- Email: [email protected]
2 radicals in the processing of organic spintronics
Organic Electronics 55 (2018) 21–25
Contents lists available at ScienceDirect
Organic Electronics journal homepage: www.elsevier.com/locate/orgel
Scavenging of galvinoxyl spin 1/2 radicals in the processing of organic spintronics
T
Jung Min Choa, Chang Eun Songb, Sang-Jin Moonb, Won Suk Shinb,∗∗, Sugyeong Hongc, Sun Hee Kimc, Sanghee Chod, Jung-Keun Leee,∗ a
Research Institute, TOPnC Co., Ltd., Daejeon, 34158, Republic of Korea Center for Solar Energy Materials, Korea Research Institute of Chemical Technology, Daejeon, 34114, Republic of Korea c Western Seoul Center, Korea Basic Science Institute, Seoul, 03759, Republic of Korea d Center for Creative Convergence, Korea Research Institute of Standards and Science, Daejeon, 34113, Republic of Korea e Department of Physics & RIPC, Chonbuk National University, Jeonju, 54896, Republic of Korea b
A R T I C L E I N F O
A B S T R A C T
Keywords: Galvinoxyl ESR Singlet-triplet Spintronics UV–Vis
In many cases of organic electronic devices, spin manipulation adopts galvinoxyl spin 1/2 radicals to achieve singlet-triplet transition for a change of spin ensembles. We report scavenging of galvinoxyl spin 1/2 radicals, both in liquid-galvinoxyl samples and in thermal processing of galvinoxyl-doped organic films, as determined by electron spin resonance and UV–Vis measurements. The two different mechanisms of galvinoxyl scavenging are very crucial in organic device processing that often encounters oxidation and high-temperature treatment (∼150 °C), in which situation most of the galvinoxyl frameworks cannot survive.
1. Introduction Manipulation of spin states in organic electronics has been one of the most interesting research topics in such diverse areas as organic photovoltaics (OPVs) [1,2], batteries, organic light-emitting diodes (OLEDs) [3], magnetic sensing [4], information storage, and quantum information processing [5,6]. An increase of power conversion efficiency (PCE) was ascribed to enhanced spin-spin interaction between the added spin radicals and the photogenerated spin 1/2 species [1,2]. The electron exchange interactions in batteries were determined by the singlet-triplet energy gap [7,8]. The change in singlet-triplet balance induced a current in OLEDs [3,5]. Singlet-triplet correlations were utilized to improve the sensitivity of magnetometers [4,9]. It even appears that nature utilizes organic spintronics in a quantum biological compass that allows organisms to sense Earth's magnetic field [10,11]. The common key ingredient here is the pairs of electronic spins undergoing reactions to a singlet or a triplet state. The polaron-polaron (p-p) pairs may undergo intersystem singlet-triplet transitions. Otherwise, the p-p pairs may undergo coherent spin manipulation with microwave pulses under electron spin resonance (ESR) conditions [3]. On the other hand, spin doping [1,2,6,12,13] has been an alternative method to manipulate such spin ensembles. Among the spin-
∗
doping materials, galvinoxyl (Gx) has been widely known for its spin 1/ 2 radicals, which are inert to oxygen, and stable [14–16]. The Gx was also used in organic radical batteries as active-electrode materials [17,18]. Typically, adding Gx was speculated to change singlet to triplet, and the singlet-to-triplet transition of p-p pairs cause the reduction of P-P recombination in the P3HT/PCBM solar cell [1,2]. However, Cho et al. [19] reported that no triplet excitons were observed by using photo-luminescence detected magnetic resonance (PLDMR) and the enhancement in PCE was also accompanied by the disappearance of Gx spins. In this paper, we report the doublet of quintets ESR hyperfine signals from Gx spin 1/2 radicals, and that the Gx spins disappear not only with oxidation in the liquid sample, but also with a high temperature annealing of up to ∼150 °C, as determined by electron spin resonance and UV–Vis measurements. These are very important results when Gx is used as a spin doping material for organic electronics that accompany oxidation and thermal processing of up to 150 °C. 2. Experimental Galvinoxyl (Gx) powder was supplied in a bottle filled with air (Aldrich). Two kinds of Gx-liquid samples were prepared using 2.1 mg of the Gx powder dissolved into 2 ml of 1,2-dichlorobenzene (ODCB;
Corresponding author. Corresponding author. E-mail addresses: [email protected] (W.S. Shin), [email protected] (J.-K. Lee).
∗∗
https://doi.org/10.1016/j.orgel.2018.01.002 Received 5 June 2017; Received in revised form 23 September 2017; Accepted 1 January 2018 Available online 09 January 2018 1566-1199/ © 2018 Elsevier B.V. All rights reserved.
Organic Electronics 55 (2018) 21–25
J.M. Cho et al.
Metal Inc.) and 0.8 mg galvinoxyl powder (∼3.3 wt%) were mixed with ODCB solution, then stirred at 450 rpm at 40 °C for 12 h in N2. After filtering the P3HT solution mixed with Gx through a 0.45 μm PTFE (Polytetrafluoroethylene) filter, the solution was spin-coated on PET (polyethylene terephthalate) substrate at 400 rpm for 60 s, then at 2000 rpm for 3 s, and finally slowly dried at room temperature (RT) for an hour in a N2 glove box. The thickness of the blend layers was ∼220 nm. ESR measurements were taken using a Bruker EMXplus apparatus (Korean Basic Science Institute) with 100 kHz magnetic-field modulation. The measurements were performed at sufficiently low microwave powers to avoid microwave saturation effects. ESR spectra were recorded for samples in standard ESR quartz tubes in dark. In order to find any variation in the Gx-effects at high temperatures (RT∼150 °C), the Gx:P3HT films have been annealed at differing temperatures (50, 100, and 150 °C) for 30 m and then measured at RT. The Gx:P3HT films were cut into small pieces and then placed in N2sealed ESR quartz tubes for the RT ESR measurements. The UV–Vis measurements were also carried out at RT on a UV2550 UV–vis spectrometer (Shimadzu Corp.) in the wavelength range of 300–1100 nm for the annealed Gx:P3HT films. The low-temperature spin density (Ns) was estimated from the double-integrated ESR intensity of the Gx signals, considering Curie's law. Low-temperature (4–150 K) ESR measurements were reported in our previous work for the Gx powder samples [19]. 3. Results and discussion Fig. 1 shows the ESR spectra obtained from galvinoxyl (Gx) liquid samples. The hyperfine signals shown in Fig. 1 (a)–(c) are known as a “doublet of quintets”. The same spectrum was reported earlier in Gx liquid samples dissolved in ethanolic solution [20]. The hyperfine splitting was ascribed to originate from four equivalent hydrogens situated on aromatic rings in Gx (Scheme 1) (to make quintets), and from the hydrogen of C-H group joining the aromatic rings (to make a doublet). The relative intensity of the lines can be predicted by using Pascal's triangle [21,22]. Here, the intensity ratio would be basically 1:4:6:4:1 for quintets, and 1:1 for the doublet (since proton spin = 1/2; I = 1/2 for hydrogen nuclei). Another point in Fig. 1 is that the “air-sample (solid line)” (which was stirred in air for 30 m, and then kept in N2 until measured in ESR) shows scavenging of signals with time at room temperature (R.T.). There is a huge difference in the decreasing rate of the ESR intensity between the “air-sample” and the “N2-sample” (dashed line). The ESR intensity of the air-sample showed a small reduction after 30 min (Fig. 1 (b)), compared to the N2-sample; and then totally disappeared after about 4 h (Fig. 1 (c)). The color of the air-samples turned to yellow with the reduction of ESR intensity from the initial dark blue. Yet the N2-samples were found to be very stable, and their EPR
Fig. 1. ESR lineshapes obtained from galvinoxyl (Gx)-liquid samples. The Gx-liquid samples were prepared using Gx powder dissolved into ODCB. The “N2-sample” (dashed line), was stirred in N2 atmosphere at 50 °C for 30 m, and was sealed into quartz tubes in N2 atmosphere; while the “air-sample” (solid line), was prepared in the same way except having been stirred in air. The spectrum was measured (a) right after the preparation; (b) after 30 m, and (c) after ∼4 h.
Aldrich). One set of samples was stirred in N2 atmosphere at 50 °C for 30 m; and was sealed into an ESR quartz tube in N2 atmosphere, which was called “N2-sample” (dashed line in Fig. 1); yet another set of samples was prepared in air, called “air-sample” (solid line in Fig. 1). To make Gx-doped P3HT (Gx:P3HT) samples, 24 mg P3HT (Rieke
Scheme 1. Chemical structure of the galvinoxyl.
22
Organic Electronics 55 (2018) 21–25
J.M. Cho et al.
Fig. 2. ESR spectrum obtained from a galvinoxyl (Gx) powder sample as purchased (Aldrich; solid line) and heat-treated (@ 150 °C; dashed line) samples. The ESR was measured at R.T. in N2 atmosphere.
spectra remained unchanged for more than 4 h and even longer; the color did not change, either. Though the precise scavenging mechanism of the air-sample is not described here, this clearly indicates the Gx radical in the ODCB experiences scavenging when prepared in air, but not in N2. Fig. 2 shows the unimodal ESR spectrum (solid line) obtained from Gx powder samples supplied in a bottle filled with air (Aldrich). Disappearance of hyperfine structure will finally lead to a unimodal spectrum as in Fig. 2. The proper distinguishing of hyperfine parameters is not always possible, because hyperlines may be broadened, and in some cases the signals in the center may overlap, and not be separated. In fact, we found that the unimodal ESR line-shape in Fig. 2 changed to hyperfine line-shape in Fig. 1, when the supplied Gx powders were well dissolved and well stirred in ODCB liquid. Meanwhile, after 150 °C annealing of the powder sample in N2 for 30 m, the ESR signal almost disappeared, as shown by the dashed ESR lineshape in Fig. 2. We noted that the g-value changed from g = 2.0054 to g = 2.0045 after the 150 °C annealing. The change of g-value from g = 2.0054 to g = 2.0045 can be attributed to a structural change that became more like an aromatic system. The g-value in a range of 2.0043–2.0047 was usually found for an unpaired electron situated on oxygen substituted to the aromatic ring–the typical structure of poliphenolic antioxidant [20]. Fig. 3 (a) shows the ESR spectrum of a Gx-doped P3HT thin film (Gx:P3HT) sample. The spectrum was obtained at RT in dark. It basically shows two peaks: one is the broad resonance (g = 2.0046) from Gx radicals, and the other sharp one is from the P3HT+ resonance (g = 2.002). The dark ESR signal from P3HT:PCBM OPV was attributed to deep trapped carriers related to oxygen or moisture [23]. The P3HT+ signal from the Gx:P3HT samples in the dark also suggests that part of it was originally charged positive [19]. The g-value (g = 2.0046) of the Gx resonance indicates that the environment of Gx radicals in Gx:P3HT sample is different from that of the pure Gx powder sample (g = 2.0054). Fig. 3 (b) shows the variation of the ESR spectra of Gx:P3HT samples for differing annealing temperatures. Each sample was annealed in N2 atmosphere for 30 m at the designated temperatures. The Gx resonance decreased as the samples were annealed from the RT to 150 °C. The Gx signal decreased distinctively even after 50 °C annealing, and completely disappeared after an annealing step at 150 °C; while the P3HT+ signal remained much the same. The result clearly shows that the number of Gx radicals decreases with increasing annealing temperature. It was previously reported that
Fig. 3. (a) The room temperature ESR spectrum obtained from a Gx:P3HT sample (solid line). Deconvolution of the ESR spectrum (dashed lines) shows a broad Gx resonance (g = 2.0046), and a sharp P3HT+ signal (g = 2.002). (b) The Gx:P3HT ESR spectra for differing annealing temperatures. The samples were thermally treated at 50, 100, and 150 °C for 30 m each and then measured at R.T.
adding Gx (∼3 wt%) in P3HT:PCBM OPV increased PCE by 18%, which was attributed to the unpaired Gx electrons that would react as spin intermediates that change singlet to triplet in the system. However, Cho et al. [19] reported that the enhancement in PCE was observed only for the post-annealed Gx-doped P3HT:PCBM samples; but at the same time, disappearance of Gx spin was also observed for the post-annealed samples. Moreover, no triplet excitons were observed upon photoexcitation of the blend by using PLDMR. Here, Gx disappeared after 150 °C annealing. If this happens, the Gx would not work any more as spin intermediates in Gx-doped materials. The Gx has been known to be inert to oxygen [14–16], and thus stable; but the disappearance of Gx signals with thermal treatment has not previously been emphasized. This is a very important result when Gx is used as a main spin doping material in organic device processing, or in organic spintronics that accompany proper thermal processing. Fig. 4 (a) shows the variation of the Gx spin density (Ns) versus differing treatment temperatures. Fig. 4(a) inset (dashed rectangular box) shows the low temperature regime, where the Gx powder specimens (●) in a quartz ESR tube were measured at the designated low temperatures. The ESR signals were reversible for varying temperatures up and down. The low-temperature measurements were reported in a previous work [19]. For the high temperature regime (RT∼150 °C), however, the 23
Organic Electronics 55 (2018) 21–25
J.M. Cho et al.
Fig. 5. UV–Vis spectra of the Gx-doped P3HT films for differing annealing-temperatures. Arrows indicate the excitation wavelength of 770 nm (left) and 880 nm (right).
environment does not change for the low temperatures. However, for the annealed samples, the g-value decreased. The change of g-value from 2.0054 (RT) to ∼2.004 (150 °C) may indicate that, as the temperature nears the m.p., the Gx structure would change to aromatic system and melt down finally. Thermal processing of organic devices usually experiences 150 °C treatment [1,2,19]. Thus we should be very cautious and ensure that the Gx radicals do not experience high temperature, where they are expected to work as spin intermediates. The effect of annealing of the Gx:P3HT solid film is shown on the UV–Vis absorption spectrum in Fig. 5. The spectrum consists of an intense band in the range between ∼350 and 650 nm and other three weak bands at around 690, 770 and 880 nm. The center-peak around 550 nm was ascribed to the absorption caused by P3HT [24,25]. The decays of the three weak bands at around 690, 770 and 880 nm show distinctive differences between the R.T. sample and the annealed ones (up to 150 °C). The peaks decreased distinctively even after 50 °C and even the peak position shifted toward longer wavelength for the ∼100–150 °C annealing. Previously, the energy transition in the absorption spectrum located at ∼860 nm and 770 nm was assigned to D0 → D3 and D0 → D4 transition, respectively, for isolated Gx radicals in differing solvents [26,27], where D0 is the ground state and Dn is the nth excited state in agreement with quantum chemical calculations [27,28]. Considering that absorption maxima for the radicals may vary according to the surrounding chemical structures (as found in different solvents) [27,29], the 770 nm (left arrow) and the ∼880 nm (right arrow) bands in our absorption spectrum may be related to those transitions for the Gx radicals in the solid P3HT films; the origin of the 690 nm is not clear, though. The peak position even shifted to longer wavelength at 100–150 °C annealing in our UV–Vis spectrum, implying any structural change in the chemical environment, which is consistent with the ESR result where the g-value shifted after the 100–150 °C annealing. Consequently, our UV–Vis absorption data is not inconsistent with our ESR data reflecting the scavenging of Gx radicals after being annealed.
Fig. 4. (a) Inset: The inset (dashed rectangular box): shows the reversible Ns measured at low temperatures. (The low-temperature measurement was reported in a previous study [19]). For the high temperature (above R.T.) regime, the Gx spin density (Ns) was estimated from the separated Gx resonance from the annealed Gx:P3HT samples that were annealed at the designated temperatures. Galvinoxyl's melting point is depicted at 158–159 °C). (b) g value vs. 1/T. The ( ) Gx(P3HT) indicates the Gx radicals in the Gx:P3HT samples, the ( ) Gx(ODCB) indicates the Gx-powder specimen wet with the ODCB, and the Gx(powder) (●) the pure Gx powder samples.
Gx:P3HT samples were annealed at differing temperatures, and then the ESR measurements were performed at RT. The ESR signal was irreversible once the sample was annealed at higher temperatures. For the Gx:P3HT samples, the separated Gx resonance was used for the convoluted ESR spectra of Gx:P3HT samples. We see that the spin density of Gx is fairly stable between room temperature and lower temperatures. However, as the annealing temperature increased from R.T. to ∼150 °C, the Ns decreased irreversibly with increasing temperature. The temperature for the complete disappearance of Ns exactly matches the melting point (158–159 °C) of the Gx. It is regarded that as the temperature nears the m.p., the Gx structure would change to aromatic system, resulting in the loss of Gx spin 1/2 radicals. The activation energy for losing Gx spins was about the order of ΔE = 10–100 meV, depending on the temperature range (RT∼150 °C). In Fig. 4 (b), The variation of g-value for the differing temperature is quite similar to that of the Ns. We see that for the same Gx radical, the g-value is different for the powder samples (●), compared to the case Gx radicals in P3HT ( ). The g-value reflects the atomic environments. The g-values obtained at low temperatures are generally the same as that obtained (g = 2.0054) at RT, indicating the electronic
4. Conclusion In this letter, we report the scavenging of spin 1/2 Gx radicals both in Gx-liquid samples, and in the thermal processing of Gx-doped organic films. We found a doublet of quintets hyperfine signal from the electron spin resonance of galvinoxyl (Gx) spin 1/2 radicals in the liquid samples. The air-sample showed scavenging of signals with time at room 24
Organic Electronics 55 (2018) 21–25
J.M. Cho et al.
temperature. The color of the air-samples turned to yellow with the reduction of ESR intensity from the initial dark blue. However, the N2samples showed persistent ESR intensity for more than 4 h, and even longer; the color did not change, either. Also the unimodal Gx signal decreased with thermal treatment increasing up to 150 °C. As the galvinoxyl-doped P3HT films were heated up to ∼150 °C, the Gx ESR signal in the P3HT:Gx samples disappeared. We note that the loss of Gx paramagnetic signals in the thermal processing of organic device is related to the melting of Gx, in consequence of the radical scavenging of Gx spins. The UV–Vis spectra for the annealed Gx:P3HT samples are consistent with our ESR results. Galvinoxyl (Gx) has been known to be inert to oxygen, and thus stable; but the disappearance of Gx signals in liquid or with thermal treatment has not been emphasized before. This is a very important issue in the processes of organic electronics or spintronics that employ Gx as spin intermediates.
[4] W.J. Baker, K. Ambal, D.P. Waters, R. Baarda, H. Morishita1, K. van Schooten, D.R. McCamey, J.M. Lupton, C. Boehme, Nat. Commun. (2012). [5] P.A. Bobbert, Nature 345 (2014) 1450. [6] A. Atxabal, M. Ribeiro1, S. Parui, L. Urreta, E. Sagasta, X. Sun, R. Llopis, F. Casanova, L.E. Hueso, Nat. Commun. 13751 (2016) 1. [7] W. Walker, S. Grugeon, H. Vezin, S. Laruelle, M. Armand, F. Wudlb, J-M. Tarascon, J. Mater. Chem. 21 (2011) 1615. [8] Y. Su, X. Wang, L. Wang, Z. Zhang, X. Wang, Y. Song, P.P. Power, Chem. Sci. 7 (2016) 6514. [9] M. Alidoust, K. Halterman, J. Linder, Phys. Rev. B 88 (2013) 075435. [10] T. Ritz, R. Wiltschko, P.J. Hore, C.T. Rodgers, K. Stapput, P. Thalau, C.R. Timmel, W. Wiltschko, Biophys. J. 96 (2009) 3451. [11] S. Engels, N.-L. Schneider, N. Lefeldt, C.M. Hein, M. Zapka, A. Michalik, D. Elbers, A. Kittel, P.J. Hore, H. Mouritsen, Nature 509 (2014) 354. [12] T. Moorsom, M. Wheeler, M.T. Khan, F.A. Ma'Mari, G. Burnell, B.J. Hickey, V. Lazarov, D. Gilks, O. Cespedes, Appl. Phys. Lett. 105 (2014) 022408. [13] C.T. Hong, W. Lee, Y.H. Kang, Y. Yoo, J. Ryu, S.Y. Cho, K.-S. Jang, J. Mater. Chem. A 3 (2015) 12314. [14] M.S. Kharasch, B.S. Joshi, J. Am. Chem. Soc. 22 (1957) 1439. [15] G.M. Coppinger, J. Am. Chem. Soc. 79 (1957) 501. [16] K. Mukai, K. Ueda, K. Ishizu, Y. Oeguchi, J. Chem. Phys. 77 (1982) 1. [17] T. Suga, H. Ohshiro, S. Sugita, K. Oyaizu, H. Nishide, Adv. Mater. 21 (2009) 1627. [18] T. Martin, D. Hager, U.S. Schubert, Adv. Mater. 24 (2012) 6397. [19] J.M. Cho, D.S. Kim, S. Bae, S.-J. Moon, W.S. Shin, D.H. Kim, S.H. Kim, A. Sperlich, S. V¨ath, V. Dyakonov, J.-K. Lee, Org. Electron. 27 (2015) 119. [20] M. Jerzykiewicz, I. Ćwieląg-Piasecka, Antioxidative and anticorrosive properties of bioglycerol, in: Gisela Montero (Ed.), Biodiesel- Quality, Emissions and By-products, InTech, 2011. [21] D. Griffiths, Am. J. Phys. 50 (1982) 698. [22] N. Bunce, J. Chem. Educ. 64 (1987) 907. [23] M.D. Heinemann, K. von Maydell, F. Zutz, J. Kolny-Olesiak, H. Borchert, I. Riedel, J. Parisi, Adv. Funct. Mater. 19 (2009) 3788. [24] P. Vanlaeke, A. Swinnen, I. Haeldermans, G. Vanhoyland, T. Aernoutsa, D. Cheynsa, C. Deibela, J. D'Haena, P. Heremansa, J. Poortmansa, J.V. Manca, Sol. Energy Mater. Sol. Cell. 90 (2006) 2150–2158. [25] H. Hintz, H.-J. Egelhaaf, H. Peisert, T. Chassé, Polym. Degrad. Stabil. 95 (2010) 818–825. [26] L.S. Degtyarev, Theor. Exp. Chem. 23 (1988) 682. [27] J. Grilj, C. Zonca, L.M.L. Daku, E. Vauthey, Phys. Chem. Chem. Phys. 14 (2012) 6352–6358. [28] W. Gierke, W. Harrer, B. Kirste, H. Kurreck, J. Reusch, Z. Naturforsch. B Chem. Sci.: Anorg. Chem., Org. Chem. 31 (1976) 965. [29] R. Edge, A. El-Agamey, E.J. Land, S. Navaratnam, T.G. Truscott, Arch. Biochem. Biophys. 458 (2007) 104–110.
Acknowledgement This research was supported by “Research Base Construction Fund Support Program” funded by Chonbuk National University in 2016. This research was also partly supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (2017R1D1A1B03028062). J. M. Cho was supported by the Commercializations Promotion Agency for R &D Outcomes (COMPA) funded by the Ministry of Science, ICT and Future Planning (MISP) (2017K000216). References [1] Y. Zhang, G. Hukic-Markosian, D. Mascaro, Z.V. Vardeny, Synth. Met. 160 (2010) 262. [2] Zhang, T.P. Basel, B.R. Gautam, X. Yang, D.J. Mascaro, F. Liu, Z.V. Vardeny, Nat. Commun. 2057 (2012) 1. [3] H. Malissa, M. Kavand, D.P. Waters, K.J. van Schooten, P.L. Burn, Z.V. Vardeny, B. Saam, J.M. Lupton, C. Boehme, Science 345 (2014) 1487.
25
Contents lists available at ScienceDirect
Organic Electronics journal homepage: www.elsevier.com/locate/orgel
Scavenging of galvinoxyl spin 1/2 radicals in the processing of organic spintronics
T
Jung Min Choa, Chang Eun Songb, Sang-Jin Moonb, Won Suk Shinb,∗∗, Sugyeong Hongc, Sun Hee Kimc, Sanghee Chod, Jung-Keun Leee,∗ a
Research Institute, TOPnC Co., Ltd., Daejeon, 34158, Republic of Korea Center for Solar Energy Materials, Korea Research Institute of Chemical Technology, Daejeon, 34114, Republic of Korea c Western Seoul Center, Korea Basic Science Institute, Seoul, 03759, Republic of Korea d Center for Creative Convergence, Korea Research Institute of Standards and Science, Daejeon, 34113, Republic of Korea e Department of Physics & RIPC, Chonbuk National University, Jeonju, 54896, Republic of Korea b
A R T I C L E I N F O
A B S T R A C T
Keywords: Galvinoxyl ESR Singlet-triplet Spintronics UV–Vis
In many cases of organic electronic devices, spin manipulation adopts galvinoxyl spin 1/2 radicals to achieve singlet-triplet transition for a change of spin ensembles. We report scavenging of galvinoxyl spin 1/2 radicals, both in liquid-galvinoxyl samples and in thermal processing of galvinoxyl-doped organic films, as determined by electron spin resonance and UV–Vis measurements. The two different mechanisms of galvinoxyl scavenging are very crucial in organic device processing that often encounters oxidation and high-temperature treatment (∼150 °C), in which situation most of the galvinoxyl frameworks cannot survive.
1. Introduction Manipulation of spin states in organic electronics has been one of the most interesting research topics in such diverse areas as organic photovoltaics (OPVs) [1,2], batteries, organic light-emitting diodes (OLEDs) [3], magnetic sensing [4], information storage, and quantum information processing [5,6]. An increase of power conversion efficiency (PCE) was ascribed to enhanced spin-spin interaction between the added spin radicals and the photogenerated spin 1/2 species [1,2]. The electron exchange interactions in batteries were determined by the singlet-triplet energy gap [7,8]. The change in singlet-triplet balance induced a current in OLEDs [3,5]. Singlet-triplet correlations were utilized to improve the sensitivity of magnetometers [4,9]. It even appears that nature utilizes organic spintronics in a quantum biological compass that allows organisms to sense Earth's magnetic field [10,11]. The common key ingredient here is the pairs of electronic spins undergoing reactions to a singlet or a triplet state. The polaron-polaron (p-p) pairs may undergo intersystem singlet-triplet transitions. Otherwise, the p-p pairs may undergo coherent spin manipulation with microwave pulses under electron spin resonance (ESR) conditions [3]. On the other hand, spin doping [1,2,6,12,13] has been an alternative method to manipulate such spin ensembles. Among the spin-
∗
doping materials, galvinoxyl (Gx) has been widely known for its spin 1/ 2 radicals, which are inert to oxygen, and stable [14–16]. The Gx was also used in organic radical batteries as active-electrode materials [17,18]. Typically, adding Gx was speculated to change singlet to triplet, and the singlet-to-triplet transition of p-p pairs cause the reduction of P-P recombination in the P3HT/PCBM solar cell [1,2]. However, Cho et al. [19] reported that no triplet excitons were observed by using photo-luminescence detected magnetic resonance (PLDMR) and the enhancement in PCE was also accompanied by the disappearance of Gx spins. In this paper, we report the doublet of quintets ESR hyperfine signals from Gx spin 1/2 radicals, and that the Gx spins disappear not only with oxidation in the liquid sample, but also with a high temperature annealing of up to ∼150 °C, as determined by electron spin resonance and UV–Vis measurements. These are very important results when Gx is used as a spin doping material for organic electronics that accompany oxidation and thermal processing of up to 150 °C. 2. Experimental Galvinoxyl (Gx) powder was supplied in a bottle filled with air (Aldrich). Two kinds of Gx-liquid samples were prepared using 2.1 mg of the Gx powder dissolved into 2 ml of 1,2-dichlorobenzene (ODCB;
Corresponding author. Corresponding author. E-mail addresses: [email protected] (W.S. Shin), [email protected] (J.-K. Lee).
∗∗
https://doi.org/10.1016/j.orgel.2018.01.002 Received 5 June 2017; Received in revised form 23 September 2017; Accepted 1 January 2018 Available online 09 January 2018 1566-1199/ © 2018 Elsevier B.V. All rights reserved.
Organic Electronics 55 (2018) 21–25
J.M. Cho et al.
Metal Inc.) and 0.8 mg galvinoxyl powder (∼3.3 wt%) were mixed with ODCB solution, then stirred at 450 rpm at 40 °C for 12 h in N2. After filtering the P3HT solution mixed with Gx through a 0.45 μm PTFE (Polytetrafluoroethylene) filter, the solution was spin-coated on PET (polyethylene terephthalate) substrate at 400 rpm for 60 s, then at 2000 rpm for 3 s, and finally slowly dried at room temperature (RT) for an hour in a N2 glove box. The thickness of the blend layers was ∼220 nm. ESR measurements were taken using a Bruker EMXplus apparatus (Korean Basic Science Institute) with 100 kHz magnetic-field modulation. The measurements were performed at sufficiently low microwave powers to avoid microwave saturation effects. ESR spectra were recorded for samples in standard ESR quartz tubes in dark. In order to find any variation in the Gx-effects at high temperatures (RT∼150 °C), the Gx:P3HT films have been annealed at differing temperatures (50, 100, and 150 °C) for 30 m and then measured at RT. The Gx:P3HT films were cut into small pieces and then placed in N2sealed ESR quartz tubes for the RT ESR measurements. The UV–Vis measurements were also carried out at RT on a UV2550 UV–vis spectrometer (Shimadzu Corp.) in the wavelength range of 300–1100 nm for the annealed Gx:P3HT films. The low-temperature spin density (Ns) was estimated from the double-integrated ESR intensity of the Gx signals, considering Curie's law. Low-temperature (4–150 K) ESR measurements were reported in our previous work for the Gx powder samples [19]. 3. Results and discussion Fig. 1 shows the ESR spectra obtained from galvinoxyl (Gx) liquid samples. The hyperfine signals shown in Fig. 1 (a)–(c) are known as a “doublet of quintets”. The same spectrum was reported earlier in Gx liquid samples dissolved in ethanolic solution [20]. The hyperfine splitting was ascribed to originate from four equivalent hydrogens situated on aromatic rings in Gx (Scheme 1) (to make quintets), and from the hydrogen of C-H group joining the aromatic rings (to make a doublet). The relative intensity of the lines can be predicted by using Pascal's triangle [21,22]. Here, the intensity ratio would be basically 1:4:6:4:1 for quintets, and 1:1 for the doublet (since proton spin = 1/2; I = 1/2 for hydrogen nuclei). Another point in Fig. 1 is that the “air-sample (solid line)” (which was stirred in air for 30 m, and then kept in N2 until measured in ESR) shows scavenging of signals with time at room temperature (R.T.). There is a huge difference in the decreasing rate of the ESR intensity between the “air-sample” and the “N2-sample” (dashed line). The ESR intensity of the air-sample showed a small reduction after 30 min (Fig. 1 (b)), compared to the N2-sample; and then totally disappeared after about 4 h (Fig. 1 (c)). The color of the air-samples turned to yellow with the reduction of ESR intensity from the initial dark blue. Yet the N2-samples were found to be very stable, and their EPR
Fig. 1. ESR lineshapes obtained from galvinoxyl (Gx)-liquid samples. The Gx-liquid samples were prepared using Gx powder dissolved into ODCB. The “N2-sample” (dashed line), was stirred in N2 atmosphere at 50 °C for 30 m, and was sealed into quartz tubes in N2 atmosphere; while the “air-sample” (solid line), was prepared in the same way except having been stirred in air. The spectrum was measured (a) right after the preparation; (b) after 30 m, and (c) after ∼4 h.
Aldrich). One set of samples was stirred in N2 atmosphere at 50 °C for 30 m; and was sealed into an ESR quartz tube in N2 atmosphere, which was called “N2-sample” (dashed line in Fig. 1); yet another set of samples was prepared in air, called “air-sample” (solid line in Fig. 1). To make Gx-doped P3HT (Gx:P3HT) samples, 24 mg P3HT (Rieke
Scheme 1. Chemical structure of the galvinoxyl.
22
Organic Electronics 55 (2018) 21–25
J.M. Cho et al.
Fig. 2. ESR spectrum obtained from a galvinoxyl (Gx) powder sample as purchased (Aldrich; solid line) and heat-treated (@ 150 °C; dashed line) samples. The ESR was measured at R.T. in N2 atmosphere.
spectra remained unchanged for more than 4 h and even longer; the color did not change, either. Though the precise scavenging mechanism of the air-sample is not described here, this clearly indicates the Gx radical in the ODCB experiences scavenging when prepared in air, but not in N2. Fig. 2 shows the unimodal ESR spectrum (solid line) obtained from Gx powder samples supplied in a bottle filled with air (Aldrich). Disappearance of hyperfine structure will finally lead to a unimodal spectrum as in Fig. 2. The proper distinguishing of hyperfine parameters is not always possible, because hyperlines may be broadened, and in some cases the signals in the center may overlap, and not be separated. In fact, we found that the unimodal ESR line-shape in Fig. 2 changed to hyperfine line-shape in Fig. 1, when the supplied Gx powders were well dissolved and well stirred in ODCB liquid. Meanwhile, after 150 °C annealing of the powder sample in N2 for 30 m, the ESR signal almost disappeared, as shown by the dashed ESR lineshape in Fig. 2. We noted that the g-value changed from g = 2.0054 to g = 2.0045 after the 150 °C annealing. The change of g-value from g = 2.0054 to g = 2.0045 can be attributed to a structural change that became more like an aromatic system. The g-value in a range of 2.0043–2.0047 was usually found for an unpaired electron situated on oxygen substituted to the aromatic ring–the typical structure of poliphenolic antioxidant [20]. Fig. 3 (a) shows the ESR spectrum of a Gx-doped P3HT thin film (Gx:P3HT) sample. The spectrum was obtained at RT in dark. It basically shows two peaks: one is the broad resonance (g = 2.0046) from Gx radicals, and the other sharp one is from the P3HT+ resonance (g = 2.002). The dark ESR signal from P3HT:PCBM OPV was attributed to deep trapped carriers related to oxygen or moisture [23]. The P3HT+ signal from the Gx:P3HT samples in the dark also suggests that part of it was originally charged positive [19]. The g-value (g = 2.0046) of the Gx resonance indicates that the environment of Gx radicals in Gx:P3HT sample is different from that of the pure Gx powder sample (g = 2.0054). Fig. 3 (b) shows the variation of the ESR spectra of Gx:P3HT samples for differing annealing temperatures. Each sample was annealed in N2 atmosphere for 30 m at the designated temperatures. The Gx resonance decreased as the samples were annealed from the RT to 150 °C. The Gx signal decreased distinctively even after 50 °C annealing, and completely disappeared after an annealing step at 150 °C; while the P3HT+ signal remained much the same. The result clearly shows that the number of Gx radicals decreases with increasing annealing temperature. It was previously reported that
Fig. 3. (a) The room temperature ESR spectrum obtained from a Gx:P3HT sample (solid line). Deconvolution of the ESR spectrum (dashed lines) shows a broad Gx resonance (g = 2.0046), and a sharp P3HT+ signal (g = 2.002). (b) The Gx:P3HT ESR spectra for differing annealing temperatures. The samples were thermally treated at 50, 100, and 150 °C for 30 m each and then measured at R.T.
adding Gx (∼3 wt%) in P3HT:PCBM OPV increased PCE by 18%, which was attributed to the unpaired Gx electrons that would react as spin intermediates that change singlet to triplet in the system. However, Cho et al. [19] reported that the enhancement in PCE was observed only for the post-annealed Gx-doped P3HT:PCBM samples; but at the same time, disappearance of Gx spin was also observed for the post-annealed samples. Moreover, no triplet excitons were observed upon photoexcitation of the blend by using PLDMR. Here, Gx disappeared after 150 °C annealing. If this happens, the Gx would not work any more as spin intermediates in Gx-doped materials. The Gx has been known to be inert to oxygen [14–16], and thus stable; but the disappearance of Gx signals with thermal treatment has not previously been emphasized. This is a very important result when Gx is used as a main spin doping material in organic device processing, or in organic spintronics that accompany proper thermal processing. Fig. 4 (a) shows the variation of the Gx spin density (Ns) versus differing treatment temperatures. Fig. 4(a) inset (dashed rectangular box) shows the low temperature regime, where the Gx powder specimens (●) in a quartz ESR tube were measured at the designated low temperatures. The ESR signals were reversible for varying temperatures up and down. The low-temperature measurements were reported in a previous work [19]. For the high temperature regime (RT∼150 °C), however, the 23
Organic Electronics 55 (2018) 21–25
J.M. Cho et al.
Fig. 5. UV–Vis spectra of the Gx-doped P3HT films for differing annealing-temperatures. Arrows indicate the excitation wavelength of 770 nm (left) and 880 nm (right).
environment does not change for the low temperatures. However, for the annealed samples, the g-value decreased. The change of g-value from 2.0054 (RT) to ∼2.004 (150 °C) may indicate that, as the temperature nears the m.p., the Gx structure would change to aromatic system and melt down finally. Thermal processing of organic devices usually experiences 150 °C treatment [1,2,19]. Thus we should be very cautious and ensure that the Gx radicals do not experience high temperature, where they are expected to work as spin intermediates. The effect of annealing of the Gx:P3HT solid film is shown on the UV–Vis absorption spectrum in Fig. 5. The spectrum consists of an intense band in the range between ∼350 and 650 nm and other three weak bands at around 690, 770 and 880 nm. The center-peak around 550 nm was ascribed to the absorption caused by P3HT [24,25]. The decays of the three weak bands at around 690, 770 and 880 nm show distinctive differences between the R.T. sample and the annealed ones (up to 150 °C). The peaks decreased distinctively even after 50 °C and even the peak position shifted toward longer wavelength for the ∼100–150 °C annealing. Previously, the energy transition in the absorption spectrum located at ∼860 nm and 770 nm was assigned to D0 → D3 and D0 → D4 transition, respectively, for isolated Gx radicals in differing solvents [26,27], where D0 is the ground state and Dn is the nth excited state in agreement with quantum chemical calculations [27,28]. Considering that absorption maxima for the radicals may vary according to the surrounding chemical structures (as found in different solvents) [27,29], the 770 nm (left arrow) and the ∼880 nm (right arrow) bands in our absorption spectrum may be related to those transitions for the Gx radicals in the solid P3HT films; the origin of the 690 nm is not clear, though. The peak position even shifted to longer wavelength at 100–150 °C annealing in our UV–Vis spectrum, implying any structural change in the chemical environment, which is consistent with the ESR result where the g-value shifted after the 100–150 °C annealing. Consequently, our UV–Vis absorption data is not inconsistent with our ESR data reflecting the scavenging of Gx radicals after being annealed.
Fig. 4. (a) Inset: The inset (dashed rectangular box): shows the reversible Ns measured at low temperatures. (The low-temperature measurement was reported in a previous study [19]). For the high temperature (above R.T.) regime, the Gx spin density (Ns) was estimated from the separated Gx resonance from the annealed Gx:P3HT samples that were annealed at the designated temperatures. Galvinoxyl's melting point is depicted at 158–159 °C). (b) g value vs. 1/T. The ( ) Gx(P3HT) indicates the Gx radicals in the Gx:P3HT samples, the ( ) Gx(ODCB) indicates the Gx-powder specimen wet with the ODCB, and the Gx(powder) (●) the pure Gx powder samples.
Gx:P3HT samples were annealed at differing temperatures, and then the ESR measurements were performed at RT. The ESR signal was irreversible once the sample was annealed at higher temperatures. For the Gx:P3HT samples, the separated Gx resonance was used for the convoluted ESR spectra of Gx:P3HT samples. We see that the spin density of Gx is fairly stable between room temperature and lower temperatures. However, as the annealing temperature increased from R.T. to ∼150 °C, the Ns decreased irreversibly with increasing temperature. The temperature for the complete disappearance of Ns exactly matches the melting point (158–159 °C) of the Gx. It is regarded that as the temperature nears the m.p., the Gx structure would change to aromatic system, resulting in the loss of Gx spin 1/2 radicals. The activation energy for losing Gx spins was about the order of ΔE = 10–100 meV, depending on the temperature range (RT∼150 °C). In Fig. 4 (b), The variation of g-value for the differing temperature is quite similar to that of the Ns. We see that for the same Gx radical, the g-value is different for the powder samples (●), compared to the case Gx radicals in P3HT ( ). The g-value reflects the atomic environments. The g-values obtained at low temperatures are generally the same as that obtained (g = 2.0054) at RT, indicating the electronic
4. Conclusion In this letter, we report the scavenging of spin 1/2 Gx radicals both in Gx-liquid samples, and in the thermal processing of Gx-doped organic films. We found a doublet of quintets hyperfine signal from the electron spin resonance of galvinoxyl (Gx) spin 1/2 radicals in the liquid samples. The air-sample showed scavenging of signals with time at room 24
Organic Electronics 55 (2018) 21–25
J.M. Cho et al.
temperature. The color of the air-samples turned to yellow with the reduction of ESR intensity from the initial dark blue. However, the N2samples showed persistent ESR intensity for more than 4 h, and even longer; the color did not change, either. Also the unimodal Gx signal decreased with thermal treatment increasing up to 150 °C. As the galvinoxyl-doped P3HT films were heated up to ∼150 °C, the Gx ESR signal in the P3HT:Gx samples disappeared. We note that the loss of Gx paramagnetic signals in the thermal processing of organic device is related to the melting of Gx, in consequence of the radical scavenging of Gx spins. The UV–Vis spectra for the annealed Gx:P3HT samples are consistent with our ESR results. Galvinoxyl (Gx) has been known to be inert to oxygen, and thus stable; but the disappearance of Gx signals in liquid or with thermal treatment has not been emphasized before. This is a very important issue in the processes of organic electronics or spintronics that employ Gx as spin intermediates.
[4] W.J. Baker, K. Ambal, D.P. Waters, R. Baarda, H. Morishita1, K. van Schooten, D.R. McCamey, J.M. Lupton, C. Boehme, Nat. Commun. (2012). [5] P.A. Bobbert, Nature 345 (2014) 1450. [6] A. Atxabal, M. Ribeiro1, S. Parui, L. Urreta, E. Sagasta, X. Sun, R. Llopis, F. Casanova, L.E. Hueso, Nat. Commun. 13751 (2016) 1. [7] W. Walker, S. Grugeon, H. Vezin, S. Laruelle, M. Armand, F. Wudlb, J-M. Tarascon, J. Mater. Chem. 21 (2011) 1615. [8] Y. Su, X. Wang, L. Wang, Z. Zhang, X. Wang, Y. Song, P.P. Power, Chem. Sci. 7 (2016) 6514. [9] M. Alidoust, K. Halterman, J. Linder, Phys. Rev. B 88 (2013) 075435. [10] T. Ritz, R. Wiltschko, P.J. Hore, C.T. Rodgers, K. Stapput, P. Thalau, C.R. Timmel, W. Wiltschko, Biophys. J. 96 (2009) 3451. [11] S. Engels, N.-L. Schneider, N. Lefeldt, C.M. Hein, M. Zapka, A. Michalik, D. Elbers, A. Kittel, P.J. Hore, H. Mouritsen, Nature 509 (2014) 354. [12] T. Moorsom, M. Wheeler, M.T. Khan, F.A. Ma'Mari, G. Burnell, B.J. Hickey, V. Lazarov, D. Gilks, O. Cespedes, Appl. Phys. Lett. 105 (2014) 022408. [13] C.T. Hong, W. Lee, Y.H. Kang, Y. Yoo, J. Ryu, S.Y. Cho, K.-S. Jang, J. Mater. Chem. A 3 (2015) 12314. [14] M.S. Kharasch, B.S. Joshi, J. Am. Chem. Soc. 22 (1957) 1439. [15] G.M. Coppinger, J. Am. Chem. Soc. 79 (1957) 501. [16] K. Mukai, K. Ueda, K. Ishizu, Y. Oeguchi, J. Chem. Phys. 77 (1982) 1. [17] T. Suga, H. Ohshiro, S. Sugita, K. Oyaizu, H. Nishide, Adv. Mater. 21 (2009) 1627. [18] T. Martin, D. Hager, U.S. Schubert, Adv. Mater. 24 (2012) 6397. [19] J.M. Cho, D.S. Kim, S. Bae, S.-J. Moon, W.S. Shin, D.H. Kim, S.H. Kim, A. Sperlich, S. V¨ath, V. Dyakonov, J.-K. Lee, Org. Electron. 27 (2015) 119. [20] M. Jerzykiewicz, I. Ćwieląg-Piasecka, Antioxidative and anticorrosive properties of bioglycerol, in: Gisela Montero (Ed.), Biodiesel- Quality, Emissions and By-products, InTech, 2011. [21] D. Griffiths, Am. J. Phys. 50 (1982) 698. [22] N. Bunce, J. Chem. Educ. 64 (1987) 907. [23] M.D. Heinemann, K. von Maydell, F. Zutz, J. Kolny-Olesiak, H. Borchert, I. Riedel, J. Parisi, Adv. Funct. Mater. 19 (2009) 3788. [24] P. Vanlaeke, A. Swinnen, I. Haeldermans, G. Vanhoyland, T. Aernoutsa, D. Cheynsa, C. Deibela, J. D'Haena, P. Heremansa, J. Poortmansa, J.V. Manca, Sol. Energy Mater. Sol. Cell. 90 (2006) 2150–2158. [25] H. Hintz, H.-J. Egelhaaf, H. Peisert, T. Chassé, Polym. Degrad. Stabil. 95 (2010) 818–825. [26] L.S. Degtyarev, Theor. Exp. Chem. 23 (1988) 682. [27] J. Grilj, C. Zonca, L.M.L. Daku, E. Vauthey, Phys. Chem. Chem. Phys. 14 (2012) 6352–6358. [28] W. Gierke, W. Harrer, B. Kirste, H. Kurreck, J. Reusch, Z. Naturforsch. B Chem. Sci.: Anorg. Chem., Org. Chem. 31 (1976) 965. [29] R. Edge, A. El-Agamey, E.J. Land, S. Navaratnam, T.G. Truscott, Arch. Biochem. Biophys. 458 (2007) 104–110.
Acknowledgement This research was supported by “Research Base Construction Fund Support Program” funded by Chonbuk National University in 2016. This research was also partly supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (2017R1D1A1B03028062). J. M. Cho was supported by the Commercializations Promotion Agency for R &D Outcomes (COMPA) funded by the Ministry of Science, ICT and Future Planning (MISP) (2017K000216). References [1] Y. Zhang, G. Hukic-Markosian, D. Mascaro, Z.V. Vardeny, Synth. Met. 160 (2010) 262. [2] Zhang, T.P. Basel, B.R. Gautam, X. Yang, D.J. Mascaro, F. Liu, Z.V. Vardeny, Nat. Commun. 2057 (2012) 1. [3] H. Malissa, M. Kavand, D.P. Waters, K.J. van Schooten, P.L. Burn, Z.V. Vardeny, B. Saam, J.M. Lupton, C. Boehme, Science 345 (2014) 1487.
25