Before discussing the dangerous chemistry of these films, let us remember how motion pictures work.  The video portion of a movie is a series of pictures, projected at a usual rate of 24 per second, although that can range from 12 to 48 per second (television video in North America is usually broadcast at 30 frames per second).  At these speeds, our eyes do not perceive the individual pictures, but rather see movement on the screen.  In the mid-19th century, instruments such as zoetropes were used to create illusions of motion from still photographs.  A limitation of these instruments was that only a small number of photographs could be used, restricting the length of any “video” to a few seconds.

Besides, the technology did not exist to capture continuous photographs for a significant length of time.  Étienne-Jules Marey’s chronophotographic gun could capture 12 frames per second, but all images were superimposed onto the same picture.  Was there a way to capture each image onto a separate surface, and then replayed later?

Different sources credit Thomas Edison or William Friese-Greene as developing the first motion camera.  Friese-Greene’s apparatus, patented in 1889, was able to capture 10 photographs per second, with the images captured on a celluloid film.  At the same time, Eastman Kodak developed a celluloid film base – cotton was dissolved in nitric acid, in a reaction catalyzed by sulfuric acid.  The cellulose in the cotton would react to form nitrocellulose:

C₆H₁₀O₅ + 2 HNO₃ ? C₆H₈(NO₂)₂O₅ + 2 H₂O

A plasticizer, often camphor (the ingredient that gives Vicks VapoRub its distinctive odour) was added to the mixture, allowing it to be stretched into thin celluloid strips, which were then coated with a light-sensitive emulsion that would capture the image. The celluloid film had the right combination for a film base: it was transparent to light, and tough but flexible, allowing it to be perforated so it could be guided through the camera, and later spooled for storage.  It was also cheap and relatively easy to produce.

(A) The structure of a monomer of cellulose. (B) The structure of a monomer of nitrocellulose – note the nitro (NO2) groups bound to three oxygens. From BioScience.

Both cellulose and nitrocellulose are polymers.  Each repeating entity is called a monomer.  The figure to the right shows (A) a cellulose monomer and (B) a nitrocellulose monomer – note that three OH groups in cellulose become ONO2 groups in nitrocellulose.  The nitrogen content of the nitrocellulose determines whether it will explode spontaneous.  Fully nitrated nitrocellulose (with three nitro groups) contains 14.2% N, and was originally intended to be used as a weapon, but this was quickly abandoned when it was realized that it was too sensitive and could be easily set off.  Any nitrocellulose with a nitrogen content over 12.6% N is considered explosive [ref].

The nitrocellulose used for motion picture films had two nitro groups per monomer instead of three, and its nitrogen content was closer to 11% – no longer prone to explosion, but still highly combustible.  (Flash paper, used by magicians when they need to burn something quickly, is another form of nitrocellulose with similar nitrogen content.)  The nitrate film was sensitive to moisture, static electricity, friction, light and heat.  This made the film, and therefore the projectionist, highly vulnerable to the hot lamp of a film projector in a small, cramped room. The exact mechanism and products of the decomposition of nitrocellulose appears to be unknown; recent publications point to the formation of nitrogen monoxide (NO), nitrogen dioxide (NO₂), oxygen (O₂) and nitric acid (HNO₃).  The formation of oxygen as a product of decomposition is quite problematic in fighting a nitrate film fire, since smothering it won’t work – the reaction generates enough oxidant to keep it going until all of the film is consumed.  A celluloid film also cannot be extinguished by immersion in water.

httpv://www.youtube.com/watch?v=AL9izOFrqbw

As movie houses became more popular in the 1910s, and the length of movies (in time, and therefore in physical length of film) increased, so did the risk of fire and, unfortunately, the occurrences of fatal accidents.  The projectionist was much busier in that era, since a single roll of film would contain no more than 20 minutes of video.  Two projectors were used to ensure a seamless transition from the end of one roll to the start of the next roll.  Unfortunately, film did catch fire quite easily, with disastrous consequences, including the Dromcolliher cinema tragedy in Ireland in 1926, the Laurier Palace Cinema tragedy in Canada in 1927 and the 1929 Glen Cinema Disaster in Scotland, each killing over 50 people, mostly children.  (The front page of the Montreal Gazette on the morning after the Laurier Palace tragedy pointed out that children under 16 were not allowed in movie theatres in the province – I’m not sure if it was due to safety concerns from film, or due to concerns of the film content on the screen.)

Despite the hazards of nitrate film, its advantages still outweighed its drawbacks, and it wasn’t until the 1948 that Eastman Kodak finally offered a viable alternative – cellulose acetate film (or safety film), which was much less flammable.  In the final decades of nitrate film, projection rooms were stripped of any wooden furniture and insulated with – wait for it – asbestos, to isolate any fires within the room.  As far as I know, projectionists weren’t entitled to danger pay.

Most nitrate films from this era are now lost or in poor condition.  Even under proper storage, the decomposition of nitrocellulose will, to quote Kodak,

yellow the film base, yellow and soften gelatin, and oxidize the silver image. Later, the base cockles, becoming very brittle and then sticky. Finally, it disintegrates completely. This inevitable deterioration is usually gradual, but elevated temperatures and humidity speed it greatly.

The sharp, pungent smell of nitric acid becomes strong, and the buildup of gases in a well-sealed container causes heat to build, making conditions rife for combustion.  This information sheet on Kodak’s website lists precautions for handling nitrate films that would rival what most chemists have to follow when handling dangerous compounds.

Some nitrate films were preserved by transferring them onto acetate film, but even those are beginning to decay, so there is now a movement to transfer all film onto digital platforms.  Unfortunately, fire has already consumed the original nitrate film negatives of many classic films.  This list of vault fires in the 20th century gives us a sense of what has been lost.  More reading on nitrate film and its dangers can be found here.

Anyone who watched the excellent 2011 film The Artist, which was set in the late 1920s and early 1930s, will remember this scene:

http://www.youtube.com/watch?v=3n-I4gBNcDo

George Valentin, shaken as the invention of talkies (films with recorded sound) would mean the premature end of his acting career, angrily unspooled the film from his collection onto the floor, and set the pile on fire with a single match.  He initially laughed as the flames quickly grew, but the intense smoke caused him to lose consciousness.  His ever-loyal dog was able to draw the attention of a police officer who reluctantly followed him to the burning house, where George was found badly burned and barely alive, clutching a film canister that remained intact in the blaze.  A symbolic act for sure, but also a reminder that in the first half of the 20th century, the most dangerous part of attending the movies was not the content projected on the screen, but the actual film that contained those pictures.

But you said that HPLC can give detection limits in the parts-per-billion range.  That seems really low…  Can we really do even better than that?

Yes we can, although we won’t get mired in little details of how to optimize HPLC.  We will, however, take a bit of time to consider the detection system. Let me remind you that the detector gives a signal that should be proportional to the concentration of the compound in whatever mixture passes through the detector. Simple spectroscopic detectors shine a light of a pre-selected wavelength onto whatever is flowing through it, and different compounds may absorb some of this light at different rates.

The UV spectrum of caffeine, with wavelength along the x-axis. Note the maximum at 276 nm. From Wikipedia.

Having the detector measure a signal at a single wavelength is fine if we are only interested in analyzing one compound, since the sensitivity can be optimized by selecting the wavelength that is best absorbed by our analyte. The figure to the right shows a spectrum of caffeine – this plot shows how light is absorbed, at different wavelengths, by a sample of caffeine.  The wavelength of maximum absorption here is 276 nm.  (The absolute absorbance value on the y-axis is irrelevant here, just the location of the highest value.)

If we were only analyzing caffeine by HPLC, we would set the detector to measure absorbance at 276 nm to get maximum sensitivity.  Maximum sensitivity means we will be able to detect smaller changes in concentration.  But when we analyze several compounds through a single chromatographic separation, we may need to compromise and select a wavelength that works well for all of them, but likely will not be the wavelength of maximum absorption for any of them.

Hmmm… I have an idea.  What if we installed multiple detectors, with each one set to measure at a different wavelength?

A diagram of a diode-array detector system.  In HPLC, the “sample cell” would be the flow of the mobile phase. From Seton Hall University.

That’s a great idea, and in fact we can do even better!  A modern solution is to use a diode-array detector (DAD).  Instead of a single wavelength of light going through the sample and reaching the detector, and light from a lamp that contains many wavelengths is sent through the sample.  A prism or a grating then disperses the white light that is not absorbed onto the detector, which measures the leftover light at many different wavelengths at once.  This approach allows measurements of hundreds of wavelengths simultaneously.  A spectrum is generated for each measurement, just like the caffeine spectrum shown above; with HPLC, a new spectrum can be generated every second or even faster.

Note that a different spectrum is associated with each peak, allowing identification of each compound as it is separated. From Sciencephoto.

A spectrum for each measurement allows the user to independently select the best wavelength to analyze each analyte in the chromatogram.  Collecting a spectrum at each point can also assist when several analytes co-elute; with a single signal per measurement it is impossible to quantify both analytes, since we cannot know how much of the signal is caused by each analyte. The spectra allows the user to mathematically distinguish between the analytes – this methodology was part of the Masters work of your humble guide, with a publication to boot.

Speaking of distinctions – doesn’t the detector see both the eluent and the analyte?

Yes, this is a significant handicap of HPLC, as it is certainly possible that the detector will capture both the eluent and the analyte, particularly if both are organic substances – they will tend to absorb light well in the ultraviolet (UV) region, between 200 and 400 nm.  This can cause a significant background, and detecting the analyte becomes a bit like reading grey text on a white background. That said, the detector accounts for the background much like a weighing scale is reset to zero after an empty container is placed on it.

I understand that you can account for the mobile phase when measuring the signal.  If only we could remove the eluent and only keep the compounds that we want to analyze.

But there is a different approach that minimizes the impact of the eluent…  Instead of using a liquid as the mobile phase, we can use a gas as the mobile phase.

A gas? How does that work?

A diagram of gas chromatography. From about.com.

The fundamentals of the technique are similar to HPLC, but instead of the liquid mobile phase, a carrier gas is used to drive the sample along the column. The column is kept in a small oven, and the column is heated after sample injection.  Higher temperatures will cause compounds to pass through the column more quickly, as some compounds will exceed their boiling point and evaporate – this is great for eliminating the solvent, but a problem when trying to separate several analytes with similar boiling points.  This is where operating a GC can become a bit of an art – finding a compromise, in terms of temperature and carrier gas flow rate, so that the analytes run through the column as quickly as possible, while making use of the stationary phase to ensure separation.  Just as with other chromatographic methods, some compounds adsorb onto the stationary phase better than others.  GC columns are usually packed with polymers as a stationary phase, and they are typically tens to hundreds of meters long.

Wow, you’d need several football fields to house that instrument…

Of course not! Think of your 50-foot garden hose… Is it a stiff 50-foot-long piece of material?

An open GC instrument – notice the wound copper-coloured column in the oven. From Intellectual Ventures.

No, I keep it wound up in the shed.

Exactly. GC columns are wound as well, so that they fit into instruments that are similar in size to HPLC instruments. The gas can still make its way around the column.

Carrier gases are small molecules such as acetylene, etc.  As you can imagine, the pressure of the gas (and therefore the flow rate) can be programmed, just like the temperature of the oven in which the column is housed, to optimize separation.

Are similar detection methods used in GC?

A diagram of a flame ionization detector. From Equipco.

UV spectroscopy is not sensitive enough for use in GC, since such minuscule quantities of sample are used.  Since the analyte is mixed with inert gases, there are other avenues for detection. A common detector uses flame ionization (FID). As the sample is brought into the flame, it is heated to extremely high temperatures, which causes the organic species to ionize. The signal that is measured by FID is proportional to the concentration of the ionized species, which will be proportional to the concentration of the organic species itself.

Is this a sensitive detection method?

FIDs tend to offer detection limits that are in the parts-per-billion (ppb) – comparable to HPLC, but with fewer complications than UV spectroscopy.  It does suffer from the same handicap that plagues single-wavelength detection in HPLC, in that it cannot distinguish the compounds that reach the detector. But it is fairly inexpensive, and works well in a wide range of concentrations.

Another approach, the thermal conductivity detector (TCD), gives a signal whenever there is a change in the gas that passes through the detector – this allows us to note the presence of any analyte with the carrier gas. Again, it is incapable of distinguishing which compounds are in the gas.

Well then, it doesn’t sound like GC offers much advantage over HPLC, and the equipment seems more complicated.

The basics of mass spectrometry. From UC-Davis.

Ah, but there is a popular and highly sensitive, but finicky, detection system that is often combined with GC.  It is mass spectrometry, or MS; the combined acronym GC-MS is often used to describe the entire system.  In MS, the molecules are bombarded with a stream of high-speed electrons. When an electron strikes the molecule with the right energy, the molecule is ionized (becoming a cation, or a positively-charged species).  This molecular ion breaks apart – we say “fragmented” – and some fragments will maintain that positive charge.  A magnet in the MS detector causes the charged particles to bend, with the degree of bending related to the fragment’s “mass-to-charge” (m/z) ratio.

Mass spectrum of caffeine. From Carnegie Mellon University.

A mass spectrum shows the relative abundance (or percentage) of fragments with each m/z ratio.  The figure on the left shows the mass spectrum of caffeine.  Note the largest peak is at m/z 194; this is the molecular mass of a caffeine molecule (C₈H₁₀N₄O₂).  From this spectrum, it is possible to figure out the fragmentation steps followed by the molecule.  For caffeine, the steps that lead to the major peaks in the spectrum are shown to the right.

How can someone know that this is the way the molecule broke apart?

Fragmentation pathways for caffeine. From Carnegie Mellon University.

Admittedly, figuring these pathways requires practice, but there are general rules in terms of how molecules fragment – and that is well beyond the scope of this post.  An experienced user, when looking at the mass spectrum of an unknown molecule, can compare the pattern of the fragmentation to databases of mass spectra to deduce the structure of that molecule. While the analytical chemists are out having fun, organic chemists entertain themselves on Friday nights looking at mass spectra to figure out whether is that white substance at the bottom of their flask.

Oh yeah, orgchem snap!

Oh yeah, I went there.

As it turns out, one of the requirements for using MS is that the sample is vaporized, and since the outflow from the GC is already in the gas phase, it can be streamed directly into the MS instrument.

Quantitative analysis is done in GC-MS similarly to HPLC; standard solutions are used to develop a calibration curve, relating the concentration of the analyte to the height of its peak in the chromatogram.  (See our last post for a reminder of calibration curves.)  When a sample is analyzed, the height of the peak for the analyte is measured, and the calibration curve is used to calculate its concentration.  The mass spectra is very convenient when several compounds are eluting at similar times, allowing unambiguous determination of which peak belongs to an analyte of interest.

So combining gas chromatography and mass spectrometry allows us to lower detection limits?

Yes.  A few examples include detection cocaine and heroin at 5 ppb and 10 ppb, respectively, and hydrofluorocarbons and hydrochlorocarbons are detected to less than 200 ppq (parts-per-quadrillion, or pg/L, picograms per litre) and some contaminants in drinking water to low ppq values, through careful application of GC-MS.  There are even examples of 1 fg of pesticide detected by GC-MS.  (1 femtogram is 10^-15 g, or a billionth of a milligram.)  GC-MS is widely considered a gold standard in forensic substance identification.

This is great, and I look forward to reading those links.  So what have we accomplished in these last few posts on Numbers of Atoms?

We have looked at various ways to separate analytes in compounds – from picking raisins from cereal flakes by hand, to using state-of-the-art equipment to distinguish between minute quantities of substances. As this series is a dynamic product, there may be more discussion on chromatography and other separation methods at a later time, but we will be changing topics in the next portion of Numbers of Atoms.

Initial reactions from Canadian scientists appear quite negative, pointing to this as yet another facet of the Conservative government “War on Science”. This government has taken measures that are hostile towards scientific research in this country, such as making significant cuts to research programs in various Crown-funded scientific agencies, the closing of the unique Experimental Lakes Area (a measure that the Ontario and Manitoba governments are trying to undo), and the censoring of government scientists that has garnered worldwide attention.  I do not see these changes to the NRC as an assault on science in Canada, but as a missed opportunity to truly grow our knowledge economy.

The changes to the NRC were proposed last year; at that time, I wrote two articles on Atoms and Numbers that considered (1) how the changes were one-sided, as businesses would have the upper hand on any new relationship with NRC researchers; and (2) the potential impact of these changes on academic researchers who are not focussed on short-term economic gains.

In the official announcement of the new NRC mandate, Minister of State for Science and Technology Gary Goodyear, declared that

There (are) only two reasons why we do science and technology: First of course is to create knowledge and push the frontiers of understanding. The second reason is to use that knowledge for social and economic benefit. Unfortunately, all too often, the knowledge gained is opportunity lost.

It is true that any knowledge acquired through research, in laboratories or on the field, means little it is not shared with a wider public. Knowledge is shared by publication in the scientific literature, and this is of great importance to others conducting research but it has little real value outside of the laboratory. Knowledge is also shared through the production of innovative products or the development of new services that can improve our lives, while also providing economic benefits to the stakeholders.

There should always be room for so-called blue skies research, research “for curiousity’s sake” that does not bring immediate or apparent benefits. Last year, Goodyear stated that the fundamental research which was previously led by NRC researchers is now done in universities – again, a somewhat valid quote that misses the bigger picture of reduced funding budgets and renewed emphasis on quantity vs quality of produced research. As I wrote last year:

The “publish or perish” mentality among young faculty members at our research institutes drives them away from research for research’s sake – tenure-track faculty will pursue research directions that can be generously funded and published quickly, in order to make the best case for tenure. In that atmosphere, there is little time or thought for tangents in the research program, or for discovering commercial potential for research (except to mention it in the research proposal as another argument for receiving funding). Even the ideal of a research institution forming new researchers is set aside, as the ambitious graduate student also knows that publishing quality (read: highly-cited) articles is the route to the best postdoctoral opportunities.

Any government is well advised to encourage and fund some blue-skies research, as the return on investment can be enormous: one fundamental discovery today can inform a hundred research programs tomorrow. But researchers should be able to articulate some potential avenues by which their research could be monetized. Some of the criticism from the academic community towards this new NRC mandate seems to come from a desire to maintain the purity of academic research, untainted from self-serving corporate interference. In the past years I’ve realized that the state of academia and economics today makes this ideal untenable. While universities are pumping out larger numbers of capable researchers, they are also in an era of declining funding (including education cuts by non-Conservative governments). This further restricts the number of new positions for researchers in academic institutions. It would certainly benefit the wider scientific community for research leaders to consider innovative directions that can lead to revenue and, eventually, more jobs for these capable scientists.  Many scientists are more loyal to their science than their country, and will quickly leave Canada for other opportunities if they need to do so.

That said, the new NRC structure falls short of a proper integration of science and business. University of Toronto chemistry professor and Nobel Prize laureate John Polanyi encapsulated my thoughts to the Toronto Star:

One should structure things so (scientists) have the freedom and responsibility to provide ideas to industry, not just receive commands.” He added that “innovation is the result of dialogue between science and business. It would be a mistake to think that industry sees ahead to the basic innovations that are going to benefit it. It sees some. But scientists who are in touch with the development of scientific knowledge will see more.

The knowledge economy can grow fastest if business and science work together. There will certainly be instances when a business needs some assistance in creating a new product, and having quick access to government-paid experts and equipment could accelerate that product to market. This may be preferable to the company hiring new staff or expanding their facilities in order to get over a small “hump” in their R&D. However, the flip side is that the NRC may become little more than an “offshore hub” within Canadian borders for low-cost research. I foresee companies offloading riskier research onto NRC researchers, which will minimize the potential setbacks for the business while creating problems for the researcher who wants a record of positive outcomes for their curriculum vitae.

Commercial innovation in science and technology should be a two-way road – certainly NRC researchers have come up with discoveries that have enormous economic potential, and they will need business experience and capital to help bring these ideas to market. Some researchers may have zero interest in embarking on a business venture, but they should be open to selling market rights to others who see an opportunity to bring it to the marketplace. Some of the best-known technology companies today – Facebook, Google, even Apple – started with a product or service that was developed that its creators before going to investors. The new NRC mandate would turn the process around – imagine Steve Jobs sitting by the phone, waiting for a businessperson to offer him a few dollars to create a combination cell phone and music player which can also be used to surf the Internet.

It is often the people on the front lines, doing the R&D work, who get the bright ideas and develop them long before the business world ever becomes aware of its potential. That is knowledge that must be harnessed: for the benefit of science, for the benefit of our society, and for the benefit of our economy.

Tyler Irving discusses the government’s changes to the NRC in his blog, and is inviting Canadians to engage in a debate over these changes. I am very interested to see what other scientific researchers think about the new NRC mandate.

The story has spread through the Internet, with accusations that the punishment was motivated by racism and sexism, or by general chemophobia.  Kiera is described as a bright and honest kid who was simply curious about what would happen when she mixed these two things together.  I can’t comment on what went through her mind that morning, or through the minds of the school authorities or the police officers involved in the aftermath. But I will offer some other thoughts on what happened.

Several scientists have stated that this incident is a blow to science, as it will discourage young people from seeking answers and having fun doing so.  We certainly want to encourage all people to wonder about how the world works, and to seek answers to those questions. Sometimes, the only way to get an answer is to conduct an experiment – what happens if I mix this and that? Not all experiments are planned with a carefully thought-out step-by-step protocol, and occasionally things end up being messy, but doesn’t that make it even a bit more exciting?

Every person is born a scientist, as they spend the first years of their lives figuring out how their body works and how the world around them works.  They are curious about everything around them, exploring dark corners, touching things, putting them in their mouths, sometimes hurting themselves but always processing new information. As adults, we need to encourage and maintain the desire and the wonder of discovery. I once told the story of a 5-year-old who used the scientific method to conduct a short but useful experiment to confirm that a wine aerator really did make wine taste better. Of course, we want to encourage children to explore in a safe environment – so parents take the time to cover electrical sockets, barricade stairwells, lock cabinets and put hazardous materials where they cannot be grabbed and put into young, curious mouths.

While some people defend the scientific ideals in Kiera’s story, it is not as simple as saying that “she was arrested for doing science”.  We do need to acknowledge that an important part of being a scientist is making use of knowledge that has already been accumulated.  A teenager who sees a sign at a gas station banning open flames may be genuinely curious to know why. Knowing the danger involved, we would point him or her to an article or a video on YouTube.  We wouldn’t condone this teenager lighting a match next to a pump, explaining that “I was just curious…”

Kiera’s experiment was, in a way, the purest form of science – she wanted to see for herself what happened if she mixed “The Works” toilet cleaner and aluminium foil, so she tried it. (EDIT: I originally stated it was Drano and aluminium foil, but the police arrest report says differently.  Drano and “The Works” would have similar ingredients; Tyler Irving explains the chemical reaction that would have caused the explosion.)  But experiments conducted without adequate knowledge can have dangerous consequences – thank goodness she used small quantities of the reactants that morning. A quick search on the Internet shows that similar mixtures can be dangerous, and that it has been used as a “bottle bomb”. She may have already known that, as she did say to the police she expected to see some smoke.  There is nothing wrong with questioning what is already known, but having seen warnings of what could happen, she should have taken the proper precautions to conduct her experiment.

On that point, though, I do wonder what the attitudes are in her school towards proper independent study by students.  Is it encouraged and nurtured, or is it condemned and seen as a annoyance to the proper functioning of the school?  If Kiera had approached a chemistry teacher about wanting to conduct this experiment under a fume hood, would that teacher have allowed it? Would that teacher have taken a few minutes after school to do the experiment with her, with proper precautions? Would that teacher have turned it into a class demonstration for all to see? Or would that teacher have shooed her away, not wanting any information about this explosive mixture to be known?

There is also a real dilemma for school administrators in a case like this, because one student’s sincere attempt to satisfy her curiosity can look just like another student’s prank. Let’s face it, kids and teenagers are fascinated by stinky things that go boom, and usually not for altruistic reasons. Kids have always caused trouble with stink bombs and other nasty concoctions, and those who want to cause a bit of trouble would have no shame in telling the teachers afterwards that “I was just doing an experiment”. A school culture that has predefined consequences for infractions makes it difficult, even pointless, for a teacher to figure out a student’s true motives, and it does become a blunt object in the rare instance where the student’s interests appear genuine.

I do think Kiera must be reinstated to her school and all charges must be dropped.  A short suspension, or maybe an essay to reflect on what she did, would have been quite sufficient, so the time she has already served more than makes up for that.  I also wonder if Kiera was already known for asking challenging questions and showing strong interest in science classes – if so, perhaps a science teacher should offer to find her a mentor to nurture her interests in science.

Ashutosh Jogalekar states that this incident “definitely succeeded in squelching independent scientific curiosity in its students”. I don’t think that will be the case. Good scientists are tenacious in the face of adversity: when research is going badly, when answers are not easily found, even when people who ought to know better seem to discourage independent thought. If Kiera truly loves science, she will see this incident as a momentary setback.  She will persevere, and she will continue to explore and learn more about the world around her.  Maybe she’ll someday work in a laboratory, contributing to our global body of knowledge.  And maybe one day she will be called upon to mentor high school students, helping them journey independently along that exciting path of discovery.

Some big-time column chromatography from the 1950s. Thankfully things are smaller today. From Wikipedia.

Our last post was an introduction to some chromatographic approaches for separating compounds in a mixture.  The methods described there were, if we can say, “manual” – just the stationary and mobile phases, the sample, and nature doing the work with gravity or capillary action.  We concluded the discussion with column chromatography, and some interesting questions on automating the process and dealing with colourless samples.

Yes. let’s consider automation first.  These manual methods seem quite involved, requiring many graduate students to work many hours a day to process everything in a lab.

It would take up a lot of time to always separate compounds on paper or thin layers, or through large columns propelled by gravity.  In previous posts, we looked at ways to separate the compounds of a mixture so they can then be used in further analysis.  Chromatography can be used strictly for physical separations – the column chromatography we just mentioned is excellent for doing that – but it can be much more convenient for that same instrument to also provide us data about those compounds, to assist in a qualitative test (is that compound in the mixture?) or a quantitative analysis (how much of that compound is in the mixture?).  Besides, in many cases we have no interest in recovering any part of the sample that we are analyzing.

Typical HPLC columns. From Hamilton Company.

Technology has certainly improved on the paper or thin-layer approaches.  For one thing, the columns used in liquid chromatography can be much smaller than the long, wide tubes required for “manual” column chromatography.  Columns used in automated systems tend to range from a few to 50 cm long, and a few millimetres wide.  These columns are packed with silica or polymer particles, which provides large surfaces over which the sample could be adsorbed.  They are also easier to keep wet; a continuing concern with column chromatography through an open tube is that the column will dry out and form cracks in the stationary phase that will ruin the separation, since the sample will simply flow through the cracks without adsorbing on the column.

Cross-section of HPLC columns. Note that analytical columns are narrower. From Waters Company.

Instead of relying on gravity to propel the sample through the column, a large pressure is mechanically applied to the sample column, forcing the mobile phase through the column much more quickly. High performance liquid chromatography (HPLC) involves applying pressures of several MPa (every MPa represents nearly 10 times regular atmospheric pressure) through the column.  The small stationary phase particles and high pressures applied onto the column allows for quick separation of the mixture.

Wow, that sounds hard.

Yes, it would be pretty hard to have several MPa of pressure bearing down on you.  Luckily, the particles in the column is very well packed together.

Har har.

Exactly.

A diagram of how HPLC works. Several solvent reservoirs can be attached to the pump solvent manager to regulate the mobile phase composition. From Waters Company.

Modern HPLC instruments are controlled by computer.  Along with controlling the pressure, we can also manipulate the composition of the mobile phase during a run.  By attaching two or three miscible (mixable) solvents to a pump (for example: water, methanol and acetonitrile), the eluent can be modified through a run to ensure the best possible separation of the compounds in the sample.

I’ve noticed, in the past few posts, that you hand-wave some explanations about how some eluents work with some analytes better than others. How is a newcomer to know which eluent, or which combination of eluents, to use?

HPLC instrumentation can become complex. From Oregon State University.

That’s an excellent question, and in most cases it comes down to an educated process of trial and error.  If the mixture contains both polar compounds (such as ionic compounds, or compounds with halogens or oxygen atoms) and non-polar compounds (such as compounds with long chains of carbon and hydrogen atoms), it may be best to run a mixture of water and acetonitrile, with acetonitrile concentration gradually increasing over time.  That will force out the polar substances first, followed eventually by the non-polar analytes.  But HPLC is so widespread that the best approach is usually to find research protocols or publications that successfully separated similar compounds to what you need to analyze, and adapt the protocol from there.

Chromatography is very reproducible when the variables are kept constant.  The composition of the stationary phase, in terms of what material it is and the size of the particles, is a crucial factor – but then, the same column is usually used throughout a series of samples.  The composition of the mobile phase, as we’ve established, is important.  The force with which the eluent passes through the column, which establishes the flow rate of the mobile phase, is also a factor. When we use a vertical column, the force bearing down on the column is gravity, which is constant.  In an HPLC setup, the mobile phase is pumped through the system at a rate that is optimized so that the entire separation of one sample can be done in a reasonable time.

The columns seem so small.  How do you get a sample into the column, and how do you know what is coming out?

The sample is loaded into a loop with a syringe, then injected onto the column. From Hitachi.

In HPLC, a very small quantity of liquid sample can be used, on the order of microliters.  The sample is sucked into a syringe, with care taken to avoid any air bubbles, and then loaded into a small loop near the entrance of the column, which ensures that just the right quantity of sample is injected onto the column itself.

The stream that exits the column is led into a detector which, well, detects when a substance is passing through.  In later posts we will look at spectroscopy, which will help us to understand how these detectors work.  For now, we will accept that the detector gives a signal that is proportional to the quantity of analyte in front of the detector at that moment.  (Once the stream has passed through the detector, it is usually run into a waste bottle.)

Chromatogram of five compounds from flavonols in Ginkgo biloba. From Kanfer Lab, Rhodes University, South Africa.

This is were we find the quantitative power of HPLC.  When the program (the flow rate, the eluent composition and the column composition) is the same, an analyte will always pass in front of the detector at the same time after injection.  A plot of the detector signal over time is a chromatogram, and the heights of the peaks or the area of the curve underneath the peak can tell us how much of that substance was in the sample.  Note that each compound has a different sensitivity in terms of its signal – in the example on our left, there is not necessarily more quercetin in the mixture even though its peak is the highest.  Also note that chromatography users record elution time in decimal time – for example, rutin is noted as having reached the detector in 8.944 minutes instead of 8 minutes 56.6 seconds.  Using decimal time makes some of the mathematics of chromatography easier to handle.

How does a chromatogram help us know how much of an analyte is in a solution?

If they are analyzing several similar samples for the same analyte – for example, a series of blood samples for particular vitamins – we can use the signal measured on the chromatogram at that analyte’s elution time to figure out the quantity of substance.  For this to work, we need to be certain that the same volume of sample is always injected onto the column – this process is automated by the instrument, which is why the sample is injected into a loop instead of directly onto the column.

A series of standard solutions – each containing a known quantity of the analyte, dissolved in a solvent similar to the samples – is run, from which the elution time of the analyte is confirmed.  The signal measured at the elution time for each solution is either calculated by the computer, or measured with a ruler on the chromatogram.  In principle, the height of the peak or the area under the curve should vary linearly to the concentration of the analyte – if we double the concentration, the peak will be twice as high.

Calibration curve for conjugated linoleic acid. From Herzellah et al, J Diary Sci, 88(4), 1301 (2005)

A calibration curve is shown here – the peak area is plotted against the concentration of conjugated linoleic acid in ppm (same as ug/mL).  The slope of the calibration curve tells us exactly how the signal will change for each unit of concentration – in this case, for every 1 ppm of linoleic acid, the area under the curve will grow by 132,743 units.

Then how do we use that with a real sample, where you don’t already know the concentration?

In this case, linoleic acid is monitored in milk samples.  Suppose the peak at the elution time of linoleic acid from our first real sample gave an area of 4,355,688 units.  Since we know from the curve that the concentration changes by 1 ppm per 132,743 units, we calculate that 4,355,688 units x (1 ppm / 132,743 units) = 32.81 ppm.

Wow, that’s not so difficult.

It isn’t.  But remember – this is the concentration of the linoleic acid in the mixture that was injected onto the column.  We may have performed our chemical steps beforehand to concentrate or dilute the linoleic acid in the mixture, so all of those steps must be accounted for into order to find the “real” concentration in the original sample.

It is essential that no other compounds elute at the same time as the analyte.  We can analyze multiple analytes at the same time, as long as we know the elution times and the slope of the calibration curve for each compound.  If we know we will analyze several compounds at the same time, our standard solutions can contain all of those analytes to make the process easier.

How sensitive are these methods?

The process I’ve described here – HPLC with a detector that measures a single signal at a time – can routinely reach detection limits in the parts-per-million (ug/mL) to parts-per-billion (ng/mL) range.  This is adequate for many applications.  In trace analysis of waters and bodily fluids, or in forensic analysis, even lower concentrations need to be analyzed.  Thankfully, innovations have improved the sensitivity of chromatographic methods even further.  And we will look at that in our next post.

Syringes for SPE. From Genpore.

Let us generalize the concept the SPE. The solid matrix in the syringe, through which the sample is passed, is called the stationary phase, because it is, well, stationary in the process. The liquid that is passed through the sample is called the mobile phase, as it moves along or through the stationary phase.

Ok, but you promised we’d look at chromatography…  Are we going to get there soon?

We’re almost there.  I just want you to consider one other aspect of SPE.  In many cases, we are trying to eliminate what remains on the solid phase – sometimes, we just throw the solid away, and work with the cleaned-up liquid.

But consider a situation where the compounds that we want to analyze are adsorbed onto the solid phase.  Why do they remain there?

Because they have a greater affinity for the solid matrix, instead of the liquid they were dissolved in.

Washing an SPE column. From Polyintell.

Exactly.  So how do we then remove those compounds from the solid phase?  We run a new solvent through it – a solvent with different characteristics from the sample we just passed through (we’ll consider what these characteristics are in a bit) –  forcing the analyte into that solvent and eventually out of the syringe.

So by changing the mobile phase, we can affect whether an analyte remains on the stationary phase?

Yep. And that observation is key to understanding the power of chromatography, as a tool not only for preparing samples, but also for qualitative and quantitative analysis.

Paper chromatography in a beaker. From ChromatographyScience.

That’s great.  So how is chromatography used?

Chromatography – from the Greek words “chroma” and “graphy” which together can mean “colour writing” – was originally used mainly for analysis of dyes or mixtures of compounds of different colours.  The first chromatographic methods used paper as the stationary phase.  In fact, that is still used today: small drops of samples were placed near the bottom of the paper, and the paper is placed into a chamber with a small volume of solvent – at a level below the drops.

Calculation of the retardation factor. From Periodni.com

The solvent climbs the paper by capillary action, and it will carry with it the various constituents of the sample, to different degrees. As the leading edge of the solvent (known as the solvent front) approaches the top of the paper, different components of the sample will have climbed different heights. The ratio of the height of the solvent front to the height that a compound climbed is known as the retardation factor R_f.  (Yes, that is the correct word in this context, but many people also refer to R_f as the retention factor.)

How can it be known if a compound will have a high or low R_f?

Generally, the polarity of the solvent will determine whether a compound will prefer to dissolve in the solvent (and result in a high R_f), or not prefer to dissolve (resulting in a low R_f). Remember that like dissolves like, so a non-polar solvent such as hexane will carry non-polar analytes much further than it will carry polar analytes. Likewise, a polar solvent such as water will carry polar analytes further up the paper.  This is the same characteristic used in SPE to choose a solvent that will remove a compound from the solid matrix.

But there is an important point here: the R_f for an analyte is consistent when using the same stationary and mobile phases. So this becomes a form of qualitative analysis – by comparing the various spots of a sample to a standard, we can quickly confirm whether an analyte is part of a mixture.

That all sounds nice, but what if the compounds are colourless?  Can chromatography still be useful?

Most certainly, although it requires some extra steps.

A silica gel plate under an ultraviolet light source. From the RSC.

Instead of paper, a thin layer of silica gel (derived from silicon dioxide) is spread onto a glass plate, to serve as the stationary phase. For colourless organic materials, the plate is placed under an ultraviolet light source, where the dots become visible.

As with paper chromatography, this approach can be used to confirm the presence of a compound in a mixture.  As with SPE, the solid phase can be rinsed with a new solvent to extract the compound back into a liquid.

That’s pretty neat.

Actually, it sometimes gets pretty messy.  *rimshot*

I really need to stop setting you up like that.

Hey, I’m just saying it as it is…

Now, let us consider how we can bring together what we’ve seen in this article.  Along with an external force (such as the syringe), let us also use a solvent to push the sample through the stationary phase.

What would that look like?

The most efficient way to do this is to pour the stationary phase into a column of silica gel. By holding the column vertical and introducing the sample through the top, gravity becomes the external force that drives the mixture down the column.

Typical column chromatography setup. From chemistry-blog.com

But consider this – the sample solution is poured into a column that has a much larger volume than the sample itself. The sample solvent and some compounds percolate through the column, while other compounds remain at the top.

How do we get those compounds out?  Sounds like the column will dry out.

And we definitely don’t want the column to dry out.  A solvent is poured into the column – preferably, a solvent in which some of the compounds adsorbed on the column will dissolve and be quickly carried down the column, while other compounds remain near the top. Once one compound has passed through the column, a different solvent (with a different polarity) is poured down the column – other compounds will be driven down the column. Repeat for all of the compounds to be removed.

For each compound that we want to recover, a different beaker is placed at the exit of the column to capture that compound in the solvent.

That’s kinda cool – but there must be a way to automate this.  And besides, how do we know that a compound has left the column if it were colourless?

Chromatography is a mature and fascinating field.  In the next post, we’ll reduce the column from a large vertical apparatus to a small column that sits horizontally on the benchtop.  We will also consider chromatography as a highly reproducible quantitative method.

CBC offers this disclaimer in their study – I wonder how many people bother to read it:

The website is not meant as a definitive source of information about hospital performance. It provides general information and does not offer medical advice or recommendations for treatment.

A stay at a hospital with a good grade will not necessarily be problem-free. A hospital with a lower grade might do well on measures that CBC has not considered.

The website is designed to help Canadians understand a hospital’s possible strengths and weaknesses so that they can ask their health care providers informed questions about their care.

Fair enough – this report card does not represent medical advice, but certainly their editors recognize that people will react to their local hospital getting an A or a D – and CBC did highlight the “top 10″ and the “bottom 8″ in a front page story.  Certainly, people will be talking about the results, and we need to establish how they were generated.

CBC gives some details of their methodology.   The data was collected by the Canadian Institute of Health Information (CIHI), a group that publishes 27 clinical and financial indicators on hospital performance.  The CIHI website is quite dry, and those objective indicators remain wide open to a subjective interpretation of what makes a “good” hospital.  The CBC panel chose to focus on a small subset of these indicators, reasoning that “many of the indicators reported by CIHI focus on specialities such as obstetrics and cardiac care, but not all hospitals offer the full range of specialized care.”

The CBC panel selected five indicators which they believe assesses the quality of general surgical and medical treatment: (1) death after major surgery, (2) surgical and (3) medical patients who experienced unexpected complications tied to nursing care during a hospital stay, and patients who were readmitted following (4) surgery or (5) medical treatment.  To put it in less eloquent terms, they are looking at whether people who come into a hospital for treatment or surgery leave the hospital in a reasonable time (or at all…), and whether they have to go back for more treatment within 30 days.

This information is certainly good to know, but are they really the best criteria for ranking hospitals across the country?  Are these the top things that most people consider when thinking about a “good” hospital?  To me, this seems like preparing a report card for automobiles based solely on which ones are involved in the fewest fatal accidents.

Another note: the methodology description points out that they use “standardization”, another way of stating that they are calculating z-values – the number of standard deviations away from the group mean, for each statistic obtained at each hospital.  This approach may be problematic – I have doubts that the number of deaths/complications/readmissions after surgeries, from all hospitals across the country, follows a normal distribution, a necessary prerequisite for this standardization.  Differences from one hospital to the next are not likely to come down to random fluctuations (from a statistical point of view); factors such as average age and average income in the community, and types of treatments available, will have much more impact on whether patients at a hospital will tend to need readmission, suffer complications, or die after surgery.  A town with an older population will have more readmissions, complications and deaths, simply because older people tend to have more health problems – hopefully the hospital staff is doing all it can to help their patients, but there is only so much they can do with the demographics of their community.  Furthermore, it should not be surprising that Toronto General Hospital has a bad grade for readmissions, since they are much likelier to deal with acute and complex medical situations, or make use of experimental or risky treatments that may require several patient visits.

Looking at the data from Prince Edward Island gives us some more insight.  Prince County Hospital’s A grade is buoyed by having reported no patient deaths within five days of major surgery – an A+ rating, as it is far below the national average of 8.62 deaths per 1,000 major surgeries. Charlottetown’s Queen Elizabeth Hospital is also below the national average in terms of mortality, but with a rate of 6.09 deaths per 1,000 patients undergoing major surgeries, it only gets an A in that category.  Prince County Hospital also had excellent ratings for readmission after surgery and readmission after medical treatment, while the Queen Elizabeth Hospital had readmission rates that were slightly above average – C grades in both cases.

Is this an indication that Summerside’s hospital is “better” than Charlottetown’s?  Should a patient in Summerside feel safer, better cared for, likelier to have their medical situation resolved during their stay than someone in Charlottetown?  Unfortunately, some may be led to that conclusion, but it simply cannot be drawn from this data.  What is known, but not described in this study, is that serious surgeries on PEI tend to be sent to Charlottetown, and higher-risk surgeries such as heart conditions are not even done on PEI; instead the patients are sent to Heart Centres in Saint John and Halifax.  These considerations are necessary to provide context to the numbers in the study.

A hospital with a record of zero deaths after major surgery is impressive, but if a death did occur – wouldn’t your impression change if it was an 80-year-old with a history of heart disease and diabetes who had suffered a serious heart attack, versus a 30-year-old who was misdiagnosed by the doctors?

As I’m sure many Canadians did, my wife and I looked through the list for hospitals that we knew.  The first hospital I checked was Southlake Regional Health Centre in Newmarket, Ontario.  Not only is it our area hospital, but I also underwent an appendectomy there in 2011.  It scored a C – we thought that was quite low, particularly after we checked the hospitals in PEI and other hospitals in the Maritimes, which all scored As and Bs.  I thought the treatment I received at Southlake was excellent – from the emergency room to the nurses and the doctors.  My three nights there were as comfortable as it could be in the circumstances.  My wife thought it was easy to get around, and the staff were very helpful and cordial to her when she visited me.

Maybe my experience was unique, and I was lucky to get good care when I was there. Maybe it was perfectly normal for Southlake – and from what I’ve heard in the community, I think the latter is much closer to the truth.  Southlake is also home to the cancer care institute for York Region (population of just over a million people).  Are some of the readmitted patients, or those who died after surgery, stemming from complications of cancer treatment – something that other hospitals in the area do not deal with?  The life expectancy in York Region is 2.5 years longer than for the entire province of Ontario – surely Southlake contributes to that in a positive manner?

At this point, I am sliding in a direction that scientists, particularly those in health care, take great pains to avoid – looking at anecdotes as a substitute for data.  But in the end, the question of the “best” hospital, or even a “good” hospital, is one that can never be answered objectively.  I love working with numbers and statistics, and it is part of my life’s vocation – but in doing that, it is important to recognize instances where the numbers do not adequately tell the story.

Rating hospitals is a difficult undertaking, and I have to wonder what it ultimately accomplishes.  There are raw, hard, objective numbers available, and yet there will always be disagreement over what statistics are most important when it comes to health care.  And even then a ranking is achieved, people will naturally counter these “scientific” ratings with their own highly subjective impressions, which are clouded by circumstances.  Except perhaps for a birth, hospital visits normally occur in stressful and unpleasant times, when emotions are very close to the surface.  A person’s impression that a hospital is “bad” can come down to an encounter with a seemingly uncaring doctor or nurse – perhaps themselves having a bad day.  We may forgive a bad cup of coffee from our favourite restaurant, but a bad cup of coffee while we are on our last nerves, waiting for life-changing news that the nurses and doctors are not telling us, can lead us to believe that the entire hospital just doesn’t care.  Most people visit hospitals infrequently, and are admitted even less often, so the sample size that they have to compare hospitals themselves is very small – and it’s not as if the hospital gets a second chance, on their terms, to make things right.

And that leads to the bigger question: is it necessary, or even useful, to rank hospitals?  Does it really matter that the hospital in the next city, or the next province, is ranked a bit better than mine – so long as my hospital provides proper, adequate and caring service to my community?  What are the benefits of pitting one hospital against another?  This study may have caused some debate and discussion about what is expected from our health care system.  Perhaps one part of this discussion will involve defining what a “good” hospital really is, and getting people mobilized to make sure their community’s hospital reach that objective.

But when the analyte isn’t so different from everything else, it can’t be easy to do the separation. 

These sieves are stacked, and can separate solid mixtures by particle size. From uta.edu.

That is true.  When a mixture of solids has different-sized particles, such as soil mixed with pebbles, a stack of sieves will separate the mixture by particle size.  The larger particles are trapped in the top of the stack, while smaller particles pass through towards the bottom.

There are situations where adding a solvent to a mixture of solids will dissolve only the analyte, or perhaps the analyte and minimal additional components.  An analysis for the salt content in potato chips can be done by first immersing a pre-weighed sample of chips in distilled water.  The sodium chloride will leach into the water, along with other dissoluble salts and some of the surface oils.  After filtering out the ensuing potato mush, the liquid mixture is then further treated to measure the salt content.

“Further treated”?  The analyte is now dissolved in a liquid.  How can it be separated from the rest of the mixture?

Certainly, separating dissolved components in a liquid mixture is more involved – particularly in a quantitative analysis, where we want to ensure that all of the analyte has been isolated.

One approach for isolating the analyte (particularly when it is dissolved in ionic form) is reacting it with another compound to form a precipitate – this is known as gravimetric analysis.  For example, a known volume of that extract of potato chip can be treated with some silver nitrate (AgNO₃).  The silver cations (Ag⁺) reacts with the chloride ions (Cl^-) to form insoluble silver chloride (AgCl).  (In this case, it must be assumed that all extracted chloride ions came from sodium chloride, and that nothing else in the solution would form a precipitate with either the silver cations or the nitrate anions.)  Although not a concern in this example, sometimes the pH of the solution can be adjusted to hasten the precipitation.

After the chloride ions in the solution have precipitated (strictly speaking, a small portion of AgCl does dissolve, but that quantity can be calculated), it is filtered out, washed, heated in an oven or ignited, cooled and weighed.

The solid is in the white crucible, being ignited. From University of Washington.

Wait a minute – ignited?  Like set on fire?

Yes, it can be ignited, but nowadays it is more common to heat the solid in an oven.  Heating above 100°C will dry the solid of any water caught in it.  Igniting it will ensure evaporation, and may also turn the compounds in the precipitate into a more stable chemical form.  The filter paper will burn along with it, but it will form a predictable quantity of ash that is indicated on the box.

That’s pretty cool.

Actually, at that point it’s still pretty hot.  *rimshot*

Oy, that was bad.

But I had to correct you, because the next step is to cool the sample in a desiccator – a container that forms a tight seal to prevent exposure to water from the air. Once it is at room temperature, the solid is weighed.  A little bit of mathematics is needed to work out the mass of analyte.

Both of these are carbonyl compounds. From CUNY Brooklyn.

I just noticed – if nothing had precipitated, you would have known that there was no chloride in the solution.

You’re right – and this logic is commonly used in qualitative analysis, when we only need a yes/no answer.  Some of these tests are used regularly in the organic laboratory.

Note the beautiful silver mirror. From San Francisco State University.

A great example is the Tollens’ Test, which is used to determine whether a compound that is known to contain a C=O bond is an aldehyde (with a hydrogen and a carbon bonded to that carbon) or a ketone (with two carbons bonded to that carbon).

This test is done with Tollens’ reagent – a silver ion coordinated to ammonia [Ag(NH₃)₂⁺] with hydroxide ions (OH^-).   A few drops of this reagent is poured into a test tube containing the test mixture. If the substance is an aldehyde, a silver precipitate forms, and if well done, it turns into a beautiful silver mirror. If the substance is a ketone, then nothing happens to the mixture.

Wow, that is also cool… I mean awesome.

It is – analytical chemistry saves the day!

Returning to quantitative analysis, another approach that is used in laboratories is the liquid-liquid extraction, sometimes known as a partitioning.  This method works best when an analyte, dissolved in a solvent, is known to dissolve much better in a different, immiscible liquid.  Immiscible means two liquids cannot mix together, such as oil and water; typically one of the solvents will be aqueous (water-based) and the other will be organic, since most compounds will dissolve considerably better in one solvent versus the other.  This approach may be useful, for example, in analyzing dissolved organic compounds in river water – those compounds could be extracted from a large volume of water into a small volume of an organic solvent, making further work much simpler.

The heptane on top separates nicely from the water below. From NOAA.

The sample and the second solvent are poured into a separatory funnel, a conically-shaped piece of glassware, and thoroughly and vigorously mixed.  Despite the fears of many students, the liquids do eventually separate again, and a significant portion of the analyte will have dissolved into the new solvent.  The shape of the funnel makes it easy to pour out the contents into two different containers.  It is understood that all of the analyte will likely not have been dissolved into the new solvent, so multiple extractions are performed and combined.

But if you know that less than 100% of the analyte will have been extracted, isn’t that a problem?

You’re right, and in a quantitative analysis, that has to be reconciled.  The fraction of the analyte that moves from one solvent to the other depends on which analyte and which solvents are used, and the volume of each solvent.  I won’t get into the calculations here (this PDF document can help if you are interested), but these parameters must be established before the extraction is done.

A distillation apparatus. From Wikipedia.

Distillation is another approach for separating liquids, although it is more commonly used with miscible liquids that have fairly different boiling points.  The mixture is heated to the temperature at which one of the liquids boils away – it does so into a tube that is cooled, so that the vapours are re-liquefied and collected.

Another option for separation is the solid-phase extraction, also known as SPE.  (That’s not to be confused with SPF, the Sun Protection Factor used by sunscreen makers to charge you more money for marginal additional benefits.)  In this case, the sample solution is forced through a solid matrix, often by lodging the solid within a syringe.  Each analyte in the solution will either remain in the solution, or adsorb onto the surface of the solid.

Syringes prepared for solid phase extraction. From General Laboratory Supply.

Oops, I think you meant to write absorb…

No, I meant adsorb.  Adsorption is the process of a molecule (or atom, or ion) sticking onto a surface, while absorption is the process by which that particle enters or goes into that surface.

If the analyte has no affinity to the solid matrix, it will have flowed through and remains in the solution – while other compounds would presumably have remained on the solid.  On the other hand, if the analyte adsorbed onto the solid phase, the solid is then rinsed with a different solvent to remove it from the matrix.

That sounds like it could be powerful method for separating components of a mixture.

In fact, it is.  SPE is but one of several methods for separating compounds in mixtures by running a sample along a solid surface. A whole family of methods are based on this principle – they are known as chromatographic methods.  When done properly, they can provide excellent ability for quantitative analysis. We’ll explore chromatography in the next post.

Before we can explore ways to determine what and how much of something is in a mixture, we need to consider that the physical specimen we will be handling is but a portion of an entity – whether it is a portion of soil from a field, a bottle of water from a lake, or a scoop of a food product processed through a factory.  That sample must be representative of whatever we want to analyze.  To ponder that, let us first consider a box of Raisin Bran.

Two scoops of raisins…

Raisin Bran? What does that have to do with analytical chemistry? And besides, I hate raisins.

Yeah, I’m not a fan of them either. But think about it – what is the promise given in every Raisin Bran commercial?

“Two scoops of raisins in a package…” Uh, how big is a scoop?

I’m not sure, and unless we break into the Raisin Bran factory, we won’t know for now. But some people have investigated the claim themselves – in fact, two researchers from Seattle, Gregory J. Crowther and Elizabeth A. Stahl, recently counted the raisins in several boxes of the product and reported the results, admittedly in a light-hearted manner, in the Science Creative Quarterly. They pointed out that different box sizes still claim the same two scoops of raisins. Are bigger scoops used with larger boxes? Or is the concept of a scoop strictly a marketing construct that is unrelated to how many raisins are in the box?

So what did they conclude?

They couldn’t draw much of a conclusion. They used three different sized boxes, two boxes of each, and counted the number of raisins in each box. There was significant variance even between two boxes of the same size. But the data would lean towards more raisins (and thus, larger scoops) in bigger-sized boxes.

Wait a minute. They counted the raisins? That seems so low-tech in the 21^st century.

It is, but counting is also the only type of measurement that is exact. When you count things, there is no uncertainty involved. When they found 381 raisins in a box, that was 381 +/- 0 raisins.

But shouldn’t a larger raisin count more than a smaller raisin?

If we are counting strictly the number of objects, then every raisin counts as 1 raisin, regardless of its size. It’s like counting people – if LeBron James and I are in a room together, we’d say that there are exactly 2 people in the room, since we each count as 1 person. We wouldn’t say that there are 2.25 people in the room, just because LeBron is 25% more person than I am.

But you make a good point – even if it is undefined to us, a “scoop” is a unit of volume, and we can fit a larger number of small raisins in the same scoop. So it would be preferable to separate the raisins and the flakes into two piles, and look at the volume of raisins. If we assume that the jingle is correct, we could take many boxes of the same size, average the volume of raisins in each box, and take that as the size of “two scoops”.

That sounds like an awful lot of work, and I don’t have students on whom to dump this task. Do we have to use the entire box of cereal?

No we don’t. You (or your student) could take a portion of the contents of the box, figure out how many raisins are in that portion, and then scale that quantity to the entire box. In other words, if we figure out how many raisins are in half of a box, we multiply by 2 to get the raisin content of the entire box. If we took a quarter of a box, we’d multiply by 4 to get the raisin content of the entire box.  And so on.

That’s better. I mean, when they take blood tests, they don’t drain all of the blood from your body, they only take a bit of it. Same thing when they want to analyze drinking water or soil.

You’re right, but done improperly, sampling creates its own problems. If we’re using a portion of that box of Raisin Bran, we first have to weigh the cereal in the entire package (not just take the mass printed on the box, which is probably not that accurate), and then weigh the sample of cereal, so we know what fraction of the cereal in the original package is in that sample. That requires two measurements, each of which will have a bit of uncertainty associated with it.

But even more importantly, the package must be very well mixed before taking a sample. If all of the raisins lay in the bottom of the box, and we took a sample from the top of the package, we’d have the impression that our cereal is rather raisin-free; likewise, if we took the bottom quarter of that box, we could think we’ve hit raisin heaven. (Or hell for you and me.) So we mix the entire package as thoroughly as we can, so that every portion pulled out of the box will have approximately the same composition of raisins and flakes.

Alright. I’m still not sure what this has to do with what analytical chemists do in the lab. You work with white powders and solutions, not breakfast cereals.

Hey now – we do sometimes work with breakfast cereals. Someone has to figure out how much of those essential nutrients are in one serving!

But Raisin Bran is a large-scale way to think about the analysis of a target (analyte) in a sample. A sample must be well mixed to ensure that the sample is representative – this is much easier to do with gases, and can be easy with liquids, but it requires extra consideration for solid samples.

And as I indicated earlier, counting is the only measurement that has no uncertainty. The ideal chemical analysis would give us the exact count of how many molecules of every chemical species are in a sample – but that’s impossible.

I know we can’t see and separate individual molecules.

And even if we could, we wouldn’t have the time to do it. A sample of 1 mg of a mixture – as much as a speck of dust – has trillions of trillions of molecules in it. Even if we could separate and count them at a rate of ten molecules a second, it would still take us approximately the age of the Earth to go through all molecules of that single sample.

Ok then, so how are we going to figure out how much of an analyte is in a sample?

We will still try to separate the analyte from the rest of the sample, as best as we can. The simplest separation approaches are physical – just like picking out the raisins in the Raisin Bran – but there are also methods that involve chemical reactions. Once we isolate the analyte, we can proceed to measuring how much analyte is there.  We will look at that in the next post.

While I was on vacation, I saw this sign posted at a Wal-Mart store:

No chemicals down the sink or drain? Does that mean I have to wash my hands over a bucket?

My chemistry-twisted brain needed some time to notice the obvious (to most) typo: it appears that “pain” is a flammable product at Wal-Mart. Instead, I focussed on what they consider to be “chemicals”: bleach, cleaning supplies and pool chemicals. (How many Wal-Marts have swimming pools?) Interestingly, lawn chemicals are not listed as “chemicals”, but as “lawn and garden products”.

Yes, I agree that bleach, cleaning supplies and pool chemicals can be dangerous products, and should be handled carefully – although I wonder if they would take the same precautions to dispose of 100% natural cleaner. But these products should have been categorized as, well, cleaning supplies. To simply call them “chemicals” perpetrates the notion that chemicals are dangerous things – so dangerous, in fact, that you are not even supposed to dispose of them down the sink!

Meanwhile, in a higher-profile takedown of chemophobia, blogger See Arr Oh discusses, in some detail, claims made in a petition created by two food bloggers seeking to ban two artificial colourings in Kraft Mac and Cheese (Kraft Dinner to us Canadians), mainly because these colourings are “artificial food dyes… man-made in a lab with chemicals derived from petroleum.” It is pointed out that the molecular structures of these dyes hardly look like petroleum products, nor do they behave like petroleum products. Furthermore, See Arr Oh counters claims that these dyes are “contaminated with known carcinogens” by showing that the National Library of Medicine toxicology database, the Merck Index and even Wikipedia contains “[no] positive studies suggesting carcinogenicity.” (One of the food bloggers, who tweets under @thefoodbabe, declared See Arr Oh’s article to be incorrect, then pushes aside some simple questions raised in the post because she gets to be on CNN, and because @SeeArrOh does not show his face on his Twitter feed. You know, science.)

While the petition is just the kind of sensational “news you can use” that the media loves to propagate without a second thought, at least CNN is willing to admit that the scientific studies on this are, at best (or worst), inconclusive. Advocating for healthy food is certainly noble and important, and it is good to question the contents of highly-processed foods; no one should think that Mac and Cheese has the nutritional content of a garden salad and a bowl of fresh fruit. But this advocacy must be backed by actual facts, not tired clichés and logical fallacies that something must be dangerous if it is a “chemical” and “artificial”.