Use origin 7/7.5 or origin pro 8

Thank you for the suggestions!

"Graph" is another free graph plotting software with large interest for Math and engineering students

just wanted to do some self-promotion for my optical spectroscopy software SPEKWIN32, but Pedro already did it.

However, one more hint: if your data format doesn't fit one of the formats accepted by Spekwin32, have a look at the *.fak or the *.csv format (see samples from the test spectra file in the download area). Try mimicking one of these data formats and you are done.

The CALC program in Open Office is free and operates much like Excel.

QTIPlot works like Origin. You may use it on different latforms, including Linux.

Friedrich

I intalled WINSPEK but I have excel files it does not open. What conversion you suggest is best and easiest for WINSPEK

Harry,

try saving your data within Excel in *.csv file format. Spekwin32 reads a wide variety of csv files. Please send me an example file, so I can have a look at it and advice how to do it best.

Friedrich

Hereby. I could not save it as CVs myself

in "Save as.. " menu item, just choose CSV as file type. Then say "Yes" until the csv file is created. Xou will have to close the Excel workbook, in order to allow Spekwin32 reading of the csv file. I tried it with your Excel file and it works right away...

http://textmagic.cachefly.net/faqimages/save%20as%20CSV%20in%20excel.png

... And this is the way it looks

Friedrich

Thanks

Still: rechecked it and my excel does not have the option to save a file as CSV?!

It is smoothed, while in the orginal the spectrum had a curious ramp at 800 nm

By the way we used copper nitrate as a reference

Do you maybe have a better reference for the region between 700 and 800 nm

2 problems with Cu(NO3)

-there is quite an increase in absorption in the lambda region

-it may or amy not be hydrated, thus weighing is less accurate

- of course your Excel can save as csv, which version do you use? Did you have a look at the link above?

- Spekwin32 wants equidistant x axis datapoints, if not, the data gets interpolated during data import. Your data set is special in the way, that it has 10 nm steps, except for the 820 - 840 nm region, where it has 2 or even 1 nm steps. This leads to the smoothed result...

- The irregularity at 820 - 840 nm is an artefact from your measurement system. Perhaps an inceompletely corrected filter change...

- The spectrum is that of aqueous Copper dication, please compare to the attached file

- To address your problems, I need to know what you are doing with it and what you need it for...

Friedrich

Thank you for the prompt reply

I was not aware anymore of the non-equidistancy of the points on the x-axis. In normal operation we use a fixed wavelentgth

The original issue was calibrating of a monitor for measuring black carbon (soot) in suspension. The soot is in suspension because it is collected in water from the atmosphere.

The abosption cell is a cell with a length of 70 cm and we wanted to know what the exact optical path was, that is if there would be internal reflections or not lengthening the path.

The instrument is an extension to the commercial instrument MARGA, with which we collect and measure on-line anions via ion chromatography..

There is a description of the black carbon but only in Dutch

I was guiding this work because of my spectrroscopy background and my main work of measuring black carbon in a proper way

So, If I understand it right, you need an optical density reference solution to find out about the true measurement pathlength of your multi-pass cell? I would measure a spectrum the Copper solution in an standard UV/VIS spectrometer in a cuvette with known path length (1cm) against water as blank solution. Thus, I know the relationship of optical density for a given pathlength for any measured wavelength. I would then dilute the solution, the dilution factor choosen to give an optical density inside the OD measurement range of your 70cm cell. Then, i would measure the spectrum in your MARGA system, and there you are: compare measured against expected OD value (for the complete spectrum or at a useful wavelength, e. g. 800nm)

So, the solution to the problems would be:

problem 1: choose any wavelength with high enough optical density, the cell pathlength is the same any wavelengths

problem 2: you don't need to know the precise concentration of the solution if you just want to find out about the measurement pathlength

Friedrich

Thank you very much:

You embarrass me, because I forgot we did what you suggest, but it did not work out well because it was done at the analytical department where the solution was not well prepared, because of the hydration. This is one reason we like another referecne material.

Still, I am very grateful for the spectrum because I was only able to find on the web old references with different absorbances

QUESTION: in what units is the Perkin-Elmer absobance you sent?

The absorbance/ optical density is unitless, and defined as -log(T/T_0). However, you can presume it was measured in a standard 1cm cuvette. Then the dimension would be "1/cm".

You will find more metal ion spectra in acqueous solution from the link below. The +.dx files are all readable with Spekwin32...

http://wwwchem.uwimona.edu.jm/courses/Tanabe-Sugano/TMspectra1.html

Friedrich

Thank you again. Still I would like to know the concentration unit; is it 1mol/liter?

I will go to the suggested website

Friedrich

On the mentioned site the concentration is specified so I have all I need

Thanks again

Friedrich

I am somewhat worried becasuse the concentration is given as ~0.02 M; that is not very precise

Harry,

no it isn't precise. And it isn't intended to be...

And you don't need to know the exact concentration of your solution, because you then would also need to know the molar extinction coefficient of acqueous Cu2+, which you presumably don't know either.

For your determination of an unknown pathlength, measure the solution first with a known pathlength (if instrumentation available) and you get there by using rule of three... Perhaps a dilution step is needed to stay in working range...

I still wonder if you are thinking about what could have been done before, or if you want to do it now or in the future?

Friedrich

First to answer your question; yes I try to re-assess what has been done. My main objective is to deduce the molar extincton coefficient of the black carbon suspension. With a good calibration solution that would have been easy.

First. It has been bothering me that the results in the 1 cm cell with copper nitrate were a fator of four different from the calculated with the molar extinction coefficients I took from literature. The value of ~0.02 M is precise enough to have a similar difference

Second. In the detection system of the MARGA the whole range of 600 to 700 nm is used for the absorption. We then get an average absoption over this range the in its own units. Problem is that the extinction of copper vaies by a factor of ten over the lambda range 600-700 nm. Thus an average absorption is not very precise.

Therefore a substance with a flatter exticntion curve over this wavelngth range would be better. This could be applied in the MARGA

Hope to have given you the proper answers

By the way: there is no method yet to calibrate the absorption coefficient of black carbon in the gas-phase. Via the water phase it seemed possible, but the uncertainty is just a litttl too high still. It is up to others to optimize it.

OK, then:

1. new questions:

- does MARGA measure by scanning the spectrum with a monochromator, or is there just a spectral bandpass for the 600-700nm region (integral emasurement)?

- do you have access to a UV/VIS spectrometer?

- I would have thougt your goal is to finally determine the black carbon concentration in your solutions, in terms of g/l not in mol/l?

2. answers/comments:

- molar absorptance of metal ions may be influenced by counter-ions. Charge is important Cu+ (green) looks different from Cu2+ (blue). There might be calculation errors (you weight the copper salt Cu(NO3)2*xH2O, but need the molar mass of Cu only...) There are also several Hydrates possible: 3251-23-8 Y water-free, 10031-43-3 (trihydrate), 13478-38-1 (hexahydrate), 19004-19-4 (hemipentahydrate)

- I suppose black carbon to consist of nanoparticels, therefore the measured attenuation will a combination of absorption and scattering (however, the scattering part won't contribute much, its carbon black after all).

- You can't give a molar extinction coefficient, because the nanoparticles are not molecules, and they don't have a definite molar mass, instead there is just a size distribution and therefore a weight distribution.

- You might use a mixture of dyes or of metal ions or any other colorants to get a steady absorbance at 600-700nm

- There might be other calibration methods usable from rubber/ carbon black industry:

www.brookhaveninstruments.com/.../ParticleSizingUsingaDisc.pdf

http://www.astm.org/Standards/rubber-standards.html

I'm sure there are a number ofmethods for quatifying carbon black aerosol, but that's more your expertise than mine isn't it?

Example: http://www.lbl.gov/Education/ELSI/Frames/pollution-Europe-f.html

Friedrich

Thank you for your elaborate answer; it will take until tomorrow for a full answer

A 1st quick answer: the spectrometer has a fixed lamda region extending from 600-700nm; it privides an integrated absorption relative to that of clean water in an own unit.

Of course we do not have a molar coefficient byt a coeffcient expressed in m2/g, the mass-specific extinction coefficient (MAES).

The reason we measure at 650 nm is that scattering is lowest there while our instrument basically cuts off at 730 nm

Reference material was never availbale until a few years ago, made at Lawrence Berkeley Lab, of which we received a suspension

I know the Broiokhaven site quite well; they do not provide a proper reference black carbon representative of soot in the atmosphere.

Well, you are right wiith the wrong choice for copper nitrate and a dye would be better. Worked with dyes myself (in dye lasers) but did not think of these. This is becasue our lab is only very good in analysis of anorganic substance of whihc we have plenty

dear Friedrich

The additional asnwers

Spectrometer: we do/did not have it and have to hire hours for analysis, which is (too) expensive, and unfortunately we do not have students to do a thorough study.

Black Carbon: the problem is that there is no standard reference material.

The material we have no can be dispersed in water, without having measurable organic material in it. Also the particle szie is rather small so that scattering is not too high

For soem time we used a black ink, Printex, but this has not the proper characteristics.

Your persistency inspired me to re-study this item and I found that a preprint came out with measurements done as we also had done. So I will approach the author for details

Your suggestion of using a dye will be pursued at my (former) lab

Harry

Dear Harry,

We use often for a quick control of the spectrometer state or for the absorption of solutions or transparent fims, filters made of a square or rectangular piece of thin wire mesh held in a simple frame positioned against the cell holder. You can control the absorbance of the wire meshes as they exist in different wire diameters. Even if the value read is not exactly the real value due to some scattering or stray reflection, their absorption is very, very stable in time and furthermore almost totally independant from the wavelength over a large domain. Another system is to use a plate of high quality silica on which a metal (W, Pt etc) is deposited through a mask containing a regular arry of small holes. Their properties are about the same s the wire mesh ones. Once again, an exact calbration of the absorbancy of these filters can be obtained with a solution of a photostable dye. But in that case, valid data can only be obtained where the dye has a sufficient molar absorption coefficient.

Dan

Dan

THank

you it might be cheaper alternative for our cheap spectrometer

Friedrich

I wondered if your company is into miniaturissation, because that is needed in environmental monitoring,and minimum soiling of the measuring cells.

No, not really. Most of our devices use "microtiter plates" as sample containers.

http://en.wikipedia.org/wiki/Microtiter_plate

Friedrich

Because of your greaat expertise in spectroscopy I thought Tecan was specifically into spectrometers