Magic Acid Blog

Getting Science into Politics with 314 Action

2018-05-03T00:00:00Z

One thing has become clear to me over the past year: our political system in the US needs more scientifically literate people. Luckily, the 314 Action group is working diligently on this. Last week, they hosted a training event in Chicago to give scientists the tools to successfully run for political office. As it turns out, many great scientists find the process of running a political campaign uncomfortable and foregin. I'll admit, the idea of asking people for money to fund a campagin sounds...unpleasant.

There were two national candidates there, but more exciting were the enthustic candidates for local and state offices: less glamorous but equally important. There was also a strong presence from graduate students and other early-career scientists. Completing a PhD and running for office is too much, but after grad-school...who knows!

I also had a chance to meet two exciting candidates in the greater Chicago area. Laura Ellman is running for state senate in the Naperville area, and Marianne Lalonde for the next Alderman (Alderwoman?) in the 46th ward uptown. Please consider donating $50 (or whatever you can) to their campaigns, so we can bring some much needed science to our local and state government.

Thoughts About Glass Electrolytes

2017-03-19T00:00:00Z

In a recent episode of one of my favorite podcasts, the Skeptic's Guide to the Universe, they discussed a recent paper from the John Goodenough lab. I figured I'd take a look at this paper and see what's inside. They propose a type of battery where there's a lithium-metal reduction on both sides of the cell. Overall, the electrochemical results they show are impressive: very high capacities with good cyclability at a decent cell potential (between 2.5 and 2.7V); though not quite as impressive as the press-release makes them sound. However, I'm not clear on how they think this reaction actually happens. Many of the clarifying details are missing, which makes it hard to properly evaluate. Regardless of the electrochemical mechanism, though, if the results can be reproduced reliably, then great!

There's really two parts to this paper. The first is the use of a solid, glassy electrolyte to overcome the limitations of using metallic lithium anodes. Dendrite formation is a major problem and, as such, modern rechargeable batteries use graphite, resulting in a lower energy density and cell potential. Their strategy also avoids the flammable, toxic liquid electrolyte, which would be nice. If the results are reproducible, this technique could be combined with existing cathodes to give a nice increase in energy density. To be clear, though, the idea of using solid or polymer electrolytes in not new.

The second part seems to be a fundamentally new mechanism for constructing batteries. Based on the schematic they've drawn in figure 4, it looks like they're oxidizing lithium metal at the anode, and then reducing lithium metal at the cathode. The anode part is fine; that's the standard discharge reaction for lithium-metal batteries. However, having the same reduction/oxidation ("redox") half-reaction taking place at both the cathode and the anode would give a cell potential of 0V; that is, the battery wouldn't be able to store any energy. They claim that:

The Fermi level of the lithium plated on the carbon–copper composite cathode current collector is determined by the Fermi level of the cathode current collector, whereas the Fermi level of the lithium anode remains that of metallic lithium, but the cell voltage is determined by the energy of the redox couple of the unreduced redox center.

The implication here is that the potential of the cathode redox couple is raised by the presence of copper. It's not clear to me why the Fermi level¹ of plated lithium would be the level of the current collector. It is true that the chemical potential of metallic lithium would be higher than that of metallic copper, and if they were proposing an alloying mechanism, then the Fermi level would be influenced by the copper. However, they're claiming that the lithium plates on to the copper. Presumably, anything beyond the first several atomic layers of lithium wouldn't really be influenced by the presence of copper underneath. A simple test would have been to replace the copper with aluminum and see what affect this has on the potential, but they didn't perform this experiment.

It's also not clear to me what the role of the sulfur, ferrocene or manganese-oxide redox center is. The idea that these redox centers are not reduced during the reaction suggests that they serve a catalytic role, which the authors confirm in the introduction. These redox centers therefore can't influence the thermodynamic potential of the cell, only the kinetic barriers that must be overcome to initiate redox². This would effectively raise the potential needed to charge the cell, but lower the potential achieved on discharge, and so doesn't explain the asymmetry between the cathode and anode potentials.

Having said all that, the images they show combined with the extremely high full-discharge capacities do seem to confirm that they have moved all the lithium through the glass from one electrode to the other. If this can be reproduced, then I have an opportunity to improve my understanding of some of these foundational electrochemical principles.

There are a few other things I think are missing from the paper. First, it's unusual that no cyclic-voltammetry (CV) was done. CV is an electrochemical technique that sweeps across a range of potentials, and the current response at each point gives information about which redox processes are occurring as well as the inherent kinetic barriers in the system. It's a standard technique for characterizing new materials that operate by known mechanisms, let alone an entirely new mechanism such as this one.

The press release claims

high volumetric energy density and fast rates of charge and discharge

neither of which are reported in the paper. The paper does claim "high volumetric energy density" (Joules per mL) in describing previous work on sodium anodes, but never actual reports the values they achieved, instead opting for gravimetric energy density (Joules per gram). To be fair, volumetric energy density is hard to calculate for a system like this, and gravimetric values are far more common. Regarding their rate capability, they show reversible electrochemistry data at 0.25C and 0.1C rates, which would corresponding to charging your car in 4 and 10 hours respectively³. Not bad by any means, but I would characterize this as "moderate". The authors even describe this towards the end as "The cell voltages and rates are acceptable.", which I agree with.

In summary, the results look promising but I await replication and better description of what's actually happening. For a paper that sounds as revolutionary as this one does, it's noteworthy that it's only 6 pages and has only 5 references, 3 of which are from the authors' own research group. This is not the first paper to report "a safe, low-cost, lithium or sodium rechargeable battery of high energy density and long cycle life." but somehow those claims never seem to be as transformative as promised. From the paper:

All that remains to be optimized is the thickness of the solid glass electrolyte, the loading and choice of the redox center in the composite cathode to provide a required voltage, and optimization of the rate of ion transfer across the cathode/electrolyte interface to obtain a desired rate performance of the stored electric power.

So what, about 5-10 years maybe...?

Footnotes

1) The Fermi Level for a species can be (non-rigorously) described as the work needed to add an electron to a material. Since battery chemistry is all about moving electrons around, Fermi levels are a key concept for understanding the underlying thermodynamics.

2) This is similar to how enzymes can only influence the rate of a bio-chemical reaction, not the underlying thermodynamic states; a second, higher-energy, reaction (eg ATP --> ADP) must be coupled to the reaction of interest in order to reverse the reaction arrow.

3) This description needs a large asterisk next to it. The authors only charged the cell partially in 10 hours but also have a significantly larger capacity than conventional materials, so it's difficult to compare this in a "your car charges in 4 hours" kind of way.

GSAS-II Introduction - Basic Profile Refinement

2015-09-01T00:00:00Z

This document will guide the user through a basic whole-pattern decomposition of a single phase using the Pawley method. This method is simpler than the Rietveld method and will provide a closer fit, however the the refined parameters are not physically meaningful so the amount of information obtained is limited. In this tutorial, I will refine a corundum scan taken on our Bruker D8 Discover ("Alice").

The diffractogram can be downloaded in .xye format. You may also want to look at the finished GSAS-II project.

Installation

Visit the GSAS-II website and follow the installation instructions for your platform (GNU/Linux, Windows, MacOS). Arch Linux users can install gsas2-svn from the arch user repository. Upon running GSAS-II for the first time, you should see a terminal window and several empty windows, similar to the screen shot below.

The terminal window is not normally used, though it can provide useful information if your refinement is not succeeding. The other two windows are the plotting window and the data tree window. They are currently empty since we have not yet imported any data.

Load Powder Data

First we must load a powder data file to refine against. First, export your data from Eva as a .xye file. In the data tree window, select Import → Powder Data → guess format from file. The first dialog to appear will ask you to choose the file containing the powder diffraction data. The second dialog asks for an instrument parameter file. If you do not have an instrument file, click Cancel and you are prompted to choose a generic instrument (eg. CuKα). You should now see your data listed in the data tree window.

Create Phase Object(s)

The next step is to create a phase object that can be refined against the powder data. There are two options for this: importing a .cif file or creating a blank phase. Once the phase has been created you should see an entry for it in the data tree window (you may have to expand the Phases entry.

Importing a Phase from a .cif File

If a .cif file is available for your material, this is the easier option. The materials project may have one you can download for your material. To import a .cif file, in the data tree window select the menu Import → Phase → from CIF file. You will be asked to associate this phase with one of more previously imported data ("histogram"); you can select your powder data, or wait to do so in the next section.

Creating a Blank Phase

In the data tree window, select the menu Data → Add phase, you will be prompted to name the phase.

Prepare for First Refinement

Before running the first refinement cycle, it's necessary to set some initial values. If you imported a .cif file, some of these may already be set for you. This section assumes you have created a blank phase.

In the data-tree window select your phase from the Phases list. A new data display window should appear. Select the Data tab. Enter your space group in appropriate text box. GSAS-II requires spacegroup terms to be separated by a space. The corundum spacegroup (R-3c) is entered as "R -3 c". Press enter and confirm your choice. You will now notice that your choices for unit-cell parameters are restricted to a and c due to the crystal system associated with this spacegroup. Enter approximate unit-cell parameters in the appropriate textboxes. Wikipedia lists a=4.75Å and c=12.982Å for corundum so let's start there. Farther down you will find a section labeled Pawley controls. Check the checkbox labeled Do Pawley Refinement.

Next switch to the Data tab. If you see a message reading ``This phase has no associated data'', you will first need to create a histogram by selecting the menu Edit → Add powder histograms. In the new dialog, select your powder data and click okay (skip this step if you associated your data when importing a .cif file). Leave the values in this tab at their defaults for now; we will come back to them later.

We will now move to the Pawley Reflections tab. Select the Operations → Pawley create menu entry to automatically create a list of reflections and Operations → Pawley estimate to have GSAS set initial reflection intensities. The refine column contains check boxes that indicate which reflection intensities will be refined. Click on the column heading to select them all and type "y" to check them.

Lastly, go back to the data-tree and under your powder data entry, select the Sample parameters entry. In the data-display dialog, uncheck the Histogram scale factor box. If checked, GSAS will refine the intensities of the reflections as a whole, which is unnecessary since we are already refining them individually.

Evaluating Initial Fit

You are now ready to run the initial refinement. In the data-tree window, select the menu Calculate → Refine (you will be prompted to save if you have not done so yet). Assuming the refinement runs smoothly, you should be presented with a dialog asking you to Load new result? (click Yes). Click on the powder data entry in the data-tree and inspect the results in the plotting window. You may have to select the Powder Patterns tab. You should see several overlayed plots showing the original diffractogram, the calculated diffractogram, the background fit and the difference between the observed and calculated diffractograms.

Chances are, your refinement is not very good at this point, but maybe it's good enough. Use the zoom button along the bottom to view several peaks close up. Along the center of the plot you will see small vertical lines representing the predicted reflections. Hopefully, your unit cell parameters are at least close enough that the predicted reflections fall somewhere within most of the peaks. If not, you may have to manually adjust the cell parameters until you're close. Something like the figure at the right should work. Select your phase and go to the General tab, check Refine unit cell. Also, for your powder data, open the Sample Parameters dialog. Change the diffractometer type to Bragg-Brentano and enable refinement of Sample displacement. This will help correct any errors due to the sample not being mounted at the exact instrument center. Now run the refinement again. You may have to re-select your powder data in the data-tree to refresh the plot.

Refinement after first pass. Shown are observed (blue), calculated (green), background (red) and difference (teal).

Refining Background

Zoom out on the plot by clicking the home button and look at how well the refinement is accounting for the background. It may be helpful to zoom vertically to get a better look. In the data-tree select the Background dialog. The default is to refine three coefficients, change this to something higher if you feel the background is not fit well. For corundum I chose 10. Re-run the refinement, if you're happy with how the background looks, you can move on.

Refining Peak Width and Shape

There are two sources of line broadening in diffraction data: instrument effects and sample effects. Instrument effects are not usually of scientific interest and are lumped together and described by parameters in the Instrument Parameters section. If you used an accurate instrument parameters file, these should not need much adjusting.

Of more interest are the sample effects. These can be found in the dialog for each phase, under the Data tab. They do not have directly refineable parameters, instead you must choose a model to describe the peak shapes and widths. Since we are doing profile fitting, the parameters in this tab do not reflect their real world analogs; for that a thorough Rietveld treatment is necessary. Select whichever model you prefer and check the corresponding box. For corundum, I used size. Sample parameters are usually described well by Lorentzian functions, but the Lorentzian/Gaussian ratio can be refined by the corresponding LGMix parameter. 1 is entirely Lorentzian.

Fit after multiple refinements.

From here, continue going until you achieve the desired fit. Each time you run the Refine command, GSAS-II performs three refinement cycles. Just because the process ends does not mean that it has found the best fit. You may need to run the Refine command several times, until the calculated patterns are constant from one refinement to the next. For the corundum example, I ended with a cell parameters of 4.75903Å and 12.99294Å.

Research Update: Mapping Artifacts

2015-07-06T00:00:00Z

I'm almost one year into my time here at UIC. I've been making steady progress on my research and have run into a strange issue. My research focus is developing a new technique to make surface maps, like the one on the right, of battery cathode materials using X-ray diffraction. The material studied in this post is $MgMn_2O_4$. By creating these maps, I am able to look for areas where the electrochemistry is different from the rest of the material. The map at the right looked promising at first; however as described below this excitement may have been premature.

Map of discharged $MgMn_2O_4$ battery cathode based on area of 18° peak.

Mapping

Each spot on the map represents one X-ray diffraction scan. The whole map is of a ½" diameter cathode that was recovered from a magnesium coin cell after being discharged. The diffractograms below show the average over all the scans in a map. The areas highlighted in blue and green represent the 2θ ranges of the expected peaks based on the known phases. The material exists in either a discharged or charged phase, however peaks were seen that don't correspond to either phase. The purpose of this experiment was the try and learn more about their origin. The peak in the 18° range is the one I focused on here. The color of each map pixel is determined by the integrated area under this peak after background subtraction. Looking at the diffractograms from individual pixels we see that indeed the peak is strong in areas colored green and not present in areas colored blue. However, the other peaks associated with this phase do not change proportionally, contrary to the usual behavior of XRD patterns. The fact that the color scale bar on the maps does not go to zero is mostly due to low-quality background fitting in some pixels.

Average of diffractograms from each scan in original map (blue) and map after 180° rotation of sample (green).

Sample Rotation

Puzzled by the strange behavior of the 18° peak, I wanted to see if this was perhaps an artifact of the instrument, rather than the sample. I rotated the sample 180° and made another map. The two are shown side-by-side below. Since the sample was rotated 180°, I expected the resulting map to also be rotated. This was not the case, leading me to suspect that this effect is either due to the instrument itself or to some aspect of the sample mounting method.

Maps and reconstructed micrographs of $MgMn_2O_4$ in original orientation (left) and after 180° rotation on specimen stage (right). Differences in spot size are due to changes made to mapping software between experiments.

The images below show how the sample was mounted. I am working at fairly low angles and the detector is set around 11° (θ-2). Because of this it is possible that the 18° peak is blocked in some areas. It is possible that the tape used to mount the specimen is obstructing the X-ray beam, though this does not appear to be the case in the photos and would require very similar obstruction upon remounting the specimen. It is also possible that the specimen itself is not perfectly flat and that some areas are obstructed by other areas of the sample. It is difficult to imagine, however, how this would result in the effect seen here; the right area would need to be blocked by something that itself doesn't contain the target peak but yet doesn't block the electrode where the 18° peak is visible (green areas). Furthermore this effect would have to be symmetrical in order to produce the same artifact after rotating the specimen.

Mounting of $MgMn_2O_4$ on µ-XRD specimen stage from front (left) and side (right).

Ultimately, I don't know the source of this artifact; I may need to run more experiments to test some ideas. Regardless of the outcome, I intend to come up with a better way of mounting specimens to replace the multiple layers of tape I'm currently using. I will also be on the look-out for this effect in future work.

Beamtime Galore

2015-04-26T00:00:00Z

My PhD research project involves using X-rays to probe the chemistry of lithium-ion battery electrodes. In order to see something with a high level of detail, you need to look at it in bright light. Analogously, I need access to very high quality X-rays. The last two weekends I traveled to two different synchrotron; particle accelerators whose sole purpose is to generate high intensity X-rays. Each one has 10-30 X-ray beams coming off of it, set up to run a variety of experiments. At Argonne National Lab's Advanced Photon Source (APS), we used diffraction to measure the strain withing a micrometer sized particle. At the Stanford Synchrotron Radiation Lightsource (SSRL) we collected the data to make three-dimensional reconstructions of the micro-structure of some samples using X-ray Transmission Tomography.

Fellow grad-student Brian May at beamline 34-ID-E at Argonne National Lab.

The Physics

Viewed from above, the synchrotron is a big ring. In the case of Argonne, it's 1.1km (3,622ft) around. Stanford's is a little smaller. Electrons travel around the ring at over 99.99% of the speed of light. Anytime a charged particle accelerates, it gives off electromagnetic light waves. [NB: "Light" doesn't necessarily mean visible light, it can also refer to microwaves, UV light, X-rays, etc.] From a physicists perspective, moving in a circle is a form of acceleration, so as these electrons travel around the ring they give off electromagnetic radiation. Because of how fast they're going, the light given off is in the X-ray region of the electromagnetic spectrum. The electrons also go through special magnets, either "undulators" and "wigglers", that give off even better light. These X-rays are very bright, quite straight and highly coherent; perfect for my research.

The Advance Photon Source at Argonne National Lab, Chicago IL.

The Chemistry

X-ray micrograph of secondary cathode particles. Each particle is ~2-5µm across.

In my research group, we look at cathode materials for lithium-ion batteries. These materials are usually crystalline, meaning the atoms form a regular, repeating pattern. X-rays are really good at probing crystal structures and so with the right experiment, we can get a lot of information about our sample. The experiment at SSRL made use of an X-ray microscope. The concept is analogous to a conventional microscope except that X-rays have a much smaller wavelength so we can, in theory, resolve much smaller distances. Also, different elements absorb X-rays at different wavelengths, so by tuning the beam we can actually collect images of just certain elements, something conventional microscopy cannot do. We looked at the secondary particles of a battery matherial. Then we rotated the sample and by taking multiple images were able to reconstruct a 3-dimensional model of where specific elements are located.

The Experience

The first reaction to being at a synchrotron was awe. These machines are enormous and incredibly complex. At first, the equipment can be overwhelming. Fairly quickly, though, I came to realize that each piece is straightforward. I certainly don't claim to understand what all (or even most) of the pieces do, but there's nothing mysterious about their operation. There's also a fair amount of downtime, since it takes time to collect all the data and much of the process is automated. This provides a good opportunity to wander around experiment hall and read the various research posters that the different beamlines have created. Each beamline is set up for a different technique, so seeing what each is capable of was a fascinating tour.

Image Formats

2015-01-24T00:00:00Z

This post is meant as a resource for helping to decide how best to send images for posting on websites. There are usually three possible options: lossly bitmaps, lossless bitmaps and vector graphics. The right decision is usually a compromise between a high-quality image that gives a clean, sharp presentation to the user, and a small file-size so a user doesn't get bored while waiting for your website to load all of the 2MB images. Using the right image formats will result in a website that loads quickly and has a consistent, sharp layout.

Image Type	Format	Example Extension	Min Resolution (pixels)
Photograph	Lossy bitmap	.jpg .jpeg	Landscape: 750×464
			Portrait: 350×475
Logos and other flat graphics	Vector graphics (preferred)	.svg .ai .eps	N/A
		.png .tif .gif .bmp	Context dependent

I design a few websites in my spare time, namely the Kalamazoo ACS local section and an ACS regional meeting. This post is aimed at contributors for these organizations.

Raster vs Vector Graphics

Image formats basically store information as a series of 1's and 0's on a computer's hard-drive. Two mains schemes are available. Raster graphics, also known as bitmaps, divide the image into a series of discrete picture elements, or “pixels”, and assign each one a number that represents its color. Deciding how many pixels to use (“resolution”) determines how small each pixel is. Any details that are smaller than the size of one individual pixel are lost. A higher resolution saves more detail but results in larger file-sizes and thus slower page loading times. When submitting images for publication, higher resolutions are generally better; I can always remove detail but I can't add it back in once it's gone.

Vector graphics, on the other hand, describe the image as being made up of various shapes, or “vectors”. A clock-face might be described as a black circle with a specific thickness; another slightly smaller circle filled with white; and various straight black lines of different lengths. The advantage here is that all the details of the image are captured perfectly. As a result, the image can be scaled to any size with no loss of quality. The catch is that the image has to be designed and saved as a vector graphic using something like Inkscape or Adobe Illustrator. A vector graphic can be converted into a raster graphic but a raster graphic cannot be converted back into a vector graphic. For this reason, please send me vector graphics if possible.

Example image showing the difference between vector graphics and raster (bitmap) graphics.

Photographs vs. Graphics (Lossy vs Lossless)

Not all raster graphics are created equal. As discussed earlier, storing a higher level of detail requires higher resolutions and so results in larger file-sizes. Techniques have been developed that compress the image into a smaller file-size; these are known as compression algorithms. They can be devided into two categories; lossy compression makes permanent changes to the image, while lossless algorithms do not. The trade-off is that lossy compression can achieve smaller file-sizes.

The JPEG standard divides the image into 8 pixel square sections and converts each section to a series of frequency components. A full explanation can be found on wikipedia. This process introduces some artifacts into the image and so it is an example of lossy compression. In photographs, these artifacts are generally not noticeable and the benefits in file-size reduction are significant. In flat graphics (logos, etc.), these artifacts become noticeable and file-sizes are usually not much better than lossless alogirithms, like PNG. A lossless image can be converted to JPEG but I cannot go the otherway around without a considerable amount of work. Please submit logos and similar graphics using PNG or another lossless format; photographs can be sent as JPEG files.

JGLCRM logo using lossless compression (top) and lossy JPEG compression (bottom) using a quality factor of 5 for emphasis. Notice the artifacts along the color edges.

A Computer for Chemistry Data Analysis

2015-01-16T00:00:00Z

I've been tasked with finding a new computer for my PhD research here at UIC. I'll be collecting multiple data frames and then processing and combining them into one image; I think it's worth the time to get this right. If you have anything to share on the issue, feel free to leave a comment. I wasn't given many restrictions other than that it must have a monitor, keyboard and mouse (for ergonomic reasons) and that “$1000 is on the high end but reasonable”. I've spent the past few days researching current technologies and have come up with some suggestions. All of the products listed will probably by obsolete by the time I'm done writing this, but hopefully the background will still be valid for a while. If you want to see specifically which parts I'm considering, jump straight to the parts list.

Requirements

The first step in finding the right computer is deciding what this computer actually needs to do. My research centers around mapping the 2D and 3D changes that occur in a battery electrode. I will use an X-ray microdiffractometer to measure the structure on a spot approximately 100µm across. This will be repeated several times in various places. Each location will result in roughly 1 MB of data. The first step will be to integrate each two-dimensional data-set into a one-dimensional spectrum, so I need my computer to be good at mathematical calculations. Since I'll be processing multiple independent spectra, I also want it to be amenable to parallel computing. I would like to leave the option open of using something like PyCUDA if I come across a nice graphics card. UIC does have a super-computer that I can use at times, so if I need more than my computer can provide I have that option.

Processor and Motherboard

Right now I'm looking at the Haswell CPU architecture, which gives me the choice of a Core i3, i5 or i7. The i7 line has hyper-threading, which is the ability to run multiple threads on each core and make good use of processor downtime; seems like something that would help me. The big difference between the different i7's is their choice of memory, DDR3 vs DDR4. This affects the choice of motherboard, and obviously memory. The Haswell-E processors (5960K, 5930K, 5820K) have either six or eight cores versus four cores in the 4790K. Remember, all the i7's have hyper-threading so they can run 12, 16 and 8 threads respectively. The 4790 and 4790K differ mostly in the sense that the 4790K can be overclocked. For an extra $30, I think I will give myself this option.

The motherboards for these processors come in two flavors. One has the LGA 2011-v3 socket, which fits the newer Haswell-E line, and uses DDR4 memory; the other uses the 1150 socket with DDR3 and processors of the regular Haswell type. The former has more cores but the parts (CPU, motherboard and memory) are more expensive. I've decided on the regular Haswell setup. A graphics card is designed to do small repetitive tasks very quickly and can be used for fast data processing. A motherboard with a PCI-Ex16 slot will make that possible in the future. Boards with the Intel Z97 chip-set support overclocking (compared the the H97 chip-set). I'm not sure if I'll use this but why not get it anyway, right?

Memory

So motherboard selection will be determined in part by whether I get DDR3 or DDR4. DDR4 is the new successor to DDR3. There are not as many options available yet and it's more expensive. It's a little bit faster and uses significantly less power. I checked out the performance tests in the video below and it looks like the speed benefits only become significant with six or more cores. This fits well with the processor section since the one that uses DDR3 has four cores and the DDR4-class processors have six. I'm leaning towards the 4-core DDR3 setup as a better value.

Hard Drive

I'm sold on using a solid-state drive (SSD) for holding system files. I love the fast boot time and snappy application loading. An SSD connected by PCI-express would be great, although I think I'll go for SATA III to save a few bucks. With the massive speed gains of SSDs compared to convention hard-drives, the interface itself is now the bottle-neck. A PCI-Express SSD would make my system a little more responsive but right now they cost several hundred dollars. To complement the SSD, I have a 1 TB conventional hard disk drive (HDD) that I will put in for data storage. Hopefully transferring data from the HDD to memory will not be too slow. If it is I may have to transfer the active project from the slow HDD to the fast SSD before performing that analysis.

Operating System

The real question here is whether I need to run anything that absolutely won't run on GNU/Linux. Those of you that know me know that I'm a big Linux fan-boy and really don't like Windows. The first thing that I'll be doing is installing Arch Linux. My plan right now is to use Python to analyze most of these data, which is platform agnostic. Some of the diffraction analysis software that is specific to our instrument will only run on Windows. Hopefully, the Windows emulator written for Linux (WINE), will be able to handle this task. If not, I might have to buy a copy of Windows in the future *shudder*.

Power Supply (PSU)

Every system needs enough power to run all the hardware inside of it. The major consideration here is the maximum power that the PSU can deliver. This number has to be greater than the power required by all the components or things will get weird. I've dealt with power supply issues in the past and they're a nightmare. There's never a light that comes on saying "hey, you're under-powered". Things will just sporadically stop working. The computer might freeze or randomly restart itself. Maybe the whole thing will just be sluggish and I won't know why. According to a handy chart I found on Newegg, I estimate I'll need 350-400 W of power. The advantage I have is that I won't be using a separate graphics card for now. The PSU I selected has a max power rating of 550W so that should give me enough breathing room. Beyond the maximum power consideration, there are some other considerations, like efficiency and noise. I will rely mostly on reviews for that information. I also like the idea of modular cables. This lets me connect only the ones I need and will reduce the clutter in the case.

My selections

Here are the parts I'm considering. I haven't included the monitor since this is just a peripheral that I can swap out at will.

Function	Part	Cost
Processor	Intel Core i7-4790K Haswell	$339.99
Motherboard	Biostar Hi-Fi Z97WE	$114.99
Memory	2x4GB G.Skill Ripjaw Series DDR3	$72.99
Hard-drive (SSD)	Samsung 850 EVO-Series SSD	$90.00
SSD Mount	Icy Dock EZ-Fit Lite	$7.99
Case	Rosewill Challenger Black ATX Mid Tower	$49.99
Power Supply	Cooler Master V550	$89.99
	Total:	$765.94

Entropy of the Universe

2014-12-17T00:00:00Z

I had the following e-mail exchange with one of my Chem 114 students, Kinza. Usually, the questions I get are specific to what we're studying in class and they tend to be easier to answer during office hours. These questions, on the other hand, were Kinza trying to understand the concept of entropy. We had covered it briefly earlier in the semester in the chaprter on thermodynamics. It's definitely a difficult concept, one I don't understand fully myself. It's more in the physics realm than chemistry but I still felt it was worth sharing. Published with student's permission.

Hi Mark,

If every reaction that occurs increases the entropy of the universe, doesn't that mean that all the reactions going on in the universe are causing the entropy of the universe to keep going up? What's going to happen to the universe as predicted by thermochemistry?

-Kinza

Hi, Kinza.

What a fantastic question! You're absolutely right. There's even an ominous name for this concept: the heat death of the universe. Basically the universe will be one uniform temperature and no energy will be left to do useful work.

"This is the way the world ends
Not with a bang but a whimper."
-TS Eliot

It's not entirely clear what effect gravity has on entropy so that does make it a little more complicated.

The book Fabric of the Cosmos by Brian Greene is worth checking out if you're still curious. It's written as a non-fiction pleasure-reading book. It describes some of the weirder parts of physics but without all the math that's normally involved (for people who may not yet have taken calculus 4). Part II deals a lot with entropy.

Cheers, Mark

Hello,

Thank you!! I've been thinking about it a lot; yes I'm still curious! So since the world is reaching an ever-rising level of entropy, doesn't that mean everything will eventually become gas since gas has the highest entropy?

So we don't know whether or not the pull of the universe expanding will eventually become stronger than internal forces like gravity? Also, is the solar system going to drift apart or do we not know what effect entropy has on the solar system yet either?

Sorry, I have so many questions and it's a lot more helpful to ask someone who understands this stuff instead of reading articles written in words I can't pronounce. Studying the stuff we learned in chemistry all semester, looking at it now, I can see that it's brushing the surface of something really huge.

Thank you for telling me about the book! I found the PDF of it online I'm going to see what it's all about when finals are over.

-Kinza

Hi, Kinza.

More great questions. I've put my answers down below. These conversations might make a good blog post. Do you mind if I post them? If not, would you prefer I include your name or leave it off?

Cheers, Mark

That's one possibility. Gas has the highest entropy compared to solids and liquids but it gets tough to compare when you're talking about exotic things, like blackholes. Gravity has some relationship with entropy that I don't completely understand.

The ultimate fate of the universe is still unknown. When Einstein came up with relativity theory, he assumed a static, unchanging universe that would not have a beginning or end. In the 1920s and 30s, Edwin Hubble and Georges Lemaître discovered that the universe is actually expanding. The effects of gravity work to slow down this expansion. If the initial expansion were fast enough, gravity would not be able to keep up and the universe would keep expanding forever (heat death). The other option is that gravity would win and the universe would contract into a singularity (the "big crunch" model). Started in 1998, evidence was found that suggested the acceleration of the universe was actually expanding. This requirs some sort of energy that is always repulsive (the opposite of gravity) and could push matter faster and faster. We currently don't what that stuff is; we call it "dark energy", which is really just a placeholder until we understand it better.

No problem at all. Questions from someone who's genuinely curious are always fun to answer; a welcome change from the "how do I do this homework" type questions I usually get. =P

Enjoy. It's one of my favorites. I first read it while I was on vacation with a friend. I would periodically look up and say things like "Wow! Did you know that from a physics perspective it's statistically more likely that all your memories are made up than that they actually happened" and they would all look at me like I was insane.

More Sunshine (Vitamin): Doctor's Orders

2014-07-02T00:00:00Z

As per my doctor's suggestion, I'm spending the next week on a sailboat. No, it's not a blood pressure or a stress thing. I had a physical exam a while ago and for the most part everything was normal. The one exception was vitamin D deficiency. The solution? Spend more time in the sun.

Vitamin D is actually a group of five vitamins: D₁ through D₅. The most relevant to human health are vitamins D₂ and D₃ and are usually what people mean when they refer to “vitamin D”. My results listed “Vitamin D, 25-hydroxy” as 18 $\frac{ng}{mL}$ and the target range was 30-96 $\frac{ng}{mL}$. An article in The American Journal of Clinical Nutrition suggests that levels below 20 $\frac{ng}{mL}$ indicates a deficiency. Clearly I need to get those levels up. But why?

The sun at 17.1 nm light. Courtesy of SOHO/EIT consortium. SOHO is a project of international cooperation between ESA and NASA.

The Importance of Vitamin D₃

Vitamin D₃ has no significant activity on its own^(a). It's just the starting material for a system that regulates the amount of calcium your body absorbs. The first step prepares the vitamin for use in regulation. Chemists often number the carbons in medium-sized organic molecules for easy reference. The one we're interested in here is carbon 25. Initially this carbon is bonded to three other carbons and a hydrogen atom. In the liver, the hydrogen atom is replaced with an oxygen and a hydrogen, a “hydroxy group”, to form 25-hydroxy D₃. This process happens readily and this molecule is the dominant form of vitamin D₃. This is why my lab work lists the concentration of vitamin D in this form.

The second step turns it into the active form. In your kidneys there is an enzyme that adds another hydroxy group, this time to the other end of the molecule at the 1 position, to form dihydroxy D₃^(b). Once it's in the dihydroxy form it can bind to and activate a protein in the nuclei and membranes of cells, called the vitamin D receptor. Each hydroxy group has a polarity that matches up with an amino acid in the vitamin D receptor. The common 25-hydroxy form is missing one of the hydroxy groups so can't fit nearly as well. In intestinal cells, activating the vitamin D receptor causes other proteins to transport calcium ions across the intestine to the blood. In bone-synthesizing cells (osteoblasts), it activates molecules that absorb calcium from the blood and build it on to the bone.

Dihydroxy D₃ (gray) inside the vitamin D receptor. The dashed yellow lines show non-covalent bonds between hydroxy groups (red) and the receptor protein.

That second hydroxylation step is closely regulated by your body. As the level of the active dihydroxy D₃ in your blood rises, the activity of the kidney enzyme goes down. High calcium levels similarly trigger low enzyme activity. There's also a hormone that causes more enzyme to be produced. In order for these regulatory steps to be effective, however, there needs to be enough 25-hydroxy D₃ as starting material. Vitamin D deficiency results in bone pain and muscle weakness in addition to being linked to cardiovascular disease, cognitive impairment, childhood asthma and cancer. The childhood illness Rickets can also be caused by insufficient vitamin D and results in soft and malformed bones due to impaired calcium absorption. One way to increase vitamin D levels is to have more in your diet, for example from Vitamin D milk or oral supplements. The other way is to get more sun.

Making Vitamin D₃ From Sunlight

The reaction mechanism for making vitamin D₃ (right) from 7-dehydrocholesterol (left). Ultraviolet light breaks one bond to form the intermediate (middle), which then rearranges to form the final product.

Vitamin D₃ is produced in the skin when exposed to ultraviolet light, like that from the sun. 7-dehydrocholesterol is present throughout the body since it's a precursor to cholesterol, which is a part of animal cell membranes. The structures of 7-dehydrocholesterol and vitamin D₃ are very similar; the former just has one extra ring. When UV light of the right wavelength (270 – 320 nm, UVB) hits that ring, it breaks one of the bonds. The resulting molecule then rotates around the remaining bonds to form the more stable conformation that matches vitamin D₃. The first step takes place immediately, while the second one happens over several days.

So how much sunlight does it take? The recommended daily allowance of vitamin D taken orally is 600 international units (IUs), about 15 μg. The amount of sun needed to reach this level is influenced by a lot of factors: latitude, season, skin color, clothing, sunscreen, etc. Generally, though, it's a matter of minutes. One article estimated that “In Boston, MA, from April to October at 12 PM EST an individual with type III skin, with 25.5% of the body surface area exposed, would need to spend 3 to 8 minutes in the sun to synthesize 400 IU of vitamin D.” This is not a question of sunbathing to get vitamin D, but more a question of trying to spend a few minutes outside every day. Also, the sun's impact on your skin can be negative, such as sunburn and increased risk of skin cancer.

As with most fat-soluble vitamins, it is possible to overdose. Too much sun doesn't cause problems where vitamin D is concerned, though. Once your skin reaches a certain amount of that intermediate pre-vitamin, the product starts to break down into other, inactive molecules. Cases of vitamin D toxicity are almost always associated with people taking too many oral supplements.

What About the Other Vitamin D's?

Up until now, I've focused on vitamin D₃. This is the one most common in vertebrates and the one we can synthesize using sunlight. The other varieties are very similar in structure, with a few double bonds or methyl groups added on here and there. They are metabolized by similar pathways to those for vitamin D₃. Due to the slightly different shape, however, they may not have the same affinity for the vitamin D receptor. Vitamin D₂ is the only other common form and is made by a variety of fungi. Research suggests it has about 25% of the efficacy of vitamin D₃.

Footnotes

There's some confusion about what “vitamin D₃” actually means. In this post, I will use vitamin D₃ to refer to the inactive, dietary form: colecalciferol. “hydroxy D₃” will refer to the metabolized 25-hydroxy colecalciferol and “dihydroxy D₃” will refer to the biologically active 1α,25-dihydroxy colecalciferol.
There are actually two dihydroxy forms. The active form that binds to the vitamin D receptor is 1α,25-dihydroxy vitamin D₃. There is also a second form, 24R,25-dihydroxy vitamin D₃, which has almost no binding activity with the vitamin D receptor.

Follow-up: I Have a Carbon Monoxide Detector. Now What?

2014-06-01T00:00:00Z

This is a follow-up post to “Why Do I Need a Carbon Monoxide Detector?”, which talks about the chemistry of carbon monoxide with oxygen-transporting proteins in your blood. I received a few questions about the relationship between carbon monoxide levels in the air and the levels in your hemoglobin. I also wanted to learn how carbon monoxide alarms respond to carbon monoxide (CO) but not other gases in the air, like carbon dioxide (CO₂).

CO Levels in Air and Blood

When I was younger, I was terrified that carbon monoxide binds permanently to hemoglobin and that the damage might never be undone. Real life is a lot more complicated than that. Small amounts of carbon monoxide are produced routinely in your body and may serve a function as a neurotransmitter. But at higher levels, it becomes very toxic. In 1965, a paper was published ("Coburn-Forster-Kane") that provided an equation for estimating levels of hemoglobin bound up with carbon monoxide ("COHb"), based on a variety of factors. In case you're curious, the equation is:

\[ \frac{A[HbCO]_t - (BV_{CO} + PI_{CO})}{A[HbCO]_0 - (BV_{CO} + PI_{CO})} = e^{-tAV_bB} \]

It's not terribly important that you understand the whole equation so I'll point out the the interesting parts. We're trying to solve for $[HbCO]_t$, the concentration of carboxyhemoglobin in the blood at any given time. First there's the presence of a term $t$, which means the duration of exposure to CO, measured in minutes. This means that, in some way, the amount of time you spend exposed to higher levels of carbon monoxide is important. Then there's $PI_{CO}$, which represents the concentration of carbon monoxide in the air you're breathing. Lastly there's $B$, which tracks how much air enters your lungs every minute. So the effect depends on how fast you're breathing. Basically, exercise and other vigorous activites make the whole process run faster. The rest of the equation takes care of things like total volume of blood and how much carbon monoxide is naturally present in the body.

Some smart people who are good at math have made graphs that show how this equation works. The one on the right has been shamelessly borrowed from CO Headquarters. It shows the per cent of hemoglobin that's in the carbon monoxide form as a function of time. Each curve is a different level of CO exposure in the surrounding air. The World Health Organization has published some information that is helpful here. Levels of COHb below 5% have no noticeable impact unless you are a member of an at-risk population (heart-disease, etc). Severe symptoms occur between 30-50% and bad things like coma, convulsions and death occur at levels above 50%. So let's say you're pregnant and are concerned about hitting the 15% COHb threshold at which the WHO reports developmental problems. This means that if you're exposed to 200ppm carbon monoxide in the air, you have about 300 minutes (5 hours) before the level of COHb hits 15%. At 500ppm this drops to 100 minutes (1 hour 40 minutes) and 1000 ppm to 30 minutes. Of course if any number of variables change (breathing rate, etc), so do these numbers. Also, if you get into fresh air, the concentration returns to normal levels over a similar time frame.

Solutions to the Coburn-Forster-Kane equation for various air carbon monoxide levels. Each line is a different level of carbon monoxide in the surrounding air.

How to Detect Carbon Monoxide

So how does any of this impact carbon monoxide detectors? A CO level of 35 ppm can be problematic but only for extended periods of time. The need to avoid false alarms is important, especially in places like hospitals where evacuation has its own problems. Cigarette smoke produces lots of carbon monoxide but you wouldn't want your alarm going off every time you light one. Alarms typically have a threshold that's a combination of CO concentration and exposure time (eg "150 ppm, 10-50 minutes"). There are several ways to achieve this but here's my favorite: the biomimetic sensor. It uses a disc of gel that absorbs carbon monoxide over time. The composition of the gel is designed in such a way that it responds to carbon monoxide in a similar way to hemoglobin, hence the term "biomimetic". It's made of an atom of metal (to simulate the iron in hemoglobin) bound to a large organic molecule. Part of the molecule holds the metal and another part, the "chromophore", changes color depending on whether the metal is bound to carbon monoxide. The color of the sensor then follows the curves in the graph above. By putting a light source on one side and a detector on the other, the darkness can be measured and the alarm triggered once it hits a certain level. This type of alarm is the most reliable but also expensive so most consumer products use other methods.

Why Do I Need a Carbon Monoxide Detector?

2014-05-25T00:00:00Z

What if I told you carbon monoxide detectors are a scam? Well I'd be lying. Carbon monoxide is a poisonous gas that gets into your red blood cells and messes everything up; knowing when it's in your house is worth the investment. It doesn't help that it's colorless and odorless either. We had some furnace troubles last winter and after a night of not being sure if we'd die quietly in our sleep, I decided to buy a CO detector ($20 at my local hardware store). In order to understand why carbon monoxide is so deadly, we first need to learn a little about how the body transports oxygen.

It starts with red blood cells. These are special cells that are really rich in a molecule called hemoglobin. It's a big curled up protein with an iron atom in just the right place. This iron atom wants electrons and holds on to the extra ones that are in a molecule of oxygen. The rest of the protein makes sure there's a space that's just big enough to fit an oxygen molecule (more on this later). Each molecule of hemoglobin has four of these iron centers. When the red blood cells are in a place with lots of oxygen, like your lungs, these four sites fill up quickly. When the cells are in an area with low oxygen, like your muscles during a workout, then hemoglobin rearranges and lets go of the oxygen so your muscle can use it to turn carbs into energy.

An iron atom can bind an oxygen molecule while the rest of the protein makes a nice pocket that keeps other things out of the way.

So what about carbon monoxide? Well it's a similar shape and size to an oxygen molecule. They have some other properties in common too, like sticking to that iron atom. The problem is that carbon monoxide binds to the iron way better than oxygen does. So much so that no matter how much rearrangement it does in your muscles, hemoglobin can't shake any CO molecules that are attached. It gets worse, actually. Once a carbon monoxide is stuck to one of the four oxygen-binding sites, it gums up the other three so they don't work properly either.

Humans (and other animals) have evolved some defenses against this problem, namely in the structure of hemoglobin itself. Remember how I said that the protein has an oxygen-shaped pocket next to the iron? One way in which oxygen and CO are different is in how they stick to the iron atom; carbon monoxide sticks out straight and oxygen at an angle. There's a pentagon-shaped part right next to where the oxygen sits. It doesn't really bother the bent oxygen much but it gets in the way of the straight carbon monoxide. This means that the oxygen can fit into that space a lot better than cabon monoxide can. Carbon monoxide still binds to hemoglobin about 250 times better than oxygen does, but it gives your body the ability to tolerate a low level of the poisonous gas.

Oxygen (left image, light-red) binds in a slightly bent configuration which doesn't get in the way of the distal hystadine (blue-green pentagon). The straight carbon monoxide molecule (right image, light-red) doesn't fit nearly as well.

Levels of naturally occuring carbon monoxide are rarely high enough to overwhelm this defense. However, incomplete burning of fossil-fuels like the natural gas in our furnace, creates artificially high levels in the immediate area and there hasn't been nearly enough time for our species to evolve any sort of response. The symptoms of cabon monoxide poisoning are pretty general (headache, nausea, etc). If you have a CO detector, on the other hand, you can do something about it. Get yourself and loved ones to some fresh air. In regular air it takes about 2-6 hours for half of the carbon monoxide to be replaced with good old oxygen. If you're given pure oxygen at 3 atmospheres, this time goes down to tens of minutes. And if you don't have a CO detector, go get one right now.

This article prompted me to write a follow-up piece: "I Have a Carbon Monoxide Detector. Now What?".

Hello, world!

2014-05-23T00:00:00Z

So I've decided to start a blog. I think I'll make it about chemistry. You know, explaining everyday phenomena in a way that's both informative and easy to understand. Who knows if I'll succeed. I'm also planning to throw some stuff in about going to grad school and computer programming.

I'll probably need a fair amount of content to post so if you have chemistry related questions, I'd love to hear about them. I've put up a thread here. Leave a comment with your question and I'll see what I can dig up. I'm thinking of starting out with a post on how oxygen gets transported from your lungs to your muscles.

Ask a Chemist

2014-05-23T00:00:00Z

Update (2018-08-27)

I've stopped answering questions posted to the "Ask a Chemist" page. I was getting asked about specifics that were outside of my area of expertise and didn't make for good blogging fodder. If you have a chemistry question, I recommend posting it on the chemistry stack exchange.

Ever wonder why water and oil don't mix? Well maybe not, but if you have anything similar that you've wanted an answer to then you've come to the right place. Leave a comment with your question and maybe I'll write a post about it.

Examples of good questions:

How does C-4 work?

Why do cars rust more in the Northern US?

Why is water wet? (seriously)

Examples of bad questions:

I couldn't think of any. I'll post some here if they get asked.

Magic Acid Blog

Getting Science into Politics with 314 Action

Thoughts About Glass Electrolytes

Footnotes

GSAS-II Introduction - Basic Profile Refinement

Installation

Load Powder Data

Create Phase Object(s)

Importing a Phase from a .cif File

Creating a Blank Phase

Prepare for First Refinement

Evaluating Initial Fit

Refining Background

Refining Peak Width and Shape

Research Update: Mapping Artifacts

Mapping

Sample Rotation

Beamtime Galore

The Physics

The Chemistry

The Experience

Image Formats

Raster vs Vector Graphics

Photographs vs. Graphics (Lossy vs Lossless)

A Computer for Chemistry Data Analysis

Requirements

Processor and Motherboard

Memory

Hard Drive

Operating System

Power Supply (PSU)

My selections

Entropy of the Universe

More Sunshine (Vitamin): Doctor's Orders

The Importance of Vitamin D3

Making Vitamin D3 From Sunlight

What About the Other Vitamin D's?

Footnotes

Follow-up: I Have a Carbon Monoxide Detector. Now What?

CO Levels in Air and Blood

How to Detect Carbon Monoxide

Why Do I Need a Carbon Monoxide Detector?

Hello, world!

Ask a Chemist

The Importance of Vitamin D₃

Making Vitamin D₃ From Sunlight