[jkulier]

Hi there,

I am just wondering what procedure I would take in preparing 1, 3, 5, and 10 ppm standard solutions from a standard solution that is 100ppm?

Would subsequent dilutions be the proper way?

Thanks in advance.

[JGK]

It's no different than the dilutions you would make if the units were µg/mL and you needed to make the same STDs from a 100 µg/mL solution

[wpenrose]

Sequential dilutions result in increasing error.

Better to start with a string of four 100 mL volumetric flasks, pipet in 1, 3, 5, 10 mL, and make each to 100 mL.

[enahs]

Sequential dilutions result in increasing error.

No.

There is more error in going from 100 to 5 then from 10 to 5, just on the physical limitations of accuracy of instruments. You also will likely have to use different instruments with different accuracies and calibrations.* If you do sequential and or serial, and thus keep the same instruments, any error can easily be spotted and adjusted for. If you make 5 different solutions, it is much harder statistically to say "this one is bad", and so you have to start all over.

Sequential is the way to go. Serial would be even better.

There are tons of other reasons. Any highly regulated lab only does sequential (more likely, only serial dilutions); and for good reasons.

* Under typical conditions.

[wpenrose]

"Sequential is the way to go. Serial would be even better. "

That would be true in order of magnitude dilutions, as in 0.1, 0.01, 0.001, etc. For smaller intervals, you use pipets of appropriate size for each dilution.

In sequential dilutions, errors are propagated and compounded along the series. A serious error in the first or second will result in error in all the following dilutions.

In single-step dilutions, you compound two errors for each one. In sequences, it's two for the first dilution, four for the second, six for the third, and so on. With typical errors of, say, 0.2%, the error in the final dilution can be quite large.

Dangerous Bill

[enahs]

The error can be detected and easily adjusted for. If you use different pipets and volumetric flasks, this is not as easy. It would take 10's to 100's of experiments (depending on how many solutions you make) to find your error and correct for.

With sequential dilutions it is quite trivial to account for the systematic error.

The other advantage of the sequential dilutions is, if there is a systematic error, it is likely to be in all samples. This means it does not change the trend of your data, just not the "correct" value. If you do it non-sequentially, it is quite easy to ruin any possible trend, because you might have errors in measurements in both directions.
 

[wpenrose]

"The other advantage of the sequential dilutions is, if there is a systematic error, it is likely to be in all samples. This means it does not change the trend of your data, just not the "correct" value. If you do it non-sequentially, it is quite easy to ruin any possible trend, because you might have errors in measurements in both directions."

Your reasoning leaves me scratching my head.

In one of the analytical classes I taught, I used this very example, individual vs serial dilutions, as an illustration of propagation of random normal error.

DB