Help with G.C basics

[Gnome]

Hi,
could some of you help me with some very basic GC basics?
My plan is to get an old GC ( not GLC) and use it to analyze gases,
mainly exhaust/combustion products, and normal atmospheric
outgassing from plastics and similar materials.
The goal is to identify those gases/chemicals.
Note that this is not tests for specific chemicals... it is search/
tests
for what is initially unknown chemicals !

1) Lets take the HP5890, with a TCD or FID,
is it suitable for the task mentioned above?

2) What would be the most suitable detector for my intended use?

3) Will a GC, i.e the HP5890 accept sample in gaseous form as
mentioned above?

4) Outgassing from plastics at room/low temperatures is in trace/ppm
range... can the HP5890 handle that?

5) The detector output. If I have understood things correctly the
output from the detector is essentially just a time domain analogue
image
of the gas distribution in the column... carrier gas = no detector
output,
no carrier gas (sample gas (chemical)) = detector output... correct?

6) What does the detector output look like? Is there one sharp peak
for each clean (same molecule) chemical? Or will there be a multiple
peaks
fingerprint type of output for each molecule type?

7) In either case my samples will contain a lot of different unknown
chemicals, and there is likely to be a 'forest' of peaks... how do we
make sense out of all that data?

8) Lets say we have a printout or datastream with some peaks
and basic data.... then what? How is this data used to identify the
chemicals? I assume some software with database search & match?
I need to stick a chemical name to the peaks!


Thanks for any help on this.
E.K

[David Stranz]

Post by Gnome
Hi,
could some of you help me with some very basic GC basics?
My plan is to get an old GC ( not GLC) and use it to analyze
gases, mainly exhaust/combustion products, and normal
atmospheric outgassing from plastics and similar materials.
The goal is to identify those gases/chemicals.
Note that this is not tests for specific chemicals... it is
search/ tests
for what is initially unknown chemicals !

Geez. Where to start? Your severe lack of even basic knowledge
makes it hard to decide.

Post by Gnome
1) Lets take the HP5890, with a TCD or FID,
is it suitable for the task mentioned above?

You need to first question whether GC in the first place is a
suitable technique before focusing on a specific instrument.

GC by itself is nearly useless for -identification- of -unknowns-.
There are specific detectors (i.e. ECD or chemiluminesence) that
can give information about whether certain elements or classes of
elements are present, but this still gives you no information as to
chemical identity.

GC is a good quantitative technique. Once you know what an unknown
substance is, know its retention time and response factors relative
to a standard reference material, then you can calculate the amount
present in the sample.

Post by Gnome
2) What would be the most suitable detector for my intended use?

A mass spectrometer. The fragmentation pattern produced by a mass
spectrometer is a fingerprint of chemical identity. Most of the
compounds you propose to invistigate are well-studied and are
present in mass spec databases. A search of the mass spectrum of a
GC peak against a database would likely give you the identity of
the analyte. However, library searching ALWAYS gives an answer,
whether the analyte spectrum is in there or not. The search simply
returns the library spectrum that best matches the analyte, and you
have to make the interpretation of whether that match is correct.

Post by Gnome
3) Will a GC, i.e the HP5890 accept sample in gaseous form as
mentioned above?

You want to use a purge and trap protocol, where the gases are
sampled for some period of time and trapped on an adsorbent. This
is then heated in the GC inlet to desorb the trapped compounds.

Post by Gnome
4) Outgassing from plastics at room/low temperatures is in
trace/ppm range... can the HP5890 handle that?

Via the purge and trap method described above. You will have to
calculate based on the flow rate, trapping time and efficiency, and
response factor of the GC what the original concentration was.

Post by Gnome
5) The detector output. If I have understood things correctly
the output from the detector is essentially just a time domain
analogue image
of the gas distribution in the column... carrier gas = no
detector output,
no carrier gas (sample gas (chemical)) = detector output...
correct?

Most GC instruments have an analog detector out signal you can
monitor. The magnitude of the signal is proportional to the amount
of analyte present at any time and the response factor of the
detector to the analyte. Every analyte has a difference response
factor, which is why you need reference standards if you want
quantitative information.

GC columns may exhibit "column bleed". As the temperature ramps
up, the stationary phase can bleed off and contributes a gradually
increasing background signal. The signal from the analytes is
superimposed on this background.

The carrier gas should not give any signal; if it does, it will
undoubtedly swamp the analyte signal and render the method useless.
Thus you cannot use any old source of carrier gas, but must use
highly pure (and expensive) GC-grade gases.

Post by Gnome
6) What does the detector output look like? Is there one sharp
peak for each clean (same molecule) chemical? Or will there be a
multiple peaks
fingerprint type of output for each molecule type?

Ideally, there is one peak per analyte. If your sample is too
complex or the analytes too similar in retention time, they cannot
be separated into clean peaks. If you overload the column, a
single analyte may split into multiple peaks because the stationary
phase simply can't hold all of it at one time.

Post by Gnome
7) In either case my samples will contain a lot of different
unknown chemicals, and there is likely to be a 'forest' of
peaks... how do we make sense out of all that data?

If you know the identities of the peaks, their response factors,
and so forth, you can calculate concentrations of each component.

If you don't know the above but have reasonably good
reproducibility between samples, then you can compare samples using
pattern recognition and chemometric techniques. This will at least
tell you if two samples are the same within specified error limits,
tell you which peaks are present (or up or down in amount) in one
but not another, but still gives you not clue as to identity.

Post by Gnome
8) Lets say we have a printout or datastream with some peaks
and basic data.... then what? How is this data used to identify
the chemicals? I assume some software with database search &
match? I need to stick a chemical name to the peaks!

There are a few databases of retention time information on specific
classes of compounds. However, just because your compound has the
same retention time under the same chromatographic conditions as
one in a reference database, that does not at all imply that you
have identified the material. It could be (and most likely would
be) purely coincidence.

Post by Gnome
Thanks for any help on this.
E.K

I'd strongly suggest to you that before you waste considerable
time, money, and effort trying to do this experiment that you spend
a whole lot of time in the library reading up on what you can and
can't do with a GC, and what you really need to do to answer the
question you are posing.

David

[Gnome]

David, thanks for the lengthy answer.
This was not what I wanted to hear... BUT I'm sure you're right on all
points
so you've probably saved me from doing a 'bad' mistake.
A MC is out of the question... Maybe I should just set the jobs out
to
well equipped and qualified experts.

E.K



................................

Post by David Stranz
I'd strongly suggest to you that before you waste considerable
time, money, and effort trying to do this experiment that you spend
a whole lot of time in the library reading up on what you can and
can't do with a GC, and what you really need to do to answer the
question you are posing.
David

[David Stranz]

Post by Gnome
David, thanks for the lengthy answer.
This was not what I wanted to hear... BUT I'm sure you're right
on all points
so you've probably saved me from doing a 'bad' mistake.
A MC is out of the question... Maybe I should just set the jobs
out to
well equipped and qualified experts.
E.K

Well, you have guts enough to think about trying, and are smart
enough to ask before you try.

Possibly you could find an environmental analysis lab that could do
these analyses, but they won't be cheap. These are typical analyses,
as far as I know. The manufacturer of your furniture might actually
have the information you want if you could find the right person to
talk to, or possibly there is information available through the
Consumer Product Safety Commission, EPA, the organization that
publishes Consumer Reports (forget the name), or state-level
organizations. Some of these things might be covered under
California's Proposition 65 law, which regulates hazardous materials
in places of business, products, etc.

[Mr. Dawg]

Post by David Stranz

Post by Gnome
David, thanks for the lengthy answer.
This was not what I wanted to hear... BUT I'm sure you're right
on all points
so you've probably saved me from doing a 'bad' mistake.
A MC is out of the question... Maybe I should just set the jobs
out to
well equipped and qualified experts.
E.K

Well, you have guts enough to think about trying, and are smart
enough to ask before you try.
Possibly you could find an environmental analysis lab that could
do these analyses, but they won't be cheap. These are typical
analyses, as far as I know. The manufacturer of your furniture
might actually have the information you want if you could find
the right person to talk to, or possibly there is information
available through the Consumer Product Safety Commission, EPA,
the organization that publishes Consumer Reports (forget the
name), or state-level organizations. Some of these things might
be covered under California's Proposition 65 law, which
regulates hazardous materials in places of business, products,
etc.
These

--- aren't ---

Post by David Stranz
typical analyses, as far as I know.

[Gnome]

David,
I got a quote from a lab for a GC/MS test with ID/QTY of specimens in
a fabric sample, $1800,-
What do you think... does it sound like a normal charge for such a
test ?

E.K

Post by David Stranz
Possibly you could find an environmental analysis lab that could do
these analyses, but they won't be cheap. These are typical analyses,
as far as I know. The manufacturer of your furniture might actually
have the information you want if you could find the right person to
talk to, or possibly there is information available through the
Consumer Product Safety Commission, EPA, the organization that
publishes Consumer Reports (forget the name), or state-level
organizations. Some of these things might be covered under
California's Proposition 65 law, which regulates hazardous materials
in places of business, products, etc.