Scalable Nanostructured Carbon Electrode Arrays for Enhanced

Feb 26, 2018 - IBM, Thomas J. Watson Research Center, 1101 Kitchawan Road, Yorktown Heights ... *E-mail: [email protected]., *E-mail: [email protected]...
0 downloads 0 Views 2MB Size
Subscriber access provided by Kaohsiung Medical University

Article

Scalable Nanostructured Carbon Electrode Arrays for Enhanced Dopamine Detection Silvia Demuru, Luca Nela, Nathan Marchack, Steven J Holmes, Damon B Farmer, George S. Tulevski, Qinghuang Lin, and Hariklia Deligianni ACS Sens., Just Accepted Manuscript • DOI: 10.1021/acssensors.8b00043 • Publication Date (Web): 26 Feb 2018 Downloaded from http://pubs.acs.org on March 1, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Sensors is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sensors

Scalable Nanostructured Carbon Electrode Arrays for Enhanced Dopamine Detection Silvia Demuru,† Luca Nela,† Nathan Marchack,& Steven J. Holmes,& Damon B. Farmer, & George S. Tulevski, & Qinghuang Lin,&,* Hariklia Deligianni &,* &

IBM, Thomas J. Watson Research Center, 1101 Kitchawan Road, Yorktown Heights, NY

10598 Present Addresses † Politecnico di Torino, Corso Duca degli Abruzzi, 24, 10129 Torino, ITALY Abstract Dopamine is a neurotransmitter that modulates arousal and motivation in humans and animals. It plays a central role in the brain “reward” system. Its dysregulation is involved in several debilitating disorders such as addiction, depression, Parkinson’s disease, and schizophrenia. Dopamine neurotransmission and its reuptake in extracellular space takes place with millisecond temporal and nanometer spatial resolution. Novel nanoscale electrodes are needed with superior sensitivity and improved spatial resolution to gain improved understanding of dopamine dysregulation. We report on a scalable fabrication of dopamine neurochemical probes of a nanostructured glassy carbon that is smaller than any existing dopamine sensor and arrays of more than 6000 nanorod probes. We also report on the electrochemical dopamine sensing of the glassy carbon nanorod electrode. Compared with a carbon fiber, the nanostructured glassy carbon nanorods provide about 2X higher sensitivity per unit area for dopamine sensing and more than 5X higher signal per unit area at low concentration of dopamine, with comparable LOD and time response. These glassy carbon nanorods were fabricated by pyrolysis of a lithographically defined polymeric nanostructure with an industry standard semiconductor fabrication infrastructure. The scalable fabrication strategy offers the potential to integrate these nanoscale carbon rods with an integrated circuit control system and with other CMOS compatible sensors. KEYWORDS: glassy carbon nanorods; array of carbon nanorod electrodes; 3D nanostructure; nano-biosensor; polymer pyrolysis

ACS Paragon Plus Environment

1

ACS Sensors 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 20

The brain is made of about 80 billion nerve cells, called neurons, that act as basic units for information processing and communication through electrochemical signals. The communication among neurons is carried out by various neurotransmitters. Major neurotransmitters include amino acids, monoamines, peptides and purines, etc. Dopamine is a monoamine neurotransmitter that regulates motivation, mood and movement. Its neurotransmission occurs among neurons cross the striatum and other sites of the basal ganglia in the brain. Dopamine dysregulation and deficiency can cause debilitating movement and mood disorders such as addiction, depression, Parkinson’s disease, and schizophrenia. The size of the neurons that carry dopamine is usually tens of microns, whereas the space in which neurotransmitters flow between the neurons is only tens of nanometers. Release of dopamine in the brain occurs on a sub-second temporal scale. The dynamics of dopamine release have been probed by several electrochemical methods. Among them, amperometry has excellent time response and sensitivity but offers poor selectivity because all electroactive species can contribute to the measurement. Signal separation of electroactive species with amperometry has not been possible without any surface functionalization. Cyclic voltammetry at slow potential scanning speeds shows improved selectivity, however, it has not been proven suitable for in vivo electrochemical measurements of neurotransmitters.1 Fast Scan cyclic voltammetry (FSCV) is a modified cyclic voltammetry which provides the ability to measure dynamic effects of dopamine release with high temporal resolution. Dopamine FSCV sensors usually employ carbon fiber microelectrodes with carbon fibers of a few microns protruding out of a fused silica or glass shaft as the sensing material. Glassy carbon (GC) is a very promising material for electrochemical measurements of dopamine due to its good electrical properties and excellent mechanical and chemical stability. Moreover, GC can be obtained by the pyrolysis of photoresist films, allowing the use of conventional photolithography to pattern complex shapes.2– 3

The electrochemical detection of dopamine and other positively charged neurotransmitters relies on their adsorption at carbon surfaces. The detection sensitivity is expected to increase with higher surface area electrodes.4 Nanostructures could drastically increase the active surface area, both minimizing tissue damage and enhancing the sensitivity of neurotransmitter detection. Among all the neurotransmitters, dopamine is one of the most studied since it is associated with disorders such as Parkinson’s disease.5–7 For future brain computer interfaces (BCIs) that will treat Parkinson’s disease, depression, obsessive compulsive disorder and neurodegenerative conditions, an array of millions to billions of electrodes to match the number of neurons in the brain may be necessary. The carbon nanorod electrode was fabricated at one end of a 1mm wide and 10mm long silicon. The electrode is connected with a single carbon wire that runs along the length of the 10mm long silicon. Further improvements in the fabrication process with reactive ion etching (RIE) to fully or partially isolate the nanorods from each other and with the use of insulation to create separately addressable electrodes per chip with a smaller number of nanorods per electrode is standard practice in integrated circuit fabrication. Multiple electrodes along the substrate length can be fabricated allowing dopamine measurements at several depth locations.

ACS Paragon Plus Environment

2

Page 3 of 20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sensors

Additionally, with co-integration of the analog electronics using CMOS on the same substrate, a single nanorod probe per electrode can be addressed and millions of electrodes can be made on a single substrate. In that case, a few thin wires embedded below the electrodes can be used to multiplex the dopamine signals out of the neural probe. It has been shown that deep miniaturization of neuronal recording electrodes enables improved spatial and temporal resolution with a large signal-to-noise ratio (electrophysiology) and opens the ability to record intracellular signals.8 For in vivo measurements in the intracellular space, the size of the electrodes needs to be scaled down to the nanoscale.8 Miniaturization of dopamine probes is essential to precisely detect local changes in neurotransmitter concentrations. Moreover, small electrodes offer several other advantages for in vivo measurements of dopamine, including reduced tissue damage, enhanced sensitivity and potential for large-scale integrated microelectrode arrays.9-10-15 Glassy carbon microfibers are the most commonly used electrode materials for dopamine detection with fast scan cyclic voltammetry.11–13 While carbon fibers cannot be easily integrated for multichannel detection, integrated arrays with several electrode sites have been demonstrated with photoresist films. Previous works have reported the patterning of micrometer-sized pillars from pyrolyzed carbon.10 Another approach that was used is oxygen plasma treatment of photoresist films to create random nanofibers and pores in the carbon film.4 Carbon nanotubes grown on fiber microelectrodes have also been used for dopamine detection.14 These previous studies provided extracellular measurements using micrometer sized carbon electrodes. However, these were not scalable to create millions of electrode measurements and could be integrated with CMOS electronics on a single shank. Future brain computer interfaces (BCIs) will need neural stimulation and recordings from millions of neurons at the same time8. Furthermore, coupling knowledge from neuronal firing with dynamic release of neurochemistry can provide new insights into neural circuit function in health and disease. In addition, electrode arrays could allow the detection of multiple neurotransmitters in different locations along with simultaneous electrophysiology measurements,14–16 increasing our basic knowledge of the human brain and allowing for correlations with behavior. Furthermore, intracellular neuronal measurements can be accomplished with nanoscale electrodes.8-18 A single novel carbon fiber nanoelectrode with a conical shape has been previously demonstrated to monitor individual exocytotic events within an artificial synapse.18 These single tip nanofibers had a diameter of 50-200nm and length of 500nm-2000nm. A carbon nanopipette electrode with a tip diameter of 250nm and length 5-175 µm was also shown previously.19 Nanostructured electrodes thus offer immense potential to enrich basic scientific knowledge by enabling intracellular neurotransmitter measurements with improved sensitivity and extremely high spatial and temporal resolution.20 In this work, using advances in nanolithography we present a scalable fabrication of nanostructured glassy carbon electrode which was used to detect dopamine with fast scan cyclic voltammetry (FSCV). The novel patterned glassy carbon material has been characterized by scanning electron microscopy (SEM), X-ray photoelectron spectroscopy (XPS), Raman spectroscopy, and fast scan cyclic voltammetry (FSCV).

Experimental methods

ACS Paragon Plus Environment

3

ACS Sensors 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 20

Fabrication of carbon nanorod electrode. The fabrication of glassy carbon nanorod electrode arrays was carried out in the Microelectronic Research Laboratories at the IBM Thomas J. Watson Research Center (IBM Research MRL). The IBM Research MRL is a Class 100, 40,000 square feet of cleanroom with an industry standard microelectronic fabrication infrastructure using 200 mm-sized wafers. 200mm Si wafers were cleaned using a dilute hydrofluoric acid dip to remove the native oxide layer, followed by immersion in a 40:1:1 H2O: H2O2:NH4OH solution at 35°C and a 100:1 H2O:HCl solution at 60°C to remove organic and metallic contamination respectively. The wafers were rinsed in DI water and dried prior to the deposition 100nm of silicon oxide using a silane (SiH4) and oxygen precursors in a chemical vapor deposition (CVD) chamber. A 30nm SiC layer was deposited at 350°C using silane and propene precursors in a plasmaenhanced CVD chamber. The phenolic polymer layer was then spun onto the wafers in two separate steps at 2375 RPM, resulting in a thickness of ~1µm (500nm for each step). Subsequently, a 15nm thick titanium (Ti) hard mask layer was deposited by evaporation. The polymer film was baked to form a physically stable cross-linked polymer network. The 15nm thick inorganic hard mask was deposited over the polymer to serve as a reactive ion etch mask for the cross-linked carbon polymer. On top of this hard mask, a pattern of photoresist pillars was formed, with a pillar dimension of 150 nm at a 300nm pitch. The pattern was formed by lithographic exposure with an ArF (193 nm) ASML 1100 patterning tool, with 0.75 numerical aperture and conventional illumination. The photoresist was developed with 0.26 N tetramethylammonium hydroxide (TMAH), and then the resist image was transferred into the hard mask using reactive ion etching (RIE) with a 500 Ws/75 Wb/4 mTorr/50 sccm Cl2/100 sccm Ar plasma in an inductively coupled plasma (ICP) reactor. After RIE of the hard mask, the process chemistry and conditions of the etch tool was changed to a 400 Ws/100 Wb/10 mTorr/100 sccm N2/20 sccm Ar/14 sccm O2 plasma, and the polymer underlayer was partially etched to a depth of 400 nm, creating an array of pillars over the remnant underlayer. The remaining underlayer served as an electrical connection for the pillars to the output electrode. After the RIE, the initial photoresist was consumed, and the hard mask material remaining on top of the pillars was removed with dilute HF. The etched organic material was then annealed under Ar atmosphere at 900°C for 10 hours to generate a glassy carbon material from the phenolic polymer. After this anneal, the carbon became conductive and was tested as an electrode material. A flowchart of the fabrication process is shown in Figure 1. Scanning electronic microscopy (SEM) micrographs of the nanorod array are shown in Figure 2a prior to the wet strip and Figure2b after the hard mask removal and after annealing. It was observed that the diameter of the polymer nanorods underneath the hard mask tapered from ~100nm at the base to ~60nm at the top. The tapering of the polymer nanorods was attributed to isotropic etching of radical species from the plasma discharge. After annealing at 900°C for 10 hours, the base and the top of polymer nanorods shrank to ~85 nm and 30 nm, respectively. Chemicals and fast scan cyclic voltammetry system. Dopamine hydrochloride (SigmaAldrich, St. Luis, MO) stock solution were prepared daily in water and diluted with Tris buffer at the desired concentration for testing. TRIS buffer solution (1 L) consists of 3.25 mM KCl, 1.2 mM MgCl2•6H2O, 2.0 mM Na2SO4, 1.25 mM NaH2PO4•2H2O, 140 mM NaCl, 15 mM Trizma

ACS Paragon Plus Environment

4

Page 5 of 20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sensors

HCl, 1.2 mM CaCl2•2H2O and it is adjusted to pH 7.4 daily with the addition of hydrochloric acid and sodium hydroxide. Potassium Chloride was purchased from J. T. Baker Inc. (Phillipsburg, NY) and Na2SO4 from Alfa Catalog Chemicals (Danvers, MA). All the other chemicals were purchased from Sigma-Aldrich (St. Luis, MO). The buffer was supplied continuously by dual syringe pumps (New Era PumpSystem Inc., Farmingdale, NY) with a flow of 1 mL/min towards the carbon nanorods electrode and the reference electrode (BASi Inc., West Lafayette, Inc.), placed in the flow cell. Carbon fiber microelectrodes 7 µm diameter, 100 µm length (World Precision Instruments, Sarasota, FL) are used to compare the response of the carbon nanorods. A bolus of dopamine is injected through a six-port switching valve (INDEX health and science, Bristol, CT). All experiments were performed in a grounded Faraday Cage (CH Instruments, Austin, TX). The fast-scan data acquisition system used was developed at the University of North Carolina, at Chapel Hill. For the detection of dopamine, the electrodes were scanned from -0.4 to 1.3 V (vs Ag/AgCl) with a scan rate of 100-400 V/s and a repetition rate of 10 Hz for the triangular waveforms. A sampling rate of 100,000 Hz has been used. The gain of the potentiostat is 200 nA/V. The response time was quantified as the time necessary for the peak current to rise from 0% to 90% of its maximum value. Electrode preparation and active surface calculation. The electrode patterning was performed through an image reversal lithographic process. A photoresist (AZ 5214) was spin-coated at 4000 rpm for 60 s, yielding a 1.5 µm thick layer. The opening defined is 30 µm x 20 µm (L x W), with an area of  = 600 μ . The wafer was then sliced into ~1 mm wide strips to allow immersion in a flow cell. While the top sensor area is defined by the photoresist, also the silicon (Si) bottom surface and sides were isolated to prevent charging in the solution. In the absence of this step, the signal immediately saturates. An effective isolation was achieved by covering the bottom surface and the sides with a thick (~1 mm) 10:1 PDMS (Sylgard) layer. Lastly, the top part of the Si strip was coated with silver paste to provide an electrical contact between the carbon nanorod layer and the potentiostat. The total active surface should consider the lateral surface of the nanorods. This is evaluated approximating each rod as a truncated cone having larger base radius 42.5 nm (Rb), smaller base radius (rb) 15 nm and height 240 nm (H), with a cylinder having radius 15 nm (r) and height 60 nm (h) on top (see Figure 2b). The number of rods in the opening (N) is calculated considering a

 pitch of p=300 nm, as N = ( +1) ( +1). The nanorods area is calculated as in Equation 1: 



 = ( + 2 ℎ +  ( +  )( −  ) +  )

(1)

This area should be added to the starting opening area, with a total active surface calculated as in Equation 2:  =  +  −   

(2)

ACS Paragon Plus Environment

5

ACS Sensors 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 20

Results and discussion Characterization of Carbon Nanorod Materials. The sheet resistance of blanket phenolic polymeric films was measured with a four-point electrical probe system. Before pyrolysis, a ~0.5 µm polymer film had a very high sheet resistance which was difficult to measure (>106 Ω/sq). After pyrolysis at 900°C for 1 hour, the film thickness was reduced to ~0.25 µm with a sheet resistance of ~315 Ω/sq and a resistivity of 7825~8000 µ Ω cm. After 10 hours pyrolysis at 900°C, the thickness was ~0.23 µm and had a sheet resistance of ~190 Ω/sq, and a resistivity of 4400 µ Ω cm. The water contact angle after pyrolysis at 900°C for 10 hours was equal to 54 deg. X-Ray Photoelectron Spectroscopy (XPS) analysis of the cross-linked phenolic polymer film was conducted to assess changes in chemical bonding as a result of the pyrolysis process. Blanket coupons were used to obtain the measurements. Survey spectra of the pre- and postpyrolysis films are shown in Figure 3c. As a result of the pyrolysis, a slight shift in binding energy was observed, as well as a change in the relative intensities of the C 1s and O 1s peaks. The changes were observed more readily by plotting the detailed spectra as shown in Figure 3a and Figure 3b. Post-pyrolysis, the full width at half maximum (FWHM) of the fitted peaks for both C and O decreased, which was attributed to a shift in surface charging of the annealed film. Pyrolysis at 900 °C resulted in a significant increase in the relative percentage of C-C bonding in the C 1s spectra (from 24% to 57%), as calculated from the areas of the fitted peaks normalized to the total area. Conversely, the C-O bonding percentage decreased from 66% to 35% upon annealing. The percentage of C=O bonds decreased slightly (10% to 7%). However, it should be noted that the fitting of this peak may have been affected by the presence of trace amounts of F contamination (either from the plasma reactor or fluoroware containers), as evidenced by the slight downward shift in binding energy. The overall composition of the film and carbon to oxygen (C:O) ratio are plotted in Figure 3d. The oxygen content of the film decreased by roughly half upon pyrolysis, which corresponded well to the bonding changes observed from the detailed spectra. In Figure 4a, the Raman spectra of the cross-linked phenolic polymer film are shown before and after pyrolysis under different conditions. The Raman spectra for the cross-linked phenolic polymer film before annealing is very broad. In all the other cases, the cross-linked phenolic polymer film reveals two characteristic bands at ~ 1350 cm−1 (D”band) and ~ 1600 (G”band) cm−1 with intensity ratio “D”/“G”~1.1. Similar spectra have been previously reported for glassy carbon materials. The ratio between “D” and “G” bands correlates with the extent of microstructural disorder. The ratio for glassy carbon is in the range ~1.2 - 1.5, while the ratio of our pyrolyzed phenolic polymer film material was slightly lower (~1.1). This can be attributed to a less disordered and more graphitic microstructure of our pyrolyzed phenolic polymer film.3 Moreover, in Figure 4b the Raman spectra of a 30 µm x 100 µm carbon microfiber and the Raman spectra of the carbon nanorods after pyrolysis for 10hrs at 900oC are very similar further confirming the effectiveness of the pyrolysis to convert the polymer to glassy carbon. Fast Scan Cyclic Voltammetry (FSCV)21 was used for the detection of dopamine at the electrode surface. The experiments are performed with a scan rate of 100-250 V/s. To avoid the potentiostat saturation (maximum current ±2000 nA), a small nanorod region was defined. The resulting total area was estimated to be ~900 µm2 with about 6700 nanorods (see Electrode

ACS Paragon Plus Environment

6

Page 7 of 20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sensors

preparation and active surface calculation). Since it has been shown in previous works that oxygen plasma can improve dopamine sensitivity,4-7-14 a 45s oxygen plasma treatment at 50 W bias power was carried out before the measurement. The background current and the dopamine oxidation current before and after 1 hour in pH 7.4 TRIS buffer is shown in Figure 5. Figure 5a shows that the background is stable before and after 1 hour of running an experiment. In general, we observed that the background was stable after repeated experiments. Moreover, a 30-minute overoxidation treatment with triangular scans up to 1.4 V at 60 Hz was performed to improve the carbon surface reactivity and to clean the surface after the first calibration.22-23 The treatments have been previously shown to enhance the adsorption of dopamine on the carbon by creating carboxylic bonds (-COO-) on the surface.22 The effect of overoxidation can be seen in Figure 5b. Overoxidation treatment increases the dopamine signal but slightly decreases the sensor time response. The time response after 1 hour in pH 7.4 TRIS buffer solution increased of the 28% (5±0.7 s versus 6.4±0.4 s, n=4 DA concentrations). The time response was compared with the DA oxidation signal from a commercially available carbon fiber with the same FSCV system (Figure S1) without the overoxidation treatment. An SEM image of the defined opening with nanorods after the surface treatments is shown in Figure S2. We obtained a time response of 4.9±0.4 s for the fiber, which is almost the same as the carbon nanorod electrode before the overoxidation treatment (5±0.7 s). The long-time response is most likely due to our flow system apparatus setup. Nevertheless, these values are similar to previous published work with in vitro calibration of carbon fibers.24 We can therefore conclude that our carbon nanorod electrode has a similar time response with a commercially available carbon fiber electrode. Different dopamine concentrations have been tested and the dopamine calibration curve and the dopamine color plots are shown in Figure 6. The electrode was calibrated before and after 1 hour in pH 7.4 TRIS buffer. Figure 6d reports the value of the dopamine oxidation peak as a function of the dopamine concentration. The calibration curve is linear (R2=0.9867) with a slope of 5±4 nA/µM (n=4 concentrations). For a signal to noise ratio equal to 3 and the noise level of the FSCV system25, we have a limit of detection (LOD) of 60±5 nM. The sensitivity of a commercially available carbon fiber with 4X larger dimension and 4X larger scan-rate tested with our system has shown a sensitivity of 7±3 nA/µM (m=7 electrodes, n=5 concentrations) and a LOD of 41±6 nM.25 Table 1 shows figures of merit of different carbon electrodes reported previously and a comparison with this work. The LOD of our nanocarbon electrode is comparable with one of standard carbon fibers.23-24-25 The current density derived from the slope of the calibration curve of the carbon nanorod electrode considering the nanorod active area (~900 µm2) is equal to 5.6±4 pA/(µM. µm2). On the other hand, the current density of the carbon fiber (area 2236.5 µm2) is equal to 3±1 pA/(µM. µm2). Hence, we have 2X current density improvement from the bare carbon fiber that has a ~ 4X bigger open area. A relatively slow scan rate of 100-250 V/s was used due to the high reactivity of the surface that resulted in the background current reaching the current limit of the FSCV apparatus. Even higher current densities could be achieved with faster scan rates for smaller electrode openings since the signal is linearly proportional to the scan rate.11 Moreover, we focus on the signal obtained for 500nM dopamine, that represents a concentration suitable for in vivo applications. It has been shown previously that dopamine transients are on the order of 50-100 nM in rats,26-28 and reached values higher than 200 nM with the

ACS Paragon Plus Environment

7

ACS Sensors 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 20

use of drugs.26 In awake monkeys with electrical stimulation, dopamine transients reached 850 nM dopamine.27 The oxidation current of the nanostructured carbon electrodes (~85 nm pillar base in a 30 µm x 20 µm opening) for 500 nM dopamine is 13.4 nA. The current density value (14.9 pA/µm2 vs. 2.7 pA/µm2) is more than 5X higher than the carbon fiber electrode at the same concentration of dopamine (results of a carbon fiber electrode shown in Table 1). Considering the state of the art electrodes for dopamine detection, the oxidation current for the nanorods at the low dopamine concentration is higher than that of a carbon fiber microelectrode (dopamine oxidation current of 6±3 nA) with 4X larger dimensions and 4X higher scan rate,24 proving the effectiveness of the nanostructure to obtain very good signals at relatively low concentrations.

The nanostructures increase the sensitivity per unit area of the detection by 2X while yielding an LOD of 60±5 nM. This is comparable to the LOD of the bare carbon fibers, with the same time response and with 4X smaller open area. As we and others have shown previously, for in vivo applications, one must use Nafion-based polymeric coatings in order achieve the desired selectivity versus interferents such as Ascorbic Acid and DOPAC.25-29 Also, a smaller open area and improvement in the choice of the insulated substrate will enable a decrease of the noise level, provide a lower LOD, and allow the integration of multiple carbon electrodes onto a smaller footprint.

Conclusions In summary, we have demonstrated for the first time a scalable fabrication of a nanostructured (less than 100 nm in size) glassy carbon electrode array (with over 6000 electrodes) and the detection of low concentrations of dopamine with fast scan cyclic voltammetry (FSCV). This nanostructured glassy carbon electrode array exhibits 2X higher sensitivity per unit area for dopamine sensing and more than 5X higher current density for low dopamine concentration compared to a micron-sized carbon fiber (4X bigger open area) with comparable LOD and time response. Using the same nano-lithographic technique already employed in this work, an array of nanostructured electrodes to combine neuronal firing and neurotransmitter detection in multiple locations along a long needle type implant can be realized in the future. Furthermore, neurotransmitter measurements within a single synapse could be possibly envisioned.16-18 In addition to the miniaturization, the carbon nanorod electrodes were fabricated on a CMOS compatible silicon platform which can be readily coupled to the back-side with CMOS electronics to drive the measurements and sophisticated multiplexing of signals with few wires. The nanoscale electrodes combined with the CMOS electronics can become a compact advanced high-resolution future BCI.8 Further improvements on the substrate insulation may allow a decrease of the LOD and the noise level of the sensor. Furthermore, combining electrophysiological and neurochemical measurements will lead to a better understanding of the human brain and its pathologies. ASSOCIATED CONTENT The Supporting Information is available free of charge on the ACS Publications website. Time response comparison with a standard carbon fiber electrode, SEM image of carbon nanorod electrode after overoxidation and plasma treatment.

ACS Paragon Plus Environment

8

Page 9 of 20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sensors

AUTHOR INFORMATION Author Contributions Experiments were designed by H.D., S.H. and Q.L. Fabrication of carbon nanorods and testing was performed by S.D., H.D., L.N., N.M., S.H. and G.T. The manuscript and figures were prepared by H.D., S.D., L.N., N.M., S.H. and Q.L. H.D. supervised all aspects of this work. All authors have given approval for the manuscript. Corresponding authors: Qinghuang Lin, [email protected], Hariklia Deligianni, [email protected] Notes The authors declare no competing financial interest. ACKNOWLEDGMENTS The authors are grateful to IBM Research in Yorktown Heights, NY and to the IBM Research Microelectronics Research Laboratories in Yorktown Heights, NY for the support of this work. The authors wish to thank E. Galligan for the excellent SEM images and P. J. Sorce for help with dicing of the samples to allow experiments in a flow cell. We are grateful to Professor R. M. Wightman and to Dr. C. McKinney at the University of North Carolina Chapel Hill, for teaching and providing many insightful discussions on the fast scan cyclic voltammetry method. We are thankful to Professor J. Cheer at the University of Maryland, for encouraging us to take on the challenge of creating nanoscale electrodes for neurotransmitter sensing.

REFERENCES (1) Johnson, J.A.; Wightman R.M.; Cyclic Voltammetric Measurements of Neurotransmitters. Electrochem Soc Interface. 2017, 53-57. (2) Ranganathan, S.; McCreery, R.; Majji, S. M.; Madou, M. Photoresist-Derived Carbon for Microelectromechanical Systems and Electrochemical Applications. J. Electrochem. Soc. 2000, 147 (1), 277. (3) Řeháček, V.; Hotový, I.; Vojs, M.; Kotlár, M.; Kups, T.; Spiess, Pyrolyzed photoresist film electrodes for application in electroanalysis. L. J. Electr. Eng. 2011, 62 (1), 49–53. (4) Yi, W.; Yang, Y.; Hashemi, P.; Cheng, M. M. C. 3D carbon nanofiber microelectrode arrays fabricated by plasma-assisted pyrolysis to enhance sensitivity and stability of real-

ACS Paragon Plus Environment

9

ACS Sensors 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 20

time dopamine detection. Biomed. Microdevices 2016, 18 (6), 1–9. (5) Qi, L.; Thomas, E.; White, S. H.; Smith, S. K.; Lee, C. A.; Wilson, L. R.; Sombers, L. A. Unmasking the effects of L-DOPA on rapid dopamine signaling with an improved approach for nafion coating carbon-fiber microelectrodes.Anal. Chem. 2016, 88 (16), 8129–8136. (6) Bucher, E. S.; Wightman, R. M. Electrochemical Analysis of Neurotransmitters.Annu. Rev. Anal. Chem. (Palo Alto. Calif). 2015, 8, 239–261. (7) Ong, X. C.; Willard, A.; Forssell, M.; Gittis, A.; Fedder, G. K. A silicon neural probe fabricated using DRIE on bonded thin silicon. Proc. Annu. Int. Conf. IEEE Eng. Med. Biol. Soc. EMBS 2016, 4885–4888. (8) Robinson, J.T.; Jorgolli M.; Park H.; Nanowire electrodes for high-density stimulation and measurement of neural circuits. Front. Neural Circuits. 2013, 38(7), 1-5. (9) Amato, L.; Keller, S. S.; Heiskanen, A.; Dimaki, M.; Emnéus, J.; Boisen, A.; Tenje, M. Fabrication of high-aspect ratio SU-8 micropillar arrays. Microelectron. Eng. Microelectron. Eng. 2012, 98, 483–487. (10) Amato, L.; Heiskanen, A.; Caviglia, C.; Shah, F.; Zór, K.; Skolimowski, M.; Madou, M.; Gammelgaard, L.; Hansen, R.; Seiz, E. G.; Ramos, M.; Moreno, T. R.; Martínez-Serrano, A.; Keller, S. S.; Emnéus, J. Pyrolysed 3D-carbon scaffolds induce spontaneous differentiation of human neural stem cells and facilitate real-time dopamine detection. Adv. Funct. Mater. 2014, 24 (44), 7042–7052. (11) Bath, B. D.; Michael, D. J.; Trafton, B. J.; Joseph, J. D.; Runnels, P. L.; Wightman, R. M. Subsecond Adsorption and Desorption of Dopamine at Carbon-Fiber Microelectrodes. Anal. Chem. 2000, 72 (24), 5994–6002.

ACS Paragon Plus Environment

10

Page 11 of 20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sensors

(12) Takmakov, P.; Zachek, M. K.; Keithley, R. B.; Walsh, P. L.; Donley, C.; McCarty, G. S.; Wightman, R. M. Carbon microelectrodes with a renewable Surface. Anal. Chem. 2010, 82 (5), 2020–2028. (13) Heien, M. L. A. V; Johnson, M. A.; Wightman, R. M. Resolving neurotransmitters detected by fast-scan cyclic voltammetry. Anal. Chem. 2004, 76 (19), 5697–5704. (14) Parent, K. L.; Hill, D. F.; Crown, L. M.; Wiegand, J.-P.; Gies, K. F.; Miller, M. A.; Atcherley, C. W.; Heien, M. L.; Cowen, S. L. Platform to Enable Combined Measurement of Dopamine and Neural Activity. Anal. Chem. 2017, 89 (5), 2790–2799. (15) Strand, A. M.; Venton, B. J. Flame etching enhances the sensitivity of carbon-fiber microelectrodes. Anal. Chem. 2008, 80 (10), 3708–3715. (16) Zhang, S.; Song, Y.; Jia, J.; Xiao, G.; Yang, L.; Sun, M.; Wang, M.; Cai, X. An implantable microelectrode array for dopamine and electrophysiological recordings in response to L-dopa therapy for Parkinson’s disease. Proc. Annu. Int. Conf. IEEE Eng. Med. Biol. Soc. EMBS 2016, 1922–1925. (17) Yang, C.; Trikantzopoulos, E.; Jacobs, C. B.; Venton, B. J. Evaluation of carbon nanotube fiber microelectrodes

for neurotransmitter detection:

Correlation of

electrochemical performance and surface properties. Anal. Chim. Acta 2017, 965, 1–8. (18) Li, Y.-T.; Zhang S, -H.; Wang, L.; Xiao, R.-R.; Liu, W; Zhang, X.-W.; Zhou, Z.; Amatore, C.; Huang, W.-H. Nanoelectrode for Amperometric Monitoring of Individual Vesicular Exocytosis Inside Single Synapses. Angew. Chem. Int. Ed. 2014, 53 (46),

ACS Paragon Plus Environment

11

ACS Sensors 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 20

12456-12460. (19) Rees, H.R.; Anderson, S.E.; Privman E.; Bau H.H.; Venton B.J. Carbon nanopipette electrodes for dopamine detection in Drosophila. Anal. Chem. 2015, 87 (7), 3849-3855. (20) Ganesana M.; Lee S.T.; Wang Y.; Venton B. Analytical Techniques in Neuroscience: Recent Advances in Imaging, Separation, and Electrochemical Methods. J. Anal. Chem 2017, 89, 314-341. (21) Zachek, M. K.; Takmakov, P.; Moody, B.; Wightman, R. M.; McCarty, G. S. Anal. Chem. Simultaneous decoupled detection of dopamine and oxygen using pyrolyzed carbon microarrays and fast-scan cyclic voltammetry. 2009, 81 (15), 6258–6265. (22) Mitchell, E. C.; Dunaway, L. E.; McCarty, G. S.; Sombers, L. A. Spectroelectrochemical Characterization of the Dynamic Carbon-Fiber Surface in Response to Electrochemical Conditioning. Langmuir 2017, 33, 7838−7846. (23) Heien, M. L. a V; Phillips, P. E. M.; Stuber, G. D.; Seipel, A. T.; Wightman, R. M. Overoxidation of carbon-fiber microelectrodes enhances dopamine adsorption and increases sensitivity. Analyst 2003, 128 (12), 1413–1419. (24) Annis, D.S.; Mosher, D.F.; Roberts, D.D.; Higher Sensitivity Dopamine Measurements with Faster-Scan Cyclic Voltammetry. Anal Chem. 2009, 27(4), 339-351. (25) Demuru, S.; Deligianni, H. Surface PEDOT:Nafion Coatings for Enhanced Dopamine, Serotonin and Adenosine Sensing. J. Electrochem.Soc. 2017, 164 (14), G129-G138. (26) Robinson, D. L., Hermans, A., Seipel, A. T. & Wightman, R. M. Monitoring rapid

ACS Paragon Plus Environment

12

Page 13 of 20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sensors

chemical communication in the brain, Chem. Rev. 2008, 108, 2554–2584. (27) Schluter, E. W., Mitz, A. R., Cheer, J. F. & Averbeck, B. B. Real-time dopamine measurement in awake monkeys. PLoS One 2014, 9(6), e98692. (28) Kishida, K. T.; Sandberg S. G.; Lohrenz T.; Comair Y. G.; Sáez I.; Phillips P. E. M.; Montague P. R. Sub-second dopamine detection in human striatum. PLoS One 2011, 6(8), e23291. (29) Vreeland, R. F.; Atcherley, C. W.; Russell, W. S.; Xie, J. Y.; Lu, D.; Laude, N. D. Porreca, F.; Heien, M. L. Biocompatible PEDOT:Nafion Composite Electrode Coatings for Selective Detection of Neurotransmitters in Vivo. Anal. Chem. 2015, 87 (5), 2600– 2607.

ACS Paragon Plus Environment

13

ACS Sensors 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 20

Figure 1: Process flow for fabrication of carbon electrodes and SEM images. a) Phenolic Polymer spin coating and baking. b) Hard mask deposition and deep UV lithography to define the nanopillars. c) RIE to pattern the hard mask. d) Oxygen/Nitrogen RIE to partially etch the phenolic polymer. e) HF hard mask removal. f) 900°C thermal annealing for 10h.

Figure 2: SEM images of the Nanorods before and after the annealing process. (a) TiltSEM of carbon nanorod electrodes after hard mask open and partial etching of the phenolic polymer. (b) x-SEM of nanorod array after hard mask removal and annealing at 900°C for 10 hours.

ACS Paragon Plus Environment

14

Page 15 of 20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sensors

Figure 3: XPS spectra of the Nanorods before and after the annealing. Detailed C 1s and O 1s spectra for (a) as-deposited polymer film and (b) after 900°C annealing for 10 hours. (c) XPS survey scans of the as-deposited and post-annealed polymer (d). Atomic composition and C/O ratio of the pre- and post-annealed polymer film.

ACS Paragon Plus Environment

15

ACS Sensors 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 20

Figure 4: Raman spectra of the Nanorods before and after different annealing conditions. (a) Raman spectra of the as-deposited polymer film, after 900°C annealing for 1 hour, after 900°C annealing for 10 hours, after 900°C annealing for 10 hours of the carbon nanorod electrode. (b) Raman spectra of a 33 µm x 100 µm carbon fiber and after 900°C annealing for 10 hours of the carbon nanorods electrode.

ACS Paragon Plus Environment

16

Page 17 of 20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sensors

Figure 5: Background and time response for Nanorods before and after overoxidation. Background current (a) and dopamine oxidation versus time (b) for 10 µM dopamine before and after overoxidation treatment. The electrode was tested in TRIS buffer pH 7.4 at 100 V/s.

ACS Paragon Plus Environment

17

ACS Sensors 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 20

Figure 6: Dopamine detection with fast scan cyclic voltammetry. Dopamine signal with carbon nanorods 20 µm x 30 µm for 500 nM dopamine (a), 5 µM dopamine (b), 10 µM dopamine (c). The color plots represent also the variation of the signal in time. Calibration plot with the previous oxidation signals (d). Experiment performed in TRIS buffer pH 7.4 at 100 V/s.

ACS Paragon Plus Environment

18

Page 19 of 20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sensors

Table 1. Literature comparison of different carbon electrodes.

LOD (nM)

Sensitivity (nA/µ µM)

Current density . (pA/(µ µM µm2))

Signal with 500nM (nA)

Current density with 500nM (pA/µ µm2)

Area (µ µm2)

Pyrolyzed carbon microelectrode21

50

4.5

9

2

4

500

Carbon fiber25

41

7

3

6

2.7

2236.5

60

5

5.6

13.4

14.9

~900

Electrodes

Carbon nanorods (this work)

ACS Paragon Plus Environment

19

ACS Sensors 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 20

For TOC only.

ACS Paragon Plus Environment

1