How can you identify an unknown substance?


         Head back to the helpdesk

         Find more tutorials

         Try some practice worksheets


Table of Contents for this page:

1. When might you come into contact with unknown chemicals in the real world?

2. Simple tests you can do

3. Chromatographic methods

  •  What is chromatography?
  •  Thin layer chromatography and paper chromatography
  •  Gas-liquid chromatography (GC, or GLC)
  •  Other exotic chromatographies 

4. Spectroscopic methods

  •  What is spectroscopy?
  •  Infrared spectroscopy
  •  UV-visible spectroscopy (UV-vis)
  •  Nuclear Magnetic Resonance (NMR) spectroscopy

5. X-Ray crystallography (a.k.a. X-ray diffraction, or XRD)

6. Mass spectrometry


1. Where might you come into contact with unknown chemicals in the real world?

It turns out that the ability to identify unknown chemicals is a pretty handy ability to have. For that reason, people have spent unbelievably huge amounts of money learning how to do it.

Let's say that you're working at a doctor's office, and some guy comes into the office gasping for air, claiming that he's been poisoned. What are you going to do? Well, depending on what he's been poisoned with, you're going to do different treatments. To maximize the effectiveness of the treatment, you need to figure out what he swallowed, or breathed, or whatever.

Let's say that you're dropping a bunch of garbage off at the city dump, and you see a leaking barrel with a sign that says:

Danger! These chemicals are some real bad stuff!

What are you going to do to get those chemicals cleaned up? It depends on what chemical they are.

Let's say that the water you're drinking turns green and starts smelling funny? How are you going to clean up your water supply? Before you can do that, you've got to figure out what's contaminating the water.

Well, you get the idea. Let's talk about some ways that scientists figure out what unknown chemicals are.


2.  Simple tests you can do

OK. I've convinced you that identifying unknown chemicals is important. Let's talk about some methods of identifying unknowns.

The following methods are pretty simple, and most can be done in even the crummiest chemistry lab. However, before we go too far, I want to interject a warning:

Don't screw around with chemicals at home! ONLY screw around with them in chemistry class!

Now that I said that, let's move to the simple methods of identifying unknowns:

  • Density: Density is equal to mass divided by volume. If you're trying to choose between two chemicals, you might want to take the density of the unknown by weighing it on a balance and taking the volume in a gradated cylinder. The chemical with the matching density is your unknown.
  • Color: It's strange to think of something like color being an indicator of what an unknown is, but this is sometimes used by people who are very familliar with the unknown. However, as you might imagine, many chemicals look the same color, so this is only handy in some very limited situations.
  • Smell: Smell can be a handy way of identifying a chemical if you've got a good nose. However, since many chemicals are extremely toxic, this is probably not a test you should use.
  • Melting and boiling points: Most chemicals have melting and boiling points that are very different from others. If you can accurately find the melting and/or boiling temperature of your unknown chemical, you can probably match that value to a table of melting and boiling points to identify your unknown. The big problem with this method: If your chemical is contaminated, this will affect the melting and boiling point. You might also find that it's hard to get really good data in a laboratory setting.

Let's look at some more conclusive and elaborate tests


 3.  Chromatographic methods

 What is chromatography?

Chromatography is when you take a chemical and dissolve it in a gas or liquid (referred to as the carrier gas or liquid). You then run the carrier with the dissolved unknown over a surface in which somebody has stuck particles of a solid (the "stationary phase"). The idea here is that when you run the liquid over the particles, some of the dissolved stuff will stick to the particles sbetter than the rest. By doing this, you can separate the compounds in a mixture, and hopefully figure out what they are.

So, what makes the dissolved chemical stick to the particles? Well, one popular thing people take advantage of is similarities in polarity. (Polarity is just a measurement of how unevenly the electrons in a compound are arranged - compounds that have a very uneven distribution of electrons are referred to as being "polar"). Since compounds that are polar like to attract other polar compounds, the chemical that is more polar in a mixture will take longer to pass over the particles, because it will like to stick to them better.

Still confused? Here's an analogy: Let's say that me and my great-aunt Gertrude go to the mall. I don't like the mall much, so I don't go into many stores. Gertrude, on the other hand, does nothing but shop, so she spends a ridiculous amount of time in the stores.

Even though we started at the same point in the mall, we won't get to the other side of the mall at the same time, because she's stuck in the stores shopping, while I spend very little time in the stores. Chemicals do the same thing: If two chemicals start at the beginning of the solid particles, the one with less polarity (think of it as the one who likes to shop less) will pass through it more quickly than the one with higher polarity (the one that likes to shop more). That's how chromatography is done.

Here's a sampling of some forms of chromatography which are commonly used:

 Thin layer chromatography (TLC)

TLC is a form of chromatography where the solid particles are just tiny pieces of silica gel (a polar compound) stuck to a piece of glass. What you do is take the mixture you want to separate and place it on the bottom of the glass plate, which only has one end in the nonpolar liquid (mobile phase). When the liquid runs up the plate, the nonpolar compounds will move farther up the plate than polar ones.

TLC is most commonly used as a test to see how many chemicals are in a mixture. Usually, if you want to separate large amounts of a chemical, you need to do another form of chromatography.

By the way, many classes use paper chromatography. This is exactly like TLC, except the solid particles are just paper fibers. Aside from that, the technique is the same.

 Column Chromatography

Column chromatography is the same as TLC, except that you fill a glass column with silica gel particles and place the mixture on top. You then push a nonpolar liquid through the column to separate the compounds. Again, nonpolar compounds pass through the column faster than polar ones.

Usually, the liquid is collected at the bottom of the column after it has passed through. With some skill and luck, you can separate compounds that are very close to each other in polarity.

 Gas Chromatography (GC)

In GC, the mobile phase which travels over the solid particles is a gas, usually something inert like helium or nitrogen. Instead of being dissolved in the gas, the mixture is usually just evaporated in the gas. The stationary phase consists of a solid powder which is stuck to the sides of the tube, very similar to TLC done in a tube.

The difference here is that in GC, the column which is used to separate the compounds is extremely long (many meters, depending on how well you want it to separate the mixture). The principle is that if you have polar particles on the side of the tube and polar compounds evaporated in the gas, the polar compounds will be stuck to the particles more and evaporate less. In practice, this works extremely well for mixtures with VERY similar polarities.

Often, GC machines have a mass spectrometer stuck to the other end of the nozzle. This gives you the ability to greater guess what your unknown is. We'll talk more about MS later.

 Other exotic chromatographies

Other chromatographic methods are also used, but usually not by high school students doing a lab. Here's a few:

  • High performance liquid chromatography (HPLC): HPLC is a method similar in principle to column chromatography, in which a liquid phase containing dissolved mixtures travels through a column containing a solid stationary phase. However, the difference here is that the columns are longer, and the pressures the liquid travels through the column are higher. For more information about HPLC, here's the website of a company that sells HPLC equipment: SciQuest.
  • Reverse phase chromatography: Reverse phase chromatography is a form of chromatography where the solvent is polar and the stationary phase is nonpolar. I'll be honest with you: I don't know much about it, so I'm not going to spout off about it.
  • Gel electrophoresis: Although it's not exactly a kind of chromatography, gel electrophoresis works under roughly the same principles. Basically, if you take a bunch of different nucleic acids, you can change the pH so they have different charges. If you then apply an electric field across a gel in which you have placed the nucleic acids, the ones that have more charge will travel to one end of the plate faster than those that don't. This is pretty handy in biochemistry, where you need to be able to separate different strands of DNA and stuff like that.

Size exclusion chromatography: Size exclusion chromatography is a form of chromatography where large particles in a mixture travel more slowly through a column because they can't get through as well as small ones. It's the same principle as in a traffic jam: Motorcycles can go as fast as they want in the worst traffic, because they can just go around other cars, since they're small. Trucks on the other hand, are stuck where they are because they're so big. The component in a mixture which travels fastest is the smallest one in SEC.


4.   Spectroscopic methods

What is spectroscopy?

Spectroscopy is when you measure the strength of light in order to figure out what an unknown chemical is.

There are two kinds of spectroscopy, absorption spectroscopy and emission spectroscopy. In absorption spectroscopy, you shine a light source at a sample containing your unknown and measure the light which passes through the sample. When you have these two values, you'll find that some of the light was absorbed by the sample - this energy corresponds to some energy level in the unknown. By knowing what energy level absorbed light, you can get clues about what the sample is.

In emission spectroscopy, you add lots of energy to a sample and then measure the light that's given off. In one fairly common kind of emission spectroscopy, you burn up a sample in a very hot flame, and then measure the colors of light that are given off. These energies correspond to the differences in orbital energy levels in the element being studied. Because every element has different energy levels, you can identify an element by this manner.

Let's look at some examples:

Infrared spectroscopy: Infrared spectroscopy (IR) is when infrared light is shone at a sample, and you measure how much of it is absorbed at each wavelength. As you might have guessed, this is a form of absorption spectroscopy.

An aside: Infrared light is light that has a wavelength of between roughly 800 nm and ~20 microns, depending on who you talk to. This light is not visible to the naked eye... IR light is actually what we would refer to as being "heat". (If something is hot, it gives off lots of IR). In IR spectroscopy, the wavelength of light is transferred into units called "wavenumbers", where 1 wavenumber is equal to 1/cm (or put in other ways, wavenumber is equal to the number of electromagnetic waves that can fit into one centimeter). It takes some getting used to.

When you see an IR spectrum, it will have a series of broad and sharp peaks. Each of these peaks corresponds to some kind of functional group in the molecule. I'll let you look those up in an organic chemistry book somewhere rather than write them all out.

So, what happens when this light is absorbed? Basically, the bonds in each molecule starts to vibrate, and the energy it takes to make the vibration occur accounts for the absorbed energy.

IR spectroscopy is generally used for organic compounds, because it's good at identifying functionality. Because it doesn't give you an exact structure of your compound, you usually need some other information other than an IR spectrum to conclusively identify a compound.

A cousin to IR spectroscopy is Raman spectroscopy. It works by roughly the same principles, although it's more involved. Suffice to say that if you're reading this, you're probably not about to jump right on a machine and start doing Raman spectroscopy.

UV-Vis (ultraviolet-visible) spectroscopy: UV-vis spectroscopy is a form of spectroscopy similar in principle to infrared spectroscopy. When you shine ultraviolet or visible light (the range for a spectrometer will usually go between about 300-1200 nm) on a sample, it will absorb some energy. The absorbed energy corresponds to some electronic transition in the molecule, such as an electron jumping from one orbital to another.

As far as I can tell, every UV-vis spectrum is different. I haven't really seen much in the way of tables of functional group absorptions, mainly because every molecule is different. This method of spectroscopy is mainly handy for conclusively identifying an unknown - for example, if you know the unknown is one of two possibilities, and you know the UV-vis spectra for both, you just take a spectrum of the unknown and match it to the one you do know.

Nuclear magnetic resonance (NMR) spectroscopy: In NMR spectroscopy, radio waves are used to excite transitions in the molecules. Basically, you take a sample and place a small magnetic field on it. When this is done, the magnetic dipoles in the molecule all align in one direction. After this is done, you "pulse" the sample with an extremely strong radio wave, causing all the dipoles to align in the opposite direction. Eventually, all the dipoles go back to the way they were, causing energy to be given off. This energy is measured, and translated into a spectrum.

For the spins to flip, you need atoms that have a spin = 1/2. Common atoms that do this are hydrogen-1 atoms (which are used in proton NMR), and carbon-13 atoms (used in 13-C NMR). There are others. Many others.

To put it plainly, NMR spectrometers are really cool. They look like giant metal balls with huge computer consoles attached. When a sample is inserted or removed, white smoke is given off by the boiling of liquefied nitrogen in the center. Why are they so elaborate?

NMR spectrometers are basically huge magnets that give off enormous magnetic fields. These fields are required to cause the dipoles in molecules to switch direction. The liquid nitrogen is used to keep the magnets cool - because they are superconducting magnets, they need to stay at very cold temperatures. Do you want to really make an NMR spectroscoper mad? Just throw a handful of staples at the magnet. Then, watch the fun as they have to pry them off, one by one.

Warning: If you throw staples at the magnet, the operator will cut your heart out and eat it.

Warning: If you have a pacemaker, stay away from the magnet! Keep credit cards and disks away, too.

NMR spectrometers are unbelievably expensive. Please, for your own safety, don't screw with the magnets or anything else, unless you want to lose your allowance for the rest of your life.

By the way, MRI (magnetic resonance imaging) machines that they have in hospitals are really just big NMR spectrometers. Why the different name? Who knows?


5.   X-ray crystallography (a.k.a. X-ray diffraction, or XRD)

XRD is when you take x-rays and shine them at a crystal of a chemical. The stream of x-rays is then refracted by the crystal into a really complicated pattern. Using powerful computers, the XRD can tell you what molecule you have.

XRD is mainly used for identifying the properties and structures of chemicals that you have already identified, although I suppose there's no reason that you couldn't use it to identify unknowns. Except for the reason that they're not all that common. And they're expensive. And they require you to have a crystal of your sample. Aside from that, no problem.

Biochemists use XRD to identify what structures large proteins have. Other people use it, too, although I have no idea what for. In my years as a chemist, it never once came in handy.

The main problem with XRD is that you need a crystal of your compound, and some compounds just don't crystallize. If your compound doesn't crystallize, then tough luck, buster, you ain't getting a diffraction pattern.


6.  Mass spectrometry (a.k.a. mass spec or MS)

Mass spectrometry is really very handy for finding out what you've got in a sample. Essentially, it works like this:
1. Using a flame or laser, rip the molecules in your sample to shreds. Generally, this produces ions with positive charge.
2. Accelerate the ions toward a plate with negative charge. There should be a hole in the plate that the ions can pass through.
3. Steer the ions toward another plate with even more negative charge. If the particles have lots of mass, they'll travel further over the plate than particles with low mass because they have more momentum. 

The data from a mass spec isn't that hard to read. Basically, you end up with a chart that gives a list of peaks and corresponding m/z ratios. "m" stands for mass, and "z" stands for charge. (An example, if you have a particle that weighs 50 amu and a charge of +2, the m/z ratio will be 25). By interpreting this chart (or better yet, having a computer interpret the chart), you can figure out what the molecule looks like, and what elements are in it.

Mass spectrometers can be added to the end of a gas chromatograph so that you can both separate mixtures and identify what the are.

Back to Mr. Guch's helpdesk!

Back to the Cavalcade of Chemistry main page!

Comments on this enormous essay? Email me at