[redskytonight820]

Hey all, this is my first post. I'm currently conducting an experiment, and I'm trying to figure out whether or not my products are what they are... but I'm new to HPLC, and so, I don't really know if my peaks are completely resolved.

I'm using a phenomenex column (50 m) w/ an I.D. of 0.20 mm. Flow rate = 1 ml/min. So far, the two peaks, correspond to two compounds: Mercapto-ethanol and aniline. k' (mercapto ethanol) = 2.0, k' (aniline) = 4.5. So the peak area of the aniline peak (from the digital analysis) is 35891 mV x sec and the peak height is 3851 mV.

So... I have no idea if these peaks are completely resolved. I tried calculating resolution:

retention time = (column length / (flow rate * (1+k') ) )

N = (retention time / peak area / height) 5.545

resolution = ((sqrt N) / 2) (k'a-k'b / (k'a+k'b+2))

and so... from the calculated resolution, how can i determine if my peaks are completely resolved?

[Telamond]

Please correct me if I'm wrong, but I believe the resolution is calculated through:

R = [tR(x) - tR(y)] / ([Wb(x) + Wb(y)]/2)

tR(x) = retention time for analyte X
tR(y) = retention time for analyte Y
Wb(x) =base width of analyte X
Wb(y) =base width of analyte Y
And if your peaks are completely resolved then R > 1,5.

Might need confirmation from someone who's passed the first course. I have my exams in 14 days. :S

[redskytonight820]

There are several equations for Rs.

Well anyway, if Rs > 1.5, then my peaks are resolved? That's good to know! My resolution came out around 6 something.

But what are the limits of this? Isn't it that if the resolution is too great, then the compounds are merely flying through the column?

So yes, thank you for the note about Rs having to be greater than 1.5. I just recalled that the Rs of the baseline is 1.5... So that's the reason why it has to be greater than 1.5 to be resolved correct?

[Telamond]

If the R = 1,0 then they're 98% separated, anything under R < 1,0 is badly separated (almost no retention). Anything over 1,5 is fine.

But I guess much of this depends on how you integrated your bases as well.

Again, you probably want to double check this. This is just what I've learned.

[Arkcon]

Your class may have specifications that require you to determine if you have the resolution you need.  However, in industry, resolution required is not a hard and fast rule, manually calculated, then applied to a chromatogram on a pass/fail basis.

Resolution, R_s, is defined in the reference I have as t₂ - t₁/0.5(w₂ - w₁).  Although I generally rely on the chromatographic software to compute it for me, rather than doing it by hand.  So I'd be using Telamond:'s definition in preference to redskytonight820:'s which can also be used to determine resolution, and is based on capacity, plates, and separation factor.  The latter coefficient is abbreviated in the formula as alpha, which contains t₂ - t₁ in it, so you're both correct.  Whew, like I said, I rely on the software to do this for me.

Some people are of the opinion that any resolution greater than 1.0 is adequate, since that requires that they are farther apart than they are wide.  However, that may not be adequate.  According to my reference* a 1.0 resolution represents a 5% overlap between adjacent peaks, an R_s of 1.5 is called "baseline" resolution, with an overlap of 1%.

All of the proceeding assumes, the peaks are approximately equal in size, and are perfectly symmetrical Gaussian curves.  Which, often, is not the case.  Many people hold themselves to a standard of R_s no less than 1.7, to account for the peaks tailing or fronting.  Although exceeding 2.0 is pretty difficult, for separations involving more than a few peaks.

There are also other standards people may hold their separation to.  A particular N, for a peak.  Or a particular k'.  Depending on the development of the method, and industry requirements.

Long story short, either you have a value you have to conform to, and your separation meets it, or you haven't.  But I don't know how to find out what's good enough, based only on a forum post.

*The Practice of Modern HPLC, Snyder, Dolan, & Jupille

[redskytonight820]

All of you are very awesome! Thanks for the *delete me*

[JGK]

As an observation, the current "industry" optimum value for resolution in methods is usually set at 2.0.