semi-quantitative analysis

[amrnasser87]

what does semi quantitative analysis mean?
how accurate is it?
is it applicable for all GC detectors?

[Peter Apps]

"semi-quantitative" means only that the analyst couldn't be bothered to do a quantitative analysis properly

Peter

[DSP007]

semi quantitative in chemistry - as plus-minus kilometr in geography.

Imagine that You are standing in the middle of the Sahara, and asking the Bedouin "Are there any water ?!" And they answer of his
1) hand somewhere- Alla akbar , many water is, only three days journey by camel ! .
And then everything will depend on what water there - if there is flowing river the size of the Nile- is something you find her and save. But if there is a small put, covered with reeds out the wind, I condolences with you ...
Its quality analisis
2) Open compass and map. " Yes of course - 700 kilometers on the nort-east , is a small pit" its a semi-quantitative analisis
3) Open compass and map " Yes of couse - 78.82 degrees on north-east , 689,3 km is a small 10 m put , 10 gallons water in day"its quantitative analisys.

[Peter Apps]

100 km plus or minus 1 kilometer is just as quantitative (though less precise) than 100 km plus or minus 1 micron.

Perhaps "semi-quantitative" is a euphemism for imprecise, or for guesswork with SI units ??

Peter

[amrnasser87]

thanks for reply

but i am not joking semi quent. means the estimation of
an uncalibrated compound’s amount or concentration
by comparison to a calibrated, known compound
that has been injected and characterized
under the same method of analysis

kindly, if any one has such information about this topic he sould share it

again thanks for reply and best regards
Amr

[Peter Apps]

I am not joking either - if you have no calibration for the compound of interest then you cannot quantify it. To pretend that you can by calling such an analysis "semi-quantitative" is, bluntly, dishonest.

This does not mean that you cannot quantify e.g. octane, from the calibrated response of a non-selective detector such as an FID to another linear alkane, say nonane, because the FIDs response to this class of compounds is well understood to be non-selective (although not to the extent that is aften claimed). But detectors with the FIDs lack of selectivity are very few and far between. It should also be possible to establish a conversion factor between detector responses to different compounds, to allow the calibration from one to be applied to the other, but the results are then quantitative (with a precision and accuracy degraded by the uncertainty of the conversion factor) not semi-quantitative.

Peter

[amrnasser87]

could you open this link . it is tailking about semiquent. but in GC/MS . it is new software approches to estimate compund quantities


http://www.chem.agilent.com/Library/tec ... 4997EN.pdf

[krickos]

Hi

To put it simply (hopefully), semi quantative is like a pH stick compared to a pH-meter (quantative) or other similar semi quantative test.

Also in my opinion certain so called limit tests (certain visual colur comparison reactions for example)are also semi quantative, you can get three result, same as standard, more than standard or less than standard.

But as Peter stated, unless you can relate the uncalibrated peak to a calibrated peak by relative response factors etc, it is only a qualified guess in best case.

then again there are exceptions or "in between" from the rule I guess, thinking on the GC-AED set up for instance, not sure if anyone makes AEDs nowdays though. Was about to by one some 10 years ago but priorities changed and then Agilent stop manufacturing it I think.

[Peter Apps]

could you open this link . it is tailking about semiquent. but in GC/MS . it is new software approches to estimate compund quantities


http://www.chem.agilent.com/Library/tec ... 4997EN.pdf

I am not sure what you think that the Agilent flyer adds to the argument - the drawbacks that they point out are certainly valid. That you can get software to do something does not make it good practice, it just makes it easier, and often leads to its uncritical application in circumstances where it is not appropriate - witness the compounds "identified" by automated MS library searches.

A result is either quantitative or it is not - semi-quantitative is equivalent to the cliched "a little bit pregnant".

A pH measurement with a dipstick gives you are range of pHs - it is quantitative.

A colour comparison go - no go test gives you a comparative result - non quantitative.

Peter

[Don_Hilton]

My encounters with semiquantitative involve measurements that give numbers that cannot be validated - there may be no standard or the procedure is not reproducible. But, the technique can give a general idea of bigger or smaller.

If I am standing in my basement and the water is knee deep, I start bailing the water. When it is down to half the height of my knee is the water half gone? With all the stuff stored in the basement, there is probably more volume filled with stuff closer to the floor - so more than half the water is gone. But I may also be standing where the floor is higher as I make the second measurement - so the water level has not been lowered as much as I have estimated.

I would not call semiquantitative numbers dishonest. But, the people generating the numbers and those using the numbers must understand the nature of the numbers and the risks involved in using them. Such numbers can be useful when there are large differences between the existing situation and the desired situation. The analytical technique can be refined while changes to the situation are being made -- rather than waiting for an analytical method to be developed.

Thus, using my semiquantitative esitmate for determining the rate at which I am draining a flooded basement, I can give a good estimate of whether this will be done before guests are expected to arrive for dinner.

[Peter Apps]

So doesn't "semi-quantitative" simply mean approximate ? And if an approximate value is all that you need to get the job done, then go for it. In your example you take two quantitative measurements - and get results of 1 knee deep and 0.5 knee deep. I do not see where "semi" comes into it.

I suspsect that "semi-quantitative" is expected to sound better than approximate, rough, or ball park, which is really what uncalibrated, extrapolated etc results are to greater or lesser degrees.

Peter

[Don_Hilton]

In the example the value obtained can not reasonably be attached to a standard unit, such as liters.

The cynic in me notes that half the world is marketing - and who would trust a scientist who gives approximate measurements when we can sell to management or others "Semi-auantitative?" Semi-quantitative is so much more scientific because quantitative (which means we scientifically measured it) is half the word!

[Peter Apps]

Don, that sums it up completely ! After all, who would buy software add-ons called ApproxImate ?

Peter

[AICMM]

All,

Just thought I would comment: Spectroscopy magazine: "Semiquantitative Analysis in ICP-OES and ICP-MS." Haven't read the article yet but I thought it kind of intriguing considering this thread.

Best regards,

AICMM

[krickos]

Hi

Yes seems like there more related to ICP analysis and XRPD.

However as I have a background in medical analysis (clinical chemistry)as well the expression also exist there as: yielding an approximation of the quantity or amount of a substance; falling short of a quantitative result.

So where do you draw the line?