I'm not sure your characterization of absorbance is valid.

Absorbance is the negative log of the transmittance. Transmittance is measured directly from the signal off the detector. This is to say that if the transmittance is 10% (0.1), Absorbance is 1. At transmittance of 1% (0.01), absorbance is 2.

Absorbance is convenient because concentration (or pathlength, or pressure in the case of gas-phase measurements) are linear with absorbance. Such is not the case with transmittance.

In the case where you are comparing two or more spectra or peaks and the absolute value of the absorbance is not critical to what is being reported, then arbitrary units should be acceptable. This is generally done to avoid unnecessary complexity in the plots of data.

Joe Van Gompel yes, this is the mathematical basis of absorbance, however, what we are discussing is why arbitrary units are used in absorbance spectra when it is discouraged to do so. Regards.

John Canham that's right! Where I see only justification for using arbitrary units is when you want to compare two spectra in which the magnitude of the absorbance does not matter, but rather the shape of the spectrum or the position of a band. However, in most cases I have seen, they use arbitrary units to represent a simple UV-visible spectrum. I am not saying that this is wrong, but I am curious why they do it, especially when IUPAC does not recognise this, and the most prestigious books on instrumental analysis do not place arbitrary units on a UV-vis absorption spectrum either. Absorbance is dimensionless! I think it is customary in research papers, and especially in areas far from analytical chemistry, such as materials characterisation, among others.

Hi!

Am I missing something? Absorbance is a ratio/dimensionless quantity so what units for this do you expect?

Could you please point to the IUPAC document that discourages use of AU?

And one more thing, maybe you are confusing Absorbance Units with Arbitrary Units?

Regards

Dear Andrzej Wasik it is precisely because absorbance is adimentional that I started this discussion. I think that this is not in doubt knowing the laws governing UV-visible molecular absorption. I am not confusing absorbance units with arbitrary units, I put that some authors use arbitrary units, and here are some examples:

https://www.sciencedirect.com/science/article/abs/pii/S0142961205002681

https://link.springer.com/protocol/10.1007/978-1-61779-953-2_3

https://www.sciencedirect.com/science/article/pii/S1319562X2100807X

From what I have perceived, in practice there is no consensus on when to use arbitrary units, so I found it interesting to read the opinion of different researchers.

Greetings and thanks for participating.

Dear Javier Ernesto Vilaso Cadre:

In two papers (1 and 3, I have no access to position 2 ) you are referring to there's just a mere mention about "arbitrary units".

Probably, this results from neglecting language precision by the authors.

Please note that axes in graphs published in: https://www.sciencedirect.com/science/article/abs/pii/S0142961205002681 are properly marked as "Absorbance", the "arbitrary units" expression appear only in captions.

The same applies to: Article Novel Biosynthesis, Characterization and Bio-catalytic Poten...

the expression "arbitrary units" has been used only once in the abstract, probably for the same reason as above.

By the way, you failed to provide the evidence of IUPAC discouraging the use of "arbitrary units".

When it comes to the origin of "arbitrary units" I'm not sure but I suppose that it has something to do with early electronic spectrophotometers. Some devices measured voltage proportional to the absorbance, others current, while another ones counted pulses or something else. That's why some early authors decided to use "arbitrary units" while reporting their findings. And I think that in majority of cases it wasn't and isn't a problem whatsoever.

Regards

Dear Andrzej Wasik once again thank you for your important participation in this discussion. I think it is appropriate to give a more complete answer so that you can understand my point of view.

Here I leave you other examples where they use arbitrary units for a UV-vis spectrum or for the absorbance value:

https://link.springer.com/article/10.1186/1556-276X-7-151

https://www.researchgate.net/publication/346966144

I think that defining the units in the figure caption or in the spectrum has the same meaning: the property plotted or reported is having arbitrary units.

One of the reasons I started this discussion is because I think that using arbitrary units in an absolute scale is incorrect. Absorbance is dimensionless and if the scale is absolute, as in common spectrophotometric measurement, then absorbance has no units. Arbitrary units are valid when:

1. We want to compare particular shape or signal of two or more spectra and therefore abandon the original scale to set a new arbitrary scale where signals can be compared. So the value of the y-axis does not matter. Anyway, there are other scales to compare spectra. Also, I have seen articles where they report several spectra in the same graph and I suspect that they use arbitrary units in the absolute scale, something that is incorrect if an arbitrary scale is not set, i.e., comparing spectra without changing the original scale makes it incorrect to use arbitrary units.

2. When the measurement cannot be obtained on an absolute scale, being the value dependent on the instrument and therefore the magnitude is not reproducible in other instrument.

On the other hand, I did not mean to say that IUPAC discourages the use of arbitrary units. I put that some researchers discourage using arbitrary units (on the absorbance absolute scale) and that IUPAC does not recognize these units for absorbance, since IUPAC recognizes absorbance as a dimensionless quantity. Here is the link: https://www.degruyter.com/document/doi/10.1351/pac199668122223/html

And lastly, this discussion stems from observing contradictions in the use of units for absorbance and this is not something that only I observed with concern. Here is a link to an ACS editorial article clarifying questions about the correct use of the different scales. This is precisely given the incorrect usage that exists in the scientific literature.

Article Absolute, Arbitrary, Relative, or Normalized Scale? How to G...

As you indicated, this is not even a major problem, but I thought it was interesting to discuss.

Regards

One reason for using arbitrary units is if you don't know what your units are, or your equipment is uncalibrated, or non-linear. Side-by-side comparisons still mean something.

The logarithmic unit described above is the bel, which should be applied to power, which I think is proportional to intensity of spectra. It is usual to use decibels in radar work, to make many useful values greater than 1 (like 3dB). Whether to use a negative sign for attenuation or assume that the word attenuation includes the negative sign causes much discussion and difference of opinion.