Chromatography Forum

Hi

I use a 2000ug/L Phenols standard from Restek for calibrating my method, with methylene chloride as a solvent and my problem is that I have trouble getting a nice regression line. I am certain it is not a fault due to my technicality. I changed the syringe I work with and it still did not work so I tried to work by weight. I weighed the syringe with a septum at the tip, the drew my standard and placed the septum again and weighed the syringe again to obtain a mass difference and when calculating back to the volume I got a bigger volume than the supposed one.

I just need to confirm with anyone who works with the same standards, is this a common problem? Are phenol far less denser? Should I change my solvent?

Thanks for your help in advance.

Hi Gugu

Did you take temperature into account when calculating volume from mass ?

How big was the discrepancy, what kind of syringe did you use, and what is the readability of the balance ?

The concentration of phenol sis so low that you would have serious trouble weighing precisely enough to see their effect.

When you say that you cannot get a good calibration - what is wrong with it ?, non-linear, scatter around the line ? How are you doing the dilutions ?

The septum on the needle tip is good practice. See; APPS, P.J. and ARCHER, M. 2010. Evaluation of the source of bias caused by losses of solvent vapour during sample preparation. Journal of Accreditation and Quality Control 15: 171–180
https://www.researchgate.net/publicatio ... reparation

"Did you take temperature into account when calculating volume from mass ?

How big was the discrepancy, what kind of syringe did you use, and what is the readability of the balance ?

The concentration of phenol sis so low that you would have serious trouble weighing precisely enough to see their effect.

When you say that you cannot get a good calibration - what is wrong with it ?, non-linear, scatter around the line ? How are you doing the dilutions ?"


Hi Peter

I did not consider temperature. I use a 10 uL manual injection micro-syringe and for a supposed 5 uL transfer I get a 5000 uL which is impossible.
NB: I took my density as 0.002g/L for a mass of 0.0102g.

The regression is bad, especially for Phenol. Instead of a coefficient of determination of 0.9999, I get something far less for phenols and the other phenolic compounds are at least around 0.990.. or 0.994.
I'm struggling to insert the pictures at the moment but I hope you get the idea.

"Did you take temperature into account when calculating volume from mass ?

How big was the discrepancy, what kind of syringe did you use, and what is the readability of the balance ?

The concentration of phenol sis so low that you would have serious trouble weighing precisely enough to see their effect.

When you say that you cannot get a good calibration - what is wrong with it ?, non-linear, scatter around the line ? How are you doing the dilutions ?"


Hi Peter

I did not consider temperature. I use a 10 uL manual injection micro-syringe and for a supposed 5 uL transfer I get a 5000 uL which is impossible.
NB: I took my density as 0.002g/L for a mass of 0.0102g.

The regression is bad, especially for Phenol. Instead of a coefficient of determination of 0.9999, I get something far less for phenols and the other phenolic compounds are at least around 0.990.. or 0.994.
I'm struggling to insert the pictures at the moment but I hope you get the idea.

Hi Gugu

The density of dichloromethane is 1.33 g/ml - where did you get 0.002g/L, which is less than the density of helium, are you confusing concentration of phenol in the solution with the density of the liquid whose volume you want to calculate ? Your mass of 0.0102 g gives a volume of 7.7 ul, so there is still a discrepancy.

A low r-squared can be due to scatter around the line, or to a non-linear relationship between quantity and signal. If you do replicate analyses hpw good is the repeatability ?

Peter

Hi Gugu

This may or may not be helpful

A technique that I devised for accurate (in terms of knowing the final dilution) dilution of a concentrated bought in standard solution was

1. Take a 5 ul or cut down 10ul disposable glass capillary microtube.
2. Tare it
3. Briefly touch one end of the tube into the standard to draw up some standard solution ( 3-5ul) by capillary action. If it is too full touch the tip on a piece of tissue to reduce the amount
4. Reweigh - place it on the pan leaving a small portion of the microtube sticking out from the edge of the pan so that you can easily pick it up after weighing - surprisingly the exposed surface area at each end of the aliquot in the microtube is so small that you can comfortably reweigh without loss of solvent/weight by evaporation during the weighing.
5. After weighing - immediately pick up and drop the microtube into a 10ml or 20ml volumetric flask containing solvent then make up to the mark, stopper and shake the flask

I have successfully used this with BP as low as hexane and really does work.

All of the sample is transferred and the displacement volume of the glass itself is so low that it does not significantly affect the volume of your new standard in the 10ml flask

Regards

Ralph

Hi Ralph

I am a huge fan of using mass rather than volume.

I looked specifically at evaporative losses from syringe needles in the paper that I referenced in a previous reply. From a 0.41 mm internal diameter needle hexane evaporates at about 1 ug / s, so if you have 3ul in the capillary, losing 1 ug/s at each end you get about 1% bias for every 10s that weighing takes. If the analytes are non-volatile they stay in the tube so a slow weighing underestimates the quantity of analytes that you add to the volumetric flask.

Peter

Hi

I use a 2000ug/L Phenols standard from Restek for calibrating my method, with methylene chloride as a solvent and my problem is that I have trouble getting a nice regression line. I am certain it is not a fault due to my technicality. I changed the syringe I work with and it still did not work so I tried to work by weight. I weighed the syringe with a septum at the tip, the drew my standard and placed the septum again and weighed the syringe again to obtain a mass difference and when calculating back to the volume I got a bigger volume than the supposed one.

I just need to confirm with anyone who works with the same standards, is this a common problem? Are phenol far less denser? Should I change my solvent?

Thanks for your help in advance.

What are the concentrations of your calibration solutions?

If you are trying to cover too much range from low to high, your calibration could be slightly quadratic, this is not uncommon.

If your ng on column is exceeding the upper limit of your column's capacity you will also have problems with linearity. Even the best made standards can still suffer from the physical limits of the instrument itself. If your calibration line looks like the points are scattered it could be from problems with the standards, if it looks curved, it is probably an instrument linearity problem.

If the syringe is off slightly on calibration and you only use the one syringe, then you would still most likely get a straight calibration curve but with an offset that you may or may not be able to measure, unless your syringe is more accurate at one reading than at another, say 1ul is 1ul but 5ul is actually 6ul and 10ul is actually 12ul, then you would probably see a curved line. But most micro syringes are more accurate than that, though you do need to consider if the marking take into account the volume of the needle or not.

Hi Peter

Thank you - That is a really interesting paper

Regards

Ralph

Hi guys

Sorry guys for the late reply. Too many holidays and too much work inbetween.

"The density of dichloromethane is 1.33 g/ml - where did you get 0.002g/L, which is less than the density of helium, are you confusing concentration of phenol in the solution with the density of the liquid whose volume you want to calculate ? Your mass of 0.0102 g gives a volume of 7.7 ul, so there is still a discrepancy.

A low r-squared can be due to scatter around the line, or to a non-linear relationship between quantity and signal. If you do replicate analyses hpw good is the repeatability ?

Peter"

Peter, I took the density of the standard as 0.002g/mL from it's quantity of 2000ug/mL.

Hi GOM

"This may or may not be helpful

A technique that I devised for accurate (in terms of knowing the final dilution) dilution of a concentrated bought in standard solution was

1. Take a 5 ul or cut down 10ul disposable glass capillary microtube.
2. Tare it
3. Briefly touch one end of the tube into the standard to draw up some standard solution ( 3-5ul) by capillary action. If it is too full touch the tip on a piece of tissue to reduce the amount
4. Reweigh - place it on the pan leaving a small portion of the microtube sticking out from the edge of the pan so that you can easily pick it up after weighing - surprisingly the exposed surface area at each end of the aliquot in the microtube is so small that you can comfortably reweigh without loss of solvent/weight by evaporation during the weighing.
5. After weighing - immediately pick up and drop the microtube into a 10ml or 20ml volumetric flask containing solvent then make up to the mark, stopper and shake the flask

I have successfully used this with BP as low as hexane and really does work.

All of the sample is transferred and the displacement volume of the glass itself is so low that it does not significantly affect the volume of your new standard in the 10ml flask

Regards

Ralph"

Thank you, I will try that to see how it turns out for me.

Hi James_Ball

"What are the concentrations of your calibration solutions?

If you are trying to cover too much range from low to high, your calibration could be slightly quadratic, this is not uncommon.

If your ng on column is exceeding the upper limit of your column's capacity you will also have problems with linearity. Even the best made standards can still suffer from the physical limits of the instrument itself. If your calibration line looks like the points are scattered it could be from problems with the standards, if it looks curved, it is probably an instrument linearity problem.

If the syringe is off slightly on calibration and you only use the one syringe, then you would still most likely get a straight calibration curve but with an offset that you may or may not be able to measure, unless your syringe is more accurate at one reading than at another, say 1ul is 1ul but 5ul is actually 6ul and 10ul is actually 12ul, then you would probably see a curved line. But most micro syringes are more accurate than that, though you do need to consider if the marking take into account the volume of the needle or not."

My calibration standards are 0.1 , 0.4 , 1, 2, 10 ppm respectively and my points are scattered.

Hi guys


Peter, I took the density of the standard as 0.002g/mL from it's quantity of 2000ug/mL.

My calibration standards are 0.1 , 0.4 , 1, 2, 10 ppm respectively and my points are scattered.

Concentration, in terms of mass per unit volume, is nothing at all to do with density unless you have a pure substance. You have a dilute solution - its density is the density of dichloromethane. If you want to calculate the volume of a liquid from its mass you have to use its density, not the concentration of what is dissolved in it. In this instance the concentration of phenol is so low that it will not make the density of the solution differ from the density of dichloromethane.

Your range covers 2 orders of magnitude. This should be OK. Unless you have a large split ratio the lower end is easily detectable and the upper end is well within overload limits unless you are doing large volume injections.

The scatter possibly points to a repeatability problem. You have to solve that before there is any point in worrying about linearity because you will not get good linearity if poor repeatablity is making the points scatter around the line.

First you need to do five replicate injections of the 1 "ppm" standard and calculate the relative standard deviation of the peak areas.

By the way, if you are working with both mass and volume using non-defined units like "ppm" is likely to cause confusion. Express everything in terms of mass per unit volume, then we all know what you are talking about.

Peter

...Peter, I took the density of the standard as 0.002g/mL from it's quantity of 2000ug/mL.
...

O tempora, o mores ! These are basics.

Hi

Please stick with weight.

Perhaps I was lucky in having a cool lab and weighing quickly the 6-8 ul of standard in methanol or ethanol using the glass capillary microtube method that I posted earlier.

I noticed no loss in weight at the 4th decimal point during the weighing procedure.

Using the syringe/septum method you need to ensure that all of the syringe content is completely rinsed into your dilution solvent

regards

Ralph

Hi Peter

"Concentration, in terms of mass per unit volume, is nothing at all to do with density unless you have a pure substance. You have a dilute solution - its density is the density of dichloromethane. If you want to calculate the volume of a liquid from its mass you have to use its density, not the concentration of what is dissolved in it. In this instance the concentration of phenol is so low that it will not make the density of the solution differ from the density of dichloromethane.

Your range covers 2 orders of magnitude. This should be OK. Unless you have a large split ratio the lower end is easily detectable and the upper end is well within overload limits unless you are doing large volume injections.

The scatter possibly points to a repeatability problem. You have to solve that before there is any point in worrying about linearity because you will not get good linearity if poor repeatablity is making the points scatter around the line.

First you need to do five replicate injections of the 1 "ppm" standard and calculate the relative standard deviation of the peak areas.

By the way, if you are working with both mass and volume using non-defined units like "ppm" is likely to cause confusion. Express everything in terms of mass per unit volume, then we all know what you are talking about.

Peter"

Oh okay I did not know about the density because looking at the SOP that was compiled by the previous analyst I figured that he used that 0.002 g/mL as density to get the volumes he calculated, which I can strongly support that he did. I will re-calculate mine then.

Hi Ralph

"Hi

Please stick with weight.

Perhaps I was lucky in having a cool lab and weighing quickly the 6-8 ul of standard in methanol or ethanol using the glass capillary microtube method that I posted earlier.

I noticed no loss in weight at the 4th decimal point during the weighing procedure.

Using the syringe/septum method you need to ensure that all of the syringe content is completely rinsed into your dilution solvent

regards

Ralph"

I am considering weight as it is much more accurate than volume thanks.

Gugu

Ralph

"Please stick with weight.

Perhaps I was lucky in having a cool lab and weighing quickly the 6-8 ul of standard in methanol or ethanol using the glass capillary microtube method that I posted earlier.

I noticed no loss in weight at the 4th decimal point during the weighing procedure.

Using the syringe/septum method you need to ensure that all of the syringe content is completely rinsed into your dilution solvent

regards"

I would like to know though, when converting your mass, did you also use the density of your solvent?

Kind regards
Gugu

Hi

Please stick with weight.

Perhaps I was lucky in having a cool lab and weighing quickly the 6-8 ul of standard in methanol or ethanol using the glass capillary microtube method that I posted earlier.

I noticed no loss in weight at the 4th decimal point during the weighing procedure.

Using the syringe/septum method you need to ensure that all of the syringe content is completely rinsed into your dilution solvent

regards

Ralph

Hi Ralph, with methanol or ethanol which have quite low vapour pressures and a high ratio of vapour to liquid volume your method will work well. But (among her other problems) Gugu is using dichloromethane which has a high vapour pressure and a lower ratio of vapour to liquid volumes. She will be better off using a syringe, but on the evidence so far I do not think that the source of her problem is actually with the measuring of quantity of standard.

Usual practice with a syringe is to load it, cap the needle, weigh, dispense, recap, reweigh. The syringe is not rinsed (with solvent ?) into the solution. What left the syringe is the loss in weight - what ended up in the solution might be slightly different.

In Gugu's case she can just as well use the syringe to measure volumes directly without weighing - her problems will not be solved by the increase in accuracy and precision that weighing offers.

Peter