[MrHappy0]

I'm slacking in fundamentals a bit. I'm trying to determine the extinction coefficient and λ_max for two Nickel (II, and III) Dithiolate complexes from UV/vis. For Ni(II) λ_max is straightforward (about 850nm) but for Ni(III) it is less straightforward. There are two prominent features and the first one I think is λ_max at about 380nm.

Now the where I'm really lost... how do you determine the extinction coefficient from just UV/vis data. I know various relations that allow you to calculate it but I don't see how they relate to this data:

A=εcl or A=log₁₀(I₀/I)=εcl

[Corribus]

From your Beer's Law equation, ε = A/cl. Typical units are M^-1 cm^-1. A is your measured absorbance value. c is your molar concentration. l is the pathlength of your cuvette, usually 1 cm for most spectrometers. Most peaks will probably be in the 10³ M^-1 cm^-1 range, though exceptionally strong peaks can be >10⁴ M^-1 cm^-1.

In principle you can do this calculation from a single UV-Vis scan. In practice, this leads to quite a large degree of error, especially because you have no way of knowing if you're in the linear range from a single measurement. It is customary to make a bunch of different solutions by serial dilution, and plot the absorbance as a function of concentration, then do a linear fit. Another huge source of error is typically in weighing your starting material. I suggest using a minimum of 10 mg, 50 is better. Make a concentrated stock solution and then carefully dilute so that your samples are in the 0.05-0.5 OD range for the peaks of interest. Repeat the measurement two more times so you can calculate an average and standard deviation.

It looks like the Ni(II) spectrum you plotted above is probably too concentrated. Once you get over 1 OD, it is not uncommon to be in the nonlinear absorption range.

[MrHappy0]

Thanks for the advice Corribus. My lab time for this exercise was limited so we weren't able to do serial dilutions/multiple measurements. Btw we already diluted to 1.5mM concentration. Seems by your advice we should have diluted more. Do you think the UV/vis is good enough for an estimated ε and λ_max for both Ni[II, and III]?


The following question might be more appropriate in another topic of chemistry but will ask anyways. So my complex is a square planar [Ni(mnt)₂] complex where there is delocalized electrons between the C=C, C-S, and Ni-S bonds. My UV/vis shows a large difference  in the data for Ni(II) and Ni(III) but my IR data doesn't show really any change for the two complexes. What might contribute to this change for UV/vis?

I'm assuming no change in IR is due the minimal geometric change with change in charge and that the UV/vis change is due to number of electrons available for excitation in the UV/vis region.

[Corribus]

Thanks for the advice Corribus. My lab time for this exercise was limited so we weren't able to do serial dilutions/multiple measurements. Btw we already diluted to 1.5mM concentration. Seems by your advice we should have diluted more. Do you think the UV/vis is good enough for an estimated ε and λ_max for both Ni[II, and III]?

It depends on the compound, but 1.5 mM is probably way too concentrated. In my experience 200-300 microM is a good place to start. That said, you can calculate the extinction coefficients from any spectrum, provided you know what the concentration is. It's just that your accuracy will likely be low. (Although, the UV range of your Ni(II) spectrum is saturated, so you won't be able to calculate an extinction coefficient in that region. Sometimes it is necessary to do different concentration ranges for different spectral regions.)

The following question might be more appropriate in another topic of chemistry but will ask anyways. So my complex is a square planar [Ni(mnt)₂] complex where there is delocalized electrons between the C=C, C-S, and Ni-S bonds. My UV/vis shows a large difference  in the data for Ni(II) and Ni(III) but my IR data doesn't show really any change for the two complexes. What might contribute to this change for UV/vis?

The origin of many UV-Vis transitions for transition metal complexes are d-d in nature - that is, electron transitions between metal d-orbitals. These transitions are typically weak in centrosymmetric complexes becauase they are la porte forbidden. Oxidation/reduction of a metal center can give rise to color changes because the electron configuration is obviously different - which will impact the type of transitions between d-orbtials. (The d-orbital splitting and energy can also change, if the ligand structure is impacted.) Honestly, the transition in the Ni(II) spectrum almost looks MLCT or LMCT in origin to me, though, and oxidation of the metal can obviously have a big impact on the appearance of such a peak. I don't know what an "mnt" ligand is. Maybe that would help me figure out what's going on.

I'm assuming no change in IR is due the minimal geometric change with change in charge and that the UV/vis change is due to number of electrons available for excitation in the UV/vis region.

If the metal-centered oxidation/reduction has no impact on the ligand structure, then the IR spectrum of the ligand could be conceivably unchanged by this electrochemical process. Are you looking at a full range of IR? I don't have a whole lot of experience with FTIR of transition metal complexes, so I can only speculate in any case.

[MrHappy0]

Corribus, when you said Ni(II) spectrum is saturated I figured you just meant both Ni (II and III). Now I know that you mean just Ni (II). What makes it look appear saturated? I will tell you what I see in Ni (II): There are two main features one band at around 850nm and one smaller one at 490nm. Since 850nm has the largest absorbance I would call this λ_max. Disregarding any errors from methods, is this correct reasoning. I don't how to come to the conclusion that it is saturated.

The following question might be more appropriate in another topic of chemistry but will ask anyways. So my complex is a square planar [Ni(mnt)₂] complex where there is delocalized electrons between the C=C, C-S, and Ni-S bonds. My UV/vis shows a large difference  in the data for Ni(II) and Ni(III) but my IR data doesn't show really any change for the two complexes. What might contribute to this change for UV/vis?

The origin of many UV-Vis transitions for transition metal complexes are d-d in nature - that is, electron transitions between metal d-orbitals. These transitions are typically weak in centrosymmetric complexes becauase they are la porte forbidden. Oxidation/reduction of a metal center can give rise to color changes because the electron configuration is obviously different - which will impact the type of transitions between d-orbtials. (The d-orbital splitting and energy can also change, if the ligand structure is impacted.) Honestly, the transition in the Ni(II) spectrum almost looks MLCT or LMCT in origin to me, though, and oxidation of the metal can obviously have a big impact on the appearance of such a peak. I don't know what an "mnt" ligand is. Maybe that would help me figure out what's going on.

---first off the complex looks like the image below. It has non-innocent ligands. As far as I have learned from literature, the oxidation of the [M-S₄]^2- complex results in little change when removing a single electron because the two electrons are delocalized between the Ni-S, S-C, and C=C bonds. The CN group is not electron releasing stabilizing the mono, and dianonic complexes and resultings in minimal change in the IR features between the two complexes. A lot of this speculation is interpretation of literature so it could be a little off.

I'm assuming no change in IR is due the minimal geometric change with change in charge and that the UV/vis change is due to number of electrons available for excitation in the UV/vis region.

If the metal-centered oxidation/reduction has no impact on the ligand structure, then the IR spectrum of the ligand could be conceivably unchanged by this electrochemical process. Are you looking at a full range of IR? I don't have a whole lot of experience with FTIR of transition metal complexes, so I can only speculate in any case.
[/quote]

----The IR I'm looking at is from 4500-5001/cm with little to no change between the two complexes.

[Corribus]

Corribus, when you said Ni(II) spectrum is saturated I figured you just meant both Ni (II and III). Now I know that you mean just Ni (II). What makes it look appear saturated? I will tell you what I see in Ni (II): There are two main features one band at around 850nm and one smaller one at 490nm. Since 850nm has the largest absorbance I would call this λ_max. Disregarding any errors from methods, is this correct reasoning. I don't how to come to the conclusion that it is saturated.

I'm referring to the 300-350  nm spectral region. See how the spectrum goes up and then makes a straight line at an OD of about 5? That's saturation. 5 OD is the absolute cutoff of many UV-Vis instruments, although in general I don't trust values over an OD of 1-1.5. If you want to calculate extinction coefficients in this region, you'll probably have to dilute substantially. For instance, to measure the extinction coefficients of the intense Soret peaks of porphyrin complexes ( around 400 nm (ish)), you have to dilute the solution almost to the point where it is almost colorless. Otherwise, the spectrometer is just saturated.
 

---first off the complex looks like the image below. It has non-innocent ligands. As far as I have learned from literature, the oxidation of the [M-S₄]^2- complex results in little change when removing a single electron because the two electrons are delocalized between the Ni-S, S-C, and C=C bonds. The CN group is not electron releasing stabilizing the mono, and dianonic complexes and resultings in minimal change in the IR features between the two complexes. A lot of this speculation is interpretation of literature so it could be a little off.

Do you have a literature source I can take a look at?

[MrHappy0]

Yes.

Nickel Dithiolenes Revisited:  Structures and Electron Distribution from Density Functional Theory for the Three-Member Electron-Transfer Series [Ni(S2C2Me2)2]0,1-,2-
Booyong S. Lim,Dmitry V. Fomitchev, and, and R. H. Holm*
Inorganic Chemistry 2001 40 (17), 4257-4262

There is more but this what I'm currently deciphering.

[Corribus]

Thanks, I'll take a look, get back to you within 24 hrs.

[Corribus]

I read through the paper. It's interesting, thanks for sharing. They don't address the nature of the spectroscopic transitions. However given the intensities, the large bands are probably MLCT or LMCT in origin - that is, oxidation of the metal and reduction of the ligand, or vice-versa. I'm sure one of the cited references addresses it, but I didn't check. That's just my best guess, given that extinction coefficients are >10,000 M^-1 cm^-1, which seems too high for a forbidden d-d transition. The fact that alleged CT transition disappears in your oxidized complex isn't so surprising. It'd be interesting if you could have some time resolved vibrational results - i.e., how do the ligand vibrations change after photoexcitation with visible light. This would tell you something about whether it is LMCT or MLCT. In the paper the CT transition appears at 771 nm for the neutral complex. In yours, it is at 850. Your cyano ligands should give rise to a lower energy ligand LUMO than in the dimethyl version reported in the paper. Assuming the metal orbitals are the same (assuming they are independent of the ligand), then an MLCT transition in your complex, for example, would be predicted to be lower energy than that of the complex reported in the paper, which is what your data shows. Another reason to assign this as a CT transition.

Regarding your vibration results: you say they don't change when the metal is oxidized. I was going to ask whether you have some evidence that the metal is oxidized and not the ligand. But based on the fact that the vibration spectrum doesn't change, I'd say this is probably good evidence in itself. Under the assumption that the metal orbitals and ligand orbitals do not significantly interact in any metal oxidation state, then it makes sense that any change in the metal oxidation state would not impact the ligand's vibration spectrum. In the paper, the authors report that there is almost no orbital interaction between the ligand and metal in the neutral state (based on calculations, the HOMO has almost no metal d orbital character), which would seem to support this finding. As the complex is reduced, the degree of ligand-metal orbital interaction in the HOMO increases substantially - and the vibration energies change as well. They don't address whether there is orbital overlap in an oxidized complex, so this paper may not be very helpful to you in the long run, since you're dealing with an oxidized complex and not a reduced one. Your results would seem to indicate that in the oxidized complex the ligand structure is the same. This supports the conclusion that it is indeed the metal that is oxidized rather than the ligand. This would appear to be a different situation than in the structure reported in the paper, because there the HOMO is almost completely localized on the ligand. Since you are oxidizing the metal here, it would appear that the situation is different in your complex... which means that while the paper is useful in a general way I don't think you can directly learn what's going on in your system using their data.

Note, that was all a knee jerk reaction based on a cursory reading of the paper you provided and the little bit of data you posted here. Clearly it's a complicated system (metal complexes always are), so take that with a grain of salt. If I did a lot more research it's possible I'd reach a completely different conclusion. Sorry for the disclaimer but I don't want to lead you down the wrong path based on a hasty analysis of a partial data set.