4 Results
4.2.3 Adjusted and additional equilibria and equations:
The binding of the complexanes as 5th ligand on the copper adamanzane complex results in new equilibria and causes the equations and calculations to be more complicated. Cu2+ is most likely lost from an
adamanzane to a ligand already coordinated as 5th ligand, since this is a much more kinetically favorable reaction than one involving free Cu2+ in solution.
First the complexanes binds as 5th ligand [Cu(adz)H2O] + Lfree⇄ [Cu(adz)L*] + H2O
] Cu(adz)][L [
] Cu(adz)L*
[
free 5l =
K , 5l for fifth ligand, 4·8
Then the Cu(II) is transchelated:
[Cu(adz)L*] ⇄ [CuL]2+ + adz + nH+
] Cu(adz)L*
[
] H ][
adz ][
CuL
[ n
t
= +
K , t for transchelation 4·9
Where Lfree is all protonation states of the complexane not coordinating copper ions and * denotes that all protonation states of L are included: L, HL and H2L.
The overall reaction is then:
] L [
] L [ ] H [
free 5l t
n L
adz K K
K K
= + 4·10
Of these KL and K5l are known and Kt is defined by eq. 4·9. The three unknown concentrations of eq. 4·9 are then defined by the measured quantities (app. A), resulting in these equations:
] L ][
adz ][
CuL [
] L ][
Cu(adz)L*
[
free 5l
L
adz K
K = K
( )
(
1 [L ])
] adz ][
CuL [
[CuL]
-C [L]
free 5l Cu L
K K
= + 4·11
In order to ease calculations and minimize error CCu = Cadz = CL = 5 mM were used in all the following experiments.
4.2.4 [Cu([35]adz)]2+ + EDTA
Transchelation of Cu2+ was measured both from [35]adz to EDTA and vice versa at 50 °C. Unfortunately, all the reaction solution had been used for measurement before equilibrium was reached. The initial spectra of [Cu([35]adz)]2+ + EDTA and [Cu(EDTA)]2- + [35]adz were defined as 5.0 mM and 0.0 mM [Cu([35]adz)]2+
respectively.
Transchelation of Cu2+ from [35]adz to EDTA and vice versa at 50 °C
y = -0.1559x + 4.9428
0 1 2 3 4 5
0 10 20 30 40
Days
mM [Cu[35 ]adz]2+ [Cu-adz]
+ EDTA [Cu-EDTA]
+ adz
Figure 4.4: Transchelation of Cu2+ from [35]adz to EDTA and vice versa at 50 °C. Start pH = 7.5.
Although equilibrium between Cu(II) chelated to [35]adz and to EDTA was not reached, it was clear that the vast majority of the copper would be chelated by EDTA. So to be able to reach equilibrium with a better measurement of [CuL], HEDTA and NTA were used as competing ligands, since their formation constants for Cu(II) chelation (to the unprotonized ligand) are much lower than that of EDTA
4.2.5 [Cu([35]adz)]2+ + HEDTA / NTA
There is a large difference between the ε910 nm of [Cu([35]adz)]2+ and [Cu(HEDTA)]-. Normally, this could be exploited to follow the transchelation of Cu(II), but if there is a large background absorption at 500 nm, no baseline correction is possible, which increases the deviation of the data (figure 4.5). Furthermore, it is very difficult to see at which wavelength the increased background ends when looking solely at the absorption.
Instead, it can be useful to look at the slope of the spectra. If all the spectra are parallel at a certain wavelength, an “isoslopic point”, the background can be considered constant for the samples at that
wavelength, but not necessarily zero. However, the background can also be considered constant between two isoslopic points, which lie sufficiently close, such as in this case 650 nm and 775 nm (figure 4.6). By
measuring the slope of the spectra, errors due to significant but constant background are eliminated.
Spectra of 5 mM [Cu([35]adz)]2+ + 5 mM HEDTA
0 0,2 0,4 0,6 0,8 1 1,2 1,4 1,6
400 500 600 700 800 900 1000 1100
nm
Absorption
0 days 2 7 10 18 23 42 HEDTA-Cu2+
Figure 4.5: Transchelation spectra of Cu2+ from 5.0 mM [Cu([35]adz)]2+ to 5.0 mM HEDTA. Also shown is the spectrum of [Cu(HEDTA)]-.
Spectral slope of 5 mM [Cu([35]adz)]2+ + 5 mM HEDTA
-0,04 -0,02 0 0,02 0,04
400 500 600 700 800 900 1000 1100
nm
Δabsorption/nm
0 2 7 10 18 23 42 HEDTA-Cu2+
Figure 4.6: Transchelation spectral slopes of Cu2+ from 5.0 mM [Cu([35]adz)]2+ to 5.0 mM HEDTA. Also shown is the spectral slope of [Cu(HEDTA)]-.
The difference in slope between [Cu([35]adz)]2+ and [Cu(HEDTA)]- is largest at 600, 725 and 850 nm (figure 4.6). Of these, 725 nm is chosen for measurement of transchelation, as it lies between two isoslopic points, which actually lives up to being isoslopic thereby indicating validity in contrast to 460 nm and 920 nm,
where the slopes cross at multiple places. Deviance from the isoslopic points is most likely due to small changes in background; not from spectrum to spectrum, but from wavelength to wavelength within a spectrum. The spectrum from day 42 seems to shift to the left in a way not explainable by background.
The stability of [Cu([35]adz)]2+ was tested in a competition assay with NTA simultaneously with the HEDTA assay and analysed in the same way (figure 4.7). The wavelength chosen for measurement of transchelation was 735 nm.
Slope of spectra of 5 mM [Cu([35]adz)]2+ and 5 mM NTA
-0,04 -0,02 0 0,02 0,04
400 500 600 700 800 900 1000
nm
ΔAbsorption/nm
0 days 2 7 10 18 23 42 Cu-NTA
Figure 4.7: Transchelation spectral slopes of Cu2+ from 5.0 mM [Cu([35]adz)]2+ to 5.0 mM NTA. Also shown is the spectral slope of [Cu(NTA)]-.
Transchelation of Cu2+ from [35]adz to HEDTA and NTA
y = -0.3718x + 5 y = -0.1415x + 4.9774
0 1 2 3 4 5
0 10 20 30 40 50
Days
mM
Figure 4.8: Transchelation of Cu2+ from 5.0 mM [Cu([35]adz)]2+ to 5.0 mM HEDTA or NTA at 50 °C shown as [Cu2+] chelated by [35]adz.
Then the change in absorption at 725 nm or 735 was plotted as a function of time, showing the transchelation as a decline in the concentration of copper chelated by [35]adz (figure 4.8). The shape of the HEDTA curve is a bit strange with the middle section curving the other way than expected for a first order reaction. It is possible that equilibrium was not reached for three reasons: The shape of the HEDTA curve, the point for 42 days in the HEDTA curve is uncertain due to the left shifted spectrum, and equilibrium is only indicated by two points.
The course of transchelation to NTA seems to be more regular, but as with the other experiments, the reaction speed decreases very close to reaching equilibrium.
The concentration of [Cu([35]adz)X]n+ at equilibrium is read as 0.80 mM against HEDTA and 1.74 mM against NTA.
Equilibrium values:
= +[Cu(adz)]
DTA]
[Cu(adz)HE 0.8 mM = 0.0008 M
[CuHEDTA] = 5.0 mM - 0.8 mM = 4.2 mM = 0.0042 M
(
1 42M (0.005M-0.0042M))
[HEDTA] -0.005M 0.0042M 0[HEDTA]
M 2
4 -1 free2 + + -1 free + = ⇒
[HEDTA]free = 7.5⋅10-4 M
[HEDTA] = 8.6⋅10-3⋅7.5⋅10-4 M = 6.45⋅10-6 M (at pH 7.8)
[
[35]adz]
free = 5 mM - 0.8 mM = 4.2 mM ⇒[
[35]adz]
= 7.0454⋅10-7 ⋅ 4.2 ⋅10-3 M = 2.9590⋅10-9 M [209]( )
(
4)
-1 20.2 -19 3
3 6
6 . 17 ]adz
[3 M 10 M
10 5 . 7 42 1 10 9590 . 2 10 2 . 4
10 2 . 4 0 . 5 10 45 . 6 10
5 =
⋅
⋅ +
⋅
⋅
⋅
⋅
⋅
−
⋅
⋅
= − ⋅ − − − −
K against HEDTA
Likewise with the NTA data:
(
1 80M (0.005M-0.00326M))
[NTA] -0.005M 0.00326M 0[NTA]
M
80 -1 free2 + + -1 free + = ⇒
[NTA]free = 1.4⋅10-3 M
[NTA] = 5.840⋅10-3⋅1.4⋅10-3 M = 8.176⋅10-6 M (at pH 7.5)
( )
(
3)
-1 16.9 -110 3
3 6
13 ]adz
[3 M 10 M
10 4 . 1 80 1 10 05104 , 5 10 26 . 3
10 26 . 3 0 . 5 10 176 . 8 10
5 =
⋅
⋅ +
⋅
⋅
⋅
⋅
⋅
−
⋅
⋅
= −⋅ − − − −
K against NTA
KAdz is dependent on pH and temperature but not on the concentration and type of competing ligand, so when pH and the temperature is the same in the HEDTA and NTA experiments, the calculated KAdz should also be the same. The difference of 103.3 M-1 between the calculated values of KAdz indicates a factor not accounted for.
[Cu(adz)L] is calculated too high as [Cu(adz)OH] is not taken into account. This is especially the case for the HEDTA calculation, why that Kadz can be expected to be too big.
4.2.6 [Cu([24.31]adz)]2+ + HEDTA / NTA
The spectra of [Cu([24.31]adz)] + HEDTA and [Cu(HEDTA)] + [24.31]adz have two isosbestic points where the spectra are not expected to change. These are marked with circles on figure 4.9. However, the absorption does increases over time at the two points. Either because of up-concentration due to evaporation or because of pH change. If the discrepancy is caused by an increase in concentration, the spectra at t = n can be corrected by multiplying with:
n at t abs.
0 at t abs.
nm 674
nm 674
=
=
5 mM [Cu([24.31]adz)]2+ + 5 mM HEDTA before alignment at 674 nm
0 0.2 0.4 0.6 0.8 1
400 500 600 700 800 900 1000 1100
nm
Absorption
0 2 23 42 60 80 102 Cu-HEDTA
Figure 4.9: Transchelation spectra of Cu2+ from 5.0 mM [Cu([24.31]adz)]2+ to 5.0 mM HEDTA. Also shown is the spectrum of [Cu(HEDTA)]-. The circles show where the isosbestic points should be.
Before alignment at 674 nm
0 0,1 0,2 0,3 0,4 0,5 0,6
400 600 800 1000
No adz pH 8 No adz pH 10 [34.21]adz pH 3.5 [34.21]adz pH 7 [34.21]adz pH 10 [35]adz pH 7 [35]adz pH 4 [35]adz pH 9
After alignment at 674 nm
0 0,1 0,2 0,3 0,4 0,5
400 600 800 1000
Figure 4.10: Spectra of [Cu(HEDTA)] at different pH (with or without adamanzanes) to the left, and the same spectra after alignment at 674 nm to the right.
The question is then, whether an alignment at an isosbestic point will affect eventual changes due to pH. To investigate that, solutions of 5.0 mM [Cu(HEDTA)] + 5.0 mM adz with different pH were measured (figure
4.10). The spectra at pH 7 – 9 are almost identical, while the spectra at pH 3.5 and 4 lie under the rest and the spectra at pH 10 lie over the rest with a slight shift to the left. There is no effect of adding adamanzanes to [Cu(HEDTA)]. If the spectra are aligned at 674 nm, they are all identical except the spectra of the pH 10 solutions, which are shifted to the left. So if pH is kept under 10, it should be possible to correct the spectra of both pH and concentration change by alignment at 674 nm.
If the correction of the spectra is valid, then the result should include two isosbestic points beside the common zero at 430 nm; one at 674 nm and one at 901 nm. This is confirmed by figure 4.11. The transchelation correlates to the change in absorption from day 0, where the [Cu(HEDTA)] spectrum is defined as 100 %, and is best measured at 600 nm, as the change is largest there and it is between two isosbestic points.
5 mM [Cu([24.31]adz)]2+ + 5 mM HEDTA after alignment at 674 nm
0 0.2 0.4 0.6 0.8 1
400 500 600 700 800 900 1000 1100
nm
Absorption
0 2 23 42 60 80 102 Cu-HEDTA
Figure 4.11: Transchelation spectra of Cu2+ from 5.0 mM [Cu([24.31]adz)]2+ to 5.0 mM HEDTA after alignment at 674 nm. Also shown is the spectrum of [Cu(HEDTA)]-. The circles show the isosbestic points.
The same correction procedure based on isosbestic points was used on the spectra of transchelation from [24.31]adz to NTA with a similar result, although the isosbestic point at 930 nm is more spread than the one at 901 nm in the assay with [Cu([24.31]adz)]2+ + HEDTA (figure 4.12).
5 mM [Cu([24.31]adz)]2+ + 5 mM NDTA after alignment at 705 nm
0 0.2 0.4 0.6 0.8 1
400 500 600 700 800 900 1000 1100
nm
Absorption
0 2 23 42 60 80 102 Cu-NTA
Figure 4.12: Transchelation spectra of Cu2+ from 5.0 mM [Cu([24.31]adz)]2+ to 5.0 mM NTA after alignment at 674 nm. Also shown is the spectrum of [Cu(NTA)]-.
The transchelation from [Cu([24.31]adz)2+] is much slower than from [Cu([35]adz)2+] and equilibrium is reached neither with HEDTA nor with NTA within 100 days (figure 4.13). Again, the rate of transchelation to HEDTA is more irregular than to NTA, but otherwise the two curves are almost parallel from day 20.
Transchelation of Cu2+ from [24.31]adz to HEDTA and NTA
y = -0.0162x + 4.9718
y = -0.0192x + 4.9624
0 1 2 3 4 5
0 20 40 60 80 100 120
Days
mM
Figure 4.13: Transchelation of Cu2+ from 5.0 mM [Cu([24.31]adz)]2+ to 5.0 mM HEDTA and NTA at 50 °C shown as [Cu2+] chelated by [24.31]adz.