TIMS Tutorial: The Basics

Monday morning I finished the first drafts of all three of my NSF graduate research fellowship application essays.  Monday afternoon\Tuesday morning I finished this week's homework set for the applied statistics course I'm taking.  But now it's Wednesday afternoon and today I feel like I haven't completed anything besides discovering new research roadblocks.  It seems like graduate school is full of bursts of fairly impressive productivity followed by lulls where the instrument doesn't work, experiments fail one after another, and the to-do list gets longer and longer but nothing gets marked off.  If anyone with vastly more science experience has tips for how to overcome lulls I'd be happy to listen!

In an attempt to do something productive and work on my science communication skills (I need all the practice I can get...qualifying/comprehensive exams are looming!) I have decided that one of the projects I'm finally going to tackle during my lab work lull is to write about the main instrument that I use in my research.  Plus, I often learn best by teaching!  I've been meaning to do this for a while and if you've seen references in my posts to "the instrument" or "the TIMS" and wondered what on Earth that was..this is the series for you!  Hopefully, today's post will provide necessary background for those of you with little or no exposure to isotope geochemistry.  Next post I'll show you how the samples get loaded onto the machine and eventually we'll talk about how the machine actually goes from the samples to a useful result and what running the TIMS is really like.

The primary instrument we use in my research lab is a "thermal ionization mass spectrometer," which we refer to as "the TIMS" for short.  [Which, incidentally, has lead to several comical instances where people overheard me telling someone that I "date garnet" or "spend a lot of time with the TIMS" and then inquired about the romance in my life after assuming I was talking about a boy! lol] Breaking down the full name of the instrument actually gives a pretty solid overview of the instrument's purpose.  The TIMS uses high temperatures to ionize (break down into individual charged atoms) the sample, separates the ions by mass using an electromagnet, and allows us to measure the relative abundance of ions at specific masses to determine isotope ratios.  If that just went over your head, don't despair yet!  I promise upcoming posts will have pictures and likely use household objects as analogies to break this overview down into manageable parts.

First...a little general background on geochronology and why I use this crazy instrument.  As an 'isotope geochemist,' my goal is to figure out the age of each of my garnet grains and then use those ages to examine bigger questions about how mountains form and the surface of the Earth changes.  Figuring out the age of the garnet is possible because garnet includes trace amounts of a radioactive element called samarium (Sm) that decays to the element neodymium (Nd) at a known rate.  It takes 1.06x10^11 (106 billion) years for enough Sm to decay so that only half of the original amount is left (called the half-life).  Also, garnet is basically never 100% pure...it usually contains little bits of other minerals that we refer to as inclusions.  To borrow an example from my adviser...you can think of it as a chocolate chip cookie.  The cookie is the garnet and the chocolate chips are the bits of other minerals, or inclusions, inside the garnet.  Many of these inclusions contain a different ratio of  Sm to Nd than the pure garnet (just like the chocolate chips and the cookie base would have different ratios of sugar to milk).  Therefore, I use the TIMS to measure the Sm and Nd ratios in the pure garnet and in the inclusions.  Since the ratios are different and the rate of decay is known, when I plot the ratios for the pure garnet and the inclusions I can define a line between two points, called an isochron.

The slope of the isochron = The age of the garnet!

By the time I am ready to walk into the TIMS lab and start an analysis, I have already been working through sample preparation and clean lab chemistry on a single garnet grain for several weeks.  Those week of work result in a set of 4 beakers, each containing a single bead of dried goop about the size of a ballpoint pen's tip.  The four beakers contain the Nd from the pure garnet, the Sm from the pure garnet, the Nd from the inclusions, and the Sm from the inclusions.  Each isochron that results in an accurate age requires all four samples to run successfully on the TIMS, usually with only one shot at getting each one right. 

This is a very simplistic overview of geochronology and mass spectrometry, especially since there are entire courses on the principles of these areas of study.  Basically, once you understand that by measuring isotope ratios of Sm and Nd we can figure out the age of garnet, you understand the purpose of using a TIMS and one of my basic research goals. The remainder of my research examines ways that we can apply these methods to new materials that have never been dated or how the ages that I get tell us about the rates of metamorphism, when sedimentary basins open and close, how metamorphic conditions change over time, and much more.  

My First Garnet Age

It's been way too long since I've written a post.  I've saved up a nice long list of topics that I can write about and adventures from the last month and half but just haven't gotten around to writing the posts yet.  I'm hoping that this post will be the first of several over the next couple of days to get more or less caught up and whittle down my idea list.  Instead of going chronologically, I'm going to start with a science update...since that's the cause of my delay in writing anyway. :-)

There has been a lot of exciting science in the past month and a half since I wrote about my first trip to WHOI to use the SEM.  I'm balancing two classes (Geodynamics as well as "Thermodynamics and Kinetics of Tectonometamorphic Processes," which is taught by my adviser) and finally getting pretty close to being independent on most of my research procedures after learning several new steps these past couple of weeks.

The biggest push for research has come over the past two weeks.  This past summer when I did fieldwork in Scotland (see earlier posts if you missed the beautiful Scotland scenery!) I collected several samples and shipped them back here.  In November I sent two of those samples off for crushing and mineral separation and they came back roughly three weeks ago.  As I was in a meeting with my adviser just over two weeks ago, we were talking about research priorities and he mentioned that he would love to have an age for garnet from one of the samples ahead of his upcoming talk at the University of St. Andrews, Scotland on his European speaking tour but he wasn't sure if it was possible.  Normally taking garnet from raw mineral to an age takes us three to four weeks, with two weeks being the absolute fastest speed possible if everything goes according to plan.  I agreed to try and plotted out every step on a calendar, knowing that I would make it with just about 12 hours to spare if I succeeded.  It was also complicated by the fact that I have never completed some of the final steps, because I just hadn't gotten there yet with test samples.

So, basically the last two weeks were two solid weeks of lab work.  It included a second trip to Woods Hole to do SEM work on the Scotland samples (and incidentally learning how to change an SEM filament...), about a week and a half of 12+ hour workdays in the clean lab in my superhero get-up, two days of learning to load samples and how to load the mass spec., and two days of learning how to run our instrument.  (Eventually I'm hoping to write a series of posts to show you some of the basic methods and our instrument, the TIMS.)  Amazingly, my sample survived and I can finally say I have my first garnet age!!  Not only was this the first time I generated an age by doing the procedures from beginning to end, I made it within two weeks (with 11.5 hours to spare), it was a ridiculously small sample (I started with only 4mg of garnet...4 sand grains basically), many of the procedures I was doing independently for the first time or I was learning with supervision for the first time, and the result was right in the expected ballpark!  If you can't tell, I have been quite pleased for the last day or two.

Now that I have my first age under my belt, I can start testing and implementing some of my ideas for changes to the procedure to improve how our lab works with very tiny samples and for further method development on the new garnet dating system I am developing.  While I'm excited about my first data point, there are many improvements to be made and many more points to be gathered.  Onward in the pursuit of science.

Grains under the Microscope: A WHOI Adventure

About a week and a half ago I had the opportunity to visit Woods Hole Oceanographic Institute (WHOI) on the southern part of Cape Cod in Massachusetts.  I took a day trip away from my usual research at BU in order to gather some test chemical data on a couple of my detrital garnet samples.  I received an invitation from a research scientist at WHOI offering me the opportunity to try out a piece of equipment we do not have in our department at BU.

Before I describe what my day at WHOI looked like, let me give you some general background on the goal of the trip and the method that I was using for those of you who are unfamiliar with the SEM.  I traveled to WHOI to use a tabletop scanning electron microscope (SEM).  A scanning electron microscope uses a beam of electrons that bombard the sample.  When the electrons collide with the material being analyzed, different types of detectors can be used to provide information about the electrons that are scattered or energy that is emitted to learn more about the sample being analyzed.  The instrument I was using is smaller than most standard SEMs and is designed to be user-friendly, to take up less space, and to allow for analysis not just of polished, coated samples (which is the standard procedure) but also of rough, uncoated samples using a lower vacuum setting.  This makes this instrument an excellent candidate for analysis of my samples, since the grains do not have to be coated or polished, which would destroy them.  The tabletop SEM is equipped with two different detectors- a BSE detector and an EDS detector.  BSE stands for backscattered electrons and measures the electrons from the beam that are scattered and reflected by the object being analyzed.  It allows for the differentiation between heavy elements and light elements, since heavy elements scatter electrons more strongly.  From this differentiation, an image of the material being analyzed is generated using grey scale that can show surface detail or different mineral phases.  Below is an example from one of the grains that I analyzed at WHOI, though this particular grain is zircon rather than my usual garnet subject.

BSE image showing a zircon grain, analyzed at WHOI 1/20/2012

Besides BSE, the SEM can also do EDS analysis, which stands for energy-dispersive X-ray spectroscopy.  The electron beam hitting the sample can trigger the release of x-rays and the energy of the x-rays released is characteristic of the major elements that compose the object being analyzed.  This makes EDS useful for preliminary chemical characterization and mineral identification.  EDS results in a spectrum showing major element peaks and can also generate atomic and weight percentages for major elements using spot analysis.  Basically, we can get a general idea of the composition of the material being analyzed.

While I was at WHOI, I analyzed twelve different detrital garnet grains.  I used grains from two different locations in Vermont that we have been using as test samples throughout the method development process.  Three of the grains were from one location and the other nine were from a second location.  The sand collected at the second location contains detrital garnet of different colors, so I was able to analyze three grains of each color (red, orange, and pink) hoping to test the ability of the instrument to differentiate between garnet with different compositions.  When doing detrital work, it is important to be able to identify populations of similar grains that likely originated from the same general area and should be roughly the same age, since a single detrital sediment can contain different populations of garnet grains that originated from different locations and were affected by different conditions.  In addition to the garnet analyses, I also analyzed many grains of any other mineral that could remotely resemble garnet under the microscope from one of my other field areas to verify that garnet is not present in the sample.  Here is a picture showing how the grains were mounted for analysis.  The grains are still attached to this mount- they are just too small to see easily in the picture and blend into the carbon sticky tape.  The grain mount is sitting in a small weigh boat to minimize contamination.

Grain mount for SEM using carbon sticky tape

Just for fun, let me give you a look at what my day of research looked like.  My commute to WHOI started around 6:30am and involved two different subway lines, a Peter Pan bus, getting picked up at the bus station and a short drive to WHOI's Quissett campus.  I arrived around 10:00am and started learning how to run the instrument and by 11:00am I was running the instrument independently.  I then spent the next five to six hours (minus a lunch break) analyzing as many grains as possible and accumulating roughly 90 computer files between BSE images, EDS analyses, etc.  Then I took a shuttle bus to Woods Hole, where I ate dinner in an awesome little cafe while waiting for my bus to Boston.  (If you ever find yourself in Woods Hole, MA..I highly recommend eating at Pie in the Sky. It was delicious!)  After a bus ride and two more subways, I arrived home around 10:30pm.  It was a long and tiring day...but definitely worth it.

Here is one final picture, showing me running the SEM at WHOI.  Not only did I learn a new instrument and gather helpful information on some of my samples, I am learning how to be an independent scientist!

Me running the tapletop SEM at WHOI

Grains of Sand

About a month ago I told you about my piles of sand.  Now I am headed for single grains of sand and you are invited to follow along with my next science installment!

My last science installment followed the steps involved in generating several vials of sand with magnetic properties close to those expected for my mineral of interest- garnet.  This process narrows the search field, but unfortunately never yields a vial of pure garnet.  That's where the next step comes in.

The next step is known as hand-picking.  It takes place under a stereoscope, which looks like this:

Stereoscope on and ready to go

Basically, it's a microscope that allows the user to examine three-dimensional samples under varying magnifications and light intensities.  This particular scope is also outfitted with a digital camera for taking pictures of the individual sand grains when necessary.

The samples, which can be seen in the vials to the right of the scope above, are dumped into a weighing boat (the blue plastic dish the vials are in- above) or a clean glass dish.  The dish and sample are placed under the stereoscope.  Then I have to use the magnifying powers of the stereoscope, very fine tweezers, and my mineral identification skills to sort the garnet grains from all of the other grains of sand, proceeding one grain at a time.

PTFE-coated Tweezers, ink pen and lab notes for scale

I have two pairs of tweezers- one made from non-magnetic stainless steel and the other pair coated in PTFE (Teflon).  Both are considered needle-point and are essentially some of the finest tipped tweezers available.  They have to be...I am using them to pick up individual sand grains!  Why two pairs you ask?  The pair coated in Teflon is intended only for use with my detrital garnet grains, to minimize the exposure of the grains to unnecessary metal.  Contact with metal objects could, theoretically, contaminate my samples and skew my results since I am working with such extremely small sample volumes.  I am pretty sure that any potential impact of metal contact will be mitigated by later processing, but it never hurts to avoid the potential problem altogether when possible!  

Ultimately, each grain of garnet is individually picked out of my piles of sand using the tweezers and placed in a new vial containing only garnet from a single sample.  Generally, garnet is fairly easy to identify beneath the stereoscope's magnification, since it is pink to red to reddish brown.  Most of the other minerals that end up with it when grouped magnetically are nowhere near this color spectrum.

Hand-picking continues until I have gathered a vial with enough garnet to be analyzed.  This can take quite a bit of time, since each grain is selected individually.  Once enough grains are selected, it's off to the clean lab!  More on those adventures will be coming soon..... 

Of course, this is nowhere near an ideal system for mineral separation yet.  My original sand contains an extremely large volume of two minerals known as quartz and feldspar.  They are so abundant, that although magnetic separation should keep them far away from the garnet, they have permeated everything and to make matters worse, the feldspar is pink!  I am currently researching additional processing steps to add to our current mineral separation methods, such as heavy liquids and/or a water table, to eliminate this problem. Hopefully there will be posts in the future about new processes and procedures that I have added to our lab's capabilities to deal with these issues.

Beyond Boston: New Hampshire IV Retreat

After a weekend away and a busy beginning to this week (first exam as a grad. student and beginning to learn partial dissolution procedures in the clean lab!), I am just now writing about this past weekend's adventure.

I took the opportunity to get out of Boston and see part of neighboring New Hampshire.  Intervarsity offered graduate students from around Boston the opportunity to go on a fall retreat to the Toah Nipi retreat center, near Rindge, New Hampshire from Friday night through Sunday afternoon of the long weekend.

View from the Beach

Toah Nipi means "wide waters" in the Algonquin language.  It has long been a meeting place where people came together, whether from the Algonquin people or attending modern retreats and camps.  In fact, a birch-bark canoe recovered from the location attests to the rich heritage of the site.  The buildings are surrounded by woods containing miles of hiking trails on all sides and are near a beautiful lake rimmed by trees just beginning to change colors.  

Becca canoeing!

Almost 200 graduate students were in attendance for the weekend event.  We had all kinds of activities spanning everything from fun and games to worship and learning.  On Saturday we had free time, which I filled with canoeing on the lake, an hour-long hike in the woods, helping to assemble apple pies for dinner, and many rounds of the game Bananagrams.  Other fun activities included a late-night game of Dutch Blitz (that I won!), group games with BU GCF, and a Jack-O-Lantern carving contest between the groups from different schools.  We also had meaningful times of worship, joined in group Bible study with other graduate students from our own schools, built friendships while having fun and around the campfires, and received invaluable advice from people who have been where we are now and are willing to share their experiences by serving on a mentor panel.  We were also challenged to seek balance in our lives and to live now like we want to live the rest of our lives.

Campfire on the beach

On Sunday, I also happened to be celebrating my birthday while on the retreat!  At breakfast they surprised me with a large square of coffee cake and candle and sang happy birthday.  I don't think I could have asked for a better way to spend my first birthday of graduate school than building friendships and being challenged to grow.

It was wonderful to be able to leave behind some of the worries and stresses of graduate school for the weekend.  Don't worry though- I am a geologist through and through.  As I sat on the beach on Sunday morning, I found myself looking down at the sand around my feet and thinking "I wonder how much garnet I could separate out from this sand in the mineral separation lab...." then proceeded to sort through a handful to find a grain to show to a friend so she could see what I am studying!

Ever the geologist...."How much garnet is here?!"

Piles of Sand

Am I doing any work or just exploring the city? What have I been up to in the lab?  I'm so glad you asked, and I have an answer for you this week.  Prepare yourself- this is a (lengthy) science update post!

I am now officially one month into my time here in Boston.  Besides adjusting to life as a grad. student, life in a new city, and my two classes for the semester (geochemistry and a mathematical modeling course), I have plenty of ambitious research goals for the semester.  In fact, I have at least four projects that I am supposed to be starting or working on at some point in the next couple months.  I have a feeling you'll hear about them over the coming months..or years.

But before I can produce large amounts of relevant data in the lab, I am currently in that joyous stage of being the newbie, the trainee, the person who has no idea how to do most of the routine procedures in my new lab and has to ask someone for help every time I come up against the next step.  Clearly there are some benefits to that situation (like being limited in what labs I have been trained in so therefore feeling no guilt at not spending every waking moment in the lab or the constant learning of new and exciting things) as well as drawbacks (like not being even remotely self-sufficient and making slow research progress).  I have been assured by my adviser that it could take me a good year to two years until I can confidently handle any and all of the research tasks expected of me- so I suppose I am doing well for a month in. It's all about perspective.

That being said, I am happy to share with you what I have accomplished this month in the lab.  Really, it all centers on making piles of sand.

Glen Clova, Scotland

The first of my four projects involves sand collected from a river flowing through Glen Clova, Scotland.  Glen Clova is a famous glacially-carved, U-shaped valley and is part of the type locality for the Barrovian metamorphic sequence.  I visited in June and collected the sample, along with my adviser and field assistant.  (If you missed my post on summer fieldwork and want to see more details on my trip to Scotland or pictures, check it out here.)  The current goal for the sample is to separate out the mineral garnet and attempt to establish an age for individual garnet grains.  This sample will be a part of the method development for my dissertation.  My overall PhD project is focused on developing a system for dating detrital (eroded and then redeposited elsewhere) garnets, which has essentially never been done and would open a whole new tool for geochronology.

The first step is taking my bucket of sand and separating out the grains of garnet from the overwhelming amounts of quartz, feldspar, and other minerals present.  While you could theoretically just sit there with tweezers and hunt through the bucket, that would be terrible and not very efficient.  Therefore, we can use properties of the minerals to separate them into different "piles."  Many geologists (and most prospectors) have gone through a similar series of steps as a part of the basic sample preparation that is my current lab activity.

The first step is to send the sand through sieves with different size mesh screens.  This divides the grains into piles with similar grain sizes.  This is important not only for later mineral separation steps, but allows me to gain an idea of the grain size distribution for garnet in this sample.  Since I am developing a new method that uses a single grain of garnet at a time, I would ideally like to find the largest grains that exist to start with, before working down to smaller grains.  Also, since my samples have much larger grains than what my lab is used to working with (finely crushed "pure" garnets), I got to determine that we need some sieves with larger mesh sizes and order them.  Thus, my first contribution to making my lab more versatile! Hooray!

Once the grains are separated by size, I have been using a Frantz electromagnetic separator to divide the grains based on magnetic susceptibility.  Most mineral grains are magnetic to at least a small degree, especially if you use a big enough electromagnet and set it appropriately.  The differences in magnetic susceptibility between minerals of different types allow sand to be separated into piles based on the degree of magnetism.

The Frantz I've been using the past couple weeks

Here's how it works:

1. Clean the machine thoroughly.  (You definitely don't want grains from the previous user's sample contaminating your sample!)
2. Turn on the magnet and set the desired level of magnetism.  Since I'm aiming for garnet, I usually start with a high setting to eliminate the least magnetic minerals and work my way towards gradually eliminating more magnetic minerals.  People who want to separate out other minerals may approach it differently.
3. Load whatever sample you wish to separate into the glass container that will feed the grains into the machine.  Usually you want to use a hand magnet on this sample before you load it, because that will remove one mineral (magnetite) that is the most magnetic and keep very magnetic grains from sticking to the magnet.
4. You can control how quickly the glass feeder container and a tray between the two portions of the electromagnet vibrate to control how quickly grains move through the magnet.  Grains fall out of the feeder (ideally basically one grain at a time so they don't influence one another) and move down a tray between the electromagnet.  The tray is tilted so that by gravity, grains prefer to move down the side away from the operator.  However, grains that are more magnetic than the magnet's current setting will be pulled up to the portion of the track closer to the operator.  The grains are then separated into two containers at the other end- the closer one being more magnetic and the one further away less magnetic.
5. Since I want garnet I continually rerun the pile collected in the container closest to me at a lower magnetic setting so that it is separated into two new piles.  This continues until I have one or two piles that have a larger proportion of garnet than any of the other piles.

Some people use heavy liquids or water tables to separate minerals by density and to reduce the amount of time required and increase the efficiency of the Frantz.  We have been discussing having me implement these strategies in our lab eventually, especially since I may have to separate relatively large amounts of sand.  We shall see what the future holds.

I also got to make another contribution to improving my lab's efficiency by constructing a "backboard" to keep my grains from bouncing out of the Frantz as they dropped out of the feeder.  Evidently the combination of larger grain size plus an old Frantz model leads to quite a few grains trying to jump ship before they head down the tray- regardless of how high or low the vibration of the feeder and tray is set.  I improvised a shield from weighing paper and label stickers- nothing like scientific McGuyvering.

That pretty much sums up month one in the lab.  So far it has been basic sample processing (all of which I actually did once upon a time as an undergrad. for a term project) but at least it's progress!  I'll share the next step for these piles of sand in an upcoming post.