Introduction to Gas Chromatography

Introduction
Chromatography is most likely the most widely
used and powerful of all the techniques of chemical
analysis. The general concept of chromatography
was developed in the late 1800’s by Mikhail Tswett,
a Russian botanist, who used the chromatographic
process to extract naturally occurring chlorophylls
in plants. After three decades without further
progress, chromatography reemerged as partition
chromatography with research by Martin and Synge in
the 1930’s. Chromatography remains a widely used
technique today because it is a simple and
inexpensive method of separating and analyzing
complex mixtures.
Chromatography works to separate components
of a substance by passing a flowing mobile phase
across a porous stationary phase. The separate
components of the mobile phase flow at different
rates because the individual species are retarded
by the stationary phase based on various
interactions, such as surface adsorption, relative
solubility, and charge. Naturally, components that
tend to stay in the mobile phase (low affinity for
stationary phase) will move faster along the
stationary phase than others. Scientists have
taken two approaches to explain the separation
process, plate theory and rate theory.
Plate theory was proposed in 1941 by Martin
and Synge, based on known information about
distillation. It seems that in distillation,
actual plates exist where vapor passes through a
liquid phase, and when they mix equilibrium is
assumed. In chromatography, the plates are
unobservable and are called theoretical plates.
Further research has indicated that the number of
theoretical plates is directly related to the
efficiency of a chromatography system or how well
the components have separated. The number of
theoretical plates can easily be calculated:

Number of plates = 16 *

(Retention Volume)2 (Length

of Column)

------------------------- =
----------------------
(Band Width)2
(Height of Plate)

Obviously, chemists seek to maximize the efficiency

of a chromatographic system, and have discovered
that a careful choice of chemistry (mobile and
stationary phases) is required to do so.
Rate theory was developed in 1956 by J. J.
van Deemter, and accounts for the dynamics of a
separation. It differs from plate theory because
it accounts for solute (mobile phase) diffusion and
flow paths. In addition, rate theory can be used
to predict the effect of these on column
performance factors, such as phase velocity, phase
thickness, and flow rates. This theory is based on
a differential equation that takes into account
many factors affecting chromatography performance,
including those listed above.

Gas chromatography (GC) was the first

instrumental chromatographic method developed
commercially because it is relatively easy to
produce a stable flow and pressure for the carrier
gas (mobile phase). It was first introduced by
Martin and James in 1952, and has quickly become
one of the most powerful and widely used methods
for the analysis of the components of a sample
mixture.
GC involves passing a gaseous mobile phase
through a porous stationary phase in a long tube
(chromatography column). In this experiment we
will be using gas-solid chromatography (gsc), which
utilizes a solid stationary phase. There is also
gas-liquid chromatography (glc) in which a liquid
stationary phase is used. The basic system
consists of a carrier gas, a sample injection port,
a separating column, and a detector. The carrier
gas is usually a pure inert gas with a very
carefully controlled flow rate. The column may be
a packed column or capillary column, each of which
could be used with any one of hundreds of
stationary phases (a great advantage of GC). Also,
many commercial GC’s have built-in detectors which
measure the retention time (time between injection
of sample and when component is detected). If
required, the detectors can quantitatively measure
the concentration of a component.
Building on the seemingly mundane use of GC
in this experiment, scientists widely use it in all
areas of science, medicine, and industry. A few
examples of GC applications include testing for
trace hydrocarbons and pollutants in air, refining
of petroleum, drug and alcohol testing, discovering
flavoring agents in foods and contaminants in
alcohols, and research with pheromones.

For our experiment, we use a small-scale GC

device. Methane serves as a carrier gas, which is
highly controlled by the natural gas tap in the
Chem lab. Our injection port is a syringe injected
into ordinary latex tubing. Our column is a short
tube of soft glass filled with activated tide
detergent which acts as a stationary phase.
Finally, our detector is a Beilstein detector, a GC
detector that emits visible light when halocarbon
components go through it. The detector consists of
a copper coil placed in a flame which reacts with
halogen atoms formed from halocarbons passing
through the flame. As halocarbons pass through,
they react with the copper to form copper halides.
Then, the halides react with hydroxide radicals to
give excited copper species like CuOH+. When the
excited copper ions return to ground state, the
result is a green-blue light emitted from the
copper. Further, quantative analysis is possible
when using some type of photodetector, such as a
CdS cell or photodiode, to transduce emitted light
into an electronic effect.
In this experiment, I expect to construct a
working small-scale GC device (see appendix A) and
begin to understand its use by measuring the
retention times of certain halocarbons, namely
Freon-21 (CHCl2F), Freon-22 (CHClF2), and
dichloromethane (CH2Cl2). After that brief
introduction to GC, I will be doing a GC separation
of mixtures of the above gases. Finally, I will do
crude qualitative analysis by GC by using a cadmium
sulfide emitted light detector and measuring the
resistance of the detector with a digital
multimeter.

Procedure
In this experiment, I followed the procedure
outlined in the Chemtrek Manual (exp.. 19--Gas
Chromatography). The ensuing remarks are a general
description of the procedure by section:

A.) Create a small-scale gas chromatograph

and Beilstein detector.
B.) “Age” the Beilstein detector and measure
the retention time for air.
C.) Measure the retention times of various
halocarbons, including Freon-21, Freon-22, and
dichloromethane.
D.) Make a homogenous mixture of several
halocarbons and separate them with GC. Then,
measure the halocarbon retention times and band
width values to determine the resolution.
E.) Add a quantitative emitted light
detector to the GC and use them to find a
calibration for the quantitative analysis of
dichloromethane. Finally, find the number of moles
of dichloromethane in an unknown sample.

Results
B). Retention time of air:
The average retention time of air is 3
seconds.

We can find the linear gas velocity of the

mobile phase from the equation:

Velocity = Length of column/Avg. retention

time.
so, Linear Gas Velocity = 30 cm/3 s. = 10 cm/s.

C). Measurements of retention times and other

information for various Halocarbons:

(Table Omitted)

* Boiling Point and Molecular Weight taken

from CRC Handbook of Chemistry and Physics, 3rd.
ed.
**Neighbor’s retention times courtesy of Pat
Bonner and Ryan Blinn.

Band width figures were determined by finding

the difference between Disappearance and Appearance
times.
(e.g. Freon-22 is 9sec. - 5sec. = 4
sec.)

D). In section D, a homogenous mixture of

Freon-22, Freon-21, and Dichloromethane was
prepared and injected into the GC device. In doing
so, I observed the following:

Indication of a halocarbon at 5 seconds, with

a maximum intensity at 6 sec.

Then, all signs of a halocarbon disappeared

until a total elapsed time of 50 seconds. This
time, the maximum intensity was indicated at 57
secs.

(Note: As I was not able to successfully

separate all three halocarbons, I only used the two
samples that had the most distant retention times.)

Thus, my retention times and band widths for

the halocarbons follow:

Retention Time Band

Width
Freon-21 5 secs. 2 secs.
Freon-22 NA NA
CH2Cl2 50 secs. 14
secs.

diff. between retention times

The resolution is given as:
-----------------------------
average of peak widths

So the resolution of Freon-21 and CH2Cl2 is:

(50 secs. - 5 secs.) / ((2 secs. + 14 secs.)/2) =

5.625
 E). Table of digital multimeter resistances
for various volumes of dichloromethane:

Table Omitted

For 0.8 ml of an unknown gas sample, the

Resistance was 57.0 K Ohms.

According to my calibration curve, the volume

of CH2Cl2 in the unknown sample is given by:
Resistance = 53.178 + -8.7661*log10(volume)
so, 57.0 K Ohms = 53.178 - 8.7661 *
log10(x)

olving for x gives a volume of 0.366 ml

for CH2Cl2

Now I must find the partial pressure of

Dichloromethane at my room temperature with the
following equation:

A
log10P = (0.2185 x ----) - B
T

where, P = partial pressure of CH2Cl2

T = 23 C = 296 K (Room Temperature)
A = 7572.3 (specific for CH2Cl2)
B = 8.1833 (also for CH2Cl2)
7572.3
so, log10P = (-0.2185 x ----------) -
8.81833 = 2.59
296K

thus P = 392 torr = 0.516 atm

Finally, we can calculate the number of moles
of CH2Cl2 at my room temperature (296K) by using
the ideal gas law and preceding information:
 PV = nRT

where, P = Partial pressure of CH2Cl2

(from above)
V = Volume of CH2Cl2 (from calibration
curve)
R = 0.0821 L atm/mol K = 82.1 ml atm/mol
K
T = My room temperature in K (296 K)
n = number of moles of CH2Cl2

using the above equation gives:

(0.516atm)(0.366 ml) = n (82.1 ml atm/mol K)

(296 K)

solving, n = 7.77 x 10-6 moles of CH2Cl2

Discussion
I am surprised by the ease of constructing a
working small-scale GC with relatively common
“ingredients.” Using the device is also rather
simple after one gathers basic information about it
from the experiment.
Flow control is one of the first aspects of
my GC that I dealt with. By doing a simple test, I
find the flow rate of air to be 3.0 cm/sec. This
seems to be in line with the data of others. For
instance, another lab member (Anup Kharode) has a
flow rate of 4.8 cm/sec, which is relatively close
to mine. The discrepancy is explained by the
tightness of the packing of the column: A tightly
packed column would cause more air retention (slow
flow rate) and vice versa. Of course, a flow rate
that is much too fast would not allow for a
significant separation of the species involved.
My next investigations deal with the specific
retention times(tR) of various halocarbons, the
afore mentioned Freon-22, Freon-21, and
dichloromethane (CH2Cl2). My retention times for
Freon-21 and Freon-22 are remarkably close to each
other (both 7 secs.), while the tR of the CH2Cl2 is
much longer: 56 seconds. This leads me to inquire
about the cause of these retention times.
If I compare the boiling points of the
halocarbons, I find that a positive relationship
exists. The boiling points of the Freon gasses are
both below room temperature (Freon-22 = -40.8 C,
Freon-21 = 9 C, Room temp = 23 C), while the
boiling point of CH2Cl2 is markedly above room
temperature (40 C). This makes sense on a
molecular level because gasses with a low boiling
point would have a higher vapor pressure and faster
moving molecules than those with a high BP The
faster molecules would naturally move through the
GC column at a more rapid pace than the slower
molecules of a gas with a high boiling point.
Unfortunately, this reasoning fails to explain why
the tR values of the Freons are so similar.
However, when I compare my data with that of
another group, I find that my tR value for Freon-21
is much too low. In this new light, my molecular
explanation successfully explains the retention
times for all three sample gasses. It is also
interesting to note that molecular polarity
correlates with tR values. The more polar,
fluorine containing Freon molecules move through
the column much faster than the less polar
dichloromethane. This indicates that non-polar
substances have a greater affinity for the
stationary phase of my GC, and have greater
retention times.
The next part of the experiment involved
mixing the halocarbons and separating them through
GC. I attempted to separate all three halocarbons
in my GC device; however, only two of them
successfully disassociated. This is accounted for,
because the retention times of my Freon molecules
are very similar, and one would not notice a
separation of these two. From this, I can conclude
that the first halocarbon indication was either
Freon-22 or Freon-21 (most likely Freon-21 because
of the exact tR and band width measured), and the
second indication was CH2Cl2. The retention times
of these halocarbons are very similar to the
heterogeneous ones.
In doing this section, I noticed that band
widths greatly increase as retention times do. The
simplest way to explain this is to compare
molecules to cards in a deck: A deck dropped right
in front of one’s feet will land in a condensed
pile. On the other hand, a deck thrown across the
room will end up as individual cards covering a
large area. In other words, the thrown deck has
more time to spread itself out (gain entropy) than
the dropped one does, just as the slower moving gas
has a greater time to spread out than the faster
one.
With further data analysis, I have calculated
the resolution between Freon-21 and CH2Cl2 to be
5.625. This has led me to inquire into the nature
of resolution, for instance, what does a resolution
of 1.0 indicate. Basically, a resolution of 1.0
means that the difference between retention times
equals the average band width.Further, as the
diagram clearly shows, a resolution of 1 indicates
that the bands come quickly in succession.
The final section in this experiment is a
quantitative analysis of GC with photodetection.
The photodetection device is a cadmium sulfide
light detector whose internal resistance increases
with increased exposure to light. Using a digital
multimeter, I was able to measure resistance and
obtain a logarithmic calibration curve for
resistance vs. volume of CH2Cl2 (as increased
CH2Cl2 volume causes an increase in light radiation
by the Beilstein detector). Using this
information, along with known information about
partial pressure and the ideal gas law, I was able
to find the total number of moles of CH2Cl2 in an
unknown gas solution.
This use of GC leads to several very
interesting applications in the scientific world.
Most obviously, one could learn amounts of
halocarbons in the air. This is very important in
light of the danger that halocarbons (especially
chlorofluorocarbons) pose to the environment.
Naturally, more comprehensive research is needed to
fully examine this subject; however, our
rudimentary small-scale GC device provides evidence
that this research is indeed possible.
As far as accuracy and precision are
concerned, I generally feel confident in my
research. Precision is ensured by the controlled
flow rates of the natural gas tap and the high
quality of instruments, such as the digital
multimeter. The only significant problem with my
data occurs in my retention times for Freon-21.
When comparing my data to that of other lab groups,
I find that my retention times for this substance
are much too low. I can only speculate about the
cause of this problem as several trials yielded the
same results, but perhaps my sample of Freon-21 was
somehow contaminated.

Conclusion
Although the Chem 15 small-scale GC device is
a rather simple instrument to construct and use, it
provides many insights into the greater
applications of industrial GC. I have learned much
about the workings of such an instrument, including
the importance of good flow control, why
separations occur, and basic qualitative and
quantitative detection. For instance, separations
occur because of many factors, including molecular
speed (boiling point) and polarity (or affinity for
the stationary phase).
In addition, I have learned about the
potential applications of GC, such as detection of
CFC’s in air or separation of illegal substances
from one’s blood in drug testing. With a little
thought, an endless array of possibilities exist
for GC use. For further research, I would suggest
experimentation with commercial GC instruments to
both compare results with our hand-made variety and
to gain additional knowledge about gas
chromatography and its uses.

References
Blinn, Ryan and Bonner, Pat. Lab neighbors.

Braidic, Jim. Lab Partner.

CRC Handbook of Chemistry and Physics, 3rd ed.

David R. Lide, editor-in-chief. Ann Arbor, MI:
CRC Press. 1992.

Jenner, David J. Chromatography as an Analytical

Tool. New York: Nuggett Publishing. 1986.

Kharode, Anup. Chem lab associate.

Thompson, Stephen. Chemtrek: Small-scale

Experiments for General Chemistry. PSU
Version. Englewood Cliffs, NJ: Prentice
Hall. 1996.

Back to Works

Taken from personal.psu.edu