Improving Performance and Uniformity of Carbon Nanotube Network

that can removes polymers only under electrodes and thus improves the performance of photodiodes without inducing new defects in device channel. T...
0 downloads 0 Views 1MB Size
Subscriber access provided by Drexel University Libraries

Functional Nanostructured Materials (including low-D carbon)

Improving Performance and Uniformity of Carbon Nanotube Network based Photodiodes via Yttrium Oxide Coating-and-decoating Ze Ma, Jie Han, Shuo Yao, Sheng Wang, and Lian-Mao Peng ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.8b21325 • Publication Date (Web): 11 Mar 2019 Downloaded from http://pubs.acs.org on March 16, 2019

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.

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 27 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 Applied Materials & Interfaces

Improving Performance and Uniformity of Carbon Nanotube Network based Photodiodes via Yttrium Oxide Coating-and-decoating Ze Ma,1 Jie Han,1 Shuo Yao, 1 Sheng Wang,*, 1 Lian-Mao Peng*, 1 1Key

Laboratory for the Physics and Chemistry of Nanodevices and Department of Electronics, Peking University, Beijing 100871, China

ABSTRACT: Semiconducting single-walled carbon nanotube (SWCNT) thin film can be obtained by conjugated polymer wrapping sorting technique followed by solution deposition and can be utilized as channel materials of field-effect transistors and absorbing layer of photodiodes. However, after deposition process, there are still polymer molecules wrapping around nanotubes, remaining between nanotubes and remaining on the thin film surface, which will cause large nanotube-electrode resistance and tube-tube resistance. Here, we demonstrate an Yttrium oxide coating-and-decoating technique that can removes polymers only under electrodes and thus improves the performance of photodiodes without inducing new defects in device channel. Treating only contact area, the average short-circuit current of photodiode increases from 9.1 nA to 10.7 nA while the average open-circuit voltage increases from 0.25 V to 0.30 V. This method also improves device uniformity significantly.

1

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 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

Keywords: carbon nanotubes, photodiode, transistors, polymer removal, uniformity INTRODUCTION Infrared (IR) sensing and imaging have a wide range of applications, including optical communication, military surveillance, automotive night vision, and food inspection. Carbon nanotubes (CNTs) are direct bandgap materials with wide spectral response range from ultraviolet to IR, high resonance absorption coefficient of up to 105 cm-1,1-2 picosecond intrinsic response time and good stability. These excellent properties make CNT an ideal candidate for building lowcost photodetectors that work at room temperature.3 Barrier-free bipolar diode (BFBD),4-5 utilizing Palladium (Pd) and Scandium (Sc) as the p-type and n-type contact electrodes respectively, is a simple yet effective device structure to dissociate photo-induced excitons6-7 into free carriers hence generate photoelectric signals.8 Significant progress has been made recently on the purification of CNTs. In particular, high purity (>99.9%) semiconducting single-walled carbon nanotubes (s-SWCNTs) can now be obtained routinely by conjugated polymer wrapping.9-15 Large area CNT network thin films can be prepared on wafers through solution deposition. It is now viable to fabricate wafer scale, solution processable and low-cost CNT field-effect transistors (FETs)12, 16-21 and photodetectors2223

with high performance. However, polymers such as poly(9,9-dioctylfluorenyl-2,7-diyl) (PFO),

poly[(9,9-dioctylfluorenyl-2,7-diyl)-alt-co-(6,6′-{2,2′-bipyridine})] (PFO-BPy) and in this paper, poly[9-(1-octylonoyl)-9Hcarbazole-2,7-diyl] (PCz),11 will still be wrapping around CNTs after 2

ACS Paragon Plus Environment

Page 2 of 27

Page 3 of 27 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 Applied Materials & Interfaces

deposition. Moreover, a certain amount of free and loose polymer molecules exists in CNT solution and will be deposited between tube-tube junctions and onto CNT thin film surface randomly. The existence of these polymers on the CNT film surface can raise the Schottky barrier between CNT and electrodes, increase tube-tube junction resistance in device channel as well as increase the difference between devices. Several techniques12, 24-27 have been developed to remove polymers from CNT thin films such as annealing and rinsing process20,

24

which effectively

improve the performance of FETs. However, such high temperature process will inevitably induce additional defects in CNTs, which in turn significantly reduce the performance of photodiodes. Unlike CNT FETs, the performance of CNT photodiodes is very sensitive to the number of defects or recombination centers in device channel. Excitons generated locally in the photodiode channel will recombine at these sites, reducing photoelectric conversion efficiency in spite of the removal of polymers. In this paper, we report an Yttrium oxide coating-and-decoating (YOCD) technique that can remove polymers from CNT thin film only at contact area of a CNT BFBD locally, leaving the device channel area undamaged, therefore do not introduce notable defects in the photodiode channel. The highest temperature of this technique is 250℃, which is much lower than annealing temperature (400℃ or 600℃) reported earlier,20, 24 which also avoids inducing more defects during such treatment. We show that the quality of the Pd-CNT and Sc-CNT contacts is improved significantly after the treatment. The average short-circuit current and open-circuit voltage of CNT

3

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 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

photodiodes is improved to 10.7 nA and 0.30 V after treating the contact area. The uniformity of devices is also improved significantly.

RESULTS AND DISCUSSION CNT network thin film (~45 CNTs/μm density) was prepared through liquid-phase immersedcoating onto an n+ Si/SiO2 substrate. Then the YOCD treatment was carried out, the scheme of which is to coat the CNT film with a thin layer of YOx, then wash it off and take away PCz molecules during the YOx decoating process. 3 nm Y was first deposited by electron-beam evaporation (EBE). An electron-beam lithography (EBL) and a lift-off process were carried out to ensure Y was coated only at the required area. Thermal oxidation at 250℃ in air was carried out immediately after the lift-off process.28 The coated YOx was removed by immersing in dilute hydrochloric acid solution (volume ratio: HCl: H2O = 1: 25) for 10 minutes. During this process, significant amount of PCz molecules are removed along with YOx by HCl because the electrostatic force between Y and PCz is stronger than the force between PCz and CNT or PCz and substrate. X-ray photoelectron spectroscopy (XPS) result confirms that YOx was removed completely since no Y 3d peak emerges at ~157 eV in the spectra after YOCD treatment (Figure S2). As a comparison, Figure S2b shows multiple Y peaks when YOx is not removed completely. Raman spectroscopy is used to characterize the changes of CNT film interface after YOCD treatment. As shown in Figure 1a, the peak marked with * at 1621.9 nm belongs to PCz, which is 4

ACS Paragon Plus Environment

Page 4 of 27

Page 5 of 27 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 Applied Materials & Interfaces

clearly visible before YOCD treatment and vanished after the process, indicating that significant amount of PCz molecules have been removed. The wetting property of CNT thin film interface is also studied to confirm the change after YOCD treatment. As shown in Figure 1b, the contact angle between water and CNT film changes from around 110° to 60° after YOCD treatment, confirming the CNT film interface changes from hydrophobic to hydrophilic. The 110° contact angle before treatment is determined by both hydrophobic PCz alkyl side-chains and hydrophobic CNTs. The contact angle between water and PCz is measured to be larger than 110°. Although the whole substrate is covered with CNTs with a density of ~45 CNTs/μm, there are still blank spaces between tubes remain uncovered. Therefore, the contact angle between water and CNT remains unknown. But it’s expected to be similar to the contact angle of water on single layer graphene (95-100°).29 After YOCD treatment, most of the free and loose polymers that cover the CNT thin film and part of the polymers that wrap around CNTs are removed, causing the hydrophilic SiO2 substrate (contact angle ~40°) contributes more to the wetting property. Moreover, atomic force microscope (AFM) is used to characterize the CNT film morphology before (Figure 1c) and after (Figure 1d) YOCD treatment. It is obviously that the small particles in Figure 1c are removed after treatment. Besides, the whole thin film surface becomes cleaner and CNTs are more visible after treatment, indicating a large number of PCz molecules are removed from CNT surface. Three horizontal lines are selected from Figure 1c and 1d and the corresponding height distributions are plotted in Figure 1e. The height distributions became smaller after YOCD treatment, confirming the removal of the excessive PCz molecules from the CNT surface. 5

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 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

In order to confirm the improvement of device performance after YOCD process, FETs with natural Si/SiO2 bottom-gates from the substrate are fabricated utilizing Pd and Sc as the p-type and n-type contact electrodes respectively. Using polymethyl methacrylate (PMMA) for masking, a thin layer metal Y lift-off technology is used to define the local area being treated. Four different kinds of devices (18 devices each) are fabricated: untreated (Figure 2a), contact area treated only (Figure 2d), channel area treated only (Figure 2g) and whole device area treated (Figure 2j). The device channel length is set as 0.5 μm which is shorter than the average CNT length (1 μm) so as to reduce effects of tube-tube junctions in the channel. The device channel width is 20 μm. All devices are covered with 180 nm PMMA during the tests since PMMA is an easily obtained and effective encapsulation material in the laboratory that protects the back-gate devices from O2 and H2O in the air without damaging the device performance in short term. The transfer curves of the devices corresponding to treated area mentioned above are shown in Figures 2b, e, h, k (p-FETs) and Figures 2c, f, i, l (n-FETs). Y-axes represent the drain current per micron channel width. The source-drain voltage Vds of p-FETs and n-FETs is -1 V and 1 V respectively. The on-state currents per micron channel width of p-FETs and n-FETs extracted from Figure 2 are showed in Figure 3. After YOCD process, the average on-state current of p-FETs and n-FETs increase significantly. We use relative standard deviation (Sr), a percentage defined as the ratio of the standard deviation σ to the mean μ, to measure the uniformity of devices. After all device area treatment, Sr reduce to ~40% of the untreated devices and become smaller than 10% for both pFETs and n-FETs, which confirms obvious improvement of the uniformity of devices. 6

ACS Paragon Plus Environment

Page 6 of 27

Page 7 of 27 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 Applied Materials & Interfaces

As we can see in Figure 3a, for p-FETs, treating whole device area results in the biggest onstate current improvement (2.49 times the original). Treating only channel area results in a larger on-state current (2.15 times the original) than treating only contact area (1.38 times the original). This indicates that the Schottky barrier between electrode Pd and CNT is relatively low, in which case treating contact area can only bring about limited improvement. On the other hand, treating channel area can reduce tube-tube contact resistance as well as reduce carrier scattering in the channel. The case for n-type FETs is very different. Figure 3b shows that the whole area treating brings about on-state current increase of 2.14 times the original. Contrary to p-FETs situation, treating only the Sc-CNT contact area results in a larger on-state current improvement (2.01 times the original) than treating only channel area (1.53 times the original). This is because metal Sc can react with water and oxygen in air, making it much more difficult to form good ohmic contact between Sc and CNTs, which explains better improvement occurs when treating contact area instead of the channel area. Same results from devices whose channel length are 2 μm (longer than the average CNT length) are also carried out and shown in Supporting Information Figure S2. The on/off ratios of all devices are showed in Figure 2c and 2d. Results show that the on/off ratios exhibit a similar pattern of change with on-state currents because YOCD treatment hardly changes the off-state currents while improves the on-state currents. After exploring the effect of YOCD technique on p-type and n-type FET devices, we use Pd-Sc asymmetric contact to fabricate photodiodes array based on CNT network thin films. The device channel length and width is 0.5 μm and 20 μm respectively. Devices are also covered with 180 nm 7

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 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

PMMA during the tests because PMMA is transparent at the wavelength of 1600 nm and doesn’t influence the excitons. The short-circuit currents and open-circuit voltages are extracted from I-V curves (I-V curves see Figure S3) of photodiodes under infrared laser illumination (1600 nm, 19.7 W/cm2) at 30 V gate voltage. The bottom-gate voltage is chosen to realize maximum short-circuit current and open-circuit voltages simultaneously. Unlike CNT split-gate photodiodes,30-34 asymmetric contact photodiodes dissociate excitons with an electric-field-assisted mechanism. When a Pd-Sc photodiode works, excitons are generated in the whole channel area but can be dissociated into electrons and holes thus contribute to output current only when generated or random-walked into effective dissociation regions near the electrodes, which is approximately 100 nm. It is clear from Raman spectra that YOCD technique (Figure 1a) and annealing technique will both induce new defects into CNTs. These treatment induced defects that cause exciton recombination and therefore significantly reduce short-circuit current and open-circuit voltage. Utilizing YOCD technique, we can choose to treat the contact area only so as to avoid inducing new defects into the device channel. According to above reasons, photodiodes treated only contact area show best performance (Figures 4b and 4c). Treating only contact area, the average shortcircuit current of 15 devices increases from 9.1 nA to 10.7 nA while the average open-circuit voltage increases from 0.25 V to 0.30 V. The extent of improvement is clearly larger than the deviation ranges between devices. The uniformity is also improved and the Sr value is reduced to become less than 5% after the treatment. Two representative I-V curves in the fourth quadrant before and after the treatment are shown in Figure 4d. To evaluate the photocurrent generation 8

ACS Paragon Plus Environment

Page 8 of 27

Page 9 of 27 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 Applied Materials & Interfaces

efficiency, we calculate the external quantum efficiency (EQE), defined as the number of carriers collected to produce the photocurrent divided by the number of incident photons 𝐼𝑝ℎ 𝑒

𝐸𝑄𝐸 = 𝑃𝑖𝑛 =

𝑅𝐼ℎ𝑐 𝑒𝜆

,

(1)

ℎ𝜈

where e, h, ν , c, λ are electron charge, Planck’s constant, photon frequency, the speed of light and the laser wavelength. The photoresponsivity, defined as output photocurrent per watt of radiant optical power input, can be extracted as 𝐼𝑝ℎ

𝑅𝐼 = 𝑃𝑖𝑛

(2)

where Iph is the output photocurrent and Pin is the incident laser power onto the device channel area.35 We calculate the average external quantum efficiency (EQE) after treating contact area to be 0.42%. And the average photoresponsivity after treating contact area is 5.4 mA/W at 1600 nm. Both the RI and EQE shows large improvement than our earlier report.23 By increasing the thickness or density of the CNT film, the performance of photodiodes may in principle be further improved. Figures 4e and 4f exhibit two typical I-V curves in dark at 30 V gate voltage (the same voltage as illumination measurement). While currents at 1 V forward bias are almost equal, the reverse bias current of the device after treating contact area is clearly smaller than current of the untreated device in logarithmic coordinates (-9.20 pA after treating contact area and -16.7 pA without treatment at -0.1 V bias). Figure 4f exhibits the same data in linear coordinates to show currents 9

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 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 27

near 0 V bias more clearly. The differential resistances are extracted from Figure 4f and depict in Figure 4g. As we can see, at small bias (between -0.1 V and 0.1 V), treating contact area results in larger differential resistance, which means the treatment of contact area contributes to the forming of lower Schottky barriers between electrodes and CNTs, resulting in not only larger photoresponsivity but also better rectification characteristics of the diode. Using YOCD technique to treat only contact area also leads to a decrease in the noise of the detector. At zero bias, the noise of a photodiode mainly originates from Johnson noise

𝑖𝑁 =

4𝑘𝑇 ∆𝑓 𝑅

(3)

where k, T, Δf and R are Boltzamann’s constant, temperature, electrical bandwidth and differential resistance. At zero bias, R changes from 6.1 GΩ to 18.4 GΩ after treating contact area, indicating 1.74 times reduction in Johnson noise current. The band diagrams of the device under zero or small reverse bias before and after treating contact area are depicted in Figure 4h. As we can see, the band bending regions near contact electrodes provide the built-in electric field to dissociate excitons.8 Before treatment, carriers need to tunnel through PCz molecules into electrodes. The existence of PCz weakened the band bending in CNT thus reduce the voltage drop in CNT, leading to smaller open-circuit voltage and short-circuit current in Figures 4b and 4c. After PCz removing via YOCD technique, the band bending and the voltage drop in CNT increases, resulting in larger open-circuit voltage and short-circuit current. Moreover, current at small reverse bias originates from carriers that tunneling through CNT barrier (labeled by arrows in Figure 4h). The removal of

10

ACS Paragon Plus Environment

Page 11 of 27 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 Applied Materials & Interfaces

PCz reduces the height of metal-CNT Schottky barriers (ΦSB,p and ΦSB,n in Figure 4h) thus increases the height of tunneling barriers at small reverse bias, which explains smaller current at reverse bias and larger differential resistance near zero bias. Furthermore, we repeated our experiments utilizing rinsing and annealing technique. The results are shown in Figure S7. Previous study showed that rinsing and annealing technique can improve on-state currents of p-FETs based on PFO-BPy wrapped CNTs.24 Our results shows that the onstate currents of both p-FETs and n-FETs based on PCz wrapped CNTs also increased after rinsing and annealing process. However, the short-circuit currents of photodiodes suffered a 9.41% decrease after rinsing and annealing process in Figure S7c. This is because rising and annealing is not a local treating technique. Raman spectrum (Figure S7d) confirmed the increase of defects after treatment, which causes exciton recombination thus decreases short-circuit currents. But the YOCD technique, a local treating technique, can avoid inducing new defects and improve the performance of photodiodes. CONLUSION In summary, we demonstrate an YOCD technique that can remove polymer molecules from CNT interface effectively. As a local treating technique, the process allows us to control the position being treated at will. Raman spectroscopy and wettability test prove the removal of PCz molecules and the change of CNT film interface after the treatment. FETs results shows that YOCD technique can improve p-type and n-type contact between metal electrodes and CNT thin 11

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 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

film as well as improve tube-tube contact in device channel. The Sr shows ~40% reduction and become smaller than 10%. The BFBD results shows that treating only contact area thus avoiding inducing defects into the channel can improve the performance of CNT photodiodes. The average short-circuit current increases from 9.1 nA to 10.7 nA while the average open-circuit voltage increases from 0.25 V to 0.30 V. The Sr is improved to less than 5% after the treatment. The improved contact lead to a larger differential resistance at zero bias, indicating 1.74 times reduction in Johnson noise current. This gives us an effective approach to fabricate CNT thin-film based photodiodes with higher signal-to-noise ratio and uniformity, pushing it to its full potential to be utilized in the field of nanophotonics36-40 and IR imaging. Further improvement is expected to be achieved by removing PCz residuals in the device channel without induce more defects utilizing solvent washing before25 or after24 the deposition of CNT thin film. EXPERIMENTAL SECTION Fabrication. The FETs and photodiodes are patterned by electron-beam lithography and a liftoff process. The p-type contact electrodes are 0.3/70 nm Ti/Pd. The n-type contact electrodes are 90 nm Sc. The test pads are 5/40 nm Ti/Au. All metals are deposited by electron-beam evaporation. Inductively coupled plasma (ICP) etching is used to remove the extra CNTs outside the channel. Characterization. Raman spectroscopy is carried out using a 488 nm laser, an 1800 line/mm grating, a microscope objective (50×) lens and a silicon detector. The spectrometer was calibrated in advance by the Raman peak from a standard silicon sample at 520.7 cm-1. The wetting angle 12

ACS Paragon Plus Environment

Page 12 of 27

Page 13 of 27 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 Applied Materials & Interfaces

measurement is carried out using an optical contact angle measuring and contour analysis system (Dataphysics instruments). Electrical measurements are carried out with a Keithley 4200 semiconductor analyzer at room temperature. Optoelectronic measurements are carried out with a super-continuous spectrum laser (NKT Company) which has a power density of 19.7 W/cm2 at 1600 nm. All devices are covered with 180 nm PMMA during the tests.

ASSOCIATED CONTENT Supporting information The Supporting Information is available free of charge on the ACS Publications website http://pubs.acs.org. Figure S1 shows UV-vis-IR absorption spectrum of the CNT dispersion. Figure S2 shows scanning electron microscope (SEM) images and energy dispersive spectrometer (EDS). Figure S3 shows AFM results of a low CNT density sample. Figure S4 shows XPS results of CNT thin 13

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 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

film on Si/SiO2 substrate before and after YOCD treatment. Figure S5 shows electrical characterization results of CNT FETs with 2 μm channel length. Figure S6 shows characterization of CNT photodiodes. Figure S7 shows the device results utilizing rinsing and annealing technique. AUTHOR INFORMATION Corresponding Authors * E-mail: [email protected]. * E-mail: [email protected] ACKNOWLEDGMENT This work was supported by the National Key Research & Development Program (Grant No. 2016YFA0201902) and National Science Foundation of China (Grant Nos. 61370009, 61621061).

14

ACS Paragon Plus Environment

Page 14 of 27

Page 15 of 27 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 Applied Materials & Interfaces

Figure 1. Characterization of the CNT thin film before and after YOCD treatment. (a) G-band and D-band of the Raman spectra of the sample excited by a 488 nm laser before (gray line) and after 15

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 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

(red line) the treatment. The PCz Raman peak is marked with *. The ratio of the height of G+ peak and D peak is 25.3:1 and 13.1:1 before and after the treatment. (b) Wetting angle between water and CNT thin film before (left) and after (right) YOCD treatment. The wetting angle changed from about 110° to 60° after YOCD treatment. (c)-(d) AFM images of CNT thin film before (c) and after (d) YOCD treatment. Scale bar: 1 μm. (e) Diameter distributions of 3 lines each chosen from (c) and (d) (marked by dashed lines).

16

ACS Paragon Plus Environment

Page 16 of 27

Page 17 of 27 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 Applied Materials & Interfaces

Figure 2. (a), (d), (g), (j) Schematic diagram of the different area being treated (red region in the figure): (a) not-treated, (d) contact area treated, (g) channel area treated, (j) whole device area treated. The golden blocks are floating, indicating the electrodes are deposited onto substrate after 17

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 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

the YOCD process. (b), (e), (h), (k) Id-Vgs transfer curves of p-FETs with different area treated according to (a), (d), (g), (j) in linear coordinates of same scale. Vds = -1 V. (c), (f), (i), (l) Id-Vgs transfer curves of n-FETs. Vds = 1 V.

18

ACS Paragon Plus Environment

Page 18 of 27

Page 19 of 27 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 Applied Materials & Interfaces

Figure 3. (a)-(b) Statistical data of on-state currents per micron channel width of (a) p-FETs and (b) n-FETs. The device channel length and width is 0.5μm and 20 μm. Color gray, red, blue and green represents devices with different area treated according to Figures 2(a), (d), (g), (j). The horizontal solid line through the box represent the average value. The increase of on-state currents and the relative standard deviations are labeled in the figure. (c)-(d) Statistical data of the on/off ratios of all devices corresponding to (a) and (b) respectively.

19

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 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

20

ACS Paragon Plus Environment

Page 20 of 27

Page 21 of 27 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 Applied Materials & Interfaces

Figure 4. Measuring results of CNT photodiodes. (a) Schematic diagram of a Pd-Sc photodiode under illumination. (b)-(c) Statistical data of short-circuit currents (b) and open-circuit voltages (c). Color gray, red, blue and green represents photodiodes with different area treated according to Figures 2(a), (d), (g), (j). The increase of on-state currents and the relative standard deviations of devices after treating contact area are labeled. The device channel length and width is 0.5μm and 20 μm. The incident laser has a power density of 19.7 W/cm2 at 1600 nm. The gate voltage is set as 30 V. (d) Representative I-V curves under illumination in the fourth quadrant of devices with no treatment (gray) and with contact area treated (red). (e)-(f) Representative I-V curves in the dark with no treatment (gray) and with contact area treated (red) in logarithmic coordinates (e) linear coordinates (f). (g) Differential resistance around 0 V bias extracted from Figure 4f. (h) Band diagrams of the device under zero or small reverse bias before and after YOCD treatment. The Schottky barrier heights and tunneling barrier under reverse bias for electrons and holes are labeled.

21

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 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

REFERENCES (1) Yang, Z.-P.; Ci, L.; Bur, J. A.; Lin, S.-Y.; Ajayan, P. M. Experimental Observation of an Extremely Dark Material Made by a Low-density Nanotube Array. Nano Lett. 2008, 8 (2), 446451. (2) De Nicola, F.; Pintossi, C.; Nanni, F.; Cacciotti, I.; Scarselli, M.; Drera, G.; Pagliara, S.; Sangaletti, L.; De Crescenzi, M.; Castrucci, P. Controlling the Thickness of Carbon Nanotube Random Network Films by the Estimation of the Absorption Coefficient. Carbon 2015, 95, 2833. (3) Arnold, M. S.; Blackburn, J. L.; Crochet, J. J.; Doorn, S. K.; Duque, J. G.; Mohite, A.; Telg, H. Recent Developments in the Photophysics of Single-walled Carbon Nanotubes for Their Use as Active and Passive Material Elements in Thin Film Photovoltaics. Phys. Chem. Chem. Phys. 2013, 15 (36), 14896-14918. (4) Wang, S.; Zhang, L.; Zhang, Z.; Ding, L.; Zeng, Q.; Wang, Z.; Liang, X.; Gao, M.; Shen, J.; Xu, H.;Chen, Q.; Cui, R. L.; Li, Y.; Peng, L.-M. Photovoltaic Effects in Asymmetrically Contacted CNT Barrier-free Bipolar Diode. The J. Phys. Chem. C 2009, 113 (17), 6891-6893. (5) Yang, L.; Wang, S.; Zeng, Q.; Zhang, Z.; Pei, T.; Li, Y.; Peng, L.-M. Efficient Photovoltage Multiplication in Carbon Nanotubes. Nat. Photonics 2011, 5 (11), 672-676, DOI: 10.1038/nphoton.2011.250. (6) Wang, F.; Dukovic, G.; Brus, L. E.; Heinz, T. F. The Optical Resonances in Carbon Nanotubes Arise from Excitons. Science 2005, 308 (5723), 838-841. (7) Perebeinos, V.; Tersoff, J.; Avouris, P. Scaling of Excitons in Carbon Nanotubes. Phys. Rev. Lett. 2004, 92 (25), 257402. (8) Wang, F.; Wang, S.; Yao, F.; Xu, H.; Wei, N.; Liu, K.; Peng, L.-M. High Conversion Efficiency Carbon Nanotube Based Barrier-Free Bipolar-Diode Photodetector. ACS Nano 2016, DOI: 10.1021/acsnano.6b05047.

22

ACS Paragon Plus Environment

Page 22 of 27

Page 23 of 27 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 Applied Materials & Interfaces

(9) Ozawa, H.; Ide, N.; Fujigaya, T.; Niidome, Y.; Nakashima, N. One-pot Separation of Highly Enriched (6, 5)-single-walled Carbon Nanotubes Using a Fluorene-based Copolymer. Chem. Lett. 2011, 40 (3), 239-241. (10) Nish, A.; Hwang, J.-Y.; Doig, J.; Nicholas, R. J. Highly Selective Dispersion of Singlewalled Carbon Nanotubes Using Aromatic Polymers. Nat. Nanotech. 2007, 2, 640-646. (11) Gu, J.; Han, J.; Liu, D.; Yu, X.; Kang, L.; Qiu, S.; Jin, H.; Li, H.; Li, Q.; Zhang, J. Solution‐Processable High‐Purity Semiconducting SWCNTs for Large‐Area Fabrication of High‐Performance Thin‐Film Transistors. Small 2016, 12 (36), 4993-4999. (12) Lee, H. W.; Yoon, Y.; Park, S.; Oh, J. H.; Hong, S.; Liyanage, L. S.; Wang, H.; Morishita, S.; Patil, N.; Park, Y. J. Selective Dispersion of High Purity Semiconducting Single-walled Carbon Nanotubes with Regioregular Poly (3-alkylthiophene) s. Nat. Commun. 2011, 2, 541. (13) Chen, F.; Wang, B.; Chen, Y.; Li, L.-J. Toward the Extraction of Single Species of Singlewalled Carbon Nanotubes Using Fluorene-based Polymers. Nano Lett. 2007, 7 (10), 3013-3017. (14) Mistry, K. S.; Larsen, B. A.; Blackburn, J. L. High-yield Dispersions of Large-diameter Semiconducting Single-walled Carbon Nanotubes with Tunable Narrow Chirality Distributions. ACS Nano 2013, 7 (3), 2231-2239. (15) Lei, T.; Chen, X.; Pitner, G.; Wong, H.-S. P.; Bao, Z. Removable and Recyclable Conjugated Polymers for Highly Selective and High-yield Dispersion and Release of Low-cost Carbon Nanotubes. J. Am. Chem. Soc. 2016, 138 (3), 802-805. (16) Wang, C.; Zhang, J.; Ryu, K.; Badmaev, A.; De Arco, L. G.; Zhou, C. Wafer-scale Fabrication of Separated Carbon Nanotube Thin-film Transistors for Display Applications. Nano Lett. 2009, 9 (12), 4285-4291. (17) Liyanage, L. S.; Lee, H.; Patil, N.; Park, S.; Mitra, S.; Bao, Z.; Wong, H.-S. P. Waferscale Fabrication and Characterization of Thin-film Transistors with Polythiophene-sorted Semiconducting Carbon Nanotube Networks. ACS Nano 2011, 6 (1), 451-458.

23

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 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 24 of 27

(18) Chen, B.; Zhang, P.; Ding, L.; Han, J.; Qiu, S.; Li, Q.; Zhang, Z.; Peng, L.-M. Highly Uniform Carbon Nanotube Field-effect Transistors and Medium Scale Integrated Circuits. Nano Lett. 2016, 16 (8), 5120-5128. (19) Yang, Y.; Ding, L.; Han, J.; Zhang, Z.; Peng, L.-M. High-performance Complementary Transistors and Medium-scale Integrated Circuits Based on Carbon Nanotube Thin Films. ACS Nano 2017, 11 (4), 4124-4132. (20) Zhong, D.; Zhang, Z.; Ding, L.; Han, J.; Xiao, M.; Si, J.; Xu, L.; Qiu, C.; Peng, L.-M. Gigahertz Integrated Circuits Based on Carbon Nanotube Films. Nat. Electronics 2018, 1, 40-45. (21) Geier, M. L.; McMorrow, J. J.; Xu, W.; Zhu, J.; Kim, C. H.; Marks, T. J.; Hersam, M. C. Solution-processed Carbon Nanotube Thin-film Complementary Static Random Access Memory. Nat. Nanotech. 2015, 10, 944-948. (22) Liu, Y.; Wei, N.; Zeng, Q.; Han, J.; Huang, H.; Zhong, D.; Wang, F.; Ding, L.; Xia, J.; Xu, H. Room Temperature Broadband Infrared Carbon Nanotube Photodetector with High Detectivity and Stability. Adv. Opt. Mater. 2016, 4, 238-245. (23) Huang, H.; Wang, F.; Liu, Y.; Wang, S.; Peng, L.-M. Plasmonic Enhanced Performance of an Infrared Detector Based on Carbon Nanotube Films. ACS Appl. Mater. Interfaces 2017, 9 (14), 12743-12749. (24) Brady, G. J.; Way, A. J.; Safron, N. S.; Evensen, H. T.; Gopalan, P.; Arnold, M. S. Quasiballistic carbon nanotube array transistors with current density exceeding Si and GaAs. Sci. Adv. 2016, 2 e1601240. (25) Yu, X.; Liu, D.; Kang, L.; Yang, Y.; Zhang, X.; Lv, Q.; Qiu, S.; Jin, H.; Song, Q.; Zhang, J.; Li, Q. Recycling Strategy for Fabricating Low-Cost and High-Performance Carbon Nanotube TFT

Devices.

ACS

Appl.

Mater.

Interfaces

2017,

9(18),

15719-15726,

DOI:

10.1021/acsami.7b02964. (26) Joo, Y.; Brady, G. J.; Kanimozhi, C.; Ko, J.; Shea, M. J.; Strand, M. T.; Arnold, M. S.; Gopalan, P. Polymer-Free Electronic-Grade Aligned Semiconducting Carbon Nanotube Array. ACS Appl. Mater. Interfaces 2017, 9 (34), 28859-28867, DOI: 10.1021/acsami.7b06850. 24

ACS Paragon Plus Environment

Page 25 of 27 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 Applied Materials & Interfaces

(27) Park, S.; Lee, H. W.; Wang, H.; Selvarasah, S.; Dokmeci, M. R.; Park, Y. J.; Cha, S. N.; Kim, J. M.; Bao, Z. Highly Effective Separation of Semiconducting Carbon Nanotubes Verified via Short-channel Devices Fabricated Using Dip-pen Nanolithography. ACS Nano 2012, 6 (3), 2487-2496. (28) Wang, Z. X; Xu, H. L.; Zhang, Z. Y.; Wang, S.; Ding, L.; Zeng, Q. S.; Yang, L. J.; Pei, T.; Liang, X. L.; Gao, M.; Peng, L.-M. Growth and Performance of Yttrium Oxide As an Ideal High-κ Gate Dielectric for Carbon-Based Electronics. ACS Nano 2010, 10 (6), 2024-2030. (29) Taherian, F.; Marcon, V.; van der Vegt, N. F. A.; Leroy, F. What Is the Contact Angle of Water on Graphene? Langmuir 2013, 29, 1457−1465. (30) Lee, J. U.; Gipp, P.; Heller, C. Carbon Nanotube p-n Junction Diodes. Appl. Phys. Lett. 2004, 85 (1), 145-147. (31) Lee, J. U. Photovoltaic Effect in Ideal Carbon Nanotube Diodes. Appl. Phys. Lett. 2005, 87 (7), 073101. (32) Malapanis, A.; Perebeinos, V.; Sinha, D. P.; Comfort, E.; Lee, J. U. Quantum Efficiency and Capture Cross Section of First and Second Excitonic Transitions of Single-walled Carbon Nanotubes Measured Through Photoconductivity. Nano Lett. 2013, 13 (8), 3531-3538. (33) Malapanis, A.; Jones, D. A.; Comfort, E.; Lee, J. U. Measuring Carbon Nanotube Band Gaps Through Leakage Current and Excitonic Transitions of Nanotube Diodes. Nano Lett. 2011, 11 (5), 1946-1951. (34) Chang, S.-W.; Hazra, J.; Amer, M.; Kapadia, R.; Cronin, S. B. A Comparison of Photocurrent Mechanisms in Quasi-metallic and Semiconducting Carbon Nanotube pn-junctions. ACS Nano 2015, 9 (12), 11551-11556. (35) Jeon, S.; Ahn, S.-E.; Song, I.; Kim, C. J.; Chung, U. I.; Lee, E.; Yoo, I.; Nathan, A.; Lee, S.; Robertson, J.; Kim, K. Gated Three-terminal Device Architecture to Eliminate Persistent Photoconductivity in Oxide Semiconductor Photosensor Arrays. Nat. Mater. 2012, 11 (4) 301– 305.

25

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 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

(36) Khasminskaya, S.; Pyatkov, F.; Flavel, B. S.; Pernice, W. H.; Krupke, R. Waveguide‐Integrated Light‐Emitting Carbon Nanotubes. Adv. Mater. 2014, 26 (21), 3465-3472. (37) Pyatkov, F.; Fütterling, V.; Khasminskaya, S.; Flavel, B. S.; Hennrich, F.; Kappes, M. M.; Krupke, R.; Pernice, W. H. P. Cavity-enhanced Light Emission from Electrically Driven Carbon Nanotubes. Nat. Photonics 2016, 10, 420-427, DOI: 10.1038/nphoton.2016.70. (38) Khasminskaya, S.; Pyatkov, F.; Słowik, K.; Ferrari, S.; Kahl, O.; Kovalyuk, V.; Rath, P.; Vetter, A.; Hennrich, F.; Kappes, M. M.; Gol'tsman, G.; Korneev, A.; Rockstuhl, C.; Krupke, R.; Pernice, W. H. P. Fully Integrated Quantum Photonic Circuit with an Electrically Driven Light Source. Nat. Photonics 2016, 10, 727-732, DOI: 10.1038/nphoton.2016.178. (39) Liu, Y.; Zhang, J.; Liu, H.; Wang, S.; Peng, L.-M. Electrically Driven Monolithic Subwavelength Plasmonic Interconnect Circuits. Sci. Adv. 2017, 3 (10), e1701456. (40) Liu, Y.; Wang, S.; Liu, H.; Peng, L.-M. Carbon Nanotube-based Three-dimensional Monolithic Optoelectronic Integrated System. Nat. Commun. 2017, 8, 15649.

26

ACS Paragon Plus Environment

Page 26 of 27

Page 27 of 27 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 Applied Materials & Interfaces

TOC

27

ACS Paragon Plus Environment