Reproducible, High Performance Fully Printed Photodiodes on

Subscriber access provided by University of South Dakota

Organic Electronic Devices

Reproducible, High Performance Fully-Printed Photodiodes on Flexible Substrates Through the Use of a Polyethylenimine Interlayer Matteo Cesarini, Biagio Brigante, Mario Caironi, and Dario Natali ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.8b07542 • Publication Date (Web): 04 Sep 2018 Downloaded from http://pubs.acs.org on September 8, 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.

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

Reproducible, High Performance Fully-Printed Photodiodes on Flexible Substrates Through the Use of a Polyethylenimine Interlayer Matteo Cesarini†, Biagio Brigante†, Mario Caironi†, Dario Natali†,‡,*

†

Center for Nano Science and Technology @PoliMi, Istituto Italiano di Tecnologia Via Pascoli 70/3, 20133 Milano, Italy

‡

Dipartimento di Elettronica, Informazione e Bioingegneria, Politecnico di Milano P.za L. da Vinci, 32, 20133 Milano, Italy

Keywords: fully-printed, organic photodetector, PEI, reproducibility, inkjet printing

Abstract. This paper investigates with a statistical analysis the issue of performance reproducibility and optimization in fully inkjet-printed organic photodetectors on flexible substrates. The most crucial process step to obtain reproducible, well performing devices with a high process yield turns out to be the printing of the thin polyethylenimine interlayer used as a surface modifier for the bottom electrode. Controlling solution composition and deposition

ACS Paragon Plus Environment

1

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 2 of 27

parameters for this layer, a 57 nAcm-2 reverse dark current was achieved, with an outstanding standard deviation as low as 15 nAcm-2, with dramatic improvements in process yield (from less than 20% to over 90%). Device performance in terms of dark currents, EQE (from 50% up to 90% at 525 nm, depending on process) and rectification (ratio between forward current and reverse current over 104 and reaching 105 in the best cases) is among the best for fully-printed detectors. Furthermore, the importance of relative humidity control in the deposition environment during the interlayer deposition on device characteristics is reported, indicating the processing conditions optimal for massive scalability. The overall interlayer optimization approach was applied to a process using universally adopted materials in the organic optoelectronics field, and thus retains relevance on a wide range.

1. Introduction The manifold advantages provided by organic electronics in photodetection are largely acknowledged:1,2 organic semiconductors provide remarkable absorption in the visible wavelength range3,4, tunable optical properties5 and unique processing options. In fact, solution processability makes these materials particularly interesting, allowing near-room temperature, low-cost deposition of thin layers on almost arbitrary and even flexible substrates. Finally, a key feature of this emerging field is the possibility to access patternable and scalable deposition techniques,6 which pave the way to large-area coverage and are potentially compatible with many already assessed industrial processes7 taken from the graphical arts. These features match today’s trend for image sensors and circuits towards low fabrication costs, lightweight, mechanical flexibility, disposability and wearability.8 Thank to these properties, which

cannot be obtained with conventional electronics, a number of possible

applications unfold: from electronic skin8 and printed wearable sensors to smart monitoring of


2

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


light-sensitive goods and integration of “intelligence” in the product package;9 from large-area imaging/detection for security to biology and healthcare.8 Furthermore, the outstanding recent advances in printed organic field-effect transistors (OFETs)10,11 can be a fecund opportunity for the integration of these appealing devices in innovative printed optoelectronic systems. When device fabrication is entirely made by scalable, low cost deposition steps, the process retains all the aforementioned technological advantages. The literature shows some compelling works in the field of printed photodetectors. The paper by Azzellino et al.12 has been among the very first showing a photodiode with all the layers are inkjet printed (apart from the bottom electrode modifier) showing convincing figures of merit (external quantum efficiency in excess of 85% at 525 nm and dark current density of few hundreds of nA·cm-2). Focusing on the best examples of fully-printed photodetectors, in the work by Pierre13 et al. organic photodetectors (OPD) with remarkable detectivity performance (6·1013 cm Hz

0.5

W

−1

) are printed on flexible

substrate by means of scalable techniques such as screen printing and blade/bar coating. In another paper, Lavery14 et al. demonstrate the possibility to inkjet print a polyfluorene based photosensor engineered to work in high illuminance range detection (> 100 klux). Eckstein et al. demonstrated semitransparent, color neutral photodiodes developed by aerosol jet-printing.15 Other recent advances in the field address broadening of the wavelength detector response using small molecules,16 and developing suitable blocking layers to lower dark currents.17 The contributions in the field are still growing,1 driven by the fact that organic photodetectors have reached detectivity performances comparable to their silicon counterparts.13 However, while improved performances are being achieved continuously, an adequate statistical investigation on the maximization of the process yield and optimization of the reproducibility of said performances among devices is generally not addressed in the literature


3


Page 4 of 27

(albeit with some notable exceptions13,18–20). This aspect is of paramount importance when the aim is for instance the development of arrays of devices for imaging applications.15,21,22 This paper shows how reproducibility and performance of fully inkjet-printed vertical photodiodes can be tuned, to a large extent, by acting on the solution composition used to deposit an interlayer of a well-known material, namely polyethylenimine (PEI), used to lower the work function of the bottom electrode. The use of aliphatic amine rich polymers, like PEI, as surface modifiers for electronic contacts, was pioneered by Zhou et al.23: this helps cutting down the device dark currents, which are mostly dependent on charge injection in this device architecture,3 and thus improving power detection limit and specific detectivity. Hereby, adopting non-toxic solvents for the printing of PEI, differently from the most common 2-methoxyethanol,23–25 we explore the relationships between the PEI ink compositions, printing parameters and devices reproducibility and performance. We demonstrate devices achieving a mean 57 nAcm-2 reverse dark current, with a standard deviation as low as 15 nAcm-2, with a process yield as high as 90%. For the most reproducible solution composition, the effect of relative humidity values during the interlayer deposition on device characteristics is reported, indicating the optimal processing conditions.


4

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


Figure 1. On the left in Figure 1a: sketch of the device vertical stack. On the right in Figure 1b: top view optical micrograph of a printed device.

2. Experimental Section The photodiode has a vertical structure (sketch reported in Figure 1) with the active layer comprised between two conductive patterns.12 Polyethylene naphthalate (PEN, Teijin DuPont) was chosen as substrate for its bendability and light weight, appealing features in large area applications, and for having a higher resistance to heat and mechanical stress with respect to polyethylene terephthalate (PET).26,27 Each layer was deposited by drop-on-demand inkjet printing in ambient conditions and it was left drying overnight in a nitrogen filled glovebox (showing an oxygen amount around 1 part per million) before printing the next one above. Electrodes are made by inkjet printing commercially available conductive inks, with a poly(3,4ethylenedioxythiophene) poly(styrenesulfonate) (PEDOT:PSS) based formulation. This allows the fabrication of a semi-transparent electrode so that light can reach the photosensitive layer from both the top and the bottom.16 Bottom contacts were printed with a Fujifilm Dimatix DMP2831, equipped with 10 pL jetting volume nozzles loaded with Clevios PJ700. The bottom contact is modified by printing PEI (Sigma Aldrich). Three types of inks were formulated and printed. One ink was based on water as single solvent, and had a PEI content of 0.1 wt% (from


5


Page 6 of 27

now on referred to as PEIH2O). It was printed varying the drop spacing between 35 µm, 45 µm and 55 µm and the layer number between 1 and 4. Two other inks were obtained, keeping the amount of PEI fixed at 0.1wt%, by mixing water, ethanol and ethylene glycol according to the following proportions: 50vol% water, 30vol% ethanol, 20vol% ethylene glycol (from now on referred to as PEI0.3); ii) 50vol% water, 20vol% ethanol, 30vol% ethylene glycol (from now on referred to as PEI0.2). PEI0.2 and PEI0.3 were printed varying the drop spacing between 35 and 45 µm, and the number of layers between 1 and 3. Ink viscosity was measured by means of Fungilab ALPHA L rotational viscometer with low viscosity LCP adapter, obtaining the following results: 1.22 m⋅Pas for PEIH2O, 3.38 mPa⋅s for PEI02, 3.29 mPa⋅s for PEI03. PEI films were characterized by means of Agilent 5500 Atomic Force Microscope operated in acoustic mode AFM; surface profiles were extrapolated using a Lyncee Tec DHM-R2200 reflection holographic microscope. The active layer is made of a blend of a poly(3-hexyltiophene) (P3HT, RR = 96.6%, MW = 65500, Merck) and [6,6]-phenyl-C61 butyric acid methyl ester (PC61BM, purity > 99.5%, Solenne) in a bulk heterojunction, a well established option for both detectors and photovoltaic cells.[3] The blend solution of P3HT:PCBM (1:1) was prepared by mixing 1,2-dichlorobenzene (68 vol%) and mesitylene (32 vol%), both purchased from Sigma Aldrich, and was stirred overnight.12 Before printing with Microfab JETLAB 4 equipped with a 40 µ m diameter DLC nozzle, the blend solution was heated at 80 °C for 10 min and filtered with a 0.20 µm PTFE filter. The solution to print the top contacts was prepared adding Zonyl FS-300 fluorosurfactant to Clevios P Jet 700N (10 wt%).12 Then the solution was filtered with a 0.20 µm PVDF filter before Fujifilm Dimatix 10 pL nozzle cartridge loading. For both contacts the length was kept under 10 squares to avoid noticeable resistive losses. The final active area was determined by the


6

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


overlap of the top and bottom strips and measured 2÷7·10-4 cm2. To have an easier visual recognition of device contacts (difficult due to the semi-transparency of PEDOT:PSS) at the probing station, while keeping ohmic voltage drops low, device current is collected through inkjet printed silver strips, obtained by using Fujifilm Dimatix with 10 pL nozzles cartridge loaded with ANP Silverjet DGP-40LT-15C. After deposition, the silver layer was annealed at 130 °C onto a hot plate in ambient atmosphere for 10 minutes, resulting in a resistivity around 10 µΩcm. Silver strip dimensions are about 80 µm in width and 200 nm in thickness. Before electrical measurements, samples were stored in glove box for at least 24 hours, to allow for desorption of possible moieties absorbed from ambient air. Optoelectronic measurements in dark and under white light were performed in nitrogen glovebox. EQE was measured by reverse biasing the photodiodes at -1 V and illuminating them by a constant power density of 4.3 mWcm-2 at different wavelengths, using a set of LEDs emitting with an average FWHM of 15 nm. The humidifier was purchased from Boneco, and the relative humidity sensors from TFADostmann.

3. Results and discussion 3.1 Polyethylenimine printed by single solvent aqueous solution composition PEI is most commonly dissolved in 2-methoxyethanol, but being this solvent toxic, we decided to use water28 as a safer option. As a starting point, we adopted a concentration of 0.1 wt%, similar to the one adopted by Min et al.28 for spin-coated organic solar cell fabrication, since a reference for inkjet deposition of the material was not found. Adopting this value makes the PEIH2O ink compatible with our printing system and allows to control the quantity of deposited


7


Page 8 of 27

material down to an amount that prevents detrimental effects related to the insulating nature of PEI.23 Printing parameters control is crucial in inkjet deposition, and their impact on device performance becomes even more important in a process step as delicate as the printing of a surface modifier coating on a contact. We varied PEI ink drop spacing between 35 µm, 45 µm or 55 µm and layer number between 1 and 4. A drop spacing of 55 µm was found to be too high to obtain a uniform deposition, even with a higher number of layers. Furthermore, single and double layer PEI resulted in non-rectifying devices, giving high dark current densities (10 mAcm-2 at -1 V) in reverse bias, unpractical for photodetection. This could be due to a deposited material quantity too low to ensure an even coverage of the whole contact surface, leaving quasiohmic paths in the vertical stack. Printing 3 or 4 layers, working devices with good performances were obtained (EQE around 90% at 525 nm and dark current below few hundreds of nAcm-2 range, more details in the Supplementary Information). But despite our efforts, irrespectively of the combination of layer numbers and drop spacing, this ink composition was indeed poor in terms of printing efficiency and fluid jetting stability, resulting in a low overall device yield. Considering as functional only those detectors showing a reverse photocurrent (under 1 mW cm-2 white light illumination) over dark current ratio (measured at -1 V) higher than 104 , and a reverse dark current density below 100 nAcm-2, these inks showed a yield below 20% (calculated on samples of 20 devices). This situation can be reasonably traced back to two main reasons. First, having to print at least 3 PEI layers to get working devices, printed films are liable to thickness non-uniformities typical of solute/single solvent system.29–31 This is confirmed by surface profile measurements (see Supporting Information), showing a sizeable coffee stain effect. A second element likely related to the issue, especially regarding fluid jetting stability and


8

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


printing failure rate, could be an excessively low ink viscosity of PEIH2O, pushed too far in the low end of the suitable range for inkjet printing (around 1 cP).31 For the aforementioned reasons, new solution compositions for the interlayer material deposition were formulated.

3.2 Polyethylenimine printed by multiple solvent aqueous solution composition 3.2.1 Ink composition rationale The adopted strategy was to engineer a ternary solvent system. In the literature, a secondary solvent32 is usually added in order to exploit the Marangoni effect. In this approach, a surface tension gradient is imposed on the liquid to compensate for the solute tendency to migrate towards droplet borders, and thus to obtain a smoother material film.29–31 The secondary solvent should have a lower surface tension and a higher boiling point with respect to the primary one. Furthermore, the additive should be miscible with the main solvent, and compatible with the material to be dissolved. Another approach consists of adding a viscous and high boiling point solvent and a low-viscosity and low-boiling point solvent: in this ternary system, a fast increase in the viscosity of the printed droplet occurs, which helps suppressing the coffee stain effect.33 We actually adopted this latter strategy. We chose: ethylene glycol as the viscous and highboiling point (197.3 °C) component, which is also completely miscible with water,34 and compatible with branched (as the one used in this work) PEI; and ethanol as low-viscosity and low-boiling point (78.37 °C) solvent. Being the three solvents completely miscible, two different compositions were formulated, both with the same overall solute concentration (0.1 wt%) and the same primary solvent (water, 50 vol%), but with different additive ratios. The formulated compositions were: i) PEI0.3, with 30vol% ethanol and 20vol% ethylene glycol; ii) PEI0.2, with 20vol% ethanol and 30vol% ethylene glycol. Multi-solvent ink showed a higher viscosity than


9


Page 10 of 27

the single solvent one, being better compatible with the rheological requisites of inkjet printing technique (see Section 2), and resulted in a noticeable reduction of non-uniformities for printed PEI films (see Supporting Information). This yielded a better interlayer, and devices with improved performance and reproducibility, as shown in next sections.

3.2.2 Multi-solvent ink performance and role of printing parameter optimization

The importance of printing parameters in performance and reproducibility is confirmed also for multi-solvent compositions. Detector performances were analyzed with respect to number of printed layers and drop spacing. A single layer gave bad devices, showing high dark currents, comparable to the non-rectifying devices mentioned in section 3.1. With 3 printed layers, PEI acted as an insulator, resulting in devices with negligible currents both in reverse and forward bias. This has to be compared to the single solvent ink case, where 3 printed layers gave working devices. In the following, we thus focus on 2 layer devices and compare 35 µm and 45 µm drop spacings. First and foremost, it should be pointed out that multisolvent inks drastically improved process yield. Applying the same criterion of Section 3.1 to label devices as working or not, the yield improved from less than 20% to over 90% (evaluated on samples of 24 detectors). For PEI0.3 solution, the measured dark current values at -1 V reverse bias were below 100 nAcm-2 (except for 2 cases on a total of 24 samples), with slightly different results between 35 µm and 45 µm drop spacings. The boxplot in Figure 2 shows good dark current performances for both drop spacings at -1 V reverse bias, with a slower mean value and dispersion in the 45


10

Page 11 of 27

µm case (35 µm: mean Jdk 75.9 nAcm-2, standard deviation 38.1 nAcm-2; 45 µm: mean Jdk 56.9 nAcm-2, standard deviation 14.9 nAcm-2). For PEI0.2 solution, a 45 µm drop spacing is definitely inadequate, resulting in high current, short-circuit devices. However, adopting a 35 µm drop spacing, PEI0.2 devices were able to give a mean reverse dark current (-1 V bias) of 68.7 nAcm-2, with a 33.6 nAcm-2 standard deviation.

150

PEI0.3, 2 Layers 2

Jdk rev [nA/cm ]

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


100

50

0

35 µm

45 µm

Figure 2. Boxplot of the current densities at -1 V reverse bias for double layer PEI0.3 devices having 35 µm and 45 µm drop spacings.

The results obtained adopting the respective optimized layer number/drop spacing combinations for PEI0.2 and PEI0.3 are shown and compared in the boxplots of Figure 3 in terms of reverse dark current and photocurrent (-1 V bias) mean value and standard deviation on 34 samples. Both compositions were able to give devices showing a photocurrent over dark current ratio between 104 and 105 under white light (incident power density 1 mWcm-2). But the composition resulting in the best compromise in terms of reproducibility and performance is PEI0.3 printed in


11


2 layers, using a 45 µm drop spacing. This ink allowed for a reverse dark current mean value (-1 V bias) of 57 nAcm-2, with a standard deviation as low as 15 nAcm-2. In turn, devices with PEI0.2 printed in

a)

2 Layers

Jph rev [mA/cm2]

200

Jdk rev [nA/cm2]

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 27

150 100 50 0

PEI0.3 45 µm

PEI0.2 35 µm

1.8

b)

2 Layers

1.5 1.2 0.9 0.6 0.3 0.0

PEI0.3 45 µm

PEI0.2 35 µm

Figure 3. Comparison between devices with interlayer printed using the best parameter recipes for PEI0.3 (2 layers, 45 µm drop spacing) and PEI0.2 (2 layers, 35 µm drop spacing), in terms of: a) reverse dark current densities, and b) reverse photocurrent densities (-1 V bias, photocurrents under 1 mWcm-2). PEI0.3 with 2 layers and 45 µm drop spacing turned out to be the best performing and most reproducible among all the tested combinations, with a 57 nAcm-2 mean reverse dark current value and a corresponding standard deviation of 15 nAcm-2.

2 layers and 35 µm drop spacing shows excellent results (best case rectification ratio of 105, with reverse dark current density of 37 nAcm-2) alternated with poor ones (worst case rectification ratio of 8·103, with reverse dark current density of 127 nAcm-2). But on average, performances are inferior to PEI0.3 and with less reproducibility (Figure 3a). Average EQE plots, showing for both PEI0.2 and PEI0.3- a value of 60% at 470 nm and 50-45% at 525 nm are reported in Figure 4. Best IV curves in the dark and under illumination are reported in Figure 5.


12

Page 13 of 27

Devices show specific detectivities, estimated under the (strong) assumption that shot fluctuations are the dominant noise source, standing at 3÷4·1012 Jones, competitive with the

Average EQE%

70 60

PEI0.2

PEI0.3

50 40 30 20 10 0 450

600

750

450

600

750

λ [nm] Figure 4. External Quantum Efficiency, with corresponding dispersion from a sample of 14(11) PEI0.2(PEI0.3) devices, evaluated under 4.3 mWcm-2 light power density at different wavelengths.

10-3

J [Acm-2]

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


Jph PEI0.2 Jph PEI0.3

10-4 10-5 10-6 10-7 10-8

Jdk PEI0.3 Jdk PEI0.2

10-9 -1.0

-0.5

0.0

0.5

1.0

Reverse Voltage [V] Figure 5. IV plot of the best PEI0.2(PEI0.3) device, in the dark reported as grey(black) curve, and under 1 mWcm-2 reported as orange(red) curve.


13


Page 14 of 27

PEIH2O

PEI0.2

PEI0.3

mean Jdk

188 nAcm-2

69 nAcm-2

57 nAcm-2

Jdk Std. dev. (%)

230 nAcm-2 (122%)

34 nAcm-2 (50%)

15 nAcm-2 (24%)

mean Jph

1.473 mAcm-2

1.05 mAcm-2

1.02 mAcm-2

Jph Std. dev. (%)

566.7 µAcm-2 (38.5%)

277.7 µAcm-2 (26.4%)

197.8 µAcm-2 (19.4%)

Mean EQE

87%

49%

46%

EQE Std. Dev.

25%

5%

8%

mean D*

3.54·1012 Jones

3.3·1012 Jones

3.35·1012 Jones

Table 1. Summary of average values and standard deviation for: reverse dark current and photocurrent (biased at -1 V), EQE, Specific Detectivity (at 525 nm). For reverse dark current and photocurrent the sample size was of 34 devices per ink composition, considering all the drop spacing/layer number combinations that resulted in working devices. For EQE and detectivity the sample size was 13, 11, 14 devices for PEIH2O, PEI0.2 and PEI0.3 respectively. Photocurrent was evaluated under under 1 mWcm-2 white light, and EQE under 4.3 mWcm-2 light at 525 nm. PEI was printed under room RH > 50%.

standard Silicon-based photodetector counterparts. These results are summarized in Table 1. The effectiveness of the multisolvent ink approach in printing a higher grade photodetector interlayer, granting improved performance and process control, can be traced back to the PEI film morphology. While on the tens of micron spatial scale single-solvent and multi solvent inks do not show striking differences (indeed being PEIH2O smoother than multisolvent PEI) as shown by AFM images (see Supporting Information), it is on the hundreds of microns scale that


14

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


advantages of the multisolvent approach become apparent. As shown by surface profile measurements (see Supporting Information), PEIH2O films are highly affected by coffee stain related features, which is not the case for multisolvent films which display a higher uniformity. As a consequence, the multisolvent produces, for the same number of printed layers, thicker films than single solvent approach. This explains why printing three layers of multisolvent ink results in device behaving as open circuits (due to the insulating nature of PEI), whereas the same number of layers printed from PEIH2O gives working devices (albeit with low yield).

3.2.3 Device time constants Device speed (at -1 V bias) is evaluated from the photocurrent decay upon a pulsed beam of LED light (100 µWcm-2 at 570 nm, FWHM = 15 nm), driven by a 2.5 ns fast rise/fall time signal with a pulse width largely exceeding the device time response. Devices were fabricated using the best printing parameter set for each interlayer ink composition. All the devices show an initial fast decrease slope, which becomes increasingly slower over time. An excellent fitting can be obtained using a triple exponential decay (Figure 6). The existence of different time constants could be related to the presence of both bulk trap states, showing a faster carrier release time, and interfacial defects with slower temporal dynamics, as studied in earlier works.35 The extracted time constants (fast, medium and slow) for each composition, averaged on a sample of 10 devices per type, are reported in Table 2. For both PEI0.2 and PEI0.3 the fast time constant is on average in the 2-4 µs range, and the medium time constant is on average about 18 µs (standard deviation about 3 µs); the slow one is about 164 µs for PEI0.2 and about 200 µs for PEI0.3 (standard deviation 13 µs and about 18 µs respectively).


15


Interestingly, both PEI0.2 and PEI0.3 show a faster response than PEIH2O, especially in the medium and slow time constants. The fact that the two slower time constants are more sensitive to the change from PEIH2O to multi-solvent ink for the interlayer, suggests the existence of slow interfacial traps -bottleneck for the overall device speed- that are more sensitive to the way PEI assembles on the bottom contact, whereas bulk traps -less influenced by interlayer depositionhave faster trapping/release dynamics.

0.3

τ2 = 38.78 µs τ3 = 300.96 µs

0.2

0.4 0.3

b) PEI0.2 τ1 = 2.77 µs

0.2

τ3 = 144.73 µs

0.1

0.1

0.0

0.0 0

200 400 600 800 1000 1200

Time [µs]

τ2 = 18.63 µs

0.5

Vout [V]

0.4

a) PEIH2O τ1 = 5.42 µs

Vout [V]

0.5

Vout [V]

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 27

0.4

c) PEI0.3 τ1 = 1.69 µs

0.3

τ2 = 18.09 µs τ3 = 181.8 µs

0.2 0.1 0.0

0

200 400 600 800 1000 1200

Time [µs]

0

200 400 600 800 1000 1200

Time [µs]

Figure 6. Fall time measurement curves (black lines) and triple exponential fittings (red curves) performed on devices from each ink composition. The plots show the fall in time of the photocurrent, transduced into a voltage signal through a fast transimpedance amplifier. Interlayers were printed with optimized drop spacing/layer number combination: a) PEIH2O with 35 µm and 4 layers, b) PEI0.2 with 35 µm and 2 layers, c) PEI0.3 with 45 µm and 2 layers). Each ink was printed under room RH > 50%. Measurements were taken illuminating devices with a 100 µWcm-2 light pulse at 570 nm, while reverse biased at -1 V. The plotted curves are taken from the more representative devices of the time constant statistic summarized in Table 2. Extracted time constants for the three devices are included in each plot.


16

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


PEIH2O, 35 µm, 4 L

PEI0.2, 35 µm, 2 L

PEI0.3, 45 µm, 2 L

Avg. Time Constant

Time Const. Std. Dev.

Time Const. Std. Dev. %

τ1 = 3.34 µs

1.55 µs

46.4%

τ2 = 28.97 µs

8.69 µs

30%

τ3 = 272.18 µs

31.9 µs

11.7%

τ1 = 2.45 µs

0.77 µs

31.4%

τ2 = 18.14 µs

2.57 µs

14.2%

τ3 = 163.94 µs

13 µs

7.9%

τ1 = 1.89 µs

0.76 µs

40.2%

τ2 = 18.34 µs

3 µs

16.4%

τ3 = 199.75 µs

17.61 µs

8.8%

Table 2. Statistical results on the time constants extracted from a triple exponential fitting of the device time response, on a sample of 10 for each interlayer ink composition. Each ink was printed using optimized drop spacing and deposited layer number, under room RH > 50%. Measurements were taken illuminating devices with a 100 µWcm-2 light power at 570 nm, while reverse biased at -1 V, collecting the photocurrent through a fast transimpedance amplifier.


17


3.2.4 Relative Humidity and optimal processing window Finally, we investigated the process window in terms of the relative humidity of the deposition environment. Advantageously, as it will be shown, the process favorable humidity window is fairly large, presenting a lower bound (excessively low RH values are detrimental). The results shown in the preceding sections are obtained for relative humidity values higher

3x104

a)

PEI0.3

2x104 1x104 0

RH = 36%

RH > 50%

Rectification ratio (a. u.)

than 50% (we tested up to 80%). At an intermediate value of RH = 36% the most noticeable

Light/Dark ratio (a. u.)

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 27

105

b)

PEI0.3

104 103 102

RH = 36%

RH > 50%

Figure 7. Boxplot showing a) photocurrent over dark current ratio at – 1 V bias and b) rectification ratio (1 V forward versus 1 V reverse current), measured under the same conditions on devices with PEI0.3 interlayer printed with 45 µm drop spacing at relative humidity f 36% (red) and 60% (blue). Data was collected from a sample of 20 devices. Photocurrent was evaluated under under 1 mWcm-2 white light.

effect is a loss of a factor 10 in the rectification ratio, mainly due to a loss in the average dark forward current. At 15% RH, every printed detector loses almost completely its rectifying capability and, ultimately, its suitability for light detection, having too high reverse dark currents. More details are reported in the Supplementary Information. Figure 7 shows the improvements


18

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


obtained in terms of rectification ratio and light (1 mWcm-2 white lamp) over dark ratio on printed PEI0.3 devices passing from RH = 36% to RH > 50%. The reason behind this could be correlated to differences in the evaporation rate of the aqueous portion of the solution. At lower ambient humidity, water (the primary solvent) tends to evaporate faster, so it can be hypothesized that PEI is not able to distribute with the best uniformity on the bottom contact. This is in agreement with the fact that the same behavior was observed in the PEIH2O composition (not shown).

4. Conclusions The present work studies polyethylenimine interlayer deposition parameters in the inkjetprinting of organic vertical photodetectors on flexible PEN substrates, finding their heavy impact on the reproducibility of device performance. The water-based, single solvent solution printing approach is compared to a multi-solvent one for this layer, avoiding the use of hazardous additives and elucidating the criteria behind ink formulation. Most notably, the multi-solvent approach gives drastic improvements in yield (from less than 20% to over 90%), which is of paramount importance for massive process scalability. From a morphological point of view, by means of multi-solvent inks the coffee stain effect which plagues films obtained by water-based inks is drastically suppressed.

In addition, in the multi-solvent approach dark currents are

reduced not only in their average value, but also in terms of their standard deviation. In fact, the best multi-solvent based device set, obtained by printing 2 PEI layers with 45 µm drop spacing from an ink based on water, ethanol and ethylene glycol (50 vol%, 30 vol%, 20 vol% respectively), yielded reverse dark current values as low as 57 nAcm-2 in mean and as low as 15 nAcm-2 in standard deviation. These values have to be compared to a mean dark current of 188


19


Page 20 of 27

nAcm-2 with a standard deviation of 229 nAcm-2 in the single-solvent based approach. The multisolvent approach greatly increases device reproducibility, and indeed not only the in terms of dark current but for all the device parameters which have been analyzed (EQE, time constants etc.). Performance reproducibility is a key parameter for the development of a robust and powerful technology. In fact, enabling the possibility of effectively simulating the electrical behavior of organic optoelectronic circuits, it opens the way to the design of complex and large systems by means of computer aided tools. Finally, the multi-solvent approach produces devices with a faster response, a result that is likely related to a better interface between the polyethylenimine interlayer and the photoactive layer. In addition, the optimal value for relative humidity during PEI deposition step is studied, conveniently finding a relatively large processing window (RH > 50%). The proposed optimization approach was demonstrated on devices based on PEDOT:PSS, PEI and P3HT:PCBM, which are widely used, commercially available model materials in the solution processed device context and therefore can be pertinent on a wide range. An accurate, statistical study on the optimization of fully printed photodetector performance reproducibility is one of the fundamental, and often neglected, aspects that can help bring this promising technology towards massive scaling.

ASSOCIATED CONTENT Supporting Information


20

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


AFM and surface profile of PEI films. Effect of drop spacing and layer number in interlayer deposition in PEIH2O. Single solvent and multiple solvent general comparison. Effect of relative humidity during PEI printing.

AUTHOR INFORMATION Corresponding Author * E-mail: [email protected]

ACKNOWLEDGMENT The authors acknowledge the help of F. Scuratti for AFM measurements.


21


Page 22 of 27

REFERENCES (1)

García de Arquer, F. P.; Armin, A.; Meredith, P.; Sargent, E. H. Solution-Processed Semiconductors for next-Generation Photodetectors. Nat. Rev. Mater. 2017, 2 (3), 16100.

(2)

Jansen-van Vuuren, R. D.; Armin, A.; Pandey, A. K.; Burn, P. L.; Meredith, P. Organic Photodiodes: The Future of Full Color Detection and Image Sensing. Adv. Mater. 2016, 28 (24), 4766–4802.

(3)

Baeg, K.-J. K.-J. K.-J.; Binda, M.; Natali, D.; Caironi, M.; Noh, Y.-Y. Y.-Y. Organic Light Detectors: Photodiodes and Phototransistors. Adv. Mater. 2013, 25 (31), 4267–4295.

(4)

Lin, Y.; Li, Y.; Zhan, X. Small Molecule Semiconductors for High-Efficiency Organic Photovoltaics. Chem. Soc. Rev. 2012, 41 (11), 4245–4272.

(5)

Lee, K. H.; Lee, G. H.; Leem, D. S.; Lee, J.; Chung, J. W.; Bulliard, X.; Choi, H.; Park, K. B.; Kim, K. S.; Jin, Y. W.; et al. Dynamic Characterization of Green-Sensitive Organic Photodetectors Using Nonfullerene Small Molecules: Frequency Response Based on the Molecular Structure. J. Phys. Chem. C 2014, 118 (25), 13424–13431.

(6)

Wu, W. Inorganic Nanomaterials for Printed Electronics: A Review. Nanoscale 2017, 9 (22), 7342–7372.

(7)

Abbel, R.; Teunissen, P.; Eric, R.; Tim, van L.; Chachois, R.; Everaars, M.; Valeton, J.; van de Geijn, S.; Groen, P. Industrial-Scale Inkjet Printed Electronics Manufacturing — Production up-Scaling from Concept Tools to a Roll-to-Roll Pilot Line. Transl. Mater. Res. 2014, No. 1, 015002.

(8)

Yokota, T.; Zalar, P.; Kaltenbrunner, M.; Jinno, H.; Matsuhisa, N.; Kitanosako, H.; Tachibana, Y.; Yukita, W.; Koizumi, M.; Someya, T. Ultraflexible Organic Photonic Skin. Sci. Adv. 2016, 2 (4), e1501856–e1501856.


22

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


(9)

Kuswandi, B.; Wicaksono, Y.; Jayus; Abdullah, A.; Heng, L. Y.; Ahmad, M. Smart Packaging: Sensors for Monitoring of Food Quality and Safety. Sens. Instrum. Food Qual. Saf. 2011, 5 (3–4), 137–146.

(10)

Perinot, A.; Kshirsagar, P.; Malvindi, M. A.; Pompa, P. P.; Fiammengo, R.; Caironi, M. Direct-Written Polymer Field-Effect Transistors Operating at 20 MHz. Sci. Rep. 2016, 6 (1), 38941.

(11)

Naber, R. C. G.; Tanase, C.; Blom, P. W. M.; Gelinck, G. H.; Marsman, A. W.; Touwslager, F. J.; Setayesh, S.; De Leeuw, D. M. High-Performance Solution-Processed Polymer Ferroelectric Field-Effect Transistors. Nat. Mater. 2005, 4 (3), 243–248.

(12)

Azzellino, G.; Grimoldi, A.; Binda, M.; Caironi, M.; Natali, D.; Sampietro, M. Fully Inkjet-Printed Organic Photodetectors with High Quantum Yield. Adv. Mater. 2013, 25 (47), 6829–6833.

(13)

Pierre, A.; Deckman, I.; Lechêne, P. B.; Arias, A. C. High Detectivity All-Printed Organic Photodiodes. Adv. Mater. 2015, 27 (41), 6411–6417.

(14)

Lavery, L. L.; Whiting, G. L.; Arias, A. C. All Ink-Jet Printed Polyfluorene Photosensor for High Illuminance Detection. Org. Electron. 2011, 12 (4), 682–685.

(15)

Eckstein, R.; Rödlmeier, T.; Glaser, T.; Valouch, S.; Mauer, R.; Lemmer, U.; HernandezSosa, G. Aerosol-Jet Printed Flexible Organic Photodiodes: Semi-Transparent, Color Neutral, and Highly Efficient. Adv. Electron. Mater. 2015, 1 (8), 1500101.

(16)

Pace, G.; Grimoldi, A.; Natali, D.; Sampietro, M.; Coughlin, J. E.; Bazan, G. C.; Caironi, M. All-Organic and Fully-Printed Semitransparent Photodetectors Based on Narrow Bandgap Conjugated Molecules. Adv. Mater. 2014, 26 (39), 6773–6777.

(17)

Grimoldi, A.; Colella, L.; La Monaca, L.; Azzellino, G.; Caironi, M.; Bertarelli, C.; Natali,


23


Page 24 of 27

D.; Sampietro, M. Inkjet Printed Polymeric Electron Blocking and Surface Energy Modifying Layer for Low Dark Current Organic Photodetectors. Org. Electron. 2016, 36, 29–34. (18)

Deckman, I.; Lechêne, P. B.; Pierre, A.; Arias, A. C. All-Printed Full-Color Pixel Organic Photodiode Array with a Single Active Layer. Org. Electron. physics, Mater. Appl. 2018, 56, 139–145.

(19)

Pierre, A.; Arias, A. C. High-Detectivity Printed Organic Photodiodes for Large Area Flexible Imagers. In Technical Digest - International Electron Devices Meeting, IEDM; IEEE, 2017; p 32.4.1-32.4.4.

(20)

Eckstein, R.; Strobel, N.; Rödlmeier, T.; Glaser, K.; Lemmer, U.; Hernandez-Sosa, G. Fully Digitally Printed Image Sensor Based on Organic Photodiodes. Adv. Opt. Mater. 2018, 6 (5), 1–6.

(21)

Gelinck, G. H.; Kumar, A.; Moet, D.; Van Der Steen, J.-L.; Shafique, U.; Malinowski, P. E.; Myny, K.; Rand, B. P.; Simon, M.; Rütten, W.; et al. X-Ray Imager Using Solution Processed Organic Transistor Arrays and Bulk Heterojunction Photodiodes on Thin, Flexible Plastic Substrate. Org. Electron. physics, Mater. Appl. 2013, 14 (10), 2602–2609.

(22)

Malinowski, P. E.; Vicca, P.; Willegems, M.; Schols, S.; Cheyns, D.; Smout, S.; Ameys, M.; Myny, K.; Vaidyanathan, S.; Martin, E.; et al. Organic Imager on Readout Backplane Based on TFTs with Cross-Linkable Dielectrics. IEEE Photonics Technol. Lett. 2014, 26 (21), 2197–2200.

(23)

Zhou, Y.; Fuentes-Hernandez, C.; Shim, J.; Meyer, J.; Giordano, A. J.; Li, H.; Winget, P.; Papadopoulos, T.; Cheun, H.; Kim, J.; et al. A Universal Method to Produce Low-Work Function Electrodes for Organic Electronics. Science 2012, 336 (6079), 327–332.


24

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


(24)

Kim, Y. H.; Han, T. H.; Cho, H.; Min, S. Y.; Lee, C. L.; Lee, T. W. Polyethylene Imine as an Ideal Interlayer for Highly Efficient Inverted Polymer Light-Emitting Diodes. Adv. Funct. Mater. 2014, 24 (24), 3808–3814.

(25)

Li, Z.; Qin, F.; Liu, T.; Ge, R.; Meng, W.; Tong, J.; Xiong, S.; Zhou, Y. Optical Properties and Conductivity of PEDOT:PSS Films Treated by Polyethylenimine Solution for Organic Solar Cells. Org. Electron. physics, Mater. Appl. 2015, 21, 144–148.

(26)

Nakamae, K.; Nishino, T.; Tada, K.; Kanamoto, T.; Ito, M. Elastic Modulus of the Crystalline Regions of Poly(Ethylene-2,6-Naphthalate). Polymer (Guildf). 1993, 34 (15), 3322–3324.

(27)

Hamano, H.; Hosoi, M.; Saeki, Y.; Kobayashi, I.; Etchu, M. On The New Type Film (Teijin "Teonex®" Film) Properties For The High Density Magnetic Recording Substrata. J. Magn. Soc. Japan 1991, 15 (S_2_PMRC_91), S2_435-440.

(28)

Min, X.; Jiang, F.; Qin, F.; Li, Z.; Tong, J.; Xiong, S.; Meng, W.; Zhou, Y. Polyethylenimine Aqueous Solution: A Low-Cost and Environmentally Friendly Formulation to Produce Low-Work-Function Electrodes for Efficient Easy-to-Fabricate Organic Solar Cells. ACS Appl. Mater. Interfaces 2014, 6 (24), 22628–22633.

(29)

Girotto, C.; Moia, D.; Rand, B. P.; Heremans, P. High-Performance Organic Solar Cells with Spray-Coated Hole-Transport and Active Layers. Adv. Funct. Mater. 2011, 21 (1), 64–72.

(30)

Lim, J. A.; Lee, W. H.; Lee, H. S.; Lee, J. H.; Park, Y. D.; Cho, K. Self-Organization of Ink-Jet-Printed Triisopropylsilylethynyl Pentacene via Evaporation-Induced Flows in a Drying Droplet. Adv. Funct. Mater. 2008, 18 (2), 229–234.

(31)

Singh, M.; Haverinen, H. M.; Dhagat, P.; Jabbour, G. E. Inkjet Printing-Process and Its


25


Page 26 of 27

Applications. Adv. Mater. 2010, 22 (6), 673–685. (32)

Jung, S.; Sou, A.; Banger, K.; Ko, D. H.; Chow, P. C. Y.; McNeill, C. R.; Sirringhaus, H. All-Inkjet-Printed, All-Air-Processed Solar Cells. Adv. Energy Mater. 2014, 4 (14), 1400432.

(33)

Friederich, A.; Binder, J. R.; Bauer, W. Rheological Control of the Coffee Stain Effect for Inkjet Printing of Ceramics. J. Am. Ceram. Soc. 2013, 96 (7), 2093–2099.

(34)

Olabisi, O.; Robeson, L. M.; Shaw, M. T. Polymer-Polymer Miscibility; Academic Press, 1979.

(35)

Arca, F.; Tedde, S. F.; Sramek, M.; Rauh, J.; Lugli, P.; Hayden, O. Interface Trap States in Organic Photodiodes. Sci. Rep. 2013, 3, 1324.


26

Page 27 of 27

Table of contents

12 10

PEDOT:PSS P3HT:PCBM PEI PEDOT:PSS

count

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


PEI multisolvent

8 6 4

PEI H20

2 0 10

100

Jdark [nA]


27

Reproducible, High Performance Fully Printed Photodiodes on

Recommend Documents