Sometimes lab work is repetitive...

...like today, when I spent the entire day measuring 1 gram of powdered sample into each of these ~120 test tubes:

But now I'm done!! Also, thank goodness for internet radio to listen to in lab!

How to use an XRF machine!

The X-Ray Fluorescence (XRF) machine is super important to those of us doing research on heavy metals in soils. The XRF has a simple purpose: to measure, in a given sample, the concentrations of the various elements. It gives accurate concentration data on most elements in the periodic table. Plus, it's quick to use, requires little sample preparation, and doesn't modify or consume the sample it tests - so the same sample can be used again for subsequent unrelated tests and procedures.

The basic idea behind XRF is to bounce X-rays off of a sample at varying intensities and measure the fluorescence from the different elements present in the sample. The more fluorescence, the higher the concentration. It looks something like this: (credit to www.cat.ernet.in)

You start, as with any test, by prepping the samples. For XRF, it's easy! XRF can be a dry test, so all you need is some soil ground into a fine powder. It's definitely time-consuming, and a little dull, but it doesn't involve lots of careful solution-mixing or anything like that.

A few of our 70+ painstakingly ground soil samples

Once the samples are ready, you need to assemble the containers to hold the samples. The containers need to have transparent bottoms which will interfere minimally with the X-ray fluorescence. For this, we use a special kind of plastic wrap called prolene, which is extremely thin and pure. It's also very delicate and clingy, and you shouldn't touch the part that will actually hold the sample, so it can be tricky to work with.

The container comes in two parts: a tube and a ring. The ring snaps snugly around the bottom of the tube. The tube is placed upside-down in a hole in a small wood plank, which just makes the top of the tube more flush with the surrounding surface, and therefore the prolene film is easier to work with.

Upside--down tube on the left, and ring on the right.

Removing a section of prolene film (carefully!)

The prolene film is placed over the ring. The wooden plank gives it something to cling to so it's easier to work with and doesn't fly away.

Finally, the ring is snapped over the tube, stretching the film across its bottom. There should be no wrinkles or dust on the film.

Once the containers are assembled, labeled, and ready to go, fill the containers with approximately 5 grams of sample each.

Empty assembled containers 

A perfect measurement :)

Then load the samples into the special rack for the XRF machine. The rack spins during analysis to switch samples. Screw the rack into the XRF machine and close the lid tightly to shield from X-rays!

The rack filled with samples

Loaded into the machine

Finally, enter the names and masses of the samples into the XRF software and press "Start". Make sure the helium gas, which is used instead of air to minimize interference, is at the correct flow rate. Check it periodically, and come back in about two hours to remove the samples and get your data!

Here's what the full XRF setup looks like:

Machine on the left and computer/software window on the right. On the left of the screen, the circles on the rack change color from red to yellow to green, indicating (respectively) untested samples, the current sample, and finished samples. On the right of the screen are the names and masses of the samples and some other information about them.

Where are our samples from?

Wuhan, China!

Sam's here!

Today my official mentor Sam, a post-doc, officially arrived with the samples she collected in China. There were kilos and kilos of... well, dirt, and we're going to need to process it all.

The majority, but not all, of the samples Sam brought back

 First we're going to be looking at small cropland samples taken from 69 field sites around Wuhan, China. Each field site has a sample taken at the surface (0-2 cm deep) and one taken approximately 10 cm below the surface. We'll be processing the shallower samples first so that we can get some data as soon as possible.

The samples we're going to process first - deep & shallow samples from 69 field sites around Wuhan, China

 By "processing", I mean that we'll need to take each sample, dry it completely, grind it into a fine powder, and place it into a labeled 50-mL tube.

Large tubes where we'll put the powdered samples after we grind them

From there, we'll be able to use the powder in the XRF machine to get heavy metal concentration data, and we'll be able to perform acid extractions on the powder and run the ICP-MS machine like I did with Sarah to get lead isotope data.

Learning to graph our data in R

The past week has been pretty calm - I haven't been in the lab at all, and we've been waiting for our samples to run and then analyzing the data. Today we finally had our complete data set, so Sarah and I worked on processing some of it in R and graphing it. Hopefully early next week we'll be able to meet with Scott (the PI, or Primary (faculty) Investigator).

In the meantime, I learned how to do some simple data analysis, check correlations, run regressions, and plot some simple graphs! I've only taken one coding class (this past spring), but it was fun to apply some of what I learned there to a completely different kind of project. Here's some of what we worked on today:

Looking at the correlations of various ions

Plots of our calcium-treatment data before removing redundant variables (you can tell there are a lot because many of the plots are straight-looking)

The lithium data had less obvious correlation

Our final calcium plot: the concentration of calcium is shown on the bottom, and the amounts of calcium and arsenic (red) sorbed onto the clay are shown on the y-axis. We had to throw out our last sample, so we don't know what it would look like at equilibrium.

Our final barium plot looks good!

Our final lithium plot shows that lithium probably doesn't help arsenic sorb onto the clay much.

Why clays? A background on my mini-project with Sarah

In order to get clean water, you have to play around in the dirt.

Many people are aware of the horrific rates of arsenic poisoning through drinking water in Southeast Asia. Professor Scott Fendorf and his team have studied this contamination process extensively in Bangladesh and Cambodia. Now my preliminary mentor Sarah is working under Scott studying a similar process closer to home,  in the Orange County Water District. Don't worry - arsenic levels in Southern California are only a tiny, tiny fraction of the levels in Bangladesh. However, a new plant in Orange County uses cleaned and purified wastewater to recharge subterranean aquifers (see the video and diagram below), and water district officials there need to be sure that the recharging process won't bring dissolved arsenic levels above the California standard maximum concentration.

Original video site here

Once wastewater is cleaned and purified, it re-enters the aquifers through two channels: infiltration basins (below, left side) and deep injection (below, right side). The infiltration basins simply allow the water to seep through the ground and into the higher-level aquifer. Below this first aquifer, however, lies a 'confining layer' made up of much less porous materials, particularly our phyllosilicate clays. 

A diagram of the different modes of the aquifer recharge process showing infiltration basins for subsurface aquifer recharge and deep injection for aquifer recharge past the denser confining layer and into the deeper aquifer.

Water must be injected through this confining layer, but the injection process has been shown to liberate some of the arsenic bound to the soil into the groundwater. Scientists at the Orange County Water Department, in collaboration with researchers like Sarah, have cleverly realized that the presence of certain ions in the injected water hampers the dissolution of arsenic into the water during recharge. In other words, if the optimal concentration of ions can be found, those ions can be put into the water during the purification process and the water can essentially shield itself against arsenic as it is injected through the clays.

The current hypothesis Sarah is working with is that positively-charged ions in the water bind weakly to the permanent negative charge on the clays (see this post of mine). This creates a "bridge" by which the negatively-charged arsenic compounds can bind weakly to the clays (called sorption).

Our project now is to figure out which ions are best at creating these bridges. We currently believe that divalent cations, i.e. alkaline earth metals (below - important because they have a 2+ charge rather than 1+) with high charge density (thus, smaller radii) will be best at helping the arsenic to sorb onto the clays and stay out of the water. When we get our data, hopefully we'll have a clue if we're right!

The periodic table helps us to think that beryllium, magnesium, and calcium will help the most at keeping arsenic out of the water.

Figure 1 & 2 Source: Fakhreddine S., Dittmar J., Phipps D., Dadakis J. and Fendorf, S. 2014. “Influence of Calcium and Magnesium on Arsenic Sorption to Phyllosilicate Clays,” In Proceedings of Goldschmidt Annual Meeting, Sacramento, CA.

SLAC and SSRL visit!

This morning Sarah took Mariela, Mariela's mentor, and me to the SLAC campus to see the SSRL: the Stanford Synchrotron Radiation Lightsource. It was so neat to see! Here's some pictures:

My temporary nametag

This curved structure is the small circle where electrons are first accelerated.

Once the electrons are moving fast enough, they enter this larger curved structure. In here, they release X-rays each time they change direction.

The X-rays are captured in channels tangential to the main structure and eventually arrive in machines such as this one, which use the X-rays for various purposes.

An example of a sample whose surface will be scanned using X-rays to determine the placement of different elements within the samples

ICP-MS

Today we ran our samples on the ICP-MS machine!
We took our samples off of the shakers first thing in the morning.  From there, we had to make our standard solutions and controls, measure our samples into ICP-MS tubes, acidify our samples, and put them in the correct order to be measured by the ICP-MS machine.

The general procedure with standard solutions is to first make a high standard, which contains each of the components to be measured at concentrations higher than the expected maximum concentration in any sample. The concentrations of each component in this high standard must be known very exactly, so prepared and certified standardized solutions are used. The high standard is then diluted a few times so that exact concentrations can be known and the ICP-MS machine can calibrate itself. For this, we needed to run around the labs trying to find the correct stock standards for all of the elements we were measuring (quite a few). We found all of them except barium, so we'll need to find a barium standard and run the samples again on the ICP-OES machine later.

Each ICP-MS tube requires about 6 mL of solution, so we had to label the tubes and measure in solution (the ICP-MS machine needs about 4.5 mL of solution to collect three data points for a particular sample). After filling all the tubes, I added a small amount of concentrated nitric acid to each tube in order to bring the solution to the optimal pH. I capped all the tubes, inverted them a few times to mix them, and then we had to make a spreadsheet indicating the measurement order of the tubes. After each group of samples, the machine needs to run a blank (in this case, just dilute nitric acid), and roughly every 10 samples a QC (quality control) sample is run made from the standard solution. That way, if the measurements are off, the machine can re-calibrate itself frequently. But because of this, it's super important that the standard solution's concentration be known extremely well!

Finally, we put our tubes in order and placed the rack of samples into the machine. The ICP-MS machine has a little robotic suction arm that dips into the samples one by one and sucks up some of the liquid. Everything is automated, and it was fun to watch. We'll get to see our data tomorrow and hopefully it'll be of good quality.

Phyllosilicates galore

Today was a good example of the varied rhythm of research. Whereas yesterday was mostly packed with lab work, today I was based in the office, reading materials on different types of clay molecules in order to better understand the work we're doing in the lab. Our reading materials were from lecture supplements to Scott's soils course (Scott as in Scott Fendorf, the faculty member who runs our project and many others). I learned all about tetrahedral and octahedral oxygen-based clay structures... which sounded scary until Sarah brought out some models!

A simple model of a phyllosilicate comprised mainly of oxygen atoms (red).

A model of the muscovite we work with. One can see clearly on this image that there are two main "layers" connected by a sheet of potassium atoms (yellow, center). This space between the layers allows the phyllosilicate to physically expand in the presence of certain substances such as water or some ions.

The important thing to understand for our work from these structures is that these clays have a permanent negative charge. Many of the non-oxygen ions originally present in the structure are exchanged through long-term processes for ions with a lower charge, giving the once-neutral clays a negative charge. This greatly affects their behavior and has implications for our research on different ions' interactions with clay in water.

We also had the first of our weekly SESUR/SURGE/MUIR lunch seminars (between the three undergraduate research programs in the School of Earth Sciences). Today Assistant Professor Nicole Ardoin came to talk to us about her work studying environmental learning behaviors. It was a good way to take a break from the lab and meet the other students doing similar work on campus this summer.

Expect the unexpected

My second day of SESUR research was a full plunge into lab work. The most important lesson I learned today was to always expect the unexpected during the research process. I went to bed last night excited to measure the difference in arsenic sorption between standard and field-sample clays in the presence of calcium ions... and arrived in the lab this morning to learn that we'd be tweaking the experiment significantly in order to focus on the effects of different types of cations on arsenic sorption on standard clays. If that was confusing, basically where I thought I would be comparing two clay types I'll now only be using one, and where I thought I'd only be dealing with one type of cation I'll now be working on an experiment with three - calcium, barium, and lithium.

It was an immediate lesson in flexibility - something Richard Nevle, the fantastic SESUR program director, had reminded us many times is essential to all types of research. Having my "mini-project" completely change course on day two certainly reinforced that point for me in a more concrete way. In the spirit of adapting, I didn't hesitate, and Sarah and I jumped right into work. We were accompanied by Mariela, a SURGE (Summer Undergraduate Research in Geoscience & Engineering, a similar program to SESUR but open to students from other universities and aimed at juniors and seniors) participant whose primary mentor is also currently traveling, like mine.

Today was a long day so that we could finish preparing our samples and give them time to shake for about 40 hours before our scheduled time on the ICP-MS (Inductively Coupled Plasma Mass Spectroscopy, and yes, I've learned to say the full name correctly!) machine on Thursday afternoon. We had to label 38 tiny bottles (and some larger ones), make 19 carefully-mixed stock solutions with different cation concentrations, pipette them into the little bottles, and (the most mind-numbingly boring part of all) measure 30 mg of clay powder into each bottle. I had no idea how small 30 mg of powder is until I spent an hour getting each sample to exactly 30 mg on the balance, adding or removing a few particles of clay powder at a time. I certainly didn't expect to do that either, but it had to be done. I wonder what unexpected things will happen for the rest of the summer...