Sie sind auf Seite 1von 9

1

Characterizing the ferrocyanide/ferricyanide redox couple using cyclic voltammetry

Cao, Austin.1 Baldazo, Ben.1 Bayliff, Kyle.1 Tauzin, Lawrence.1 Link, Stephan.1

Department of Chemistry, Rice University. Houston, TX. 77005.

INTRODUCTION

Cyclic voltammetry is an electrochemical technique that measures the current that


develops in an electrochemical cell, which can provide useful information about the reduction
potential, the concentration of species, or the reversibility of the reaction.[1] While an applied
potential is ramped up and down at a constant rate, the current at the working electrode is
sampled against the reference electrode, producing a voltammogram trace. Relevant data can
then be extracted from the peaks of both the cathodic and anodic currents. In practice, cyclic
voltammetry is an extremely cost-effective method for characterizing electrochemical behavior
of redox species. Frequently, it is used as a tool to identify analyte species and determine
concentrations in solution. Even in a biological environment, cyclic voltammetry can be used to
understand chemical events, like the release of neurotransmitters from neurons, with minimal
disturbance of the system.[2]

In this experiment, we will study the ferrocyanide/ferricyanide redox couple, which is a


nearly reversible electrochemical reaction. The anodic current is produced as the following
oxidation occurs at the working electrode:

Fe(CN)6-4  Fe(CN)6-3 + e

When the potential reverses, the corresponding reduction reaction will then produce the
cathodic current. By generating a standard curve using solutions with different concentrations of
potassium ferricyanide, we can then estimate the concentration of an unknown solution. Then,
we will calculate the formal potential, the anodic/cathodic current ratio, and the diffusion
coefficient for this particular reaction.

METHODOLOGY

The experiment was carried out according to the protocol provided in the CHEM 368
Laboratory Manual.[3] The electrodes were polished and cleaned were being placed into the
2

assembly and attached to their respective wires. The glassy carbon tip was the working electrode,
the platinum wire was the counter electrode, and the silver chloride tip was the reference
electrode. Solutions of 1, 2, 5, 8, and 10 mM potassium ferricyanide in 0.1 M potassium nitrate
were prepared. In the chi1200C software, a background scan was conducted with only potassium
nitrate. Scans for each ferricyanide solution, including the unknown sample, were conducted
multiple times to ensure that the voltammogram traces remained the same over time. The last
scan for each solution were saved as text files. Between solutions, the beaker and electrodes were
rinsed with solvent. The text files were saved on a flash drive for further analysis.

DATA

Voltammogram traces for solutions with different


concentrations of potassium ferricyanide
5.00E-05

3.00E-05

1.00E-05
current (A)

0 0.1 0.2 0.3 0.4 0.5 0.6


-1.00E-05

-3.00E-05

-5.00E-05

-7.00E-05
potential (V)

1 mM 2 mM 5 mM 8 mM 10 mM

Figure 1. Comparing voltammogram traces of solutions with different concentrations of


potassium ferricyanide. Data was exported from chi1200C and plotted in Microsoft Excel.
Background trace was subtracted from all data points that were plotted.
3

Voltammogram trace for unknown solution


2.50E-05

1.50E-05

5.00E-06
current (A)

-5.00E-06 0 0.1 0.2 0.3 0.4 0.5 0.6

-1.50E-05

-2.50E-05

-3.50E-05

-4.50E-05
potential (V)

Figure 2. Voltammogram trace of potassium ferricyanide solution with unknown


concentration. Data was exported from chi1200C and plotted in Microsoft Excel. Background
trace was subtracted from all data points that were plotted.

Voltammogram traces of 2 mM potassium ferricyanide at


different scan rates
2.50E-05

1.50E-05

5.00E-06
current (A)

-5.00E-06 0 0.1 0.2 0.3 0.4 0.5 0.6

-1.50E-05

-2.50E-05

-3.50E-05
potential (V)

20 mV/s 50 mV/s 100 mV/s 200 mV/s 500 mV/s

Figure 3. Comparing voltammogram traces of 2 mM potassium ferricyanide at different


scan rates. Data was exported from chi1200C and plotted in Microsoft Excel. All traces were
conducted with the same sample.
4

RESULTS & CALCULATIONS

Data from chi1200C software was collected as text files and imported to Microsoft Excel.
For all the data collected at 100 mV/s, the background trace was subtracted at the corresponding
values for voltage. The data comparing different scan rates was not corrected for background
since we only took a background trace at 100 mV/s. To calculate the concentration of the
unknown solution shown in Figure 2, we must create a standard curve from the data we collected
at different concentrations. The ipa and ipc were collected from each trace using the MIN and
MAX functions in Excel. To create the standard curve, we used the ipa values and plotted them
against solution concentration, shown in Figure 4 below. Concentration was left in mM units,
and current was converted to µA.

Linear relationship between solution concentration


and peak current
40

20 y = 3.2473x + 2.7339
R² = 0.9676
peak current (µA)

0
0 2 4 6 8 10 12
-20

y = -5.5928x - 2.6565
-40 R² = 0.986

-60

-80
concentration of ferrocyanide (mM)

ipc ipa

Figure 4. Solution concentration and peak current show a linear relationship. Peaks from
were collected from Figure 1 and plotted against solution concentration. Linear regression was
generated with the trendline function in Excel.

Lines-of-best-fit and the LINEST results were generated, with an R^2 > .95 for both
lines, indicating strong fit for our approximations. The following equations describes the
relationship between concentration (mM) and ipc, ipa (µA):
5

𝑖𝑝𝑐 = 3.2473 ∗ 𝑐𝑜𝑛𝑐𝑒𝑛𝑡𝑟𝑎𝑡𝑖𝑜𝑛 + 2.7339

𝑖𝑝𝑎 = −5.5929 ∗ 𝑐𝑜𝑛𝑐𝑒𝑛𝑡𝑟𝑎𝑡𝑖𝑜𝑛 − 2.6565

The unknown solution has ipc = 20.459 µA. Averaging the above equations, we
approximate the concentration of potassium ferricyanide to be 5.6476 mM ± 0.5170 mM (all
error values are standard error).

Next, we approximate the diffusion coefficient (D) using the Randle-Sevcik equation:

𝑖𝑝 = 𝑘𝑛3/2 𝐴𝐷1/2 𝐶𝑣 1/2

To calculate D, we will plot v1/2 against ip, and set the slope equal to 𝑘𝑛3/2 𝐴𝐷1/2 , where k =
2.72e5, n = 1, A = 𝜋 ∗ 0.152 = 0.07069 𝑐𝑚2 , and C = .002 mol/dm^3 * (dm^3/1000 cm^3) =
2e-6 mol/cm^3. Figure 5 below shows the plots for both ipa and ipc against v1/2. All values were
converted to standard units A and V/s as dictated by the Randle-Sevcik equation.

Linear relationship between scan rate and peak current


0.00003
y = 2E-05x + 2E-06
0.00002 R² = 0.9911

0.00001
peak current (A)

0
0.00E+00 1.00E-01 2.00E-01 3.00E-01 4.00E-01 5.00E-01 6.00E-01 7.00E-01 8.00E-01
-0.00001

-0.00002
y = -4E-05x - 2E-06
-0.00003
R² = 0.9979
-0.00004
scan rate (V/s)

i_pc i_pa

Figure 5. Scan rate and peak current show a linear relationship. Peaks from both anodic and
cathodic currents were collected from Figure 2 and plotted against scan rate. Linear regression
was generated with the trendline function in Excel.
6

Two lines-of-best-fit and LINEST results were generated for the data, with an R^2 > 0.99
for both, indicating extremely strong fits. The slopes were then set equal to 𝑘𝑛3/2 𝐴𝐷1/2, to find
Dc = 3.84e-7 ± 1.15e-9 cm2/s and Da = 9.39e-7 ± 6.45e-10 cm2/s.

Next, we calculated the formal potential E°’ using the following equation:

𝐸𝑝𝑎 + 𝐸𝑝𝑐
𝐸° =
2

In order to get values for Epa and Epc, we used the ipa and ipc values determined earlier and
manually searched the data for the corresponding E values. Across all 11 traces, we calculated an
average E°’ = 0.1832 V ± 0.0004 V.

We can also show the difference between Epa and Epc for all traces:

Table 1. Difference in potential for different solution concentrations.

Conc. (mM) abs(Epa - Epc) (V)


1 0.088
2 0.099
5 0.131
8 0.162
10 0.179

Table 2. Difference in potential for different scan rates for 2 mM solution.

Scan rate (mV/s) abs(Epa - Epc) (V)


20 0.081
50 0.088
100 0.101
200 0.116
500 0.144

𝑖𝑝𝑎
Finally, we know that the ratio should be approximately 1 for a reversible, diffusion-
𝑖𝑝𝑐
𝑖𝑝𝑎
controlled reaction. Across all 11 traces, we calculated an average 𝑖 = -1.552 ± 0.029.
𝑝𝑐
7

DISCUSSION

We predicted the concentration of the potassium ferricyanide in the unknown solution


using the standard curve generated in Figure 4. As the concentration increased, the current
peaked at greater values for ipc. This is in agreement with the Randle-Sevcik equation, which
indicates that more ions in solution would result in a larger current for a given potential
difference. The Randle-Sevcik equation also demonstrates that concentration (C) and peak
current (ip) are in a linear relationship, which allowed us to generate a line-of-best-fit for the
standard curve in Figure 4. The R^2 values > 0.95 taken from the LINEST Excel function
confirms this linear relationship. We also found a standard error of .5170 mM which is just 10%
of the calculated value of 5.6476 mM, indicating a significant trend for our data. We will discuss
sources of error and linearity of this plot later in the discussion.

We also calculated the diffusion coefficient using the plot of scan rate vs. current in
Figure 5. The standard errors for our pair of D values were calculated by propagating the SE for
slope and y-intercept from the LINEST results. They are relatively small, with Dc = 3.84e-7 ±
1.15e-9 cm2/s and Da = 9.39e-7 ± 6.45e-10 cm2/s. Once again, this linear relationship is
supported by the Randle-Sevcik equation, which indicates that peak current should be directly
proportional to the square root of scan rate.

Because ipa and ipc are always different for our traces, Da and Dc are different. Our
calculations showed that the current values have a ratio of -1.552, which is significantly different
from the value of 1 that we expected. A perfectly reversible reaction would result in anodic and
cathodic current peaks that have the same magnitude, and thus diffusion coefficients that are the
same. However, ferrocyanide/ferricyanide does not exhibit a perfectly reversible reaction. One
possible factor is the inconsistent presence of metal plating on the electrodes. We observed metal
particles on the tip after the experiment, which become the responsibility of the next group to
remove through polishing. This indicates that the redox reaction is not perfectly reversible.
Because the reaction relies on the ions perfectly transferring from solution to electrode and back,
any build-up throughout the cycle can hurt our assumption of reversibility.

We can also compare our experimental calculations for the diffusion coefficient with one
reported in literature, as D = 0.62e-5 cm2/s.[3]
8

|𝑡ℎ𝑒𝑜𝑟𝑒𝑡𝑖𝑐𝑎𝑙 − 𝑒𝑥𝑝𝑒𝑟𝑖𝑚𝑒𝑛𝑡𝑎𝑙|
𝑝𝑒𝑟𝑐𝑒𝑛𝑡 𝑒𝑟𝑟𝑜𝑟 𝑓𝑜𝑟 𝐷𝑎 = ∗ 100 = 84.9%
𝑡ℎ𝑒𝑜𝑟𝑒𝑡𝑖𝑐𝑎𝑙

𝑝𝑒𝑟𝑐𝑒𝑛𝑡 𝑒𝑟𝑟𝑜𝑟 𝑓𝑜𝑟 𝐷𝑐 = 93.8%

Our experimental values differ from the theoretical value by almost 2-fold. There are
many aspects of the experiment that may be improved to both decrease error and increase
linearity in our two linear regression plots. One issue effecting linearity is the changing
concentration of ions in solution, as the concentration won’t be at the marked concentration the
entire redox cycle. This prevents us from being able to assume a direct relationship between
current and scan rate, as concentration is a variable in the Randle-Sevcik equation and may
subsequently affect the peak current values. It also simply skews the data for comparing current
and solution concentration. Another issue affecting linearity are the surface areas of the
electrodes, which we assumed to be perfectly flat and circular. Any ridges or imperfections on
the surface area could potentially hinder the redox reaction or even increase the available surface
area (a wavy surface has more surface area for ions to react). This would in turn affect the peak
current, as dictated by the Randle-Sevick equation. Continuing from our discussion on why our
estimations of D from the anodic and cathodic currents are different, we see that imperfections in
the electrochemical cell can add random error to our data. Factors like the presence of bubbles
under the electrodes or whether the electrodes were completely submerged can potentially
increase our standard error and skew the measurement away from the theoretical values.

CONCLUSION

In this experiment, we used cyclic voltammetry to characterize the redox reaction


between ferrocyanide and ferricyanide. In the first part, we compared peak currents of scans
conducted in solutions of variable ionic concentration. We found a linear relationship between
these two measures that could be used to predict the concentration of an unknown sample with
relatively low standard error. In the second part, we compared peak currents of scans conducted
at constant concentration but different scan rates. Using the Randle-Sevick equation, we plotted
peak current vs. v^1/2, which allowed us to determine the diffusion coefficient from the slope of
the graph. Lastly, we offered a few ways that future experiments may be able to further decrease
error, and potentially characterize electrochemical systems with better accuracy and precision.
9

REFERENCES

[1]
Nicholson, R.S.; Shain, I.; Theory of stationary electrode polarography. Single Scan and cyclic
methods applied to reversible, irreversible and kinetics systems, Anal. Chem., 1964, 36, 706.

[2]
Wightman, R; Probing Cellular Chemistry in Biological Systems with Microelectrodes,
Science, 2006, 311, 5767.

[3]
Link, S; Tauzin, J; CHEM 368 Laboratory Manual. 2018.

Das könnte Ihnen auch gefallen