do experiments in which we can prepare molecular states that cannot be reached with a single photon. Furthermore, the total energy absorbed can be enough to produce ions. This opens a new avenue to the chemistry of ions, a field of rapidly rising interest because of the discovery of interstellar ion-molecule reactions and because ions are major species in the plasmas (glow discharges) of nuclear fusion. Two-photon ionization has been used to detect specific molecules in difficult environments, like those found in explosions and in flames. Thus, nitric oxide, NO, which is an ingredient in smog, can be easily measured in a flame by counting the ions produced by a finely tuned laser probe. The probe is tuned so carefully that only the desired molecule, NO, can absorb light energy. However, the most spectacular instance of multiphoton excitation came with the development of extremely high-power CO2 infrared lasers. One of the most surprising scientific discoveries of the 1970s was that an isolated molecule whose vibrational adsorptions are in close vibrational harmony (near resonance) with the laser frequency could absorb not two or three but dozens and dozens of photons. In a time short compared with collision times, so many pho- ~ ~ ~ tons can be absorbed that ~ E _ chemical bonds can be broken QUASI ~ I entirely with vibrational exci- CONTINUUM ~ Ml is: S _~ ration. This unpredicted be- ~ --Gus''' _ ~ ~ RESONANT ~_L'~!~~r;~. ~ havlor 1S commonly called mul- ABSORPTION_ ~ --,.~; is. DISSOCIATIVE tiple photon excitation to dis- - .~. - - CONTINUUM tinguish it from two-photon ' T (multiphoton) excitation. ~ ~ -'- ~ '' - . - This behavior stunulated a 32 large group of studies on energy 34s~6 hV 3~` flow within excited polyatomic SF6 ~ hV ~ SF6 molecules. Many un~molecu- ~ hV 34sF ,, tar breakdowns and rearrange- ments have been triggered using multiple photon excitation. Yet, the understanding gained from this phenomena may be over- shadowed by the importance of its practical uses. Infrared absorption depends upon vibrational movements whose frequencies are quite sensitive to atomic mass. As a result, the tuned laser can be used to break up just those molecules containing particular isotopes, leaving behind the others—a new method for isotope separation. For example, deuterium is present at 0.02 percent in natural hydrogen. Yet, by multiple photon excitation, this tiny percentage can be extracted using trifluoromethane, CF3H. The process has been shown to have a 10,000-fold preference for exciting CF3D over CF3H. This could be of considerable importance as a source of deutenum since '`heavy water," D2O, is used in large quantities in some nuclear reactors. ISOTOPE SEPARATION THROUGH MUTT I PHOTON ho 34SF $$* EXCITATION nhy 3.SF`,nt 1' 3 SF5 ~ F 123

124 INTELLECTUAL FRONTIERS lN CHEMISTRY Even restore significant is sulfur isotope separation through excitation of sulfur hexafluoride, SF6. This gaseous compound gave the first convincing evidence that multiple photon excitation really occurred so rapidly that collisional energy transfer could be avoided. The successful use of SF6 for sulfur isotope separation could have heavy significance in human history. The gaseous substance that has always been used in the difficult processes used to separate uranium isotopes is uranium hexafluoride, UFO. Because SF6 and UFO have identical molecular structures, they have similar vibrational patterns. Thus, multiple photon excitation might offer a new and simpler approach to isolation of the uranium isotopes that undergo nuclear fission. It depends, of course, upon finding a sufficiently powerful and efficient laser at the lower frequencies absorbed by UFO. It will bring more general access to the critical ingredients of nuclear energy and, unfortunately, nuclear bombs as well. The dangers of increased proliferation of nuclear weaponry can only be increased by such access. Mode-Selective Chemistry When two molecules collide with each other, the violence of the collision may cause their atoms to rearrange to form two new molecules (i.e., a reaction may occur). Such an outcome al- most always requires that the molecular collision involve some minimum energy- enough to break some of the bonds in the reactants in order to form the new bonds in the products. This minimum en- ergy, the activation energy, de- termines the rate of the reac- tion and it accounts for the dramatic effect of temperature on reaction rates. However, the question of whether a reaction will result from a molecular collision turns out to involve more than just whether there was enough energy. There is also a ques- tion of whether the collisional energy is in the right form. To understand what this means, consider a bedspring thrown against a wall. As it bounces off, it has energy of several types. It will be moving through space, which is energy of the old-fashioned kinetic en- a! ~ VIBRATION STRETCH I NO $ BEND I GIG H TRANSLATION ~ C H H C \ ' H H ROTATI ON . - ~i,~ .. ' ,,. ' ,,. ' ,,.. ,, ,,.—,,~, ~ . ,, 7 .. 7 ,^~ ~ i< ~ ~~ ~ ,. .. ~= ~ i' ,. ~ ,,. ~ ~ ,. /",~,,r,~t " Z 4, ~ ,. ~ ,. ~~~, ~ ,. " , " ,. ~ , ,"' ," ' j'C ~,~ H7-C ~ H_ ~ " C `. r ~ an H ~ C _ _ \~ H MOLECULES TRANSLATE ROTATE AND VIBRATE LIKE A BEDSPRING THROWN AGAINST A WALL

~-A. CONTROL OF CHEMICAL REACTIONS ergy type. This is called translational energy. In addition, the bedspring will be tumbling in space. This, too, is a form of kinetic energy called rotational energy. Then, the spring will be twisting and vibrating to and fro. This vibrational energy consists of both potential and kinetic energy. Molecules carry energy in exactly the same ways. Whether we are talking of bedsprings or molecules, the directions of translational motion, the axes of rotation, and the spring connections (in molecules, the bonds) are called degrees of freedom. The total energy in a collision is the sum of all of these forms of energy translational, rotational, and vibrational from both molecules. Chemists have long wondered whether it matters which degree of freedom carries the energy in a reactive collision. If all of the energy is in translational energy, the molecules are near each other only a short time. If the same amount of energy is brought to the collision mostly as vibration, the molecules move toward each other slowly, but now the bonds that must be broken are vibrating rapidly. Is this more or less effective? Only since chemists have acquired lasers has it been possible to seek an answer to this fundamental question. With high-power, sharply tunable lasers, we can excite one particular degree of freedom for many molecules in a bulk sample. As long as this situation persists, such molecules react as if this particular degree of freedom is at a very high temperature while all the rest of the molecular degrees of freedom are cold. The chemistry of such molecules has the potential to show us the impomnce of that particular degree of freedom in causing reaction. This is called mod~e-selective chemistry. Both unimolecular reactions and molecular beam studies of bimolecular (two- molecule) reactions escape this problem. Unimolecular reactions involve only one molecule, so collisions are not required. At sufficiently low pressures, the effects of selective excitation on reactivity can be studied. The beam experiments sidestep the problem by giving each excited molecule only one chance for collision and by noticing only those collisions which result in a reaction. Nevertheless, mode- selective reactions are not readily coming from such experiments. Apparently the problem is that vibrational redistribution takes place within molecules even without collisions (IVR). This problem is of such basic importance to molecular dynamics that it will be one of the most important study topics for the next decade. There is, however, evidence for two-molecule mode-selective chemistry in certain solid inert-gas environments. In this situation the environment is so cold (IOK) that the reactive molecules are held immobilized. They are "frozen" in a prolonged, cold collision and rotational movement has been halted. For example, fluonne, F2, and ethylene, C2H4, suspended in solid argon at lOK do not react until one of the vibrational motions of ethylene is excited with a resonantly tuned laser. Then it is found that the most efficient vibrational motions are those that distort the molecular plananty. This is plausible because this type of distortion changes the molecular shape "toward'' the nonplanar, ethane-like structure of the final product. Theoretical Calculations of Reaction Surfaces Schroedinger~s wave equations of quantum mechanics have long been known to describe all chemical events. Yet quantum mechanics has been used in chemistry 125

126 INTELLECTUAL FRONTIERS IN CHEMISTRY to ~4 ~ ~8 ~ ,C=C~ + ,C=C~ ~ ~ ~3 Quantu m v4+~8 Yield ~ T 0010 coos HE ED /~]J3+V~o _~ _d ~ I 1. 1 600 1 800 2080 Photon Energy (cm~~) ho HE , -C=C- ~ F2 ~ ~—C C—F D' AH I O K. in D H Sot id argon The Reaction Rate Depends SELECT I V ELY On The Mode Excited NEW REACTION PATHWAYS ~ ,C=~` ~ ~3 tD,C=~H V 10 only qualitatively, or with se- vere approximations, because the equations have been too difficult to solve except for the simplest molecules (like H2 and H21. Modern computers are changing this. With today's computers, the structure and stability of any molecular com- pound with up to three first- row atoms (carbon, nitrogen, oxygen, fluorine) plus various numbers of hydrogen atoms can be calculated without ap- Droximations. This capability opens to the chemist many sit- uations not readily available to experimental measurement. Short-lived reaction intermedi- ates, excited states, and even energy barners to reaction can now be understood, at least for small polyatomic molecules. Our increasing understanding and control of chemical reactivity is providing us with new reaction pathways in synthetic chemistry that are sure to lead to new products and new processes. Again, powerful instrumental techniques play a central role. Synthetic chemists are now able to identify rapidly and accurately the composition and structure of reaction products. This greatly speeds up the development of new synthetic approaches. Organic Chemistry Organic chemistry today involves three areas of emphasis. The first area concerns the isolation, characterization, and structural determination of substances from nature. New natural products are thus identified alkaloids and terpenes from plants, antibiotics from microorganisms and fungi, peptides and polynucleotides from animal and human sources. Chromatography permits purification and char- acterization of substances present in only trace amounts from complex mixtures. Thus, workers in pheromone chemistry regularly separate out microgram amounts of these biologically potent molecules. The next challenge lies in determination of their composition, overall structure, and three-dimensional stereostructure. Here nuclear magnetic resonance, mass spectroscopy, and X-ray crystallography fill essential roles. Using proton NMR, only 100 nanograms (10-7 grams) of a substance can provide crucial information about the number and types of molecular

IV-A. CONTROL OF CHEMICAL REACTIONS linkages. With only 100-picogram (10-~° grams) amounts, mass spectrometry contributes by furnishing precise molecular weights up to 13,000 and, through the fragmentation patterns, providing revealing clues to substructures. Then, if 10 micrograms or more of a crystalline matenal become available, every stereochem- ical detail of structure is displayed through X-ray spectroscopy, such as interatomic distances, bond angles, and any mirror-image relationships present. Physical organic chemistry is the second major area of emphasis; it seeks to relate changes in physical, chemical, and spectroscopic behavior of organic compounds to changes in molecular structure. It deals with the detailed pathways by which reactants become products—it predicts what intermediate species or structures are present and determines how the reaction pathway is influenced by solvent environment, catalysts, temperature, and pH. It provides a theoretical framework with which to predict behavior and useful synthetic routes toward materials not yet known. Synthesis, the third area, is a process of inventive strategy. Two contemporary challenges it faces are to add to the availability of useful natural products and to synthesize new and useful substances not found in nature. Thus, thousands of pounds of ascorbic acid (Vitamin C) are synthesized annually at purities suitable for human consumption so that society can have an abundant supply of this healthful substance. Smaller amounts of 5-fluorouracil, an artificial drug that is extremely effective in treating certain cancers, are synthesized for prescription use. Meeting such challenges has required creative evolution of the philosophy of organic synthesis. Only a few decades ago, synthetic strategies were based on clever choices from a set of already known reactions. Like moves in a chess match, the range of logical reactions was defined in advance. With the development of mechanistic reasoning, in which reactions are classified according to the mecha- nism by which they work, it has now become possible to invent new reactions for specific synthetic goals. Organic synthesis has had powerful success using this reasoning process. At the same time, there has been an imaginative and fruitful expansion in the settings in which reactions are conducted. An example is solid-phase peptide synthesis in which amino acids are added in sequence to produce a desired peptide. This is all carried out under covalent attachment to an insoluble polymer support. Such polymer-bound peptide synthesis is already being applied to synthesis of important hormones and bioregulatory peptide substances. A quite different dimension now being explored is pressure. Equilibrium can be shifted to favor products with specially compact structures, and activation bamers can sometimes be affected to selectively speed up a desired process. A step in the synthesis of alkavinone, used in the synthesis of certain drugs, provides an example. At 15,000 atmospheres and room temperature, quinone will react with a correctly structured butadiene ester to form the desired bicycle ester. This process completely avoids undesired alternative structures that would be obtained if high temperature were used instead of high pressure as the control variable. Nowhere has progress had more far-reaching significance than in our growing ability to control molecular complexities in the third dimension. This frontier, stereochemistry, can be divided into issues of surface shape (topology) and 127

128 INTELLECTUAL FRONTIERS IN CHEMISTRY /CH3 \C'C~ ] Buted~ene Esters /0N o J: ~ 5.000 ¢~1 No Reaction ! o Qulnone ~ ~~ ~ 5.000 Cq~ + [5: atm. o Alkev~none a Drug Precursor H O Act/ MACH o Catalyst Zn(BH~)2 ~ ~ O ~ Cal [~,^ o RA I Sl N G PRESSU RE CA N SELECTI V ELY SPEED UP A DESIRED REACTION "handedness," the first being called "relative" stereochem- istry and the second "ab- solute" stereochemistry. The production of a particular mo- lecular topology already re- quires artful control of molec- ular relationships in space dur- ing reactions. However, this spatial control does not usually extend to relationships that dif- fer only in a mirror-image sense (in handedness or chiral- ity). When right- and left- handed molecular structures are possible, most chemical re- actions will produce a mixture of the two. Of course, a left-handed glove will not fit a right hand, so it cannot serve the function of a right-handed glove. It is the same in nature, where this "handedness" aspect of molecular structure assumes critical importance. Biological molecules must have proper topological conformation (relative stereochemistry), but for them to be functional, nature also insists upon a particular handedness (absolute stereochem- istry). A molecular "right-handed" glove can play a crucial role in a biological reaction while its "left-handed" counterpart will be totally ineffective or, worse, may introduce undesired chemistry. Though stereochemistry has been recognized for almost a century, major advances have been made within the last decade. In one technique, an extra molecular fragment of defined handedness is attached to a reactant. This "chiral auxiliary," properly placed, can govern the handedness of products that come from that reactant. The auxiliary is then removed from the product and reused in another cycle. Certain stereospecific propionates are synthesized in this manner and later used as precursors for making other biological molecules. Even more exciting is the use of asymmetric (chiral) catalysts to direct the handedness of the products. Asymmetric reduction is now a key step in the industrial synthesis of the important anti-Parkinson's disease agent Adopt. A more general application has been the development of asymmetric epoxidation through asymmetric catalysis. When an oxygen atom is inserted equally into either face of a carbon-carbon double bond to produce an epoxide, two mirror-image-related products result. With inexpensive and recyclable chiral catalysts, it is now possible to prepare whichever one of these two stereoisomers is needed. The resulting stereospecific epoxide can be used in many synthetic pathways, carrying along and preserving the left-right character. In a major application of this method, all of the six-carbon sugars that occur naturally have been synthesized perfectly with nature's preferred handedness.

IV-A. CONTROL OF CHEMICAL FACTIONS The significance of these new frontiers of organic syn- thesis can be seen in health applications. For example, the prostaglandins are a family of fatty acids that contain 20 car- bons and include a five-mem- bered ring. They seem to affect the action of hormones and, thereby, have important effects on the body, ranging from reg- ulation of blood flow to stimulating childbirth. We now know the structures of several of them, and we understand both their biosynthesis and their laboratory synthesis. Their synthesis in nature begins with polyunsaturated fatty acids that are a natural requirement in the diets of mammals. Su~pnsingly, these same polyunsaturated fatty acids are useful starting points for synthesis of another family of molecules, the leukotnenes, that have great potential for a variety of drug uses, including control of asthma. The ability of chemists to synthesize chemically modified prostaglandins and leukotrienes for biological testing is a triumph of synthetic organic chemistry. Equally far-reaching accom- plishments are connected with the synthesis of safe compounds for birth control (e.g., 19-norste- roids and 18-homosteroids), new antibiotics (e.g., modified cephalosponns and thienamy- cins), and drugs for hyperten- sion (Aldomet@) and ulcers (cy- metidine, Tagamet~. 129 ALLYLIC ALCOHOL EPOZIDATION EFFECTED BY LEFT HANDED CATALTST oW LEFT HANDED ~0H EPOZIDE COW / \ EPOZIDATION EFFECTED / \ BT RIGHT HANDED CATA LY ST '~'OH RIGI}T-HANDED EPOZ I DE NOW THE CATALYST CAN FIX THE DESIRED "HANDEDNESS OF THE PRODUCT Inorganic Chemistry There is great intellectual ex- clement now In Inorganic chemistry, much of it at the interfaces with sister disci- plines: organometallic chemis- try, bioinorganic chemistry, solid-state chemistry, biogeo- chemistry, and other overiap- ping fields. For example, there is growing awareness of the crucial roles played by inorganic elements in biological systems. Living things, far from being totally organic, depend sensitively on metal ions drawn from throughout the Penodic Table. Particular metal ions have critical OH Em'` H ,~o_o'H H OH ~~' ~ COOH POOH H/~ 1 -H2O ~_,COOH ENZ7MAT I C ~ LTA HYDROLTSI S' Hip—e<C 5 H 1 1 I H H H - H OH ~ r H OH H OH ~COOH C ~ HS 5 11 1 R R = Glutathione LTC R = Cys-gly LID H~COOH R=Cystelne LTE H LEUKOTRIENE B = LTB BIOSYNTHESIS OF LEUKOTRIENES

130 INTELl~CTUAL FRONTIERS IN CHEMISTRY roles in such essential life processes as transport and consumption of oxygen (iron in hemoglobins, absorption and conversion of solar energy (magnesium in chlorophyll, manganese in photosystem Il. iron in ferrodoxin, and copper in plastocyanine), communication through electrical signals between cells (calcium, potassium in nerve cells), muscle contraction (calcium), enzyme catalysis (cobalt in Vitamin Bib. This has led to a burst of research activity in the inorganic chemistry of biological systems. We are beginning to learn about the structures that surround the metal atoms and how these structures enable the metal atoms to react with such sensitivity to changes in pH, oxygen pressure, and to electron donors and acceptors. The answers to many of these crucial questions will come from the active field of organometallic chemistry. The molecule makers of the field use the latest spectro- scopic and X-ray diffraction techniques to unravel unexpected bonding patterns and structures. An example is provided by the large family of "sandwich" compounds that stemmed from the discovery of ferrocene, a compound in which an iron atom is placed between two flat C5Hs rings. Metal atoms can bond through conventional electron pair sharing, as in the gaseous molecule TiC]4 which is used in the manufacture of pure titanium metal for aircraft construction. In addition, because of the many vacant ~ orbitals in the elements from the middle of the Periodic Table (the "transition" elements), metals can act as electron acceptors (Lewis acids). Thus, in the compound Fe(CO)5, iron pentacarbony1, each O carbon monoxide molecule do- c to 111 o Iron pentacarbonyl Fe(CO )5 3d As 4p Pea .~0~0 0 OOOi dsp 3 Iron hesachioride ion _,1 FeCD6 ' 3d 4s 4p Fe 2 @~100 O 000 TRANSITION METALS USE VACANT VALENCE ORBITALS TO BOND TO ELECTRON DONOR LIGANDS nates a pair of electrons to a vacant valence orbital of the iron atom to form a stable struc- ture shaped like two pyramids. The carbon monoxide molecule, and any groups that take their places, are called "ligands." Some or all of them can be re- placed by other electron donors (Lewis bases) such as nitric ox- ide, NO; ammonia, NH3; halide ions, F-, Cl-, Br-; water, H2O; cyanide ion, CN-; and many more. A wide range of com- pounds results. For some metal atoms, even nitrogen, N2, can be convinced to take a ligand position, thus making it more susceptible to reaction ("activations. That is one of the ways in which organometahic chemists are attempting to discover new catalysts to "fix" nitrogen (convert N2 into NH3 for fertilizer use). The key to further progress is to understand the reaction mechanisms of these molecules. Through clever choice of attached groups (ligands) and control of metal atom oxidation states, organometallic chemists have prepared remarkable com- pounds that show selective reactivity toward molecules previously thought too inert to participate in useful chemical transformations. For example, a saturated hydrocarbon is one with no carbon-carbon double or triple bonds, so it is relatively unreactive. Now researchers have discovered rhodium and iridium compounds with

IV-A. CONTROL OF CHEMICAL REACTIONS phosphine (PR3) or carbonyl and pentamethyl cyclopentadienyl ligands that can attack the CH bonds of methane and cyclopropane. The challenge is to couple this important new reaction with other wed-known transformations so that saturated hydrocarbons can be used as feedstocks. The direct conversion of methane to methanol by such a process could have a tremendous impact on the world energy situation. In quite another direction, recent experimental developments permit us to study in the gas phase the weakly bound cluster complexes called "van der Waals" molecules. These are clusters made up of two or more molecules, all of which have comoleteIv satisfied bonding 3.20 D situations. The remaining in- teractions that such molecules exert toward one another are much weaker than normal chemical bonds. Nevertheless, these interactions are ex- tremely important; such "van der Waals" forces account for deviations from ideal gas be- havior, condensation of gases to liquids, and solubility. Such complexes can now be prepared and studied spectro- scopically in cryogenic (iow- o 11 C - ,C: 18 o H 0- C—0 A; 350^ 1.91~ ' F / H '`64 N N Or 131 1 .85D O / H C A r C. . . . . . . 1 2.84 ~ \ 11 349& to N 11 3.47 o VAN DER WAALS MOLECULES- THE WEAK INTERA CTIONS THAT GOVERN SOLUBILlTY. GAS IMPERFECTION, LIQUEFACTION ~ ~ ~ _ ~ temperature) matrices and under molecular beam conditions using supersonic jet cooling. These techniques have provided a wealth of information, including molecular ~eometrv. vibrational amplitudes. dinole moments. and ease of energy ~ , , ~ , ~ , _ ~ ~ . ~ . . .. ~ ,~ .. ... .. . . movement trom one part ot the complex to another. Information flue tills 1S important to the development of detailed theories of reaction rates and the prediction of reaction pathways. Further studies should help to explain such phenomena as condensation, solubility, and adsorption. At the solid-state/inorganic chemistry intersection is the opening field of com- posite structures. A composite is made up of two or more materials used together to take advantage of some of the properties of each. Multilayered ceramics for interconnections between semiconductor chips are now being fabricated as well as nonmetallic electrical conducting substances composed of alternating layers. Another new class of materials of considerable interest is the ultrafine filamentary composites. Filaments smaller than a human hair (500-1,000 A thick) are uniformly distributed throughout another material, which leads to dramatic changes in material properties. The challenge for the future will be to obtain a full understanding of such material interactions so we can design and synthesize new materials with properties to order. Selective Pathways in Organic Synthesis Selectivity is the key challenge to the organic chemist to make a precise structural change in a single desired product molecule. The different reactivity in each bond type must be recognized (chemoselectivity), reactants must be brought · Ar

132 INTELLECTUAL FRONTIERS IN CHEMISTRY ~b~, ADA M ANTANE ADAMANTADINE LAB CURIOSITY ANTIVIRAL AGENT together in proper orientation (regioselectivity), and the de- sired three-dimensional spatial relations must be obtained (ste- reoselectivity). The degree to which this type of control can be achieved is shown in the synthesis of the substance ada- mantane, Cohn. This unique molecule resembles in struc- ture a 10-atom "chip" off a diamond crystal. In a laborious synthesis, it was finally pro- duced in a many-step process, but in only 2.4 percent yield. Recent research in the synthesis of polycyclic hydrocarbons now allows production of adamantar~e in one step in 75 percent yield. Then a surprise practical payoff came when it was discovered that adding a single amine group to adamantane gives adamantadine (l-amino-adamantane), which is an antiviral agent, a preventive drug for influenza, and a drug to combat Parkinson's disease. Cycloaddition to make five-membered rings becomes important for a wide range of applications ranging from novel electrical conductors to pharmaceuticals (e.g., antibiotics and anticancer compounds). An example is the ring closed by a rhodium catalyst to form a critical pre- cursor to thienamycin. In this case, the five-membered ring contains a nitrogen atom. The final product proves to be a relative to penicillin and an im- portant drug in the battle against infectious diseases. At another extreme, large ring compounds have been exceptionally difficult to synthesize. Their structures are complicated by functionally important left/right-handed structural geometries (chiral centers). Their wide-ranging biological properties—from pleasant fragrances for perfumes to antifungal, antitumor, and antibiotic activ- ities- make large ring synthe- sis a useful and interesting chal- lenge. An example is erythro- mycin, CHIN, which can be shaped into 262,144 differ- ent structures derived from the many possible ways to couple the right- and left-handed- ness at chiral centers (2 i~ = 262,1441. Twenty-five years OH OH ~ to ~ ~S~NH3 - N_( Nit COOR COO TH IENAMYCINE CATALYTIC CLOSURE OF FIVE-MEMBERED RINGS H OH :e Me ~ O ~O~ NMe. HO TWO Jot, Me Me Me Met Me ERYTHROMYCIN ONCE CONSIDERED 'HOPELESSLY COMPLEX''

IV-A. CONTROL OF CHEMICAL REACTIONS ago, this compound was judged to be "hopelessly complex" by R. B. Woodward, who won the Nobel Prize for synthesizing molecules as complex as quinine and Vitamin By. Today we can hope to work to such a goal, in part because of the development of specially designed templates that bring together the end atoms of a 14-atom chain to form a 14-membered ring. This provides the structural framework of erythromycin, and it has already resulted in the synthesis of a number of ingredients of musk, a scented compound used by animals to communicate and by humans to make perfume. Crossing Inorganic/Organic Boundaries As indicated earlier, the traditional line between organic and inorganic chemists is disappearing as the list of fascinating metal-organic compounds continues to grow. Furthermore, research in developing new inorganic substances has provided a sue sing reward in their frequent applicability in organic synthesis. The borohydrides provide an example. These are reactive compounds of boron and hydrogen that are electron-deficient from a bonding point of view. But these borohydrides have proved to be valuable as selec- tive, mild reducing agents in organic synthesis. Silicon and transition metal organometallic com- pounds give other examples. Silicon compounds, for example, are used to fold a long molecular reactant precisely as needed to synthesize the mol- ecule cortisone. Now this valuable medicine can be made in fewer than 20 steps, at a yield 1,000 times higher than was achieved in the earlier, 50-step process. Cortisone is weD known in the treatment of arthn- tis. Unfortunately, experience showed that relief could be temporary and that continued use of corti- sone had undesired side effects. These developments made the new silicon-assisted synthetic routes aB the more valuable. Several cortisone analogs were pre- pared arid tested for their medical effectiveness. One such product, predn~sone, is more effective than cor- tisone, even when used in much smaller doses, with the result that side effects are much reduced. CH2OH c=o O~<OH , . H H Cortisone CH 20H C=0 O~OH Prednisone LESS ARTHRITIC PAIN, SMALLER DOSES Organometallic compounds furnish important intermediate steps in many organic reactions. Organometallics are electron-r~ch, and because of this, nature accomplishes a lot of its electron transfer through these compounds. OrganometaDics are easily oxidized by both inorganic oxidants and organic electron acceptors in solution and at electrode surfaces. It has been important to establish how these compounds make and break carbon-to-metal bonds rapidly, selectively, and with stereospecificity. Recent theoretical advances have been based upon the closeness of approach of the reactants at the time of electron transfer. In this picture, each reactant is considered to consist of an "inner sphere," occupied by the electron donor or acceptor (the metal atom), and an "outer sphere," occupied by the~ligands. Electron transfer reactions are classified according to the extent of penetration of these inner and outer spheres. 133

134 Normal Carbon Id Angle 1 09°, 1 20° 1 80° Yet all of these 'strained" compounds leave been _yr~thesized! C C CY CLO PRO PA N E 1 one: 60 Cal' Cal 1,~c-1-c C~ ~ CtJBANE Flee sol c c/l l#57c: ICY / PROPELLA NE tone 90°~/ ~ 6 cts ) | |: three -90°' ( 2 C's ) | Chic '' ~~l C C TETRAHEDRANE | three bo°/ (4 C's) INTELLECTUAL FRONTIERS IN CHEMISTRY Pathways Using Light as a Reagent Another promising chemical pathway is connected with the use of photons in chemical synthesis. Many natural products and complex molecules of medical importance involve high-energy ("strained") molecular structures. "Strained" molecules are those with uncomfortable or unusual bond angles. In traditional synthetic procedures, the aggressive reagents needed to force molecular reagents into uncomfortable geometry tend to threaten the fragile product. Photochemistry has been remarkably successful in avoiding this difficulty. The reason for this success is that absorption of light can change the chemistry of a molecule dramatically. After excitation, familiar atoms can have unexpected ideas about what constitutes a comfortable bond angle; functional groups can have drastically different reactivities; acid dissociation constants can change by 5 to 10 orders of magnitude; ease of oxidation-reduction can be drastically altered; and stable structures can be made reactive. The energy absorbed by the molecule puts its chemistry on a high-energy "hypersurface" whose reactive terrain can be totally unlike the ground state surface below, the one that chemists know so well. Many examples can be given to illustrate the possibilities. Most dramatic are those that involve cyclic structures that require unusual ("strained") bond angles around carbon. Thus, rings containing three or four carbon atoms are relatively unstable and, hence, difficult to synthesize. At first, they were looked for just because they were chemical oddities. Now we know that many biologically active molecules or their synthetic precursors con- tain such strained rings as essential structural elements, so their synthesis has assumed great practical importance. These unusual, energy-rich structures are natural targets for photon-assisted synthesis. The photon provides extra energy, and it places the reaction on a hypersurface where uncommon bond angles can be the preferred ge- ometry. Using these principles, chemists have made many molecules of bizarre structure. Appro- priately named cubane is an example: eight carbon atoms are placed symmetrically at the corners of a perfect cube. Once formed, the molecule is sur- prisingly unreactive. Prop ellane also involves eight carbon atoms, now in a structure made up of three squares sharing a side. Even more amazing is the family of tetrahedranes whose central structural element looks like a three-sided pyramid. Each corner carbon atom is simulta- neously bound to three others at 60° angles to form four interlinked, three- membered rings. C C Bach Carbo2 Elas C C CY CLOBU TA N E ~ one:gO Ad

IV-A. CONTROL OF CHEMICAL REACTIONS As mentioned above, these photochemical syntheses have proved to be much more than an intellectual chemical chess game. All these syntheses store energy in chemical bonds (the reactions are endother- mic). The energy can be recov- ered later for its own use or to energize subsequent synthetic steps to form other desired, energy-nch molecules. Among the important biological molecules already prepared photochemically are the alkaloid atisine, several mycine antibiotics, and precursors of Vitamin D3. To take advantage of the benefits offered by light-assisted pathways, chemists need to become as familiar with the energy geography of the multidimensional reaction hypersurfaces as they are with the ground reaction surfaces upon which stable molecules react. Lasers wiD be a powerful aid in this exploration. Already it is known that a 1 percent change in the wavelength for the exciting light (from 3,025 to 3,000 A) can double the yield in synthesis of provitamin D3 (a precursor to Vitamin D31. In the formation of the hormone mentioned above, the combination of wavelength control o (3~000 A) and low temperature (-21°C) can quadruple the product yield. - HO~) S 1 n - Hydrosyprevitamin O HOW— I n - Hydrosyprovltam1n D, TUNED LASER IRRADIATION DOUBLES THE EFFICIENCY OF THIS STEP TOWARD YITAMIN-1)3 SUPPLEMENTARY READING Chemical & Engineering News "Laser Vaporization of Graphite Gives Sta- ble 60 Carbon Molecule" by R.M. Baum (C.&E.N. staff), vol. 63, pp. 20-22, Dec. 23, 1985. "Chiral Boranes Could Launch Third Genera- tion of Organic Synthesis" by S. Stinson (C.&E.N. staff), vol. 63, pp. 22-23, Aug. 5, 1985. "Work on Polymer Models of Enzymes Forges On" (C.&E.N. stab, vol. 63, May 27, 1985. "Inorganic Macromolecules" by H.R. All- cock, vol. 63, pp. 22-37, Mar. 18, 1985. "Method Synthesizes Chiral Boranes in 100% Optical Purity" by S. Stinson (C.&E.N. staff), vol. 62, pp. 28-29, Mar. 26, 1984. "Technique Allows High Resolution Spec- troscopy of Molecular Ions" by R.M. Baum (C.&E.N. staff), vol. 62, pp. 34-35, Feb. 20, 1984. "Selective Laser Excitation Promotes Reac- tion" (C.&E.N. staff), vol. 61, pp. 25-26, April 11, 1983. "Cat Chemistry Spurs Cluster Catalyst Work" by J. Haggin (C.&E.N. staff), vol. 135 60, pp. 13-21, Feb. 9, 1982. Science "Molecular Beam Studies of Elementary Chemical Processes" (Nobel Prize Ad- dress) by Y.-T. Lee, vol. 236, pp. 793-798, May 15, 1987. Metals and DNA: Molecular Left-Handed Complements" by J.K. Barton, vol. 233, pp. 727-734, Aug. 15, 1986. "Methylene: A Paradigm for Computational Quantum Chemistry" by H.F. Schaeffer III, vol. 231, pp. 1100-1107, Mar. 7, 1986. "Theory and Modeling of Stereo-selective Organic Reactions'' by K.N. Houk et al., vol. 231, pp. 1108-1115, Mar. 7, 1986. "Selenium in Organic Synthesis" by D. Liotta and R. Monahan III, vol. 231, pp. 356-361, Jan. 24, 1986. Scientific American "Predicting Chemistry from (Molecular) To- pology" by D.H. Ronway, vol. 255, pp. 40-47, September 1986. "Quasicrystals'' by D.R. Nelson, vol. 255, pp. 43-51, August 1986.

Jack and the Soybean Stalk Perhaps a modern explanation for the amazing size of Jack's fairytale beanstalk can be found in brassirzolide. This remarkable chemical is an extremely effective plant hormone that can double the growth of food plants, by both cell elongation and cell division. Only recently have chemists been able to isolate, identify, and then synthesize this valuable substance so that it can be used to increase the world's food supply. Plant hormones have already revolutionized agnculture. They allow us to coerce cotton plants to release their cotton balls at harvest time, command fruit trees to cling to their fruit, induce Christmas trees to keep their needles, and order stored potatoes not to sprout. Brassinolide now can add to this list, and it is active in quantities of less than one-billionth of an ounce! Chemists play a crucial role along the long and arduous research road from discovery to use of a new plant hormone. For example, brassinolide is found in minute quantities in the pollen of the rape plant (Brassica rap us L.~. To isolate enough chemical to study, researchers laboriously collected pollen brushed off the legs of bees who had been cavorting in the rape plants. From 500 pounds of pollen so gathered, chemists were able to extract only 15 milligrams of brassirzolide, an amount as small as a grain of sand. Prom this they were able to grow a single tiny crystal, so that a chemical crystallog- rapher could analyze the molecular structure with x-ray d~ract~on. Just as X-rays penetrate an arm to reveal broken bones, they penetrate a crystal and reveal the geometrical arrangement of the atoms in brassinolide. The chemists were surmised to discover an unprecedented seven-atom nag within the molecule, a feature that must be essential to the function of this beneficial compound. With this key information, ~ synthetic chemists have now made several close relatives of brassinolide, and MU of, soybeans, and other vegetables. ' ~ This advance involved the knowhow ana ~nteracuon or plant ana Insect `~,, physiologists, organic chemists, and chemical crystallographers from many {' different laboratones. It shows that mental effort is as good as magic beans. 'fir Mavbe better. Jack! agncultura1 scientists are evaluating them in greenhouse production of potatoes, ) , ., ~~ I _^ '_ _ ~ it- -~\ 3~_~: it, . ~ ~~ d . . . . . .

IV-B. DEALING WITH MOLECULAR COMP~FXllY IV-B. Dealing with Molecular Complexity As detailed in earlier sections, natural products are enormously useful in meeting society's needs. These chemical substances include regulators of plants and insect growth, agents for communication among insects, pesticides, antibiotics, vitamins, drugs for cardiovascular and central nervous system diseases, and anticancer agents. As we develop these products, chemistry becomes a key science at every stage: natural products must be detected, chemically isolated, structurally charac- terized, and then synthesized as a final proof of structure. Chemical synthesis also provides sufficient amounts of important natural substances needed for biological testing. Chemical synthesis can also improve upon what nature has provided. Many natural products have useful biological properties, even though they are not ideal for our needs. For example, natural thienamycin has excellent antibiotic proper- ties, but the molecule is unstable and therefore unsuitable for use in human medicine. A synthetic chemical substitute has provided a stable molecule which promises much as an agent for fighting infectious disease. Thus, synthetic chemists have been able to follow a lead provided by a natural product to design and prepare a new molecule with even better biological and chemical properties. As was emphasized in the discussion of biotechnology, our understanding of macromolecules has provided new insights into their function in biological systems. These new insights have come from structural studies, synthetic alterations, increased understanding of the relationship between molecular structure and function, and the techniques of molecular genetics. SYNTHESIS AND BIOSYNTHESIS Modern synthetic techniques now provide access to molecules of complexity and structural specificity that were completely out of reach two decades ago. Synthesis of tailored peptides and nucleic acids of substantial size molecules widely useful in molecular biology and biotechnology has become routine. At the same time, our ability to understand and influence the synthetic processes of living organisms is advancing rapidly. All of this is based upon our impressive and growing power in synthetic and biosynthetic chemistry. Synthesis of Natural Products Over the last two decades, the synthesis of natural products has consistently advanced to new levels of molecular complexity. Chemists are now addressing a major challenge of organic chemistry, the synthesis of only one desired form of a mirror-image pair. In nature, many biological molecules can take different geomet- ric forms that are mirror images of each other. Each form is called a stereoisomer, and often only one form is biologically useful. Every carbon atom that has four different groups bonded to it gives rise to mirror-image pairs. Such a carbon is called a chiral atom or a chiral center. The synthesis of the polyether antibiotics is a prime example of the way in which chemists have met the challenge of stereoisomers. Monensin, a compound produced by a strain of bacteria called 137

138 INTELLECTUAL FRONTIERS IN CHEMISTRY Streptomyces cinnamonensis, is perhaps the best-known example from among a group of about 50 naturally occurring polyether antibiotics. Three polyether antibiotics (monensin, lasalocid, and salinomycin) are currently in use for control of infectious parasitic disease in the poultry industry (coccidiosis). Monensin has an American market of about 50 million dollars annually. Mor~ensin presents a huge challenge to synthetic chemists: 17 chiral centers are present on the backbone of 26 carbon atoms, which means that, in principle, 2~7 or 131,072 different stereoisomers exist for the antibiotic. Thus, to achieve the synthesis of monensin, it is es- sential to have a high degree of stereoselectivity. The successful total syn- thesis of monensin and its structural relatives (lasalocid, salirlomycin, and narasirl) in- volved revolutionary break- throughs. Until these achieve- ments, it was uncertain that a stereo-controlled reaction could effectively be realized in flexible noncyclic molecules. Encouraged by these results, chemists have now extended this approach to the synthesis of another group of antibiotics known as ansamycins. However, the most dramatic developments have been made in the chemistry of pay toxin. Palytoxin, a toxic substance isolated from marine soft corals of the genus Palythoa, is one of the most poisonous substances known; intravenous injection of only 0.025 micrograms into a rabbit can cause death. Pioneering investigations by organic chemists in Japan and Hawaii led to suggestions for the overall structure of palytox- in that indicated the unique- Iless of its structural complex- ity and molecular size. When synthetic chemists set their sights on the synthesis of paly- toxin, they were opening a new page ir1 the history of organic chemistry. This monster molecule con- tains 128 carbon atoms, 64 of which are asymmetric (chiral) centers. These centers, com- bined with seven skeletal dou- ble bonds, give palytoxin over two sextillion stereoisomers (2,000,000,000,000,000,000,000 = 2 x 102~! The basic structure had established the stereogeometry of 13 of the centers leaving 51 yet to be learned. Hence, the first step toward synthesis was to establish the stereochemistry of palytoxin. HO Me Me Me M~'CH2OH HO2C Me MONENSIN Me t 3 1 072 ~ IFFERENT STEREO ISOMERS THIS ONE IS EFFECTIVE! on 011 01! ~ Old OH o O _ ' ~ _ ' ~ ~ ~ ~ 'O ~J ~ X ~ N'——OH Oll Oll OH Old Oi] On 011 <'011 011 ~0t~or' O~Q~O~ 41~ (~ oll 011 OH 0~! ~ OH t`' 011 110 ~ Oil o 1 :23 / ~ 0~] 0~1 Oll Nli: ~ —No OH PALYTOXIN

IV-B. DEALING WITH MOLECULAR COMPLEXITY Enormous barriers stood in the way of the determination of the stereo structure of palytoxin. Chemists had only tiny amounts of the final product and, because it was not in crystalline form, X-ray analysis was of no use. Furthermore, nuclear magnetic resonance could not be conclusive because palytoxin is structurally too complex. However, organic synthesis was ready for the challenge, based on the experience gained with the polyether antibiotics. The researchers began with careful degradation of palytoxin to break it, chemically, into more manageable fragments. This degradation had to be gentle, so that each fragment would retain the stereochemistry it has in the parent molecule. Then, each fragment was synthesized in all of its isomeric forms to discover the shape that matched the natural product fragment. The process required that 20 different degradation frag- ments be synthesized, each HOW · . . . · ~ In its various stereo~somer~c forms, to identify the natural structure. The success of this endeavor has raised the sights of synthetic organic chemists everywhere. HOW nO<~NH it/ TYROSINE (' ).UORPHINE Biosynthesis of Natural Products Natural products have for many years played a central role in the development of or- ganic chemistry. The fact that many human medicines for the relief of pain and treatment of disease come from nature places a great deal of emphasis on this area of research. Mor- phine, for pain, and penicillin and erythromycin, with antibi- otic properties, are all found in nature. One approach is to H HO C ` - 'CO. H H. N' `1' CO H H N t' CO, H NH2 AMINOADIPIC ACID ~ By- 0 - ~, H SH H2N`/ Or OH CYSTEINE H I H2 N `: CO2H VALINE :'CO, H PROPION'IC ACID HN~SH r I of Nip CO H o my,' ~ N'' OH OH I ": POOH ~ HOW '~0 "/'040: ~ 0~'0 0~' OH OCH, (3) ERYTHRO$MYCIN NATURAL PRODUCTS POINT TO NEW SYNTHETIC PATHWAYS ~ of H.N, , ~_Sx 0~! CO H (2) PENICILLI.N study the ways in which natural compounds are actually formed in nature. Only recently has it become possible to test experimentally these possible biosynthetic pathways so that many of them are now reasonably well understood. The major experimental tool for biosynthetic investigations has been the use of isotopic tracers of the common elements, carbon (TIC, i4C), hydrogen (2H, 3H), nitrogen (ON), and oxygen (~70~. In this method, the isotope of an element is substituted for the natural isotope at a particular place in a reactant molecule. Then the isotoDe is looked for after reaction has taken place. using instrumental ~ ~ , c techniques. In this manner scientists can map reaction pathways. Extensive chemical degradations are no longer needed to locate sites of isotopic labeling because this task has been revolutionized by the development of stable isotope 139

140 c2H5 c2 ~5 H3 C \r INTELLECTUAL PRONTlERS IN CHEMISTRY NMR and the availability of high- resolution NMR spectrometers. Such NMR techniques have permitted the determination of the biosynthetic pathways that lead to certain powerful poisons that are produced by fungi and contaminate grain and other foodstuffs. These toxins, such as aflatoxins and trichothecin derivatives, pose major economic and public health threats. Recombinant DNA technology provides another set of potentially powerful new tools for the study of biosynthetic pathways. The antibiotics monensin and erythromycin, both discussed above, provide excellent examples. These two substances are structurally and stereochemically among the most complex natural products. Beyond the basic building blocks for each antibiotic (the simple sub- stances acetate, propionate, and butyrate), little is known about the details of the pathways by which these polyoxygenated, branched-chain fatty acids are assem- bled. Recent advances in the understanding of the genetics of the Streptomyces bacteria, along with the development of promising approaches to cloning for these organisms, have now made it more possible to unravel and perhaps control biosynthetic pathways at the genetic level. The Chemical Synthesis of DNA Section IlI-F on biotechnology described how nature encodes ire the molecular polymers of DNA the information needed to generate a living organism. A repeating skeletal chain of sugar-phosphate ester linkages provides a stable backbone upon which a message can be written using an alphabet of the four , cytosine, aIld guanine (A, T. C, and G). These nitrogen-rich cyclic amines are chemically bonded to the sugar groups in a sequence that carries the information. Although these amines are called "bases," in fact, each one couples the ability to form hydrogen bonds acting as an electron donor (a "base,') with the ability to form hydro- gen bonds acting as an elec- tron acceptor (a "proton donor" or "acidly. This hydrogen bonding capability furnishes the mechanism for amines: adenine, thymine Acid Base Bond (proton donor) (elect/on donor) Energy _ _ ~ to—H---- (3 it. . - \ H H H ~ - ~ N—H ~ -O ,0 2EI5 C2H5 N—H-----0~ /H cH3 C—N CH3 ~ cH3 , .. . . . water water 5 l~cal/mole dlemyl d1e~y1 2 3 amine emer teal fmole methyl methyl 3 6 acetam~de acetamtde Iccal /mole - .. , . ~. ~ _ . . . . . replication. The DNA double HYDROGEN BONDS - WEAK BUT IMPORTANT helix structure is held to- gether by hydrogen bonds be- tween each amine "acid/base" of the first strand with a matching or complemen- tary ~ base/acid'' amine on the second strand. Then reading the message of the DNA molecule can be accomplished merely by making and breaking these relatively weak hydrogen bonds without danger of breaking the stronger (sugar- phosphate) bonds of the template strand.

IV-B. DEALING WITH MOLECULAR COMPLEXITY The first chemical synthesis of a gene, about 15 years ago, required many person-years of effort. The remarkable (and continuing) progress since then permits synthesis of a similar size gene by a single researcher ire 2 weeks. There have been several syntheses of the gene for insulin in industrial laboratories, and a noteworthy synthesis of the gene for interferon in the United Kingdom. Each of these products shows promise for major medical and commercial value. The recent synthesis of the gene for the enzyme nbonuclease was designed to permit later alterations of the gene, making it possible to deliberately change the physical and chemical properties of this protein. Much progress is still needed. The yields of individual steps in DNA synthesis are still too low to permit routine synthesis of long molecules of DNA. State-of-the-art methods now can prepare gene fragments over 100 base pairs long, but we would like to deal with fragments 10 or 100 times longer yet. Costs for commercial custom synthesis of DNA molecules are coming down, but they can still exceed $200 per nucleotide. These synthetic nucleotide polymers of limited length are called oligonu- cleotides (from the Greek oligos, meaning few). Commercial machines for synthesizing DNA have only begun to meet the needed requirements for durability and depend- ability. Meanwhile, there is great excitement about the examples that are appearing. Synthetic oligonucleotides have been used to clone medically valuable proteins such as Factor ~ (a blood fraction used in the treatment of hemophilia) and commercially important proteins such as renin (used in the manufacture of cheese). The next decade will see continued efforts to alter the structure of enzymes to make them more useful in industry, to alter the structure of proteins and peptides to make new pharmaceuticals, and to uncover new knowledge concerning genetic regula- tion and human disease. STRUCTURES OF MACROMOLECULES The structures of the giant molecules of living systems- the proteins and nucleic acids—offer challenges just like those encountered for smaller natural products. We must first know which atoms are bonded to which In order to descnbe the covalent molecular structure. Then we must learn how the chains of these large polymers are onented in space, because the biological properties of the proteins and nucleic acids are intimately connected to their three-dimensional structures. This is especially true for proteins, whose remarkable range of biological functions has been descnbed in Section IlI-F. Following are some of the characteristics of proteins that allow them to be effective in areas ranging from digestion of food to blood transport of oxygen, and from contraction of muscles to antibody protection from viruses and bactena. These characteristics define some of the biological frontiers in which chemistry will play a central role. Proteins Have Complex Three-Dimensional Shapes That Relate to [Determined] Biological Function Chemical research of the past two decades has revealed that proteins have highly intricate three-dimensional forms and that these forms are cntically related to the 141

142 INTELLECTUAL FRONTIERS IN CHEMISTRY specific biological functions of each protein. A protein chain consisting of hundreds of linked amino acids takes on a three-dimensional architecture (called a confor- mation) that is determined by its particular amino acid sequence. For example, collagen, a protein that gives strength to skin and bone, has the shape of a rod. Antibodies are Y-shaped molecules with cavities that recognize foreign substances and trigger subsequent reactions for their efficient disposal. X-ray crystallographic studies have given valuable information about their architecture. Enzymes have clefts called "active sites" that bring reactants together and permit the formation of new chemical bonds between them. Thus, proteins have definite conformations that are at the heart of their biological roles. Major advances have been made in viewing these conformations using X-rays, neutron and electron beams, and other probes that enable us to "see" proteins magnified more than one million times. Clarifying these protein conformations shows us how biological functions are accomplished. We need to know much more about how proteins recognize specific sites on DNA and how they influence them. Additionally, we want to learn how peptides interact with receptor proteins to produce physiological changes in organisms. For example, the body produces a series of peptides called endo~phins, compounds that have painkilling and tranquilizing effects. Understanding of how the binding of these peptides to proteins on the surfaces of cells can lead to powerful changes in mood and consciousness will be a step toward unraveling the mysteries of brain function. Proteins Are Highly Dynamic Chemical studies of the past decade have also shown that proteins are highly dynamic molecules. Proteins change their shape while performing their functions. For example, light changes the conformation of rhodopsin, a protein in the retina, as the first step in vision. This structural change occurs in less than a billionth of a second. Such rapid changes in protein molecules can now be detected by using pulsed lasers. Another useful approach in the anal- ysis of protein dynamics involves cooling a protein to very low temperatures so that individual steps in its action are slowed down to permit more leisurely HE study. BACTERIORHODOPS IN Proteins Display Recurring Structural and Mech- anistic Themes Even the simplest cells contain more than 5,000 kinds of proteins. Yet we are finding that structural and mechanistic themes seen in one protein fre- quently recur in others. For example, there is a close relationship between the enzymes thrombin (for blood clotting) and chymotrypsin (for digestion). Moreover, the structures of many proteins have stayed the same over long evolutionary periods. There is surprisingly little difference, for example, between human and mouse hemoglobins. Enzymes work in complex organisms in much the same way that they do in simple ones. This knowledge is now being used to unravel disease

IV-B. DEALING WITH MOLECULE COMP~XI~ mechanisms, devise new diagnostic tests, and develop novel drugs and therapeutic strategies. Structural Studies on Dihydrofolate Reductases and Their Inhibitors Dihydrofolate reductase (DHFR) is an enzyme present in all living creatures, from bacteria to mammals. DHFR acts on d~ihydrofolate, which is a necessary ingredient in the complex chemistry of DNA synthesis in cells. Quite some time ago, it was noticed that feeding of folic acid, a source of dihydrofolate, actuaBy encouraged the growth of tumors present in laboratory animals. Rapidly dividing cells, like those found in a tumor, require equally rapid DNA synthesis and ample supplies of com- pounds like dihydrofolate. Hop- ing to find an antagonist that would block and reverse this effect (an "antifolate"), investi- gators set about synthesizing and testing many chemical ana- logs of folio acid. This approach paid off with the discovery of aminopter~n and later of metho- trexate. Amazingly, the essen- tial difference between these compounds and folate itself was simply the substitution of folate's 4-hydroxy! group by a 4-amino group. Thereafter, it was determined that methotrexate acts by inhibiting the enzyme DHFR. In fact, DHFR binds methotrexate so strongly that inhibition is essentially irreversible. This slows tumor growth by interrupting the action of DUFF and thus interfenug with DNA synthesis and cell division. Today, methotrexate is in widespread and effective clinical use for treatment of childhood leukemia, chono- carcinoma, osteogenic sarcoma, and Hodgkin's disease. Meanwhile, other, more distant analogs of folic acid were synthesized and tested in great numbers, including the substituted 2,4-diaminopyr~midines. This program led to the discovery of an agent useful against bacteria, called tnmethonnm. and ......... ...... .......... ; ; : NOT EFFECTIVE . . . . . i , . ; ; . i . . , .. .. :... 1~°: CO—NH—C3H 5—CASH ,.. ;..:i .:. :. .. ~ .. I......... ;: EFFECTIVE; ; l ASH N~Or CH2—~ ~CO—NH—C3H5—CASH H2N N N METHOTREXATE COOH INHIBITION OF TUMOR FORMATION LITTLE CHANGES CAN MATTER A LOT one elective against protozoa, premithamine, among others. All of these aIltifo- lates act by inhibiting DHFR. In some cases they are highly specific about which species of organism they work on. For example, trimethopam has about 100,000 times greater affinity for binding the DHFR of E. cold (Escherichia colic bacteria than for vertebrate DHFR. This fact makes trimethoprim safe for use as an antibiotic since it strongly prefers the bacterial enzyme. A decade ago, study of several DHFRs by X-ray crystallographic methods was begun to determine the molecular-structural basis for their action and to point the way toward a logical, structurally based approach to drug design. This X-ray crystallographic approach has begun to pay off. So far, the structures of DHFRs from three widely differing species, namely, the two bacteria E. cold and 143

144 Methotresate and NADPH Held in place by an enzyme INTELLECTUAL FRONTIERS IN CHEMISTRY L. cased (Lactobacillus caseiJ and the chicken (representative of vertebrates), have been determined. In addition, these enzyme structures have been examined as they appear when various mole- \\ \\Wl~; it:' by' ~ \\ \~o_~09 X-RAY CRYSTALLOGRAPHY CAN REVEAL COMPLEX MOLECULAR STRUCTURES cures are bound to them. The most striking feature seen on comparing DHFR mol- ecules from the different orga- nisms is the close similarity in their overall foldings. Clearly, the general molecular structure of the enzyme was highly con- served during millions of years of evolution, even though only about 25 percent of the amino acid sequence remained un- changed (80 percent among the vertebrates, however). DHFR provides an excellent mode} for studying how similar enzymes stimulate the rather unreactive nicotinamide nucleo- tides NADH and NADPH. Bio- chemists who study metabolic pathways have long recognized that the nicotinamide nucleo- tides serve as a kind of aB-pur- pose ox~dation-reduction currency, furnushing a way to exchange electrons in biolog- ical reactions. Now we are finding that the stereochem~cal aspects of its placement in DHFR facilitate hydride transfer through hydrogen bonds. Frontiers in the Chemistry of Genetic Material In higher organisms (including humans) the percentage of nucleotides in a strand of DNA which actuady code for the sequence of amino acids in proteins is estimated to be only about 5 percent. What is the role of the remaining 95 percent? Recently, it was discovered that another type of information is coded in the sequence of DNA nucleotides. Apparently, information concerning the different conformations or shapes that DNA can take is stored there as well. How are such conformational changes brought about? They happen around single bonds where relatively free rotational movements are possible. In ring structures, such rotations tend to pucker or wrinkle the ring into shapes that are not flat (nonplanar conformations). There is usually an energy barrier between the two (or three) energetically comfortable structures, called conformers, that result from such rotation. But the barriers can be small enough so that transfer between these structures can be relatively easy at room temperature. In sharp contrast to the stereoisomers of a molecule, the conformation of a molecule can be determined by

IV-B. DEALING WITH MO~F:CUI~R COMPLEXITY secondary interactions, they may change in re- sponse to their environment, and two or more conformers can be present at once in dynamic equilibrium. Many conformational characteristics have been Chalr Conformer Boat Conformer discovered for the nucleic acids. For example, puckering of the furanose ring, which is common to both DNA and RNA, leads to flexibility in their backbones. Furanose is the 5-carbon cyclic sugar found in the backbone of nucleic acids. A number of different conformations can be assumed by this ring, but the most prominent is called the C2' enclo conformation. This conformation has been considered to be characteristic of DNA nucleotides, while another one, the C3' endo conformation, was more frequently found in RNA nucleotides. We must learn more about the energy barriers between these two conformations. It is now thought that the energy barriers separating different conformations are lower for the deoxynucleotides than for ribonu- cleotides. In the three-dimensional structure of a transfer RNA found in yeast, which has 76 nucleo- tides, the majority were found to adopt the C3' endo conformations. This has a significant effect on the spacing of certain phosphate groups. The phosphate- phosphate distance is close to 6.7 A in the C2' endo conformation and less than 5.6 A in the C3' endo conformation. Thus, changes in sugar pucker make the polynucleotide backbone elastic, so it can accom- modate different confo~ations. We need to know these conformations more precisely, how easily they can interchange, and how they affect biological func- tion. For almost 30 years, DNA has been known to adopt two different nght-handed conformations, A-DNA and B-DNA. They are caped nght-handed because . . . . . . _ . Syn POSITION OF GUANINE DEOXYGUANOSINE A S I N Z—DNA Q DEOXYG UAN OSI N E AS IN a'~ C3 endo S u 9 or Pucker A n t i POSITI ON Of GUANI NE B—DNA C2 endo Sugar Pucker SUBTLE DIFFERENCES MATTER the L,NA spew twists to the nght. lathe A conformation is one in which ah the deoxynucleotides have the C3' endo conformation, while in B-DNA ad of the nucleotides have the C2' endo conformation. However, this simple classification into two possible nght-handed conformations has now been changed considerably as a result of single-crystal diffraction analyses. Surprisingly, some of these analyses revealed the presence of alternating C3' endo and C2' endo conformations with alternating distances between phosphates. That led to the discovery of left-handed DNA conformations in the laboratory. Polynucleotides were deliberately linked together so that puline bases and pyrimidine bases alternated with each other. Such a molecule adopts a conformation In which the punnes take the C3' endo conformation, while the altercating pyrimidines take the C2' endo conformation. This structure is caped Z-DNA; it twists to the left and exhibits an irregular tertiary structure. 145

146 INTELLECTUAL PRONTlERS lN CHEMISTRY I' Z Di\A LEFT-HANDED Mc M i nor _ G roove 3 DNA RIGHT-HANDED NATURE DOESN 'T ALWAYS TURN RIGHT Structure and Function in Biochemistry Structure determines properties and properties determine function. Thus, from the simplest molecules like ethyl alcohol, to molecules with the exquisite and vaned architectures of proteins, their molecular structure is inextricably related to their function as drugs, antibodies, biological catalysts, hormones, transport agents, cell surface receptors, structural elements, or muscles that convert chem- ical energy into work. A prime question we wish to answer is how the structure of a protein might determine-its function. One approach is to generate many structural variations of a protein in a controlled manner by precisely altering the order of its amino acids. With this method, the exact three-dimensional structure of a protein can be fixed to permit a logical analysis of the structure/function relationship. Today we have procedures that allow us to move toward this objective. Modern molecular biology has taught us how to place almost any piece of DNA into a microorganism and thereby cause it to synthesize the protein that this DNA encodes. At the same time, modern organic chemistry has enabled us to rapidly and easily synthesize sequences of nucleotides that constitute pieces of genes. These pieces of genes can then be used to change the prescribed sequence of bases in the gene for the parent protein. Thereby, a modified protein with an altered sequence of amino acids can be generated, and a structure and function never before available can be produced. This method for creating specific mutations of normal proteins is formally termed oligonucleotide-directed mutagenesis. This can lead to proteins with any structure we may desire. In addition, once a single molecule of the gene for that protein has been prepared, the protein itself can be produced forever after in microorganisms in whatever quantities may be desired. At present, the overall view of the nucleic acids is that they are conformationaDy active. It is also now thought that the weD- known right-handed B-DNA structure is likely to be in equi- libnum with a number of other structures, including left-handed Z-DNA. The focus of much chemical and biological re- search will be on the nature of these conformational changes. We need to know more about how these conformational changes are affected by their environment, by modifications in the molecule, or by alter- ations in the nucleotide se- quence.

IV-B. DEALING WITH MOLECULAR COMPI~XITY These techniques focus on the creation of a mutant protein with a predetermined amino acid sequence. Such approaches are useful for {earning the properties and functions of a protein altered in a specific manner. An alternate approach is to create a large number of structural variants, decide which ones show desired properties, and then go back and determine the structures of those desirable proteins. Such random mutagenesis can be allowed to take place anywhere in the gene of interest or, in order to better control the possible properties of a protein, can be restricted to a particular region of the gene. At present, oligonucleotides can be synthesized in 98 percent yield at the rate of one base every 5 minutes. Improvements here could make the rapid synthesis of entire genes (rather than just oligonucleotides) a routine procedure, and thereby greatly speed up the creation of new proteins. Great improvements can be foreseen in chemical and biochemical techniques for determining base sequences in nucleic acids and amino acid sequences in proteins. Presently, an automated instrument called a gas phase sequenator can reliably determine about 60 consecutive amino acids (called residues) from the amino end of a protein. Use of tandem mass spectrometry or other novel approaches might allow the complete sequence to be established for a protein of several hundred residues by automated techniques. Gene Structure and RNA Splicing The combination of a number of recent advances has yielded startling insights into the gene structure of man and other complex organisms. These advances include the ability to combine DNAs from different organisms, the ability to discover which DNA segments encode specific proteins and to isolate them, and the ability to determine the nucleotide sequence of long pieces of DNA. This new knowledge has raised many questions and opened new areas of research. To find the DNA segment that contains a single gene from the total genetic material of a human cell is like finding the legendary needle in the haystack. The sequences that specify any one particular gene are about one-millionth of the total genetic material. The solution to the problem was to use recombinant DNA techniques to distribute pieces of human DNA into well over a million rapidly dividing bacteria, and then to grow each bacterium separately to give an entire colony of offspring of the single bacterium. Then the colony of bacteria containing the gene of interest is identified by some diagnostic technique that tests for the desired gene function. Each rapidly growing bacterial colony produces billions of identical copies of each gene that can then be isolated as a chemically pure substance. This process is called cloning. So far, DNA segments for well over 100 different human genes have been purified by this method. A similar number have been isolated from a few other vertebrate species, such as the mouse, and a greater number have been isolated from simpler organisms such as yeast. GIobin is a protein found in the blood ingredient hemoglobin. The DNA sequence that codes for the gIobin protein is interrupted in places by sequences that do not code for the protein. This is typical in the genes of eukaryotic cells (celIs containing a nucleus~the coding region is interrupted by one or more stretches of noncoding DNA, called intervening sequences or introns. Introns have also been called 147

148 PART OF THE HUMAN B-GLOBIN GENE - GLY — GLU -G-G-T-G-A-G- - ALA — LEU — GLY - ARG ........................ -G- C- C- C-T-G- G-G- C- A -G-G;T-T-G-G-T-A-T-C-A- : a.~. ~ -A -G -G-T-T-A-C-A-A-G-A -C-A-G-G-T-T-T-A-A -G-G-A- ;Z :-G-A-C-C-A-A-T-A-G-A-A-A-C-T-G-G-G-C-A-T-G-T-G- O' I_ ~ -G-A-G-C-A-G-A-G-A-A-G-A-C- · Z s : · · · -~-C-C-T-T-A -G-G~ C- T- G- C- T- G- G- T- G - - VAL - TYR — G-T-C-T-A-C - INTELLECTUAL FRONTIERS IN CHEMISTRY , LEU - LEU — VAL INTRON SPACERS ARE EXCISED TO GIVE MESSENGER RNA - := :o := · Z . ~ - "nonsense codes," but it has been discovered that they may have many important func- tions. Introns are found in most genes that code for mes- senger RNA and in some genes that code for transfer and ribo- somal RNAs. In all cases that have been studied, the introns are copied along with the neighboring coding sequences as part of a large precursor RNA. The introns are then re- moved by a cleavage process called RNA splicing, which re- sults in a functional RNA mol- ecule with a continuous coding region. For example, there are two introns in the human giobin gene. After they have been removed, the resulting messenger RNA is transported from the nucleus to the cytoplasm for translation into gIobin protein. The phenomenon of RNA splicing is common in celIs with nuclei, eukaryotes, but is thought to be absent in celIs without well-defined nuclei, prokaryotes. It is the only major step in gene expression in which eukaryotes ar~d prokaryotes di~er significantly. Because of this it is interesting to examine just how RNA splicing regulates and affects the expression of genes. In addition, the possibility that introns in the genetic code might be responsible for the evolution of eukaryotic genes is being explored. The impact on society of future research on gene structure and gene expression will be enormously beneficial. Many human diseases are the result of defects in gene expression. Information about the nature of the genetic changes in cancer cells may yield new avenues for pharmacological treatment of cancer. The process of aging is poorly understood; it is possible that some of the more destructive aspects of this process are controlled by the activity of a few gerle products so that identification of the functions of these genes may lead to better treatments for aging patients. SUPPLEMENTARY READING Chemical & Engineering News "Experts Probe Issues, Chemistry of Light- Activated Pesticides" by R.L. Rawls, (C.& E.~. staff), vol. 64, pp. 21-24, Sept. 22, 1986. "Anticancer Drug Cisplatin's Mode of Ac- tion Becomes Clearer" by R. Dagani (C.& E.N. staff), vol. 63, pp. 20-21, Dec. 16, 1985. "Electrochemical Techniques Benefit Bioa- nalysis" (C.& E.N. staff), vol. 63, pp. 32-33, Jan. 14, 1985. "Penn Chemists Synthesize Complex Natu- ral Antibiotics" bv R. Dacani (C.&E.N. sta~, vol. 62, pp. 17-19, Oct. 15, 1984. "Potentiometric Electrode Aims to Measure Antibody Levels" by R. Rawls (C.& E.N. staff), vol. 62, pp. 32-33, Apr. 2, 1984. Science "Long Range Electron Transfer in Heme

IV-B. DEALING WITH MOLECULAR COMPLEXITY Proteins" by S.L. Mayo, W.R. Ellis, R.J. Crutchley, and H.B. Gray, vol. 233, pp. 948-952, Aug. 29, 1986. "Transformation Growth Factor Biological Function and Chemical Structure" by M.B. Sporn, A.B. Roberts, L.M. Wake- fied, and R.K. Assoian, vol. 233, pp. 632- 634, Aug. 8, 1986. "The Intervening Sequence RNA of Tetrahy- 149 mena is an Enzyme" by A.G. Zang and T.R. Cech, vol. 231, pp. 470-475, Jan. 31, 1986. Chem Matters "Natural Dyes" pp. 4-8, December 1986. "Autumn Leaves" pp. 7-10, October 1986. "Lipstick" pp. 8-11, December 1985.

Something for Nothing Grandpa used to say, "There's no free lunch!" That was his way of saying that you never get something for nothing. But now, Grandpa, we're not so sure! The recent discovery of high-temperature superconductors has everyone talking about amazing visions, like trains riding on air and electrical energy transmitted from Nevada to Alaska with no losses on the way. Kamerlingh Onnes, a Dutchman, started it all in 1911. When he cooled metals to low temperatures, the electncal resistance, which limits conductivity, smoothly dropped with temperature. Theorists explained that a current flow requires electrons to move through the metal crystal, but as they move, they keep bumping into the vibrating metal atoms, losing energy and generating heat. If the crystal is cooled, these lattice vibrations are diminished, so there are fewer collisions to slow down the electrons. The theory confidently indicated that the resistance would reach zero only when the temperature reached the unattainable "absolute zero." But when Onnes cooled mercury to liquid helium temperatures, he got the surprise of his life. At 4.2K, the resistance suddenly dropped so low he couldn't measure it. Below that critical temperature, Tc, an electrical current, once started, kept going for weeks, months, even years. The resistance, which usually stops such a current, had truly become zero. The metal had become a superconductor. As new superconductors were discovered, the highest Dowry value of Tc slowly crept upward. The world's record reached lSK in 1941 with the discovery of the two-elemer~t (binary) superconductor niobium nitnde, NbN. Another binary, NbGe, had led the field with Tc = 23K since its discovery in 1973. Here progress stopped cold (pun intended). Then, in 1986, the lid blew off. First, a quaternary copper oxide compound was found to become superconducting at 37K. In the next few months, rumors flew, suggesting possible TCS at 40K, 52K, 70K, 94K, and even 240K. Working around the clock, scientists in the United States, Europe, and Japan recognized finally that certain four-element copper oxides with the layered perovskite crystal structures were true superconductors with Tc near 94K. The breakthrough was an yttrium-barium com- pound, YBa2Cu3Ox, with a noninteger x near 7.4. Soon it became clear that yttrium could be replaced by six or seven other lanthanide elements, while strontium or calcium could take the places of some of the barium atoms. Where do we go from here? Anyone who feels that electrical bills are too high can be cheered} about 20 percent of the electrical energy moved around the country is wasted in the copper transmission lines. That's enough energy to light up the whole West Coast. Since liquid nitrogen, at 77K, is a cheap coolant, superconductivity is now affordable. From tiny motors to the enormous turbines in hydroelectric plants, a new era can be foreseen. Ire our most powerful computers, heat dissipation limits circuit size, hence computer capacity. This problem disappears, along with the resistance, if the connectors are superconducting. But the most advertised expectations concern new uses of superconducting magnets. They will surely reduce the cost of a medical NMR whole-body imager. And superconducting magnets might levitate whole trains so they ride on a practically fr~ctionless cushion of air. So, getting the resistance down to zero really does give us something for nothing. And by the way, Grandpa, want to get in on the free lunch? 150 N. fW ~ suP~c~!l ~ EXPRESS ._ .~ -:Y ~'7 WU 4~' .:

IV<. NATIONAL WE~-BEING {V-C. National Well-Being Research across the whole range of chemistry contributes to a better environ- ment and to sustained economic competitiveness. But certain research areas are key to progress in these realms. For example, the surface sciences, with their implications for new heterogeneous catalysts, furnish a wellspring of critical importance to economic progress. Condensed-phase chemistry and new separa- tions techniques also can be expected to contribute fruitful new dimensions. Next, the new frontiers in analytical chemistry support and contribute to advances in all other areas of chemistry. Analytical chemistry is the cornerstone upon which our monitoring and management of the environment is built. Finally, nuclear chemistry was nurtured in the World War IT Manhattan Project, and its influence continues to be of prime importance, since the worId's energy needs may involve nuclear reactors (despite ChernobyI), and world peace presently is based upon a precarious balance of nuclear arms. In each of these areas there are opening frontiers and rewarding intellectual opportunities to be pursued. CHEMISTRY AT SOLID SURFACES The surfaces of metals and ionic solids are, by nature, chemically reactive. The reason is clear the bulk crystal is based upon a structure that gives each interior atom the best possible chemical bonding to neighboring atoms around it in all three dimensions. At the surface, however, the atoms have unsatisfied bonding capacity since the neighboring atoms are missing in at least one direction. Hence, this is a region of special chemical behavior, and one of unusual interest to chemists. The importance of this special behavior simply cannot be exaggerated. Corrosion occurs, of course, at iron surfaces, with obvious bad effects on many useful structures, from the lofty Elide! Tower to the lowly nail. Estimates are that corrosion costs the U.S. economy billions of dollars annually. At aluminum surfaces, the rapid reaction that takes place on exposure to air forms a protective and quite inert oxide coating. Hence, we can safely have the convenience of aluminum foil in the kitchen, despite the fact that aluminum is flammable at sufficiently high temperatures. By far the greatest importance of surface chemistry is that it bestows extremely effective catalytic activity upon some surfaces. This capacity of a solid surface to speed up chemical reactions by many orders of magnitude without being consumed is called heterogeneous catalysis. Its great value as the basis for commercial processes of immense economic value has been noted in Sections IlI-B and ITI-C. It furnishes one of the most important and active frontiers of chemistry. Heterogeneous catalysis is not new. What is new is the array of powerful instruments, developed over the last 15 years, that at last provide experimental access to the chemistry on a surface while that chemistry is taking place. Without such techniques, catalysis has remained over many decades a fairly mysterious art. Now we have instruments with which to characterize precisely the nature of the catalyst surface and to study molecules while they are reacting there. Now we are accumulating the store of quantitative data needed for catalysis to move from an art

52 INTEL' ACTUAL FRONTIERS IN CHEMISTRY to a real science. The intellectual challenge to understand the chemical behavior of molecules on a surface has propelled surface science into the mainstream of fundamental research in most departments of chemistry and chemical engineering. The instrumentation of the surface sciences will be described in Section V-C. Some research highlights and productive frontiers will be described here. The Structure of Solid Surfaces We have already discussed, in Section IlI-C, the role of specific metallic surfaces In the catalytic restructuring of hydrocarbons to produce gasoline. As a second example, research on the catalytic production of ammonia from elemental nitrogen and hydrogen is of comparable importance. This is because NH3 is a critical fertilizer component, so it helps determine (or limit) the world food supply. At elevated temperatures, N2 and H2 can react to form NH3 on perfect crystals of an iron catalyst. The effectiveness of a catalyst depends upon how rapidly each surface site can adsorb reactants, encourage them to rearrange chemically, and then release the products so that the site can begin the process again. The iron crystal face designated (l,1,1) is about 430 times more active than the closest- packed (1,1,0) crystal face and 13 times more active than the simpler (l,0,0) face. It is now believed that the rate-limiting step is the rupture of the strong nitrogen- nitrogen bond of N2 (225 kcal/mole) and that this occurs with an activation energy near 3 kcal/mole on the (1,0,0) face but with nearly zero activation energy on the specially active (l,1,1) surface. Because of such influences on catalytic action, surface structures are attracting much research interest. Small particles tend to display many different surfaces, depending on how they are prepared. As the metallic particle grows, it becomes more like the bulk material and tends to favor surfaces without terraces and kinks. Interestingly, atoms in the surface layer may be located closer to adjacent atoms in the second layer than they would be if they were located deep inside the crystal. Even more drastically, because of the incomplete bonding of surface atoms, they may seek equilibrium positions different from the packing in the bulk material in order to improve their bonding. Such "surface reconstruction" has been found for platinum, gold, silicon, and germanium. Another important question that can now be experimentally explored is the chemical composition of the surface. Even the purest samples will have some impunties, and these may noticeably affect some properties of metals and semi- conductors. A crucial question is how much a given impurity prefers to concentrate at the surface. The difference in bonding between host atoms and impurity atoms explains why the bulk poItion of the material tends to reject the impurity. This same difference may cause the impurity to be a welcome addition to the surface, where host atoms alone cannot satisfy their bonding capability. There are cases in which impurities at the parts per million level are so concentrated at the surface that they can cover it completely. This strongly affects the chemistry at that surface. Of course, this issue is always present in alloys composed of two or more elements. There is excess silver at the surfaces of silver-gold alloys, excess copper at copper- nicke! alloy surfaces, and excess gold at gold-tin alloy surfaces. Some metals that do not readily dissolve in each other in bulk are found to mix in any proportion on a

IVY. NATIONAL WE, [~-BEING surface. Experimental data and understandings are especially needed at this tune when a variety of binary and ternary substances are under study because of their interfacial electrical properties. In summary, determination of the atomic structure of surfaces and surface composition is basic to understanding the wide variety of surface properties now finding important practical applications. They are the starting point for advancing corrosion science, heterogeneous catalysis, lubrication, and adhesion, as well as for producing new surfaces with novel electronic properties. Adsorbed Molecules; Chemical Bonding at the Surface For many decades, the strength of binding of an adsorbed substance on a surface was measured by the ease of its removal on warming. Some substances are easily removed at temperatures near or below room temperature. Such a situation is traditionally called "physisorption"; the adsorbed substance keeps its molecular shape and is bound to the surface only by weak forces, such as van der Waals or hydrogen bonding interactions. Other substances are much more tightly held by the surface and can be removed only by heating to much higher temperatures perhaps 200 to 600°C. Here, covalent bonding to the surface is involved and the molecular structure of the adsorbed substance is probably different from what it was before adsorption. This situation is called "chemiso~ption," and it is almost always involved at some stage in any heterogeneous catalysis. Thus, understanding of the molecular structure and chemical proper- ties of chemisorbed molecules lies at the heart of heteroge- neous catalysis. Among small molecules, car- bon monoxide on metal sur- faces has histoncally received the most attention, largely be- cause its spectroscopic proper- ties allow the detection of small numbers of CO mole- cules on a surface. This is for- tunate, for one of today's most pressing problems is the conversion of coal to useful hydrocarbon feedstocks, usually accomplished via carbon monoxide. Many catalytic schemes use carbon monoxide as an intermediate in the form of "syn gash a mixture of CO and H2 derived from coal (see Section IlI-C and Table IlI-C-21. A second key system is ethylene adsorbed on catalytic metal surfaces. It has been known, from its thermal behavior, that ethylene chemisorbs on platinum and rhodium catalysts. Now, we can add information about the structures that are formed on the surface through direct observation of the vibrational frequencies of the adsorbed species. Direct observation of these frequencies through infrared absorption spectroscopy is sometimes possible, but the introduction of electron ~ CCH CCH2 ~ ETHYLENE ON it, ~. RHODIUM CCH3 ~ I WHICH STRUCTURES ARE PRESENT? 153

54 ETHYLENE ON RHODIUM ( I, I, I ) CATALYST 1 (~ ~ _ 1 1 1 1 1 1 1 1 1 1 1 1 1 1 ~ 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 $ ~ : ~ ' ~ 1 1 ,-- ~ 1 000 2000 3000 VIBRATIONAL FREQUENCY 137 C 37 C MOLECULAR FINGERPRINTS REVEAL THE REACTION PRODUCTS ON SURFACES INTELI:FCTUAL FRONTIERS IN CHEMISTRY energy loss spectroscopy (EELS) has greatly accel- erated such studies. The characteristic molecular frequencies are imprinted on the energy distribution of electrons bounced off the metal surface. These frequencies provide a fingerprint that is readily interpreted by a chemist experienced in relating infrared spectra to molecular structures (see Sec- tion V-C). For ethylene on rhodium, the EELS spectrum plainly shows that, after adsorption, the ethylene molecule has been structurally altered even at room temperature. Then, on warming 50°C or so, the spectrum begins to change still more. By the time the temperature has changed by 100°C, the spectrum shows that reactions have taken place and the hydrocarbons now present on the surface have new structures. These EELS spectra reveal, then, which of the possible surface structures (C2H3, C2H2, C2H, CH3, CH2, and CH) are present at a given temperature and, hence, the sequence of their formation as the temperature is raised. Such intimate knowledge of the chemical events taking place on the catalyst surface furnishes the basis for a detailed understanding of the catalytic dehydrogenation and hydrogenation of ethylene, which is important in many chemical processes. Coadsorption on Surfaces Chemistry on surfaces takes on a new dimension when two substances are adsorbed on the same surface. Then attention shifts from the interaction of the adsorbate with the surface to the interaction of two different molecular species when they share the special environment provided by the surface. The first way in which this interaction can occur is when one adsorbate changes the special environment encountered by the second adsorbate. For example, a clean molybdenum metal surface will break down the sulfur-containing molecule, thiophene, C4HsS. However, if elemental sulfur is coadsorbed, it chemisorbs quite strongly at the active sites needed for thiophene decomposition. Thus, sulfur "poisons" the catalyst for this particular reaction. This is of great importance because thiophene is an impurity we want to remove from gasoline. As a second example, carbon monoxide is physisorbed on rhodium, as shown both by its ease of removal on warming and its vibrational frequency on the surface, which is close to that of gaseous carbon monoxide. If, however, the rhodium is 50 percent covered by coadsorbed potassium, CO becomes chemisorbed instead. The EELS spectrum shows a CO vibrational frequency appropriate to a bridged structure, with a frequency indicating the presence of a carbon-oxygen double bond. Under these conditions, hydrogenation of CO is encouraged, and this leads to the production of desirable higher molecular weight alkalies and alkenes (hydrocarbons possessing one or more double bonds). (See discussion of "syn gas" in Section ITI-C.)

IV'. NATIONAL WELL-BEING Still to be mentioned, of course, is the direct reaction between the two adsorbates. In the future, this will be seen as the origin of most of the new chemistry that can take place in this special reaction domain. An obvious example has already been cited, the hydrogenation of ethylene (C2H4~. When hydrogen adsorbs on platinum or rhodium, the H2 molecule is split and the two atoms are separately bonded to the metal atoms. Now, when ethylene is coadsorbed, it does not encounter H2 at all. Instead, it finds individual hydrogen atoms attached to the surface. Plainly, if coadsorbed hydrogen and ethylene react, they will follow a reaction path characteristic of the actual species on the surface and governed by activation energies that are different from those associated with a gas phase encounter between H2 and C2H4. CONDENSED-PHASE STUDIES Many challenges facing chemistry, solid-state science, earth science, biochem- istry, and biophysics involve the ability to understand and manipulate the proper- ties of condensed phases liquids and solids. Chemistry is central here, since these properties result directly from the interatomic and intermolecular forces between the atoms and molecules present in these phases. Optical and Electronic Properties of Solids Over the past 15-20 years high pressure has proved to be a powerful too} in the study of electronic phenomena in solids. Compression pushes molecules closer together, and this increases the overlap among adjacent electronic orbitals. Since different types of orbitals have different spatial characteristics, they are affected to different degrees. This "pressure tuning" makes pressure a powerful tool for characterizing electronic states and discovering electronic transitions to new states with different physical and chemical properties. Many examples of electronic transitions that show pronounced response to high pressure have been found. For example, it has been possible to use high pressure to convert substances that are normally insulators into electrical conductors. This has been done for nine elements and for about 50 compounds. One application is fast electrical switches without make-and-break contacts. Also, the first organic superconductor showed its superconductivity under pressures between 6,000 and 18,000 atmospheres. Visible color changes can also be caused, as has been shown for several compounds known as anile, spiropyrans, and bianthrones (photochro- mic-thermochromic transitions), and for about 30 ethylene-diamine complexes (electron-transfer transitions). Such pressure studies are showing us how phos- phors absorb light of one color and reradiate light of another color, and are helping us increase the efficiency of a variety of laser materials. Liquids Many of the fundamental processes of nature and industry take place in the liquid state. The rate of movement of molecules in solution can limit the speed with which a chemical reaction can occur, a nerve can fire, a battery can generate current, and chemicals can be punned and isolated. A properly chosen liquid solvent can accelerate a chemical reaction by a millionfold or slow it down by a similar amount. 155

156 INTEr~r ACTUAL FRONTIERS IN CHEMISTRY Molecules in liquids can be highly efficient agents for storing or transferring energy. The very structure of liquid water determines our planetary environment and influences the course and nature of all biochemical processes essential to life. The structure and dynamics of a wide range of fluids, from liquid hydrogen to molten sili- cates, can be investigated by a number of spectroscopic tech- niques, such as X-ray and neutron diffraction, nuclear magnetic resonance, and laser Raman and light scattering. Among the newer experimen- tal approaches, pulsed laser excitation techniques are par- ticularly powerful. On a pico- second time scale (10- ~2 seconds), we can sense the freedom of movement of a sol- ute molecule held in its solvent cage. Now we can watch fun- damental chemical events as they take place: how two io- dine atoms combine in a liquid to produce an iodine molecule; how electrons released in liquid water become trapped, or solvated; how energy placed in a solute molecule like nitrogen or benzene is transferred to its solvent environment. Quite a different opportunity area is connected with the melting of small clusters of metal atoms. We have a variety of new experimental methods for producing and studying small metal clusters, as well as the theoretical tools with which to interpret the results. We can look ahead to an understanding of how the change from the fluid liquid state to the rigid solid state emerges as cluster size increases toward bulk amounts. Furthermore, the computer can keep track of the energy and randomness associated with each arrangement, so thermodynamic data can be calculated for comparison with experiments, and then for predictions under conditions out of reach expenmentally. ~ fOI03 ~ ~ & To LASERS LET US MEASURE FAST CHANGES IN THE SOLVENT CAGE Critical Phenomena For any fluid, there is a characteristic temperature and pressure above which the liquid and gaseous states are identical. Fluid behavior under these "critical conditions" can differ markedly from normal behavior and give rise to new phenomena. The past 20 years have seen a revolution in our understanding of such critical phenomena. Undoubtedly, the most important single theoretical advance in our understanding in the last 15 years has been the development of the new mathematical technique called the "renormalization group" approach. It has

IV<. NATIONAL WELL-BEING shown promise for quantitative description of fluid properties and their dependence upon molecular shapes and forces. The past 15 years have seen the beneficial use of critical phenomena in a variety of applications. Critical point drying is now a standard sample preparation method in electron microscopy. Further, there are remarkable changes in the solvent power of a liquid near its critical point. These are at work, for example, in the removal of caffeine from coffee for cadeine-free instant coffee and in the extraction of perfume essences. In addition, there are valuable research applications in liquid chroma- tography. Chemistry of the Terrestrial and Extraterrestrial Materials The Earth's geochemical phenomena involve complex mixtures, frequently with a number of crystalline and glassy (amorphous) phases, and they may take place at extremely high pressures and temperatures. Recent advances in high-pressure technology have made studies possible that duplicate conditions near the earth's core. In recent years many earth scientists have studied the "geochemical cycles" of elements that is, the changing chemical and physical environment of a given element during such natural processes as crystallization, partial dissolving, change of mineral structure (metamorphism), and weathering. These processes may lead to concentration (e.g., ore deposits) or dispersion of an element. The geochemical cycle of carbon has provided a focus for the reawakened field of organic geochemistry. Research on the stability, conformation, and decomposition reac- tions of fossil organic molecules has led to greater understanding of the origin and composition of coal and other organic deposits. Such knowledge has obvious value that extends from guiding our exploration for new fossil fuel deposits to helping us decide how to use the ones we have. Meteontes are of considerable chemical interest because they include the oldest solar system materials available for research and they provide samples of a wide range of parent bodies some primitive, some highly evolved. Meteorites carry records of certain solar and galactic events and yield data otherwise unobtainable about the genesis, evolution, and composition of the Earth and other planets, satellites, asteroids, and the Sun. Unusual isotopic percentages of many metals and gaseous elements, and compositional data particularly trace elements—have shed light on stages of the formation, evolution, and destruction of the original parent body or asteroid where the meteorite originated. Within the last decade, the study of meteorites has been dramatically advanced by the recognition that if these projectiles from outer space land on the Antarctic ice sheet, they are immediately entombed in an inert environment and permanently refrigerated, stopping chemical changes. The question, of course, is how does one find these meteorites in the wide and forbidding spaces of this hostile region? Nature provides an astonishingly convenient answer. The Antarctic ice sheet is a vast glacier, so it gradually flows northward, carrying the meteorites with it. Over thousands of years, snow that fell near the South Pole finally reaches the end of the glacier where the ice begins to evaporate. Here at the glacier edges, the meteorites are dropped in great numbers, essentially never having been exposed to terrestrial life forms, erosion, or weathenng. Since this discovery, more meteorites have been 157

158 . METEORITES: AN ANTARCTIC TREASURE TROVE I7vTEr r ~.cTuAL FRONTIERS IN CHEMISTRY collected (in the last decade) than over all of history before. The chemical and physical analysis of this meteorite trea- sure trove has only just begun. ANALYTICAL CHEMISTRY Characterization of atomic and molecular species their structures, compositions, etc., is called qualitative analytical chemistry. The measurement of the relative amount of each atomic and molecular species is called quantitative analytical chemistry. Both areas contribute to and benefit from the current rapid progress in science. Basic discoveries from physics, chemistry, and biology are providing new methods of analysis. In return, these new abilities are central to research progress in chemis- try, other sciences, and medicine, as well as to a wide range of applications in environmental monitoring, industrial control, health, geology, agriculture, defense, and law enforcement. Further, the lO-fold growth of the analytical instrumentation industry to $3 billion in sales worldwide has been led by the United States with its nearly $1 billion positive balance of trade in this area. A key factor in this growth has been the incorporation of computers into analytical instrumentation. The benefits here are circular; modern computers have evolved through advances in solid-state technology. In turn, these advances have critically depended upon the ability to analyze quantitatively the concentrations of trace impurities in silicon, the key element in current computer technology. Now microprobe analyzers using computer imaging techniques are answering questions critical to making microcircuitry even smaller, which will produce computers that are faster, more reliable, and cheaper. Analytical Separations Analyses of some complex mixtures are possible only after separation of the mixture into its components. Then, a variety of identification and quantitative measurement schemes become effective that would be confusing or impossible if applied to the unseparated mixture. Hence, devising new separations for use in an analytical context is an active field of research. There is no single technique more effective and generally applicable than the chromatographic method. The basic principle depends upon the fact that each molecular species, whether gaseous or in solution, has its own characteristic strength of attachment to, and ease of detachment from, any surface it encoun- ters. The differences in these attachment strengths can furnish a basis for separation. The differences can depend upon heat of adsorption, volatility, interaction with the solvent, molecular shape (including stereogeometry),

Rae. NATIONAL WEr r-BEING charge, charge distribution, and even functional chemistry. Great ingenuity has made it possible to use the whole range of molecular properties for analytical separations that can require only tiny amounts of material. The different instrumental methods of chromatography will be discussed in Section V-C. For this discussion, a few illustrative examples will show the potential. In liquid chromatography, a solution of the mixture of interest passes through a column loaded with a suitable particulate material. For example, if an aqueous solution of pigments (such as those contained in carrot juice) is slowly passed through a tube containing small lumps of a suitable resin, the various pigments pass through the tube at different rates. The pigments that attach most weakly to the resin wash through fastest, and the ones that attach most strongly come out last. This provides a vivid example because we can actually see the different colors of the carrot juice pigments once they are separated. Of course, the method works to separate all sorts of compounds, whether colored or not. Under the best conditions, liquid chromatography can separate and reveal the presence of as little as lo-~2 grams of a substance in a mixture. For gaseous samples, the technique can separate literally thousands of components such as are found in flavors, insect communication chemicals (pheromones), and petroleum samples. It is even possible to separate compounds that differ only in isotopic composition (e.g., deuterium instead of hydrogen!) by this method. Two-~unensional chromatography can give additional specificity, resolution, and sensitivity by coupling with techniques such as electrophoresis, which involves the movement of substances in the presence of a high electric field. For example, two~unensional electrophore- sis can sort 2,000 blood proteins at once by separating a mixture spot of the sample linearly under one set of conditions, and then using another set of conditions to separate further the initial line of spots at right angles. Spot locations and amounts can be measured quantitatively with computenzed scanning based album] . . . —' ~'~ ~CT;;T~ _a01~ps~n tQnsf~rrin ' ] ~ ,~ ,, ,,,,, .,— ,, >~ : *I :t :~ - ~ ;~- Alga ;~=, ... - ;- ~ _ on National Aeronautics and Space Administration computer programs developed for satellite pictures. Optical Spectroscopy The intellectual opportunities in this field, which introduce a variety of valuable analytical techniques, can be illustrated by two notable achievements of the last decade: the incorporation of computers as an essential part of most instrumentation, and the detection of single atoms and molecules. "Smart" commercial instruments now include microcomputers preprogrammed to carry out a wide variety off expen- mental procedures and sophisticated data analyses. The more powerful computers of :~ ... .~ . ~ .~.~- i; ~ ; ; ~: Am.. ~ - ~~ - 4 ~ ~ . a ~ , ; ~ : ;~ ~ ;.—, PROTEIN GEL PATTERN HU MAN MYALOMA SERUM 5~Q . . . 159

160 INTELl~CTUAL PRONTlERS lN CHEMISTRY ~.Llp,l w7 '(~L cC14 CFC13 700 750 800 850 CO2 . ~~r~CO2_ )~-~-'''''"1 ~ ~ !r CF2C12 950 looo lose ~ loo ~ (cm~l) ~N2O . art_ 1150 1200 1250 1300 THE INFRARED SPECTRUM SHOWS ATMOSPHERIC POLLUTANTS EVEN AT NIGHT the future will digest huge vol- umes of data from spectroscopic methods (especially Fourier transform and two-dimensional methods) much more efficiently. This will further improve resolu- tion, detection limits, interpreta- tion, spectral file searching, and immediate presentation of the results with three-dimensional color graphics to permit direct human interaction with the ex- periment. Intense laser light sources are revolutionizing analytical opti- cal spectroscopy. An immediate benefit is increased sensitivity. In special cases, resonance-en- hanced two-photon ionization using tuned lasers has achieved the ultimate sensitivity: detec- tion of a single atom (cesium) or molecule (naphthalene). Achievements in laser-induced fluorescence are approaching this same incredible limit. Laser remote sensing, such as for atmospheric pollutants, is effective at distances of over one mile; fluorescence excitation and pulsed laser Raman are particularly promising. In these latter methods, a laser pulse is emitted in the direction of the sample, which might be a smokestack plume. Then the time that it takes the fluorescence or Raman signal to return (at the speed of light) is measured to determine how far away the sample is. Thus, the signal not only tells us what substances (pollutants) are in the sample but also permits us to track them as they move away from the source. The ability of a laser to emit a precise wavelength means there is the potentiality for the identification of one component in a mixture (without need for separation). Yet this selectivity is sometimes defeated because atomic and molecular absolutions can be much broader in wavelength than the laser line width. However, the resulting overlap can be eliminated by the wavelength narrowing that occurs at extremely cold, cryogenic temperatures. This cooling can be achieved for gaseous molecules by passing them through a nozzle to bring them to supersonic velocities. In an alternate approach, molecules can be embedded in a cryogenic solid, such as solid argon, at temperatures near that of liquid helium (a process ceded matrix isolation). These two complementary techniques me ze interference by rotational and vibrational abso~p- tions and improve detection sensitivity and diagnostic capability. Mass Spectrometry This method involves separation of gaseous charged species according to their mass (see Section V-B), and it offers unusual analytical advantages of sensitivity,

IV-C. NATIONAL WELL-BEING specificity, and speed (10-2-second response). All of these attributes make for an ideal marriage to the computer. In the celebrated Viking Mars Probe, mass spectrometry was the basis for both the upper atmosphere analysis and the search for organic material in the planetary soil 30 million miles from home. Such sensitive soil sniffing to detect hydrocarbons might become a fast method for of! exploration. A special tandem-accelerator/mass spectrometer can detect three atoms of ~4C in low atoms of TIC, which corresponds to a radiocarbon age of 70,000 years. The broad applications of mass spectrometry include the analysis of elements, isotopes, and molecules for the semiconductor, metallurgical, nuclear, chemical, petroleum, and pharmaceutical industnes. In tandem mass spectrometry, one mass spectrometer (DISC) feeds ions of a selected mass into a collisional zone where impacts cause fragmentation into a new set of fragment ions for analysis in a second mass spectrometer (MS-~. This technique, abbreviated MS/MS, offers a particularly promising frontier for analysis of mixtures of large molecules. "Soft" ionization that avoids extensive fragmen- tation is used first to produce a mixture of molecular ions. From this mixture, one mass at a time is selected by MS-l, and it is more vigorously fragmented to produce an MS-~l spectrum that characterizes the structure of that one component. High speed and molecular specificity are important features of MS/MS. It is a powerful tool for analysis of groups of compounds sharing common structural features. It is particularly effective in removing any background signal caused by the contaminant species usually present in biological samples. It is now possible to determine the sequence of peptides with up to 20 amino acids and, in some instances, with sample sizes as small as a few micrograms. Combined ("Hyphenated") Techniques There is a growing appreciation for the extra benefits of using these computerized instruments in combination, such as the mass spectrometer coupled to a chroma- tograph (gas or liquid, GC/MS or LCtMS) or to another mass spectrometer (MS/MS), or these coupled with the Fourier transform infrared spectrometer (GC/IR, GC/IRtMS). High-resolution MS gives one part per trillion (1/10~2) analyses for the many forms of dioxin (TCDD) to see if the toxic form is present in human milk and the fatty tissue of Vietnam war veterans. GC/MS is necessary for the specific detection of 2,3,7,8-TCDD, the most toxic dioxin isomer. GC/MS is used routinely for detecting halocarbons in drinking water at concentrations far below the toxic level, polychIorobiphenyIs (PCBs), viny} chloride, nitrosamines, and for detecting most of the Environmental Protection Agency's list of other priority pollutants. MS/MS with atmospheric pressure ionization can monitor many of these contaminants continuously at the parts per billion level, even from a mobile van or helicopter. The high specificity as well as sensitivity of these methods make them especially promising for detecting nerve gases, '~yellow rain," and natural toxins in foodstuffs (10-~ g of vomitoxin in wheat) and plants (Astragalus or "Ioco weedy. Metabolites found by GC/MS have led to the identification of more than 50 metabolic birth defects in newborn infants where early identification is critical in preventing severe mental retardation or death. One of the most exciting intellectual 161

162 lN~TE.r I EcTu~ FRONTIERS IN CHEMISTRY frontiers is the possibility that routine profiling of human body fluids can detect disease states well before external symptoms of those illnesses appear. Electroanalytical Chemistry Electrochemistry has a long history of analytical applications, beginning with pH meters. Today, pulse voltammetric techniques permit detection of picomole quantities (10~ i2 moles). Solid-state circuitry, microprocessors, miniaturization, and improved sensitivity have made possible continuous analysis in living single cells (with electrode areas of a few square microns). Electroanalytical methods are also useful in such difficult environments as flowing rivers, nonaqueous chemical process streams, molten salts, and nuclear reactor core fluids. SEPARATIONS SCIENCES 1.8 1.6 1.2 Separations Chemistry Separations chemistry is the application of chemical principles, properties, and techniques to the separation of specific elements and compounds from mixtures (including mineral ores). It takes advantage of the differences in such properties as solubility, volatility, adsorbability, extract- ability, stereochemistry, and ion properties of elements and mol- ecules. As an example, the rare earth elements neodymium (Nd) and praseodymium (Pr), impor- tant in laser manufacture, must be separated from a mineral called monazite. A difficult part of this extraction is the separa- tion from cenum, which is chem~caDy similar. Photochem~- cal studies show that this sepa- ration cart be greatly enhanced by selective excitation to take advantage of the different chem- istnes of the elements under photoexcitation . The availability of cntical and strategic matenals to U.S. industry and the military is dependent in many instances on the development of practical, econom- ical chemical separations methods. Table IV-C-] shows our dependence on imports for some critical metals and minerals. For example, almost 90 percent of our use of platinum, in great demand as a catalyst, comes from imports. Mining of the major platinum source in the United States, in Stillwater, Montana, has not yet begun. A second important example concerns our access to uranium. About 13 percent of the nation's electncal energy is denved from nuclear energy, and a much larger OPTIMUM ~ \ , _ / O _ ~ J ~5 1.4 _ 2 _ ~ _/ 1.0 INdl/ /1~1 1 | UNDER SELECTIVE ~ / RADIATIVE EXCITATION \ ,,,, I,,,, 1,,,, 1,,,, 1,, .5 2.0 2.5 3.0 HCI CONCENTRATION AT ALL MC' CONCENTRATIONS SELECTIVE EXCITATION FAVORS NEODYMIUM

IV'. NATIONAL WELL-BEING percentage than that is utilized in the industrialized Northeast. Chemical sepa- rations are vitally important in the nu- clear fuel cycle, beginning at the uranium mill where low-grade uranium ores (typi- cally only 0.1 to 0.3 percent U3Os) are treated in selective chemical processes to produce a concentrate of more than 80 percent U3Os. Then, further refinement, based on transfer from one solvent to another (solvent extraction), or formation of the volatile fluoride, UFO, produces a uranium product pure enough for use in nuclear fuel manufacture. Then, after re- TABLE IV-C-1 U.S. Import Dependence, Selected Elements (Imports as Percentage of Apparent Consumption) 1950 1980 Manganese Aluminum (bauxite) Cobalt Chromium Platinum Nickel Zinc Tungsten Iron (ore) Copper Lead 77 71 92 100 91 99 37 80 35 59 97 94 93 91 87 73 58 ~4 22 14 <10 moval from the reactor, the highly radio- active fuel is subjected to a selective chemical process to separate uranium and plutonium from the fission products for recycling or for weapons use. This step is a remarkable feat of chemistry and chemical engineering because the aim is to separate two similar elements, uranium and plutonium, from each other and also from the highly radioactive fission products, which include about half of the Periodic Table. All of this must be done in a remotely operated plant which, by robotics, handles tons of materials so radioactive that they cannot be approached by a human being. These are only a few examples of the many ways we depend upon separations chemistry. Future availability of many of the critical elements listed in Table IV-C-l will depend, sooner or later, upon developing new chemical mining or separations processes that permit us to use low-grade domestic ores and the salt solutions (brines) that are found in geothermal wells. These developments will require research advances across a wide front, mainly focusing on the action of solvents and all of the properties of the liquid state that affect solvent power. NUCLEAR CHEMISTRY Since the days of the Curies, chemists have played a key role in the fundamental exploration of radioactivity and nuclear properties, as well as in nuclear applica- tions to other fields. Thus, the 1944 Nobel Prize for the discovery of nuclear fission went to a chemist, Otto Hahn. Then, the 1951 Nobel Prize for the discovery of the first elements beyond uranium in the Periodic Table, neptunium and plutonium, went jointly to a chemist, Glenn Seaborg, and a physicist collaborator, Edward McMilian. Most of the advances in our understanding of the atomic nucleus have depended strongly on the complementary skills and approaches of physicists and chemists. Furthermore, the applications of nuclear techniques and nuclear phe- nomena to such diverse fields as biology, astronomy, geology, archaeology, and medicine, as well as various areas of chemistry, have often been, and continue to be, pioneered by people educated as nuclear chemists. Thus, the impact of nuclear chemistry is broadly interdisciplinary. 163

164 INTEL~iCTUAL FRONTIERS IN CHEMISTRY Studies of Nuclei and Their Properties Particularly exciting advances have been made in extending our knowledge of nuclear and chemical species at the upper end of the Periodic Table. In the last 15 years, elements 104 to 109 have been synthesized and identified, often by ingenious chemical techniques geared to deal with the very short half-lives of these species (down to milliseconds). In addition to these new-element discovenes, many new isotopes of other elements beyond uranium have been found, and the study of their nuclear properties has played a vital role in advancing our understanding of alpha decay, nuclear fission, and the factors that govern nuclear stability. Fission research in particular has been quite fruitful. For example, the "nuclear Periodic Table" identifies particular stable pro/on-neutron combinations ("closed shells"; one of these is the tin isotope ~32Sn (50 protons, 82 neutrons). Changing this nucleus by only one nucleon gives a dramatic change in the nuclear fission behavior, both in the distnbution of fission products obtained and in their kinetic energies. Furthermore, the study of spontaneously fissioning isomers among the heaviest elements has led to the important realization that the potential energy surfaces of these nuclei have two specially stable regions. This, in turn, opened the way to a new approach to calculating such surfaces the so-called shell correction method. Further exploration of the limits of nuclear stability is clearly in order, both at the upper end of the presently known nuclei and on the neutron-nch and neutron-poor sides of the region of stability defined by the stable nuclei found in nature. Newly discovered nuclear reaction mechanisms, based upon accelerating heavy nuclei as bombarding particles, promise to give access to more neutron-nch, and therefore much longer lived (minutes to hours), isotopes of elements with Z > 100 than have been available. This should open the way to more detailed investigations of the chemistry of these interesting elements at the upper end of the actinide series and beyond. The quest for so-called "superheavy'' elements, i.e., nuclear species in or near the predicted "island of stability" around atomic number 114 and neutron number 182, has not been successful so far, but this exciting goal is still being pursued. Space Exploration The wide range of applicability of nuclear techniques is demonstrated in the exploration of the Moon and our companion planets during the past two decades. For example, the unmanned Surveyor missions to the Moon provided the first chemical analyses of the Moon. They employed a newly developed analytical technique that utilized the synthetic transuranium isotope 242Cm. The analyses identified and determined the amounts of more than 90 percent of the atoms at three locations on the lunar surface. These analyses, verified later by work on returned samples, provided answers to fundamental questions about the composition and geochemical history of the Moon. Nuclear techniques also played an important role in the chemical analyses performed by Soviet unmanned missions to the Moon, and in experiments designed to seek life on the surface of Mars by the U.S. Viking missions. Similarly, isotopic distributions were important results in the analyses of

ILIAC. NATIONAL WELL-BEING returned lunar samples and of meteorites, making possible clarification of the history of the Moon and meteorites. Isotopic Composition Ever since the discovery of the isotopic composition of the chemical elements, it has been assumed that this isotopic composition is essentially constant in all samples, an assumption that provides the basis for assigning atomic weights. The only exceptions involved elements with long-lived radioactive isotopes. Since 1945, however, humans have affected the atomic weights of several elements (e.g., Li, B. U) under some circumstances. More fundamentally, it has been discovered that the solar system is not composed of an isotopically homogeneous mixture of chemical elements. Even for an element as abundant as oxygen, variations of the isotopic abundance have been noted for different parts of the solar system. Such isotopic variations have now been established for several chemical elements and provide clues to the processes that gave rise to the chemical elements, as well as to the conditions that existed at the birth of the solar system. A startlingly large isotopic variation was discovered in the uranium of ore samples from the Oklo Mine in Gabon (West Africa) in 1972. Unusually low isotopic abundances of uranium-235 in these ores led to the astonishing conclusion that, 1.8 billion years before the first man-made nuclear reactor, nature had accidentally assembled a uranium fission reactor in Africa! This reactor was made possible by the higher 23su concentration (~3 percent instead of the present-day 0.7 percent) at the time. Mass spectrometric analyses of various elements in the Oklo ore proved that isotopic compositions labeled them unmistakably as fission products. It also made it possible to deduce such characteristics of the reactor as total neutron flow (] .5 x 102' neutrons cm-2), power level (~20 kW), and duration of the self-sustaining chain reaction (~106 years). An important practical result of the Oklo studies is the fact that most fission products, as well as the transuranium elements produced in the reactor, did not migrate very far in 1.8 billion years. This has a clear relationship to the possibility of long-term confinement of radioactive waste products in geologic formations. Nuclear Chemistry in Medicine Nearly 20 million nuclear medicine procedures are performed annually in the United States (radioactive iodine thyroid treatment is one example). Advances in nuclear medicine depend crucially on research in nuclear and radiochemistry. For example, great progress in our knowledge of the chemistry of the element technetium in the past decade will clearly lead to much more elective applications of radioactive technetium, 99Tc. This is the most widely used radionuclide, because the chemical properties of technetium compounds give them therapeutic activity. For example, technetium tends to concentrate in bone and particularly in cancer- ous bones, providing important diagnostic power. Another important example is the development of especially rapid ways to incorporate into molecular structures short-lived isotopes that emit positrons. Two examples are the carbon isotope ~C, with a 20-minute half-life, and the fluorine isotope, OFF, with a 110-minute half-life. Both are produced through cyclotron 165

166 INTEL r.F`CTUAL FRONTIERS IN CHEMISTRY bombardment. These nuclei are then placed in such compounds as '~F-2-deoxy-2- fluoro-D-glucose and 1-~C-palmitic acid in a time short enough to permit their use in positron emission tomography (PET), which is analagous to X-ray tomography (CAT scan). The positron technique is finding new clinical applications in studies of the nervous system and the heart, known as neurology and cardiology. Stable isotopes, in conjunction with NMR spectroscopy, also have important applications in medicine. With tic, 2H, '5N, and ~70 tracers, NMR spectroscopy of humans will allow new insights into the molecular nature of diseases, provide a noninvasive method for their early detection, and make possible studies of metabolic processes in living subjects. This has led to one of the most exciting developments of the last few vears. large object imagine. In this technioue. a ~ , ~ ~ ~ computer stores the NMR signals that result when an object as large as a human is slowly moved through the magnetic field of the NMR sample space. Then the computer reconstructs a three-dimensional image of the object, showing the location and local concentration of the atoms whose NMR is being measured. Thus, the presence and chemical form of key elements can be mapped in entire human organs in living patients. These powerful, noninvasive techniques were literally undreamt of 15 years ago. They have arisen in response to demands for ability to study via NMR ever larger biomolecules and working biological systems. SUPPLEMENTARY READING Chemical & Engineering News "Vibrational Optical Activity Expands Bounds of Spectroscopy" by S.C. Stinson (C.& E.N. staff), vol. 63, pp. 21-33, Nov. 11, 1985. "Progress Reported in Coupling LC and MS" (C.& E.N. staff), vol. 63, pp. 38-40, May 20, 1985. 4'New Chromatography Columns Cut Need for Sample Preparation" by W. Worthy (C.& E.N. staff), vol. 63, pp. 47-48, Apr. 29, 1985. "New Methods for Trace Analysis of Man- ganese" (C.& E.N. staff), vol. 63, pp. 56-57, Jan. 14, 1985. "Microsensors Developed for Chemical Analysis" (C.& E.N. staff), vol. 63, pp. 61-62, Jan. 14, 1985. "New Laser System Far Surpasses Mass Spec for Surface Analyses'' by W. Worthy (C.& E.N. stair, vol. 62, pp. 20-22, Oct. 8, 1984. "New Detectors for Microcolumn HPLC" (C.& E.N. staff), vol. 62, pp. 39-42, Sept. 17, 1984. "New Methods Shed Light on Surface Chemistry" (C.& E.N. staff), vol. 61, pp. 30-32, Sept. 12, 1983. "Archeological Chemistry" by P.S. Zurer, vol. 61, pp. 26 44, Feb. 21, 1983.