Chemical Forums
Chemistry Forums for Students => Analytical Chemistry Forum => Topic started by: Golden_4_Life on May 11, 2010, 11:19:56 AM
-
Anyone ever work with an FTIR unit?
Seems like the prepared sample goes into a "disc plate" instead of a quartz cuvette as I'm used to doing with the UVVis. If so, then how does the disc plate hold/contain a liquified sample? How is the measurement done? My first time looking into doing this
-
You can get attachments For FTIR that will take differing sample types Gases, liquids, solids, films etc
-
Some instruments use NaCl plates to hold the liquid. You just put a small drop on one plate and then smush the other plate against it.
-
hmmmm though these procedures are simplistic, evidently they have worked for other investigators. Thanks will try soon ;)
-
for liquid samples you have 2 ways
1 - using the discs like what said above and this is so good which has little noise and highly clear
2 - use the quartz cell but the problem with it that it has high noise .
there is advanced type of measurement for FTIR that is the ATR unit which uses diamond crystal to measure instead of cells and it is the best for both solids and liquids and the samples has no thing to do to prepare it it works by the Kublka Munk diffusion theory
-
You can get fluid cells for FTIR. They are made from NaCl or KBr, but with a sealing spacer (usually teflon) between the plates to control the thickness. Fluid cells work well, but some of them can be very hard to clean. Try to see how to clean to cell before you buy it.
When you fill a fluid cell, you usually elevate one side. A pencil placed under one side to elevate works well. This stops air bubbles from staying in the cell.
We used this method to detemine silicone oil washed off pharmaceutical parts.
The other option is to use NaCl salt plates with a space between them. We would die a o ring out of aluminum foil and use it as a spacer between two salt plates. Works well for qualitative work, but not so well for quantitative work.
-
thanky all are under consideration
-
I have obtained 'qualitative" data on the FTIR of PDMS which I extracted via methylene chloride, evap'd and reconstituted with the same. The FTIR signal are clearly shown at wavenumbers: 1259.2 (a sizeable singlet peak), and at 1015 and 1095 (appearing as two doublets). The issues that concern me are:
1) when I added double the amount of dimethicone I got a ~4x amount of signal compared to the 1x amount of dimethicone extracted by sample weight of the lotion. This preliminary data suggests that the response to PDMS on FTIR is not linear---perhaps second order kinetics?
2) The singlet peak appeared in my reagent blank and on the sod.chloride plates only (though with a significantly less absorbance signal).
So:
1) Is the response to PDMS conc. via the FTIR detector "linear" or "quadratic"?
2) Why is there a signal for PDMS in my reagent blank and on the plates only? I KNOW I did not contaminate the solvents---but I must now presume that the silicone signal is coming from the glassware within which the samples were contained.
DOes this seem feaslible--my speculations?
-
Several options.
The first is a method suggestion. The analytical method we use is for a 350/360 viscosity PDMS. If you are using a different PDMS, you may get different answers.
We wash the parts with petroleum ether. The petroleum ether is then evaporated off.
A known amount of toluene (usually 5 mls) is added and swirled to dissolve the PDMS.
The toluene/ PDMS is then transferred to a liquid cell and the IR is taken.
You want to use the sharp PDMS peak at 1254. Above it, you will find a sharp toluene peak. Draw a baseline for both peaks and measure the peak height. Calculate the ratio of peak height PDMS to Peak height toluene. Quantitate PDMS by plotting the concentration of PDMS vs (PDMS/ toluene peak height ratio).
Your fluid cell needs to be relatively thin. And the method is relatively sensitive. If anything, try diluting the silicone.
You can also analyze for PDMS using AA. You wash off the silicone with the appropriate solvent and aspirate directly into the AA flame using a silicon lamp. This requires a bit of care. You need to cover all the beakers with parafilm. High flame from the AA and flammable solvents require a lot of safety measures.
Good luck.
-
1) Is the response to PDMS conc. via the FTIR detector "linear" or "quadratic"?
Does PDMS stand for polydimethylsiloxane? If so, I regularly got r > 0.999 on a 3-point calibration curve using the USP standard and the peak around 1259.
Draw a baseline for both peaks and measure the peak height. Calculate the ratio of peak height PDMS to Peak height toluene. Quantitate PDMS by plotting the concentration of PDMS vs (PDMS/ toluene peak height ratio).
Does your FTIR software do this for you, or do you you do it manually? It was a source of frustration back when I had to run this test that I couldn't get the software to report absorbance at a particular wavelength. Not to hijack the thread, but could you share what software you used and how you liked it?
-
Ok, I have seen the peak at 1260; however, this same peak appeared in my reagent blank and plates only blank though they were smaller in height. So, then using the 1260 peak is a questionable comparator--unless I ''subtract out" the signal from the reagent blank no?
In contrast, the peak doublet at 1014 and 1091.5, which is clearly visible does NOT show up in either the reagent blank nor plates only blank. Arguably, this peak is the more justified as a comparator no?
To the lotion sample that contained 10%w/w PDMS, I treated with acid, then MeCl2, then separated layers, then evap'd off the solvent, then reconstituted with solvent (10 mL), then assayed about 1 small drop onto the plates.
The signal is not linear--though there was a qualitative difference between the 1x conc sample vs. the 4x conc sample.
I believe some of the PDMS signal is being contributed to by the glass erlenmyere flask (which contains Silicon) no?
Why would th use of Toluene instead of MeCl2 remove the aberrant signal?
I believe tha the PDMS signal from the doublet is the more defensible (and analytically valid) comparator because it provided a clean background and the PDMS signals were defelctions of it no?
-
Ok, I have seen the peak at 1260; however, this same peak appeared in my reagent blank and plates only blank though they were smaller in height. So, then using the 1260 peak is a questionable comparator--unless I ''subtract out" the signal from the reagent blank no?
In contrast, the peak doublet at 1014 and 1091.5, which is clearly visible does NOT show up in either the reagent blank nor plates only blank. Arguably, this peak is the more justified as a comparator no?
To the lotion sample that contained 10%w/w PDMS, I treated with acid, then MeCl2, then separated layers, then evap'd off the solvent, then reconstituted with solvent (10 mL), then assayed about 1 small drop onto the plates.
The signal is not linear--though there was a qualitative difference between the 1x conc sample vs. the 4x conc sample.
I believe some of the PDMS signal is being contributed to by the glass erlenmyere flask (which contains Silicon) no?
Why would th use of Toluene instead of MeCl2 remove the aberrant signal?
I believe tha the PDMS signal from the doublet is the more defensible (and analytically valid) comparator because it provided a clean baseline/background.
-
SOrry to "quote" myownself--I did not intend to be "talking to myself" heheh.
At any rate the questions remain:
1) Is some of the vibrational signal coming from deformations in Si that are due to the glass erlenmeyer flask that the PDMS sample extract is contained in? If so, then how to get ride of this--use plastic containers? Suggestions please?
And,
2) Is the vibrational signal at 1259 of less "analytical rigor" than the doublets at 1014 and 1094 simply because the signal at 1259 is present in the negative controls? Keeping in mind in your replies that the neither reagent blank nor plates only blank give a deformation signal in the doublet range.
In theory, the vibrational spectra of the Si-O bonds (waggling, bending, stretcing) is signified at wavenumbers 1014, 1094.
The deformations in the Si-CH3 bonds are signified at wavenumber 1259.
This is important because I must defend these results before group meeting on Wednesday.
-
I used glass beakers, volumetric pipets and flasks with never any problem.
I think the linearity problem might be from just puting a drop on the plates. It seems to me like there could be some variability in the amount of sample that ends up in the IR beam. When I did this test I used a flow cell like Marquis described. If you don't have one you could try measuring out 20 uL (or whatever is appropriate as long as it's the same) with a microsyringe or micropipet, and be careful when you put the plates together that the whole sample stays in the center.
If that's not it, then I would look at the sample prep.
But really, the AA method sounds much more fun.
-
Got a bunch of topics, so hopefully this makes some sense.
The broadness of the Si-O bonds makes quantitation much more difficult. That's why we chose the 1259 peak, which represents the Si-CH3. Glassware (silicates) contain Si-O bonds, but not Si-CH3. We had no interference from erlenmeyer flasks, as long as they are cleaned well.
The individual salt discs are ok for qual work, but will give problems with quantitation. A flow cell of some kind is an absolute necessity for quantitative work.
On Trains point on software. We often had to break down and use a ruler to measure the heights. Some of the software claimed to do it. But as Train mentioned, there were usually baseline problems. As for scan parameters, our scans were set up for 1150 to 1450 reciprocal cm and we measured the PDMS peak at 1259 and the sharp toluene peak at ~1380.
As for solvents, the method works with a number of different solvents. You will see versions of this method with carbon disulfide as the solvent. I don't like CS2. To me, there is little difference in toxicity between toluene and DCM (methylene chloride). I can't speak for the DCM method, but have had 20 + years using the toluene method.
Hope this helps.
-
Well this certainly does help out, and I sincerely appreciated the timely and well-considered replies.
For my part, at the 1 p.m. meeting I will tell them:
1) IT is possible to use the DCM/acid extraction method to recover the PDMS analyte. Also, that toluene is another solvent that has been shown to be effective.
2) Use of the FTIR to quantify PDMS presents a whole 'nother realm of issues, namely,
a) the sod.chlor. plates: it is difficult to sometimes get the drop to be right in the center of the plate and the variance in the "volume" of a "drop" is another counfounder.
b) suggest we order/purch. a "flow cell" (not sure from whom or price$?) but this has been shown by other investigators (quoting this online source ;) of forum writers Train, Marquis) to be effective.
c) will try to use a pipetteman to repeatably deliver 10uL of sample instead of relying on drops.
3) Problems to resolve: a) signature peaks to use for quantifying is a dilemma---when I tried using the 1014/1091 doublet it was effective; however, when I used the 1260 singlet it also appeared in my negative controls. So, some "method devel" is indicated here.
I think they will be pleased to at least see effort/progress notwithstanding the problems encountered.
-
Offering a 'mole snack" to whomever can convincingly explain to me why I see the 1259 cm-1
peak in my "reagent blank" and "sodium chlor plates only" blank. The solvents I used were opened the day of the test and I do not believe them to be contaminated with PDMS.
-
This is kind of an off chance, but still..
Are you using gloves when running this test? You might check and see if any ingredient in the glove could generate a false peak in this range.
We were using the white/ opaque gloves for this test. It turned out that there was a vinyl plasticizer in the gloves that threw a peak at 1259. When we switched to the purple gloves, the problem went away.
It seems simple, but the glove problem took a lot of time and effort to solve. It also took GC/MS, GPC, and AA.
Good luck.
-
Did you try brand new salt plates? They shouldn't be absorbing at 1259 unless there's some contamination. According to my textbook, "Reasonable care must be taken in handling salt cells and plates. . . . Care should be taken to prevent contamination with silicones, which are hard to remove and have strong absorption patterns." (emphasis added)
I always flushed with copious amounts of solvent and took the background with the flow cell in the instrument filled with diluent.
-
First off a summary of the meeting: 1) I explained to the group (upper mgm't and sales director) that "yes, we can assay for dimethicone using the FTIR that we already have; however, we are at best semi-quantitative for now". I showed three slides of FTIR scans taken at three different concentrations and using DCM as the extraction solvent--and that the Negative Controls showed a small peak also. I told them that I may have not removed the residual PDMS from the plate or that there was inherent co-signaling or (as you'all cited yesterday that either trace levels of PDMS were in the gloves--as the 1014/1091 spectra did not show contamination in the NegCtl. THey seemed pleased with this explanation. I told them that I will also use toluene as xtrc'n solv in a future run as other Investigators have used it with success and it may offer us another ''angle" on the problem. 2) One of the senior managers asked me "would you say that you finish this testing and give us a final assessment in 3 months time?" I replied "yes, but I will need to purchase resources--namely the flow cell and some toluene". He said "sounds good. We have had to send-out our lotions for dimethicone so it seems like we might be able to get it done in-house". So overall, it was good. Also, interesting how managment seemed to be less "interested" in the science and more concerned about the bottom line of "when" I could get finished. But there is some very interesting science in play here. Marquis and Train--ENJOY your Mole Snaks--I have clicked "give" on your profiles. Thansk again I will be positng more updates.
G-4-L
-
You asked where to get the specialty equipment (in this case a fluid cell) for your IR.
I would recommend asking the instrument manufacturer, for a start. It doesn't happen often, but every once in a while a manufacturer will make an accessory that only one instrument can use. Got caught in that with an ATR, once.
As for the fluid cell. Try to get one where the spacers can be changed. It may turn out that one thickness of liquid will work out best. In which case you change the spacer.
Also, some of the fluid cells come molded together. You couldn't change the thickness without destroying the cell.
The fluid cells that allow you to change spacers usually allow you to clean all parts of the fluid cell. That can be very helpful. Especially when tracing down a "ghost" peak.
-
Thanks for that tip Marq!
I have recv'd the 'go ahead' from upper management to purch a flow cell--your advice is appreciated.
-
I will be posting up some images of new data via FTIR on simethicone, may need help from the forum ''brain-trust" with some of the interpretation/analysis--is it legit to post raw data images and tables on this website?