Tuesday, June 30, 2015

Error Bars

The ICP data from the sodium/potassium trial had extensive error bars due to the large range of sodium and potassium concentrations tested. It was impossible to measure a difference of a few ppm over top of a background concentration of 300 ppm. To resolve this I dialyzed some of the nanoparticles from the previous dialysis with water to clear out the background salt and leave only the gold nanoparticles with anything bound to them. The ICP data for those samples had much smaller error and very interesting results. It showed that there was 1 Na for every 4 sulfurs and 1 K for every 7 sulfurs. Going forward I am going to repeat this process with the water dialysis a few more times get more data and confirm that the data is repeatable.
Before Water Dialysis

After Water Dialysis

Wednesday, June 24, 2015

Sodium v Potassium: Round 1

On Monday, I started making a brand new batch of PSS coated NPs. I still have a good portion of the last batch but I wanted to make sure that the data we were seeing was reproducible. On Tuesday, I characterized these particles and ran equilibrium dialysis on them with NaCl and KCl in parallel. This way, I hope to compare the absorption of Na and K ions. Today, I characterized the resultant nanoparticle samples and ran all the samples in the ICP 4 times with 2 different calibrations. If the data is reproducible I hope to see 4 Na per S like last week's run. The charge on all of the dialyzed nanoparticles since I modified the protocol to dilute with the dialysis buffer has been negative, which matches our original expectations.

June 24th - Data Analysis

For the past week or so I have been working on analyzing the data I gathered from ICP-AES. So far I have found two interesting trends: as the PEG percentage increases (so, DNA-DNA spacing decreases), the charge neutralized by hexamminecobalt(III) increases and the charge neutralized by magnesium decreases, and as the magnesium concentration increases, the charge neutralized by magnesium increases and the charge neutralized by hexamminecobalt(III) decreases. These two trends are not surprising, and correlate well with the expectations of the DNA system. What is even more interesting though, is determining the concentration of magnesium needed at each PEG percentage so that hexamminecobalt(III) and magnesium equally charge neutralize the DNA (shown in the graph below).

Friday, June 19, 2015


Yesterday, I ran another equilibrium dialysis on the nanoparticles using NaCl. I modified my protocol to dilute the final nanoparticle product with NaCl instead of pure water. Today, I characterized these nanoparticles. They were negatively charged this time which is what we would have expected but we still don't know why the other two samples are positively charged. This afternoon I am running 4-5 ICP runs on the samples, which I have diluted. This will hopefully give me good statistical data to analyze.
Just watching the autosampler do all my work for me...

Tuesday, June 16, 2015

Too much sodium!

Yesterday, I ran another equilibrium dialysis with the same parameters but I doubled the number of spins to 20.

Today, I characterized the resultant nanoparticles, prepared new calibration samples, and ran all of the dialysis samples in the ICP. I diluted the nanoparticle sample to attempt to lower the concentration range the machine was reading. There were a few interesting results. Firstly, both this sample and the last sample are reading positive charges. The polymer coat is negatively charged so the only possible explanation is that the excess sodium (3:1 Na:S in the first sample and 6:1 Na:S in the second) is causing the surface to become positive. The very presence of an excess of sodium is counterintuitive, one would assume that the polymer would only absorb enough sodium to neutralize its charge. Going forward I will be doing more ICP and dialysis to get more data. 

Friday, June 12, 2015

The Files are Breeding Like Tribbles!

On Thursday, I ran ICP on the uncoated nanoparticles and the coated pre-dialysis nanoparticles. Since  then, I have spent most of time trying to analyze the ICP data and reading literature on polyelectrolyte deposition. As a result, I have 7 excel documents open on my computer and even more pdfs!

Next week, I plan to run dialysis on the same nanoparticle sample with KCl as the salt buffer. I will then run ICP on those samples so that I can compare them to the nanoparticles that underwent dialysis with NaCl.

Thursday, June 11, 2015

June 10th - Compiling Data

These past few days I have been working on compiling the ICP-AES data from multiple runs. Below is one set of the graphs that compares the results from each run. Because I am investigating the competitive binding between hexamminecobalt(III) (which has a charge of +3) and magnesium (which has a charge of +2) to DNA at varying levels of osmotic stress, the graphs below demonstrate how much of the DNA charge is neutralized by each respective species at differing PEG values. It is interesting to note all runs seem to have the same trend, and hopefully with even more data the numerical values for the trend will become more solidified. Tomorrow, I plan to run another set of samples through the ICP-AES and compare its results as well.

Wednesday, June 10, 2015


Yesterday, I finished making and characterizing a large batch of PSS coated NPs. I then tried to run equilibrium dialysis. I modified my parameters, decreasing both time and rotation speed, to hopefully correct the loss of concentration. I successfully recovered about 1.7 nM which is a big improvement over last week.
Today, I characterized the new nanoparticle sample, prepared new calibration standards, and ran ICP on the equilibrium dialysis samples.
Tomorrow, I plan to do analysis on the ICP data I have acquired and possibly run ICP on the non-dialysized PSS nanoparticles for comparison.

Monday, June 8, 2015


On Friday, I ran equilibrium dialysis on my PSS coated nanoparticles using a centrifugation system. Unfortunately, a significant amount of the nanoparticles stuck to the mesh in the filter causing a significant reduction in concentration. This was a problem since we needed a relatively high concentration in order for the ICP to be able to detect sulfur, an element found in the PSS coat.

Today, we came up with two solutions. First, we purchased a different type of dialysis device that floats in water and allows equilibrium to be reached over time.

The new dialysis device will take over a week to arrive so, in the meantime, I concentrated the nanoparticle solution from Friday's dialysis and ran it in the ICP. In the concentrating process, we lost a good deal of volume so we had to use a couple tricks to get the machine to use up less sample. From my calculations, we were not very optimistic that the machine would be able to detect the sulfur. However, much too our surprise, the sample had more sulfur than even our highest estimate. Tomorrow we will look at the official ICP report and it should be able to tell us something about the effectiveness of the dialysis and the ion atmosphere around the nanoparticles. Also, I will try to work out some better parameters for our existing dialysis system.

Friday, June 5, 2015

June 5th - Lab Work

After successful preparation of calibration standards yesterday, I made 86 samples (below) to run through the ICP-AES next week. Next week I will also be preparing more calibration standards and samples.

Wednesday, June 3, 2015

Polystyrene Sulfonate

Today, we ran our calibration samples in the ICP-AES to get an idea of how precisely we made our samples and to learn how to use the machine. After we finished running the samples and analyzing the results, I visited Professor Thompson's lab in the Chemistry department where he outlined the process we will follow for the next couple weeks. Today, I diluted a solution of spherical gold nanoparticles he prepared, and treated them with polystyrene sulfonate and sodium chloride. This mixture needs to sit over night so the polystyrene sulfonate has time to absorb onto the nanoparticles surface. Tomorrow, I will centrifuge down the treated nanoparticles and begin characterizing both the treated and untreated samples.

June 3rd - Summer Research Begins!

Looking forward to an interesting and exciting summer! These past few days I have been organizing and gathering background information by reading scientific papers so I can begin research soon. Today we ran standard samples through the ICP-AES to learn how the machine works as well as determine the precision of our standards. Hopefully later this week I will be able to run some samples through the ICP-AES and begin to look at ion binding competition in DNA-arrays!

Monday, June 1, 2015

The epic quest begins...

Today was the first day of summer research! Ambitious plans were laid, literature was read, and calculations were made. My project will focus on the interactions between gold nanoparticles and DNA. My goal for the next couple weeks will be to synthesize polyelectrolyte coated gold nanospheres and characterize their features. Then I will run equilibrium dialysis and atomic emission spectroscopy (ICP-AES) on these nanospheres with different concentrations of ions. I'm looking forward to an exciting summer!

Wednesday, July 23, 2014


I averaged all of the ICP runs from the past two weeks and compiled them onto one Excel spreadsheet. I then began looking at the Na/Co ion competition more thoroughly using the article Prof. Andresen and John Giannini collaborated on a few summers ago, "Ion Competition in Condensed DNA Arrays in the Attractive Regime" as a guide. I tried to use the ion binding model used in the article to analyze the relationship between the number of ions near the DNA arrays and the solution's ion concentrations. This involved finding a constant, ξ, to relate the two sets of numbers (aka. the squiggle). After rearranging the formula given for a simplified ion binding model, I found a ξ for each data point. I took a rough average and an actual average and compared the values found using the ξ against the data from the ICP (shown below).
As you can see, the model only fits the middle three points. This could be due to my own error in calculating each point, the assumptions I made when using the model, the interactions due to Cl- that I neglected, or the model might not accurately represent this data. 

Tuesday, July 22, 2014

The Long Post

I haven't been blogging lately because I broke the USB wifi adapter on the computer I've been working on. I accidentally stepped on it. So, here's a run down of everything I've been doing lately, with pictures. Also, the magnetic tweezers are almost finished, minus a working flow cell.

As of today, this is what the Tweezers look like:

As you can see, the lab is not as much of a mess as it has been in the past. Besides that, let's take a tour of each part of the MT, now that it is completed (more or less).

In general, MT is actually an inverted microscope with magnets jammed into it. It is inverted because in a normal microscope, the light source comes from the bottom. However, in MT, the light source comes from the top down. 

This is the light source. It is a red Thorlabs LED with a wavelength of 625nm. The LED has been collimated by an aspheric condenser lens, f=20mm. In MT, it is important to have a collimated light source with a low coherence length, as this allows for a better generation of diffraction rings around the beads, which is essential in taking measurements. The long, ventilated part on top of the LED is the heat sink. The LED is being held by a kinematic mount attached to our rail system. This allows for greater control over the positioning of the LED and direction of the light source.

Directly under the light source are two motors and the magnet setup. Here you see the rotary motor, which has been coupled to a sled moved by a stepper motor, which in turn has been mounted on the rails. The rotary motor holds the magnet holder, and allows for rotation of the magnets. By rotating the magnets, we can apply a torque to the magnetic beads, and thus, the DNA (or whatever is attached to the bead). The stepper motor that controls the sled allows for moving the rotary motor up and down in space, and therefore controls the strength of the force being applied to the beads (magnet closer to the bead, stronger force).

This is the magnet holder, and attached to the magnet holder are the magnets. They are two Neodymium magnets.

Here we can see the full-middle setup of the MT. Underneath the rotary motor is the XY table. We will couple our flow cell (still in production) to the top of this. The XY table allows manual movement of the flow cell in X and Y directions (i.e. horizontal to this picture.

This is the view underneath the XY stage. Not shown here is the microscope objective, which is a Nikon 100X oil-immersion objective. The objective attaches to the objective holder, which has been coupled to another sled controlled by a stepper motor. This stepper motor allows for control of the focus of the magnified image of the sample cell. By moving the objective closer or further away from the sample, we can change the focus of the image.

Light exiting the objective is reflected by this 45 degree mirror. I have momentarily removed it from the MT setup in order to make access to the underside of the MT easier. The mirror reflects the light to the camera setup.

And this is the camera setup. It consists of a lens tube, a 100mm aspheric focusing lens, and a JAI-Pulnix CCD camera. The light collected in the tube is focused onto the CCD, which then transmits black and white images of the sample to the computer.

This is the entirety of the MT setup. So, how does it work, and what is it's purpose? You'll just have to wait for another blog post...


Monday, July 21, 2014

Last Week...

Today, I tried to normalize the cobalt-only data several ways. I'm still not sure if I did it right or if WinLab did it correctly on the computer. Both ways I normalized it lead to larger error bars (not by much but still). Overall they all look pretty much the same. Because of this, I made a PowerPoint of my overall results for this set of data off of the original data instead of the reprocessed. A few days ago I thought my error bars were great overall, then today I realized I forgot to multiply one set by three so now they are not that good. Hopefully they are good enough when I add a few more runs of samples onto it in the next three days that have the extra buffer.
All three of the ways I tried to look at my data. All pretty much the same except for slightly different error bars.