A Portrait of the Scientist as a Young Woman Read online

  In the early 1800s, Michael Faraday made fundamental discoveries about electricity and magnetism in a simple laboratory; he could have done them in his kitchen. Around the same time, Charles Lyell published his Principles of Geology, which contained the first clear, supported argument that geological processes are slow and uniform and extend over far greater reaches of time than had been previously imagined by Western natural historians and philosophers. Some had thought the world had been in existence for only a few thousand years, while others posited ages in the tens of millions, none longer than about 400 million years (Lord Kelvin), still only one-tenth of the Earth’s real age. Lyell drew upon travels and observations from all over the world to make his arguments, which spanned three volumes. Without that understanding of the immensity of time, geology made no sense, and could not be studied as a science. Lyell’s work greatly influenced Darwin, and helped geology begin its long, slow climb from descriptive science to hypothesis-driven science.

  Already, as a freshman, I had the idea that the real goal was doing, not studying about. MIT is dominated by its research culture, and has as many graduate students, who spend all their time doing research, as it has undergraduates. It’s normal, almost expected, for undergraduates to have research jobs with faculty. Sitting in the kitchen area of our dorm hallway in the late afternoon after classes, I watched some students trail in after sports practice, and others trail in after time in their labs. Finally, at the end of the first semester of my freshman year I summoned courage that surprises me to this day, and called the famous professor Nafi Toksöz. Nafi came to MIT in 1965, the year I was born, and founded the Earth Resources Laboratory, where he and his team made fundamental discoveries about earthquakes, plate tectonics, and natural resources. He could have stood on his high pedestal and been intimidating, but instead, he greeted the world with a warm, humble smile, I soon learned. On the phone I asked him if he had a research position for a freshman with no experience of any kind. Some experience, actually: I had been a post office clerk.

  Nafi took my call, and listened to me, and he hired me and set me to work writing code for the New England seismic network. These seismometers, scattered around New England, were used for monitoring earthquakes, and also for detecting Soviet nuclear tests. We could tell the difference between the two by the shapes of the waves recorded by the seismometers. So now I was both an Earth scientist and doing something about that terrifying nuclear world of Hans Bethe.

  For hours each day I would sit in the Earth Resources Lab building, surrounded by computer terminals and the great big reel-to-reel tape machines of that era. The next room had an under-floor cooling system for the big computers, and we sometimes put sodas under the floor tiles to make them cold. Our room was filled with the sound of key taps and tape spooling, the tables covered with continuous-sheet printer paper with those perforated edges. There I worked with a couple of undergrads and a couple of grad students and a couple of staff scientists, led by the kind, smiling, and supportive Mike Guenette. Mike taught me how to code in FORTRAN and set me to work writing a program that located the town closest to the coordinates of any given earthquake our New England network detected. That team of kind, supportive people became my home port at MIT. We went to dinner at Mary Chung’s in Central Square and ate the searingly hot Suan La Chow Show. We shared our lives in that magic way which happens most often at that age.

  Since I lived on campus not far from the lab, I often drew evening duty changing the paper rolls on the seismographs. I felt strong and even a tiny bit important as I walked, holding the building key in my pocket, across dark Ames Street and over to the Earth Resources Laboratory. I loved turning on the lights in the dark room and knowing how to stop the recording, and remove and replace the foot-wide cylinder of paper with its many seismic wiggles.

  One night I turned on the lights to discover that the tiny pen had drawn huge waves across the heat-sensitive paper, waves so large that they crossed over all the other wiggly lines and bumped against the limit of the pen’s available movement. There was a literal red phone for such emergencies, which I picked up that evening for the first time to call the lab manager at home and tell him that a big earthquake had just been recorded. No one yet knew where. This was the giant 1985 Mexico City earthquake, which showed up as huge rolling waves and spikes on the seismic monitoring network all the way into New Hampshire.

  Here I had found a new team that felt like family. This group and the work we did together gave meaning to what I learned in the classroom, and unlike completing problem sets, achieving goals with Nafi’s team felt real. I began to feel I had found my footing a bit, as I completed the year and went home to Ithaca for the summer.

  In my sophomore year, beginning the studies of my major, I was in a class called Igneous Petrology. Our professor, Tim Grove, was serious and rigorous and organized with pages of notes, but also approachable: he wore Birkenstocks to class with his proper shirt and trousers. He was rigorous, though, and would look at us unsmilingly through his wire-rimmed glasses when we fumbled an answer.

  One day Tim announced he was looking for an undergraduate to work on some research. I eagerly volunteered. Later that week I met Tim in his office on the twelfth floor of the tall MIT Building 54, more commonly known as the Green Building. The office had a microscope, a computer, piles of papers and books, and Tim’s modest metal desk. The window had a sweeping view over the Charles River to Boston, but Tim’s blind was often down. Tim told me he had an idea that was risky; the required experiments would be time-consuming and might or might not work, so the project was perfect for an undergraduate, who would learn all about scientific process and how to do experiments, without any downside if no results materialized. High risk, high reward? No surprise and no problem if the project failed in the end? I was in. I started asking questions. Tim answered a few, and then said, coldly, “Questions are fine as long as there are not too many of them.” I felt a chill, and realized it came from a memory of my father, and his temper, as sudden and irreversible as a fall into a chasm, requiring just one wrong step.

  Tim introduced me to what became my first great scientific obsession: conducting high-temperature and high-pressure experiments that mimicked the interiors of the Earth and the Moon, and allowed me to interpret what the temperature, pressure, and composition of those interiors are now, and what they were in the past. That felt like real science to me, teasing out truths of our universe from seemingly unrelated bits of rock, and learning about places too remote for us to ever visit. I could work at the processes and become an expert. I could learn something about the Earth no one else knew.

  Tim’s project concerned the most common rock-forming mineral on Earth, feldspar. In many igneous rocks—rocks that have solidified from a magma—two feldspars coexist: one with high potassium, and one richer in a mixture of calcium and sodium. The exact compositions of each—that is, their relative proportions of potassium, calcium, and sodium—vary with the temperature and pressure at which they formed. Tim thought we might be able to heat powdered rock at set pressures and temperatures, extract the resulting rocks, and measure the potassium, calcium, and sodium in their feldspars. By doing this over and over at different pressures and temperatures, we would build a database of the compositions of feldspars formed at a range of pressures and temperatures. Then, anyone with a natural rock containing two feldspars could measure their compositions, and compare them to our database to discover the pressure and temperature of their natural rock’s formation. This kind of calibration is called a geobarometer and a geothermometer. We were setting out to make the first feldspar geobarometer and geothermometer. Scientists could take rock samples and find out the temperature and pressure of their formation, answering questions like, how deeply was this rock buried in the crust before it was thrust up into a mountain?

  Tim explained that the furnaces we needed for our experiments didn’t exist in his lab yet, and we would build them together. For several months, I was Tim’s apprentice
. In one room of the lab, tables and old soapstone counters were filled with dissecting scopes, squat glass evacuated storage bells, polishing sandpapers, acid bottles, boxes of gold tubing, a cardboard barrel of barium carbonate (not for rat poison, but for a special kind of high-pressure experimental apparatus), welding machines, and hot-pot furnaces over which teetered shelves of heavy books. In another room, the high-pressure and -temperature furnaces were packed as close as could be designed. To get to the back of the lab where Tim and I would work, I had to turn sideways and slide between the lab bench and the big disk of the pressure gauge for one of the furnaces.

  Tim followed that most correct and fundamental rule in science: You must know how your tools work. You should build them yourself. When he hired me, he looked at my hands: Was I a tool user? Yes, I was—the calluses from building and riding my bike were still there. He showed me how to connect high-pressure tubing from the water reservoir to the valve in front of each bomb and furnace, and how to connect the valve to the bomb’s screw cap to that thinner, flexible high-pressure tube, itself as thin as a wire.

  We built a dozen cylindrical high-temperature ovens, lying on their sides and opening like clamshells, into which would be laid smaller metal cylinders—the cold-seal bombs—each one about the size of a thin baguette. Each bomb had a tube drilled into its center, where we would insert a tiny volume of powdered rock welded into a gold capsule about a centimeter long. One end of the bomb was closed, and the other end had a finely machined screw cap connected to one of those wire-thin metal water tubes. The metal tubes connected in turn to a high-pressure water system. The metal bombs were pressurized from within with that water, and they lay in furnaces as hot as 1,650 degrees Fahrenheit.

  In the vernacular of today, what could go wrong?

  Those high-pressure fittings were fussy. They had to be tightened just the right amount and not too much. Tim showed me how to read the pressure gauge for the water reservoir; if it was falling, there was a leak somewhere. Sometimes the leaks were pinholes in the high-pressure wires, and extremely hot, pressurized water would spurt out in an almost invisibly fine stream. Everyone in the lab—myself and the other students—was aware at all times of the danger of the furnaces. Certain kinds could explode, and pressure lines could open, material could burn, and the metal parts of the furnaces could crack under the pressure with concussive bangs like gunfire. There had been all sorts of disasters, both small and large, though because of our training and care and some luck, no one had ever been hurt. Tim told me to call out “Noise!” in the moment I dropped a tool, so that its clatter on the floor would not cause too bad a startle for the others in the lab, all tuned to a high pitch by the challenges of the apparatus.

  When we were ready to do the first round of experiments, Tim explained how to request pure mineral samples from the Harvard collection to supplement what he already had in the lab. These mineral names still read like poetry. There was Amelia albite from Amelia Court House, Virginia; Crystal Bay bytownite from Crystal Bay, Minnesota; Lake Harbour oligoclase from Baffin Island, located in the Canadian territory of Nanavut; the Sannidal andesine from Sannidal, Norway; Hugo microcline from the Hugo pegmatite, South Dakota.

  I ground them to powder and measured them to the ten-thousandth of a gram in order to make exactly the starting compositions we wanted. Then, I learned how to make the tiny gold capsules, hammering the capsule into the correct width, and welding in the tiny measure of rock powder and droplet of water using an arc welder with a hand-sharpened carbon tip. The welder had an electrical short in it, and sometimes when I stepped on the foot pedal to start the current, the machine shocked me in my eye socket through the welding glasses my face was pressed against. Having a steady hand under these circumstances took special determination.

  Finally our first ten experiments had sat baking in their ovens, up at pressure, monitored every day, for six full months. One by one, using welder’s gloves, we lifted the bombs from their fiery oven beds. We held them lid-down and rapped them with a great wrench to loosen the experiment within its narrow drilled cavity, allowing the experiment to fall into the cooled head of the bomb, which quenched the experiment to room temperature quickly and without allowing any additional reactions to occur. Each little gold capsule in turn tumbled out of its bomb after the lid was unscrewed. One or two were discolored and wrinkled, and Tim explained that those had almost certainly burst during their long bake and would be useless. The others we carried reverently into the microscope lab. One at a time we would open them under a binocular microscope, each glittering capsule snipped open over a clean petri dish. So much labor had gone into them that we treated each one as delicately as a butterfly, more precious than the gold that wrapped them.

  I watched intensely over Tim’s shoulder as he clipped the end of the first gold tube and upended it in the petri dish; he was lit like a star actor on the stage by a focused light and enlarged under the dissecting microscope. A drop of water, and a tiny pile of sand poured out. “Ugh,” Tim grunted.

  “What? What happened?” I asked.

  “Nothing,” he said. “Nothing happened. That’s the same rock powder and water you welded into that capsule six months ago.” Well, we thought, perhaps that furnace had a faulty temperature sensor and it hadn’t been at a high enough temperature. Our frustration grew as one after the other produced the same disappointing heap of powder, unchanged from when it was first welded into the capsule. Those six months had not been enough; the temperature and pressure had not been enough. No new feldspar minerals had crystallized.

  Losing all those experiments was no small failure. The time, the cost, the care . . . in all, the better part of a year was gone with nothing to show. But at the time, I knew no better—this is what experimental science is, I thought! We’ll try again, said Tim, at higher pressures and temperatures. And so, we started again, from the very beginning.

  By midway through my junior year, thanks to the higher reaction rates at higher temperatures, the experiments were working. When I snipped open a gold capsule under the dissecting microscope a gratifying tiny pebble of feldspar, newly crystallized in the oven from its powdery starting materials, rolled out into the petri dish. These I mounted in epoxy, polished by hand using increasingly fine sandpapers and wet polishing compounds, coated with vapor-deposited carbon, and analyzed in an electron microprobe.

  Oh, that electron microprobe. The instrument filled a whole room, and the operator (me!) sat in front of an angled bank of screens, knobs, dials, and switches. On one side was the big sample chamber, held at high vacuum, and chilled with a Dewar of liquid nitrogen that I would refill for each measurement run, often spilling a shower of skittering liquid nitrogen droplets across the floor. A microprobe measures the atomic composition of solid materials by shooting a column of electrons at the surface of the sample. The electron gun is just a hot tungsten filament in the vacuum chamber. The electrons stream off the filament, are focused by electromagnetic lenses, and strike the grounded sample in a beam as small as a micron across, a hundred times smaller than the smallest thing you can see with your eye, smaller even than bacteria. The electrons add energy to the atoms of the sample, causing the electrons in those atoms to rise to higher energy orbits around the atomic nucleus. When each electron loses that added energy, it sinks back to a lower orbit, and releases an X-ray. Those X-rays fly away from the sample, and some of them strike special crystal sensors in that sample chamber. Each atom releases X-rays of characteristic energy, and so by counting those X-rays, we can know how many atoms of each kind are in the sample.

  The audacity to invent such a machine! I wondered at it then, and still do now. The electron microprobe allowed me to measure the exact composition of the feldspar minerals that formed in my experiments, and, in the end, allowed Tim and me to produce our feldspar geothermometer and geobarometer.

  But first, I had to learn how the machine worked. The friendly, brilliant, über-nerdy lab manager who knew every in and out of that mac
hine sat with me for hours and days while I learned and practiced. First, I watched and listened. Then, I operated the instrument while he watched. Then, I operated the instrument with him outside the room. Finally, after a couple of months, I could fill the microprobe’s liquid nitrogen tank; prepare my sample; coat it with carbon (its own exciting and error-prone activity, involving another vacuum chamber, a graphite rod, and electricity); put it into the microprobe; bring the probe sample chamber back to vacuum, calibrate, and standardize the probe so its measurements were reliable and replicable; and finally, measure my own experiments by myself, sometimes in marathon overnight sessions when the instrument cost was lower because of decreased demand (to keep the facility running, user time is paid for on research grants).

  Those nights were the first times I felt like a real scientist. The probe room was windowless and refrigerated, and completely dark while operating, so that I could better see the relatively dim images formed by bouncing electrons off the sample. Not only was I measuring composition with a focused electron beam, but I could also “see” my sample using electrons, instead of photons. The electron beam rastered back and forth across the sample to build up an image, and rather than seeing colors in visible light, I was seeing composition using electrons. The denser the atoms, the more electrons were bounced back, and the brighter the image was. Bright = dense, and dark = less dense. I understood the machine, could use it with some confidence, and was building a database of information about feldspar behavior that no one else had ever done. Like a lightbulb with a loose connection, from time to time my sense of confidence would flicker on and shine, and then, flicker off.