Dumb question, but how do chemists use Rf values in chromotography?
Eric Lucas
"Theron Kousek" <the...@charter.net> wrote in message
news:v4695gs...@corp.supernews.com...
> Hi folks:
>
> Pardon my dumb, ignorant question but how do chemists use Rf values? In
> what situations and how are they useful to chemists?
>
> Just finished my 2nd lab today which was paper chromotography. The
> worksheet had us calculate the Rf values based on the
> solvant distance and cation distance. That went well and I was able to
> calculate these values.
This is a really on-the-ball question, that indicates you're thinking about
what you're doing. That type of curiosity will help you to really
understand what others are just swallowing and regurgitating for tests. It
also points out the frequent gap between the theoretical and the practical
aspects of chemistry. In principle, Rf can give you information about
identity of a compound, its polarity and solubility, etc., but that works
best when you're comparing an unknown compound and a known compound on the
same plate. It's good to know how it's calculated and what it means.
However, in practice, the Rf is very dependent on exactly how the stationary
phase is prepared--method of drying of hydroscopic adsorbents like silica
and alumina, state of protonation (acidic/basic/neutral) of amphoteric
adsorbents like alumina, state of purity of the mobile phase, etc. Hell,
I've even seen significant Rf differences on silica TLC plates with and
without ZnS fluorophore, and between silica TLC plates on glass, plastic and
aluminum substrates. Old literature references used to list TLC Rf as one
of the many methods of characterization of newly synthesized compounds, but
I don't see that much any more. The Rf on TLC is related to how a compound
will separate on a chromatography column (gravity, flash, MPLC, etc.).
However, to be useful, you need to get used to how a certain type of TLC
plate behaves and how it relates to a particular type of column packing, and
stick with that type of plate and column packing. Deviations introduce
uncertainties that will make the numbers hard to interpret without direct
comparisons.
Eric Lucas
Bob
Just to add a bit to that... (And I agree it is a good question.)
With luck, under controlled conditions, Rf is a rather reproducible
number that helps to identify compounds. It does measure the nature of
the interaction between the compound and the chemicals of the plate
and the solvent. But it is also useful simply empirically. If you
analyze different mixtures of the same chemicals, you could measure
the relative amounts of the Rf=0.8 component and the Rf=0.4 component.
(This does not mean that there is only one possible compound at 0.8,
but that may be a reasonable interpretation in a specific context.)
bob
Bruce Hamilton
>Pardon my dumb, ignorant question but how do chemists use Rf values? In
>what situations and how are they useful to chemists?
They are very useful, as there are a wide variety of TLC methods,
and some may be useful for one application ( such as quickly comparing
ability of totally different solvent systems to achieve a complex
separation ), and effectiveness of systems can be quickly compared.
Rf values can quickly identify whether changes have improved the
separation of compounds without waiting for slow solvent systems
to reach the top. My major use of TLC is to fine-tune separations,
and Rf values enable me to do that without worrying about how far
up the plate the solvent front has travelled. Before the advent of
small columns ( flash chromatography, solid phase extraction,
filled pasteur pipettes, and HPLC ) preparative chromatography was
often performed by a smear along the TLC origin and the bands were
scraped off the plate and extracted.
One critical parameter that affects my use of TLC Rf values is
saturation of tank. If the TLC tank is fully saturated, the data
more accurately reflects behaviour in column chromatography using
the same substrate ( usually silica or alumina for many organic
samples, but paper is still used for some materials ) and solvent
system. Fully saturated systems tend to run faster because solvent
evaporation is reduced.
Most organic solvent systems in tanks aren't fully saturated
( filter paper helps - but it needs to closely surround the plate
and be well saturated with elution solvent ). In many cases using
partially saturated tanks, the TLC separation can be quite different
to a similar column system. It's often useful for TLC systems to
not be saturated, as the solvent evaporation can improve the
separation - but the analysis may take longer.
Transferring methods to systems such as HPLC is an important use of
TLC with Rf values, as behaviour and separation of molecules can be
compared to published Rf data.
Paper chromatography often used for samples like dyes, where visual
examination is easy. The big plus for TLC is that you see all the
sample you added providing you use an universal detection reagent.
Mass balances from HPLC and Column chromatography often fail because
users haven't varied the mobile phase sufficently to ensure all
loaded material has eluted off the column.
For many investigations, I prefer to use an Iatroscan ( TLC with
a flame ionisation detector ), as it can quantify even the completely
retained organic material. Because the Iatroscan uses silica or
alumina rods, TLC methods and Rfs are quickly transferred.
Bruce Hamilton
Oscar Hur
1. Believe or not, I have seen the lab notebooks from a postdoc
recording Rf values of 2 and 3?!?! Of couse, it has become a joke
around the department.
xyzbb...@uclink4.berkeley.edu (Bob) wrote in message news:<3e434d55...@agate.berkeley.edu>...
Uncle Al
>
> Hi folks:
>
> Pardon my dumb, ignorant question but how do chemists use Rf values? In
> what situations and how are they useful to chemists?
>
> Just finished my 2nd lab today which was paper chromotography. The
> worksheet had us calculate the Rf values based on the
> solvant distance and cation distance. That went well and I was able to
> calculate these values.
>
> purpose of finishing the lab but for my own knowledge,
> I was just curious to their purpose in a real-world situation?
>
> Finished 2 labs now, about 7 to go plus a lab midterm and lab final. I'm
> still intimidated by the labs (still not comfortable performing them)
> but my comfort level increased very slightly due to surviving 2 labs so far.
> I hope this comfort level continues to increase? When
> I read the manual last week about this lab, I was clueless what it was.
> After doing more digging on the internet, I found some interesting sites on
> it and read up on it. It became more interesting. After finishing the
> lab (no matter what my final score is), I came to the conclusion that it was
> quite an interesting experience. Next week is a fractional crystalization
> lab. Don't know anything about it yet. Any advice as to which process
> is better (ie, fractional or paper)???
We've got somebody competent here! Nobody tell Homeland Severity or
the Department of Education. This one child could destroy a lifetime
of Liberal diversity!
Rf is an extremely useful measure if the peaks are narrow and
reproducible. A silica gel plate is not the best venue here, though
Rf ratios are a good indication of how well you are separating your
components and why (changing Rf with changing eluent properties). GC
and especially HPLC, with their very narrow lines and reproducible Rf
values (certainly against an internal standard) allow identification
of peaks by Rf. A sample of petroleum has a unique fingerprint of
certainly hundreds and possibly thousands of peaks. This is nice if
you wonder who spilled it (source) and how long it has been around
(loss of volatiles, differential microbial degradation). Ditto drug
or toxic agent ID in forensic investigations. Getting info on one
chemical species in the complex mess that is blood is a small miracle.
Look up "theoretical plates" in a p-chem txt. Crystallization is one
theoretical plate - though it can be a potent one if equilbrium
distribution is really skewed. Pumping solvent out of an oil is a
real pain (Henry's law), whereas crystallization of the syrup may
exclude all the solvent all at once. For column work (distillation or
chromatography) you care about the height of a theoretical plate. A
foot-long Vigreaux column run properly may have three plates. A three
inch long silica gel plate may have a dozen or more. A silica gel
column has wider plates but more of them. GC will have thousands of
plates in the column, HPLC tens of thousands or more for the same
column length (how does the mechanism differ for gas-solid vs
liquid-solid)?
Fractional crystallization vs. column exchange is well compared by
lanthanide isolation. Look up the discovery of the lanthanides vs.
their contemporary production.
--
Uncle Al
http://www.mazepath.com/uncleal/
(Toxic URL! Unsafe for children and most mammals)
"Quis custodiet ipsos custodes?" The Net!
Repeating Decimal
significance? At least spell out the complete term. When I see it, I think
radio frequency.
Bill
Bruce Hamilton
>For us less than professional chemists, what is Rf?
Rf is the ratio of the distance travelled by a peak divided by the
the distance travelled by the solvent front. Hence it will be less
than 1 for single dimension TLC. It is used in systems where the
solvent front is detectable, usually TLC, paper chromatography,
and column chromatography when the solvent composition remains
constant and you are going to dismantle the column to obtain
sections with fractions of interest.
Obviously, if the solvent reaches the top or end of the column, Rf
loses all significance and column volumes often become the unit of
choice ( eg for HPLC and most column chromatography ). Also, Rf is
of little to no use for systems where material bonds to the column
and is eluted by a change in conditions or solvent, such as
ion-exchange or with gradient elution ( changing solvents or column
comditions).
For TLC on silica ( the most popular plates because many compounds
degrade on alumina - which has separations that are also very
sensitive to the degree of hydration ( Brockman activity )), the
Rf value is fairly reproducible if the same commercial plates
are used, often related to Merck silica gel 60 with or without
a fluorescent indicator.
However, as noted in an earlier post, the amount of saturation of
the chamber has a profound effect on the solvent front, and I've
found that if a published separation isn't reproducible in a fully
saturated tank, the separation is often improved in a partially
saturated tank, but it takes much longer. As saturation decreases,
Rf values tend to vary more, especially for complex solvent systems.
TLC separations on silica are often optimised to yield Rf
values between 0.4 - 0.8 for the main peak of interest, as less
polar material travels faster and more polar material stays at
the origin or travels slower. However sometimes any Rf is OK
if you are just trying to separate a couple of closely-eluting
compounds. TLC is very popular for screening natural products
and determining the effectiveness of large scale chromatography.
Generally, for speed, TLC is conducted only in one dimension,
but you can also achieve improved separations but taking a
square TLC plate and putting one spot in the corner. The plate
is run in one solvent system, carefully and gently dried, and
then rotated 90 degrees so the first run is now the origin,
and placed in a second solvent sytem.
This 2D technique can help resolve difficult separations but,
because there is a drying step and residual solvents, the Rf
values of the second run are much more variable, and these days
it's easier to put the material through a column or SPE system.
TLC with good detection ( FID, Iodine, UV, charring ) is still
a quick qualitative technique for finding out what how all your
sample behaves. HPLC, GC, Flash, SPE, and other column
chromatography systems only detect what elutes, and can
greatly mislead when applying results to scaleup.
I've seen panic-stricken production chemists because they have
several kilogram columns containing 50g of highly expensive
( $10,000/g ), cytotoxic, and unstable material that are blocked
because insoluble material precipitated on the column. A TLC
revealed that an oil was sitting near the origin, and a long
night was spent helping a more polar solvent to diffuse into
the column to restart flow.
Bruce Hamilton
Repeating Decimal
B.Ham...@irl.cri.nz wrote on 2/7/03 9:00 AM:
> Repeating Decimal <Salm...@attbi.com> wrote:
>
>> For us less than professional chemists, what is Rf?
>
> Rf is the ratio of the distance travelled by a peak divided by the
> the distance travelled by the solvent front. Hence it will be less
> than 1 for single dimension TLC. It is used in systems where the
> solvent front is detectable, usually TLC, paper chromatography,
> and column chromatography when the solvent composition remains
> constant and you are going to dismantle the column to obtain
<snip>
Thank you for the explanation. It makes previously nonsensical posts
meaningful even I do not fully appreciate all the finer points.
Bill
F. Isla
> ( filter paper helps - but it needs to closely surround the plate
> and be well saturated with elution solvent ). In many cases using
> partially saturated tanks, the TLC separation can be quite different
> to a similar column system. It's often useful for TLC systems to
> not be saturated, as the solvent evaporation can improve the
> separation - but the analysis may take longer.
What do you mean by tanks that are fully saturated? By the way could
someone explain to me why we take an IR for solid and liquid samples
of the same product?
Bruce Hamilton
wrote:
>> Most organic solvent systems in tanks aren't fully saturated
>> ( filter paper helps - but it needs to closely surround the plate
>> and be well saturated with elution solvent ). In many cases using
>> partially saturated tanks, the TLC separation can be quite different
>> to a similar column system. It's often useful for TLC systems to
>> not be saturated, as the solvent evaporation can improve the
>> separation - but the analysis may take longer.
>
>What do you mean by tanks that are fully saturated?
Consider how a TLC is run. First you decide on the solvent and you
place a small amount in the botton of a tank/chamber/beaker and
place a cover over the top. The air above the solvent will only slowly
reach an equilibrium with the liquid. That can greatly be enhanced by
adding filter paper running up the walls, the solvent will quickly run
up the paper and help saturate the chamber. The smaller the chamber,
the quicker to saturate. Once the chamber is saturated, you add the
TLC plate and the solvent will run up the plate.
If the chamber is not saturated ( do't put the filter paper in ), then
the plate acts as the filter paper and volatile solvents will
evaporate off to saturate the chamber, so your final solvent front is
actually behind where the solvent front would normally be, and your
sample Rf value will be different. Unsaturated chambers also can cause
mixed solvent systems to change composition because the more volatile
solvents rocket off first. Another aspect of presaturating the tank is
that solvent vapour can condense on the TLC plate surfaces before the
solvent front arrives, which may affect the activation or
chromatographic properties of the separation.
If you are interested, most TLC books explain the importance of
saturating the tank, and a couple of good books on TLC are
Thin Layer Chromatography by Justus G. Kircher ( who was a senior
scientist with the Coca Cola company ) J.Wiley and Sons. !978,
old but good for applications. Another old but good book is one
called ( I think ) TLC by E.Stahl.
A more recent book on technique and theory, but without extensive
applications is Applied Thin-Layer Chromatography - Best Practice and
Avoidance of Mistakes by Elke Hahn-Deinstrop ( Wiley VCH 2000
ISBN 3527298398 ). Very useful for people playing with plant
materials.
> By the way could someone explain to me why we take an IR for solid
> and liquid samples of the same product?
You haven't said what the product is, but some compounds have
different IR spectra according to crystalline structure. If the sample
is dissolved in a liquid ( instead of dispersed in a Nujol mull ) a
different spectrum can result. Alternatively it could be to look
at impurities masked by the liquid, or even the carrier ( if as a mull
).
Bruce Hamilton
F. Isla
carried out an IR with the solution. Then we evaporated it, got the
solids and did another IR. It looked pretty similar though the peaks
were not exactly at the same places. Would it be right to say that the
solids showed a more defined peak than the liquid?
Uncle Al
>
> B.Ham...@irl.cri.nz (Bruce Hamilton) wrote in message news:<3e4421a6....@newshost.comnet.co.nz>...
> > On 7 Feb 2003 12:05:03 -0800, pay...@nottingham.ac.uk (F. Isla)
> > wrote:
> >
> > > By the way could someone explain to me why we take an IR for solid
> > > and liquid samples of the same product?
> >
> > You haven't said what the product is, but some compounds have
> > different IR spectra according to crystalline structure. If the sample
> > is dissolved in a liquid ( instead of dispersed in a Nujol mull ) a
> > different spectrum can result. Alternatively it could be to look
> Sorry, the product is a homoserine lactone. After cyclization we
> carried out an IR with the solution. Then we evaporated it, got the
> solids and did another IR. It looked pretty similar though the peaks
> were not exactly at the same places. Would it be right to say that the
> solids showed a more defined peak than the liquid?
The solids have broadened peaks compared to dilute solution. You can
differentiate intra- vs. intermolecular hydrogen bonding vs.
concentration, and perhaps see deformation from lattice packing in the
solid.
Nujol mulls are pretty crappy. A (dry) KBr pellet doesn't have
background.
Bill Walker
> This is a really on-the-ball question, that indicates you're thinking about
> what you're doing. That type of curiosity will help you to really
> understand what others are just swallowing and regurgitating for tests.
You bet. Hang in there Theron.
> In principle, Rf can give you information about
> identity of a compound, its polarity and solubility, etc., but that works
> best when you're comparing an unknown compound and a known compound on the
> same plate. It's good to know how it's calculated and what it means.
I think that ID works ONLY when compared to a standard on the same plate.
Even then, there are a ton of assumptions that are made.
> Hell,
> I've even seen significant Rf differences on silica TLC plates with and
> without ZnS fluorophore, and between silica TLC plates on glass, plastic and
> aluminum substrates.
Yeah, I can do 3 TLC's of the same sample, and get 3 completely different
looking results.
Might be due to plate manufacture, differences in hydration, or just bad
technique on my part.
> Old literature references used to list TLC Rf as one
> of the many methods of characterization of newly synthesized compounds, but
> I don't see that much any more.
I wish they'd report it though - it would help _SO_ much in reproducing the
reported results.
Bill