Gold electrodes are often not suitable for dopamine measurements as dopamine creates a non-conducting polymer layer on the surface of the electrodes, which leads to increased amount of electrode passivity with the gradual increase in voltammograms measurement. This work presents the impedance spectroscopy and cyclic-voltammetry comparative study for dopamine detection with two modifications for the surface of Au electrodes; cysteamine and mercaptopropionic acid for thermally bonded and ultrasonically welded microfluidic chips, respectively. The effects of optimized tubing selection, bonding techniques, and cleaning methods of the devices with KOH solution played crucial role for improvements in dopamine detection, which are observed in the results. Furthermore, comparison for the modification with unmodified chips, and finding the unknown concentration of dopamine solution using flow injection techniques, is also illustrated.
Cyclic Voltammetry (CV) is famous for its simplicity for the functionality and for the high detailed information content for potentio-dynamic electrochemical measurement. Cyclic voltammogram trace is obtained when the current at the working electrode is plotted graphically against the applied potential for that respective counter electrode. CV analysis gives the best study not only for the electrochemical properties of an analyte but also for the oxidation and reduction patterns and rates for specific species deposition on the electrode in a sample of chemical solution. CV can also be used to determine the diffusion coefficient of an analyte, and the formal reduction potential of an analyte, which can be used as an identification tool. In addition, because concentration is proportional to current in a reversible, Nernstian system, the concentration of an unknown solution can be determined by generating a calibration curve of current vs. concentration.
The waveform of the voltage applied to a working electrode in CV is triangular shaped (i.e., the forward and reverse scan). Since this voltage varies linearly with time, the scan rate is the slope (V/s). An example of a CV is shown in Fig. 1. The peak shape of the reductive and reverse oxidative current vs. electrode potential curve (I-E) in Fig. 1 is typical of an electrode reaction in which the rate is governed by diffusion of the electroactive species to a planar electrode surface [1, 2, 3, 4].
Besides CV, “Electrochemical Impedance Spectroscopy (EIS)” technique is a reliable way to measure the solution resistance and capacitance developed on the electrode systems due to deposition of monolayer on electrode-electrolyte interface. Resistance R(ohmic) and capacitance C (double layer capacitance), gives blockage at the electrode surfaces by giving more impedance readings along with Wargburg resistance at low and high frequencies. EIS can read each component of the Randles model individually; however, analyzing the value at different frequencies and correlating the enormous data quickly. Impedance measuring is a complex phenomenon and it is one of the electrochemical measurements that is being used in number of fields such as biosensors, corrosion measurements and surface coating evaluation, fuel cells and many more.
One of the major contribution for surface modification of electrodes in an electrochemical analyte solution was demonstrated by Heiskanen et. al. , in 2006, where it was observed that MPA, possessing weak acid functionality, decreased the rate of dopamine polymerization resulting in least electrode passivation of Au electrode. It was also found that modifications of microchip electrodes with mercaptopropionic acid (MPA) did not only improve dopamine electrochemistry but also significantly increased the storage stability of the transducers.
Apart from the CV analysis measurements, an impedance spectroscopic study of the interaction between thiol-modified Au electrodes and Saccharomyces cerevisiae was presented first in 2008 by Heiskanen et al. , in which monolayer coverage was reached after 20-28 h of cultivation, and was observed as 15 percent decrease in the real capacitance of the system by cysteamine-modified Au microelectrodes techniques. It was also seen that after an addition of Saccharomyces cerevisiae cells at pH 7.2, the obtained value of R(ct) showed over 560 percent from the value obtained on the same thiol-modified electrode [8, 9, 10, 11, 12].
The main objective of this presented work is to give a better understanding of the comparative study of targeted dopamine detection with two modifications for the surface of Au electrodes i.e., (1) cysteamine and (2) MPA for thermally bonded and ultrasonically welded microfluidic chips, respectively. The influence of both bonding techniques along with the selection of optimized tubing for the fabricated microfluidic device and cleaning functions with KOH has been illustrated for better calibrations and effective results. Additionally, results for comparison of the modification with unmodified chips, and finding the unknown concentration of dopamine solution using flow injection techniques, is also presented.
Materials and methods
Electrodes were fabricated on an injection molded COC substrate having dimensions of 50 mm diameter and thickness of 2 mm by the following fabrication steps:
First, the substrate was spun coated with a 10 μm thick layer of positive photoresist AZ (4562 Standard) to be used as an etching mask. Then UV exposure through a chrome mask using Mask Aligner was applied and was developed afterwards. Reactive Ion Etching (RIE) was then carried out with a power of 150 W; pressure of 300 mTorr, gas composition of 80% of Oxygen and 20% of Nitrogen, to create grooves for embedding metal structures of dimensions (depth: 5 μm for ultrasonically welded chips; and 900 nm for thermally bonded chips). The resist was stripped off in an acetone by ultrasonic bath. The resulting substrate was again spun coated with AZ (4562 Standard) positive photoresist and was then exposed in UV and was developed further. E-beam evaporation method was used to deposit first a 20 nm Titanium layer for adhesion and support purpose and then finally a 200 nm thick layer of Gold (Au) was deposited for electrode functionality. Lift-off process in acetone using ultrasonic bath was used to strip off metal coated photoresist. Figure 2 shows the schematic of the electrode fabrication.
The complete electrode part of the chip is fabricated on a Cyclic Olefin Copolymer (COC) substrate and is explained in Fig. 3. Electrode patterns are clearly seen on the substrate.
Silicon mold fabrication
A silicon wafer substrate mold was fabricated to create a negative replica of nickel shim using injection molded technique.
The steps to create Si mold are given as follows: Si wafer substrate was first spun coated with 1.5 μm thick layer of AZ MIR (701 Standard) at 4600 rpm and was pre-baked at 90 °C for 60 seconds. UV exposure of 169 mJ/cm2 was provided using hard contact recipe of mask Aligner. Then it was developed further with TMAH surface puddle for 60 sec to get rid of excessive photo-resist. The resulting substrate was then dry etched first for 5 μm and for 5 minutes and substrate surface was further smoothened by using Plasma Asher for 30 minutes at 400 ml/min O2, and 70 ml/min N2 at a power of 100 W to make it ready for metal deposition for creating shim. Figure 4 shows the schematic of Si mold fabrication for Nickle shim.
Nickle Shim was then made out of Silicon Mold substrate fabricated with the steps mentioned as follows:
Seed metallization of Nickle Vanadium (NiV) alloy was carried out at the specifications of 1000 seconds of time, 5 mtorr of pressure, and power of 157 W using a sputter system. Electroplating of Nickle was proceeded for up to 6 hours and 13 mins of deposition procedure to get the required thickness for the Ni shim.
The silicon wafer was removed afterwards using wet etching KOH process in fume-hood with 25 wt % of KOH at 80 °C for 6-8 hours. Figure 5 shows the Nickle shim fabrication steps.
Device fabrication by injection molding
Nickle shim was then used to fabricate required microfluidic devices of COC substrate using injection molding techniques as described further.
Once the Nickle shim was created and separated from the silicon wafer, it was cut with 1064 nm laser micro machining tool using 255 mm optics with a laser power of 1.9 W at 10%. After cutting, the Nickle shim was cleaned in water with ultrasonics for 10 minutes and was rinsed and wiped. The main task of polymer injection molding was then carried out in Topas 5013L10 polymer using injection molding tool for positive replication of structures on polymer devices. Figure 6 shows injection molding steps from the Nickle shim.
Bonding methods of microfluidic chips
Two methods for Bonding of Microfluidic chips were used for the devices made above with channel dimensions of 400 μm in width and 80 μm in height. Those two techniques are detailed below:
Ultrasonic welding: 20 kHz Telesonic USP4700 welder tool was used to bond the two chips together to make a full real time device for testing with bonding specifications as 1 atm welding pressure, 0.35 seconds hold time, with energy directors: base 100 μm and height 8 μm.
UV assisted Thermal Bonding: First, UV exposure for 30 seconds using 5000-EC series UV curing Flood Lamp system was given and then the two chips were thermally bonded with the specifications as (1 kN, 125 °C for 10 min) using PW20 hydraulic press tool.
Results and Discussion
Effect of cleaning procedure on analyte signal
Two chips, which were thermally bonded and ultrasonically welded were first tested before any cleaning of them and it was observed that on different scan rates for cyclic voltammetry (CV), the peaks for current were in the range from 12 to 14 μA, respectively.
However, after cleaning with KOH solution, not only current values were improved but also the sensitivity was observed close to the ideal scenario. Figures (9, 10, 11, 12) show the CVs for both chips which were bonded using the different methods as mentioned above.
Electrochemical impedance spectroscopy (GAMRY instrument, Ref: 600, Potentiostat/Galvanostet/ZRA) results also demonstrate that cleaning had a clear effect on both types of bonded chips in a matter of reducing initial overall impedance.
In addition, the effect of KOH cleaning on both devices show less deviation from the ideal behavior as compared to their respective uncleaned states. Figures (13, 14) show these vivid impacts of KOH treated surface of electrodes.
Influence of surface modification on charge transfer properties of dopamine
Comparative study of unmodified and MPA modified ultrasonically welded chip is demonstrated in Figures (15, 16). It is found that mercaptopropionic acid reduced the rate of dopamine polymerization. MPA possesses a weak acid functionality, which had the greatest effect on the dopamine electrochemistry by decreasing electrode passivation, as well as improving reversibility and sensitivity closer to ideality. Besides that, results for the thermally bonded chip with cysteamine modification in Figure 17 also shows a good understanding for less electrode passivation and better reversibility of cyclic voltammograms. EIS results in Figure 18 show comparison of two modifications on electrochemistry of dopamine.
Flow injection analysis
Effects of Different Tubings: Tubings having different dimensions were tested to select an optimal type, which can prove to be best suitable for the flow-injection method for the developed microfluidic chip. Three tubings with dimension (500 μm diameter, 15 cm length), (750 μm diameter, 15 cm length) and (500 μm diameter, 30 cm length) as show in Figure 19, were examined at 750 flow rate for their respective functionalities. The result in Figure 20 shows that the first mentioned tubing not only improved the functionality of flow rate but also provided with the highest current values in the flow injection process.
Unknown concentration detection of dopamine and ferrycyanide by known concentration spectra: Figures (21, 22) show the unknown concentration detection for both solutions i.e., ferricyanide and dopamine by using known concentrations of the respective analyte solutions using cyclic amperometry in injection flow analysis. Some known concentrations are first evaluated before checking the unknown concentration of solution in order to detect the value for that required unknown concentration of an analyte.
A comprehensive study for different surface modifications for electrodes, different bonding techniques and flow injection analysis is presented to detect the best optimal detection of dopamine on electrodes. It is found that modification with the mentioned chemicals not only improved the peak current values for dopamine detection but also provided great sensitivity and reversibility of dopamine cyclic voltammograms. Influence of various tubings and cleaning with KOH effects also helped significantly in ameliorating the outcomes.
Authors deeply thank Arto Heiskanen, Ada-Ioana Bunea and Vida Krikstolaityte for their demonstration and help. Authors also want to thank DTU Nanotech Department for providing the well-equipped research environment, which enabled them to finish the project.
Pothos E.N., Davila V, Sulzer D. “Presynaptic recording of quanta from midbrain dopamine neurons and modulation of the quantal size.” Journal of Neuroscience. 18(11) (1998): 4106-4118. https://doi.org/10.1523/JNEUROSCI.18-11-04106.1998
Patel, B. University of Brighton. https://i.ytimg.com/vi/1f92vGOridg/maxresdefault.jpg
Zachek, M.K., et al. “Electrochemical dopamine detection: Comparing gold and carbon fiber microelectrodes using background subtracted fast scan cyclic voltammetry.” Journal of Electroanalytical Chemistry. 614(1) (2008): 113-120. https://doi.org/10.1016/j.jelechem.2007.11.007