Jimmy Stewart: A Chemist's Everyman

A few weeks ago I attended one of the most inspirational and entertaining talks I've ever witnessed, given by Daniel G. Nocera (from M.I.T) on his wireless "artificial leaf" technology, published recently in Science. Apart from presenting elegant and relevant research (self-healing catalysts, anyone?), Nocera proved to be a very charismatic fellow, and provided many insights into the strategies he used for developing this technology (analysis of technology cost per unit weight and energy storage methods proved to be more interesting than I would have guessed).

The real "moment" of the talk, though, was when he presented a brief excerpt from the 1938 film "You Can't Take It With You", starring Jimmy Stewart:

This was really one of those rare moments when you realize how monumental a particular problem is. That someone in a film 73 years ago explained--in a mildly condescending way--the difficulty of accomplishing artificial photosynthesis and the exciting possibility of using it for large-scale energy applications is truly amazing.  That it has taken us this long to fully understand how photosynthesis works and how to reproduce it is humbling, and yet still impressive.  This really is one of the greatest problems we face, both from a sustainable energy standpoint, and from a pure research standpoint. The fact that it is now so close after so many years (Nocera has partnered with the Tata Corporation in India to make the push towards production and wider applications) is extremely exciting.

The other, perhaps more humorous and salient, point of the video is the very last line. What did Jimmy Stewart's character end up doing after years of relentlessly pursuing such a noble and complex goal? Throwing in the towel, and heading off to work in "this strange thing called banking".

Times sure have changed, haven't they?

Subscribe to the comments for this post

Sigma-Aldrich: "All-U-Can-Eat" styrofoam with every purchase!*

*Substitute for weird grey flaky stuff at no extra charge!

Hi, everyone. Apologies for my absurdly long absence from the blog--I've been extremely busy hammering out an enormous project and writing grant applications to the Canadian government for the past few months.

Today's post is going to be a short one, but it's a problem that's been vexing me for the entire time I've been at grad school. As the person at my lab who is charge of ordering reagents, and partially responsible for the inventory and cataloging of them, I get to see first-hand how much packaging goes into a Sigma order. I think I can speak for everyone when I say that the policies they use for determining "adequate packaging" are absurd to the point of being humourless (figures 1 & 2).

Figures 1 & 2. This is a 25g bottle of 1-decyne that we received today from Sigma, shown for scale with the size of the box and the amount of styrofoam used to ship it. This is far from the worst case of overpackaging I have seen.

Now, I understand several things:

1. These are indeed hazardous chemicals (sometimes)
2. A breakage or leak in transit would be a "non-trivial" (cf. dangerous and embarrassing) problem
3. The company is responsible for making sure that I receive the items intact and in perfect condition
4. There are numerous regulations regarding the transport and handling of these materials, both domestically and across international borders

HOWEVER. When receiving items from Strem or Alfa, the packaging isn't nearly as excessive (that is, they usually ship multiple items individually wrapped in a single box, unlike Sigma, who generally place a single item in a single box for an entire order, unless there are a series of similar items, such as various 500 mg bottles of pybox ligands). Both Alfa and Strem ship from the U.S., so I know that it isn't simply a border issue, and indeed, when things get shipped to us by Sigma from Oakville, Ontario, the reagents are still swaddled in enough layers of plastic pillows, styrofoam, grey fabric, or that weird flaky stuff to survive getting thrown out of an airplane. I've received orders where individual 1g bottles of reagents are packed in their own boxes, resulting in an entire garbage can full of packing junk after ordering only 5-10 g total of actual materials.

It may be a funny coincidence that I work in a "green chemistry" oriented laboratory, but all comedic weepiness aside, this packaging offends me every time I make an order. The amount of styrofoam alone that goes into the garbage here every week could probably fill a hot tub, and that stuff never goes away. I've found an aftermarket use for the grey fabric, and the boxes can be recycled easily, but the rest of it just goes straight to a dump.

So, I pose these questions to fellow chemists and other scientists in the States:

Does Sigma-Aldrich use the same asinine packaging policies when shipping domestically?

Does it enrage you, having to swim through a sea of styrofoam just to find your starting material?

Do you fantasize about sending back the box-o-styrofoam with a note encouraging them to re-use it?

Do you fantasize about collecting a year's worth of Sigma-foam and filling your swimming pool with it?

What is that weird greyish flaky stuff?

Nick

Subscribe to the comments for this post

Open-source chemistry: still on the frontier

For most organic chemists (or synthetic chemists in general), if you were to ask them which program they use most on a day-to-day basis, there would be one overwhelmingly popular answer: ChemDraw.  The ChemBioOffice suite from CambridgeSoft is like the Excel of the chemistry world: a program so good, and so widely used that thinking of a replacement is often considered folly.  While the program is supported on both Mac and Windows, Linux support remains elusive, and likely will for a long time.

About two computers ago (how sad that this can actually be understood as a measurement of time), I had my first encounter with Windows Vista.  After about a week, I decided to format the computer and give Ubuntu Linux a go.  The version out then was Feisty Fawn (7.04), and while I had to tweak a lot of things to get the OS working the way I wanted, it was actually a very positive experience.  I was finishing up my second year of undergrad, and hadn't had too much experience with chemical drawing, so I figured that the plethora of open-source chemical drawing software would be sufficient.  In the long run, it turned out I was wrong.

Simply doing a search in the Ubuntu Software Installer for chemical drawing software turns up quite a few results, often with confusingly similar names (Xdrawchem, GChemPaint, JChemPaint, Chemtool, ChemSketch, Marvinsketch, BKChem, to name a few).  There has also been a pretty long-standing effort to get ChemDraw to work under Wine, a windows API emulator for use in Linux systems.   This seems to have been mostly unsuccessful, but since I haven't been using Ubuntu for the past two years, my ear hasn't exactly been to the ground on this issue.

The main problem with most of these open-source replacements for ChemDraw is that, while they have the basic functionality down (i.e. being able to draw simple chemicals) they lack a lot of the flourishes that make ChemDraw such a pleasure to use.  For one, none of the programs I've tried has templates like ChemDraw's (where you can simply load an ACS or RSC template and not have to worry about bond lengths/widths, typeset, or font size at all).  Some of them even lack curly arrows, meaning that electron pushing is essentially impossible, unless you were to ink them in by hand (which I admit to doing once or twice).  Some have template molecules like carbohydrates and amino acids, most don't.  Some support a vast array of image export formats, some don't.  While some of these aren't hugely important to, say, an undergrad drawing structures for lab reports, they make writing actual papers a joke because the formatting is never right and small features like aligning objects have to be done by hand (#firstworldproblems).

So what's an Ubuntu-lover to do?  The impetus for this post is that I recently got that four-year-old laptop working again, and threw the latest version of Ubuntu on it.  In the past four years, you'd think that the chemistry drawing software would have improved by leaps and bounds.  Turns out...not so much.  However, there is light at the end of the tunnel.

GChemUtils is an ongoing project that is attempted to provide an open-source suite of chemistry software to Linux users.  They recently merged the GChemUtils and GChemPaint projects, and so far it looks like they have a fairly good feature set on the go.  The program has been actively developed by one very dedicated programmer, with help, since late 2002.  Lately, maintenance releases have been coming out about every month or so, and most of the changes are related to bug fixes as opposed to feature expansion, but the program is still getting stronger all the time.

Marvin is another suite that appears to be getting very good.  One excellent feature contained in their drawing program, MarvinSketch, is searching through ChemPub and ChemSpider for drawn structures.  Marvin has many of the structure editing features of ChemDraw, including (some) templates, cleanup, and valency checkers.  It also comes with a powerful set of visualization tools, and even molecular property calculators (which, while not something to depend too heavily upon, can be great first estimates for medicinal chemists).

One big difference between these two programs is the development team behind each.  Marvin is backed by ChemAxon, a company dedicated to the development of an array of cross-platform chemistry software.  GChemUtils is developed by a total of seven people, and as far as I know has abolutely no corporate backing.  As such, Gchemutils could really use some help. Whether or not you have programming experience, their homepage says they are looking for feature testers, web developers, programmers, bug hunters, translators, help with documentation, and even just brainstorming.

So while open-source and Linux-friendly chemistry software development is ongoing and vibrant, what seems to be missing is that little bit of refinement and feature diversity that many open-source projects seem to lack.  Unlike programs such as OpenOffice.org, which are almost perfect mimics of the program suites they replace, most of the chemical drawing software I've come across on Linux seems to need a push in the right direction.  Being a relatively niche market, it's my guess that a lot of companies aren't interested in spending time on an open-source program for (organic) chemists who use the least-popular operating system type on the market.  But if more weight gets behind these projects, one of the few remaining barriers to Linux adoption could be torn down.

To freedom!

Subscribe to the comments for this post

Yes, it's been done: coffee flavour chemistry

When I was much younger, and my interest in chemistry was just beginning to influence my thoughts of post-secondary school and (a lifetime away) a career, there were hints of my destiny that came in the form of somewhat perverse interests.

One of these was my profound interest in the chemical constituents of coffee. A non-chemist simply sees a cup of coffee for what their nervous system and digestive see it: a black liquid that tastes bitter, inhibits your appetite, and gives you about a two hour energy boost. However, being the child of two coffee snobs who also happened to be career scientists, I had different notions.

What makes a good cup of coffee? People have their preferred brands, countries of origin, or even vintages, if you have that much money to throw away. There are even those who eat an obscenely expensive bean that has been digested and excreted by the Asian palm civet. But can you actually taste the difference? And more importantly, what makes up the difference in flavour?

Most coffees are simply described in terms of acidity, roast, and some vague notes about other flavours (vanilla, caramel, berries, chocolate). This, to me, was completely unsatisfactory. Little did I know that coffee flavour chemistry is a legitimate field of study, as is the study of flavours and scents in general.

In my slightly-more-recent searches on the topic, I stumbled across a book entitled Coffee Flavour Chemistry; a better match for my childhood fascination could not have been conceived. Though now somewhat out of date, the book represented a major achievement in the field in 2002 when it was written by Ivon Flament, and contains some very interesting information and a comprehensive review of the work done in identifying coffee's chemical components. It identifies over one thousand compounds present in both green and roasted coffee beans, describing their aromas and their significance in the overall flavour of the coffee. Intuitively, the most important constituents of coffee flavour are the ones with the highest "signal-to-concentration" ratio, or how easily they can be detected at a certain concentration. Some of the most important constituents can be seen below, with their described flavours:

The majority of these compounds are of course formed via the well-known Strecker degradation and the Maillard reaction.  These, as well as other important aspects of the roasting of coffee beans, are discussed within the book.

Perhaps what I found most interesting are the techniques used to determine what are called "odour activity values", which essentially amount to a quantification of how strong a compound's odour is at a given concentration.

The methods used for finding these values are actually gas-chromatographic olfactometry, which is exactly what you think it is: they run a GC, and have someone sniffing the end of the column, who presses a button each time they smell a compound. I laughed at this when I read it, not because it's unreliable (quite the contrary), but because imagining running a column with a panel of "sniffers" at the end instead of a mass spectrometer is quite an image indeed.

The original method (GC-olfactometry) was purely qualitative in terms of odours, but later methods known as CHARM (a proprietary technique involving dilution-to-threshold) and AEDA (aroma extract dilution analysis, a similar method) have overshadowed it in recent years. Yet another method, named with a classic, groan-inducing acronym is "GC-SNIF", which stands for "gas chromatography-surface of nasal impact frequency", in which panels of test subjects are used to determine how "smellable" a compound is at a single concentration, producing averaged and normalized curves for the detection of smells.

Apart from being a funny image, these methods of analysis (which are almost always used in conjunction with purely analytical methods) illustrate an interesting point about flavour chemistry. While GC or LC data, as well as mathematical methods such as canonical analysis or principal component analysis are useful for predicting simple properties, the use of human organoleptic testing is essential to actually understanding the results.

Another salient point that was mentioned was the perception of quality. In a 1986 paper, Liardon and Spadone discovered that while the degree of coffee roasting was correlated to a large number of compounds and their concentrations, the quality of the coffee did not appear to be correlated to any of them. One of the few quantitative differences that could be established based on quality was that between green C. arabica and C. robusta beans. Robusta tended to have higher concentrations of methanol, acetone, pyridine, methylpyrazine, and furfural, and also seemed to be unique in that they contained methyl formate, t-butyl alcohol, and furfuryl alcohol, not found in Arabica strains. Robusta beans are widely perceived to be inferior in flavour and aroma to arabica beans, which is why many coffee packages will declare themselves to contain "100% arabica beans" to avoid confusion and distinguish themselves.

So what does this tell us? After skimming through the book I came away with two lessons: the first is that while computers and statistical analyses become more and more powerful every day, it seems there is usually a place for subjective human-generated data.  Without olfactory analyses from panelists, much of the work on coffee and its constituents would have been completely useless. The second is that while this data is essential to understanding the importance of certain compounds in generating a specific flavour, it is almost worthless when trying to establish a causal link between specific compounds and the perceived quality of the coffee (edit: this is not strictly true for identifying fundamental flaws in the bean due to parasites, mould, or poor growing conditions, which can all be identified by screening for certain compounds). For the most part, as everyone has heard so many times in their life: there's just no explaining some people's tastes.

However, I think for most of us in grad school, the fact remains that the most important compound in coffee is one which contributes almost nothing to its odour profile:

Subscribe to the comments for this post

The Wiley Interscience Blues

Hello, everyone!  Since this is my first post on Chemistry Blog, I should introduce myself.  My name is Nick, and I'm a Ph.D. student in organic chemistry at McGill University, in Montreal.  Mitch contacted me via the chemistry subreddit, and I'll be writing a few articles with what I hope is a unique perspective.  In advance, I would ask that you excuse my Canadian spellings; the letter "u" will pop up a lot more often than you're used to.

As anyone who regularly reads scientific journals may have noticed, Wiley redesigned some of their website earlier this year.  Mid-way through the summer, they slicked up their Interscience pages to look more "Web 2.0", and in the process, broke integration with one of my favourite things, which is Zotero.  Zotero was previously mentioned on the site quite some time ago, as one of several reference management programs available to modern researchers.  Given that it's free, absurdly easy to use, efficient, fast, allows proxies, and acts as a bridge between OpenOffice and Firefox (with downloadable reference formats), I unabashedly support the abandonement of every other reference management system in favour of it.  Zotero makes collecting references and writing papers a breeze, and a whole lot more enjoyable than any other option I've tried.

What Wiley did to break Zotero's flow was very simple.  Instead of having direct links to actual PDF files as part of their abstract pages (as nearly every other online publishing website does), they now direct you to a PDF file within an "iframe", meaning that Zotero is not able to "see" the PDF as an actual PDF.  This allows them to place a highly annoying "Wiley Interscience" bar at the top, including your institutional logo, and links to citing articles, abstract, and supplementary info, as seen blow.

This would be okay, except that with Zotero absolutely none of those links are necessary.  When you do the one-click save on an abstract it automatically generates a snapshot of the abstract page, including links to all that information.  Normally, it also saves a copy of the PDF, but Wiley has now made this significantly more complicated.  You must now either save the iframe page as a snapshot (including the annoying header and useless links), or download the PDF separately, import into Zotero, then delete the original download to avoid having duplicate copies on your hard drive.  So basically, instead of a one-click save, you now have an option of a four-step non-PDF download (via the "add item" button, seen above at the bottom left), or a five-step (take snapshot, navigate to "pdf", download, import, delete) rigmarole.

Compare this to ACS Publications, or ScienceDirect, where you click once on the address bar icon, and get all the above done in about 5 seconds (see below), or even ThiemeConnect, where you simply have to add the PDF as a separate item, and Wiley's "site improvements" actually begin to look like a big step backwards.

I've e-mailed Wiley about this twice, and it seems that their support staff have no idea what Zotero is, or why this is important, and don't seem to care.    Ultimately this isn't a huge issue, but I would really love to see a return to the old functionality; as it stands right now I cringe every time I see a paper I want hosted by Wiley Interscience.

Nick

Subscribe to the comments for this post