Angewandte A Journal of the Gesellschaft Deutscher Chemiker
International Edition
Chemie www.angewandte.org
Accepted Article Title: Analysis of Molecular Orientation in Organic Semiconducting Thin Films Using Static Dynamic Nuclear Polarization Enhanced SolidState NMR Authors: Katsuaki Suzuki, Shosei Kubo, Fabien Aussenac, Frank Engelke, Tatsuya Fukushima, and Hironori Kaji This manuscript has been accepted after peer review and appears as an Accepted Article online prior to editing, proofing, and formal publication of the final Version of Record (VoR). This work is currently citable by using the Digital Object Identifier (DOI) given below. The VoR will be published online in Early View as soon as possible and may be different to this Accepted Article as a result of editing. Readers should obtain the VoR from the journal website shown below when it is published to ensure accuracy of information. The authors are responsible for the content of this Accepted Article. To be cited as: Angew. Chem. Int. Ed. 10.1002/anie.201707208 Angew. Chem. 10.1002/ange.201707208 Link to VoR: http://dx.doi.org/10.1002/anie.201707208 http://dx.doi.org/10.1002/ange.201707208
10.1002/anie.201707208
Angewandte Chemie International Edition
COMMUNICATION Analysis of Molecular Orientation in Organic Semiconducting Thin Films Using Static Dynamic Nuclear Polarization Enhanced Solid-State NMR
Abstract: Molecular orientation in amorphous organic semiconducting thin film devices is an important issue affecting device performances. However, to date it has not been possible to analyze the “distribution” of the orientations. Although solid-state NMR (ssNMR) can provide information on the distribution of molecular orientations, the technique is limited because of the small amounts of sample in the devices and the low sensitivity of ssNMR. Here, we report the first application of dynamic nuclear polarization enhanced ssNMR (DNP-ssNMR) to orientational analysis of amorphous phenyldi(pyren-1-yl)phosphine oxide (POPy2). The 31P DNP-ssNMR spectra exhibited a sufficient signal-to-noise ratio to quantify the distribution of molecular orientations in amorphous films: the P=O axis of the vacuum-deposited and drop-cast POPy2 shows anisotropic and isotropic distribution, respectively. The different molecular orientations reflect the molecular origin of the different charge transport behaviors.
Organic thin-film semiconducting devices, such as organic lightemitting diodes (OLEDs), organic solar cells (OSCs), and organic thin-film transistors, are expected to be the next generation devices owing to their light-weight, cost effectiveness, and flexible properties. Thus, organic semiconductors have been extensively studied over the last few decades.[1] The properties inherent to these devices, such as charge carrier mobility, light emission, and light out-coupling, depend on the intra- and intermolecular structures, including the orientation of organic molecules in the devices.[1d, e] Organic molecules in OLEDs and OSCs are often in an amorphous state, which limits the application of diffraction techniques based on X-rays and neutrons for detailed analysis of the orientation of organic amorphous films. The molecular orientation can be determined by angular-dependent photoluminescence measurements,[2] and variable angle spectroscopic ellipsometry[3]; however, these methods provide only average values, such as the order parameters. There are currently no appropriate methods for revealing the “distribution” of molecular orientations in an amorphous aggregate.
[a]
[b]
[c]
Dr. Katsuaki Suzuki, Shosei Kubo, Dr. Tatsuya Fukushima, and Prof. Dr. Hironori Kaji* Institute for Chemical Research Kyoto University Uji, Kyoto 611-0011, Japan E-mail: [email protected] Dr. Fabien Aussenac Bruker BioSpin 34, rue de l’Industrie, 67166 Wissembourg, France Dr. Frank Engelke Bruker BioSpin Silberstreifen, 76287 Rheinstetten, Germany Supporting information for this article is given via a link at the end of the document.
Figure 1. a) Molecular structures of POPy2 and bTbK. b) Schematic representation of sample preparation. The samples on glass (SiO2) or polytetrafluoroethylene (PTFE) are inserted into a 5-mmø quartz tube.
Solid-state nuclear magnetic resonance (ssNMR) is a powerful technique for analyzing the structure and dynamics of materials with atomic resolution, with both crystalline and amorphous morphologies.[4] Thus, it may be possible to perform a detailed analysis of the structure of semiconducting materials in organic devices. However, most ssNMR studies are based on bulk samples[5] and reports on thin film samples are rare.[6] One reason for this lack of previous studies is that the limited amount of organic material in the devices results in a low signal-to-noise (S/N) ratio. To enhance the sensitivity of the NMR signal, dynamic nuclear polarization (DNP) enhanced ssNMR (DNP-ssNMR) has recently attracted considerable attention. [7] In DNP-ssNMR experiments, radicals dispersed in a sample are polarized by microwave irradiation leading to a high electron polarization. The resulting electron polarization is transferred to the 1H population in the sample. In most cases, the enhanced 1H polarization is further transferred to other nuclei, typically by cross polarization (CP). Although most recent DNP-ssNMR measurements have been performed under magic angle spinning conditions,[5g, 7] Bechinger’s group reported static DNP-ssNMR measurements, which enabled insight into the structure of membrane polyproteins in lipid bilayers.[8] Here, we performed static DNP-ssNMR measurements for a semiconducting organic material, phenyldi(pyren-1yl)phosphine oxide (POPy2) (Figure 1a). POPy2 have electron transport properties,[9] and is thus frequently used in OLEDs. As shown later, we have detected different electron transport behaviors depending on different preparation methods, vacuum-
This article is protected by copyright. All rights reserved.
Accepted Manuscript
Katsuaki Suzuki[a], Shosei Kubo[a], Fabien Aussenac[b], Frank Engelke[c], Tatsuya Fukushima[a], and Hironori Kaji*[a]
10.1002/anie.201707208
Angewandte Chemie International Edition
deposition and drop-casting methods (Figure 1b). We selected a bisnitroxide radical, bTbK[10] as the polarizing agent, (Figure 1a), because our thermogravimetric measurements (Figure S1) revealed that bTbK can be sublimed before decomposition. The concentration of radicals is the main factor affecting the signal enhancement, because strong electron-electron exchange couplings among radicals are reported to decrease the DNP efficiency.[11] Thus, we used a radical doping concentration of 0.25 wt%. Electron paramagnetic resonance (EPR) spectra of the sublimed bTbK doped in POPy2 showed essentially the same spectra as the drop-cast spectra, as shown in Figure S2. The EPR signals were well-resolved, indicating that the radicals in the sample films did not feature any notable electron-electron exchange couplings. Amorphous POPy2 and bTbK (0.25 wt%) layers were formed on glass (SiO2) substrates. The thickness of the SiO2 substrate was 30 μm. The thickness of the vacuum-deposited POPy2 layer, 1.5 μm (ca. 52 μg for each SiO2 substrate), was thinner than the typical sample thickness required for time-offlight (TOF) experiments (normally, thicker than several μm), which is a most frequently used method to measure photocurrent charge transients. As shown in Figure S3, no distinct peaks were found in the XRD profiles for vacuumdeposited and drop-cast films, confirming that these layers are amorphous states. The Halo patterns for both the films are similar, indicating that XRD measurements do not provide information on the molecular orientation in amorphous films, even qualitatively. Details of sample preparation and measurements are provided in the Supporting Information. The POPy2 films on the substrates were set perpendicular to the external magnetic field, B0 (Figure 1b). We performed 31P CP DNP-ssNMR experiments under static conditions to obtain chemical shift anisotropy (CSA) spectra, which provide information on 31P=O orientations. Figure 2 shows the 31P CSA spectra of POPy2 with and without DNP enhancement. According to reports from Bechinger’s group,[8] we calculated the DNP enhancement factor (εon/off) to be the ratio of the integral signal intensity of the CSA spectra with and without DNP enhancements. The values of εon/off for the vacuum-deposited and drop-cast films on SiO2 (12 sheets for the vacuum-deposited films; 15 sheets for the dropcast films) were 3.0 and 2.0, respectively (Figure 2a and b). Figure 2c shows the 31P CSA spectrum of a single sheet of a POPy2 thin film with a measurement time of ~16 h. Compared with the stacked samples in Figure 2a and b, the εon/off value increased to 6.0. The diferrence in εon/off for stacked samples and one sheet sample is probably due to the difference of cooling efficiency of thin-film. The one sheet sample is more effectively cooled, leading to a higher εon/off value. Through the DNP enhancement, we successfully obtained a CSA spectrum with a sufficient S/N ratio from a sample of even a single sheet. Note that the POPy2 films on polytetrafluoroethylene (PTFE) substrates (10 sheets for the
Figure 2. 31P CSA spectra of POPy2 with (black) and without (gray) DNP enhancement. a) Vacuum-deposited on SiO2 (12 sheets), b) drop-cast on SiO2 (15 sheets), and c) vacuum-deposited on SiO2 (1 sheet). The normalized CSA lineshapes with and without microwave irradiations were the same as shown in Figure S4.
vacuum-deposited films; 15 sheets for the drop-cast films) exhibited εon/off values of 10 and 9.3, respectively (Figure S5a and b). These values are much higher than those on SiO 2. This difference in the values of εon/off for the samples on SiO2 and PTFE is likely caused by the different degrees of sample heating under microwave irradiation of SiO2 and PTFE substrates.[8a]
This article is protected by copyright. All rights reserved.
Accepted Manuscript
COMMUNICATION
10.1002/anie.201707208
Angewandte Chemie International Edition
Figure 4. Photocurrent transient curves of a) the vacuum-deposited POPy2 and b) drop-cast POPy2.
To quantify the molecular orientation distributions, the experimental CSA spectra were fitted by the Legendre moment expansion approach.[4a, 12] The spectra simulated by assuming a Gaussian-type orientation distribution function showed slightly poor reproducibility (Figure S6). Because the CSA spectra of POPy2 have an axially symmetric pattern, the orientational distribution function p(θ) was defined as an expansion of Legendre polynomials Pn(cosθ) with the following equations (1)– (3). 𝑝(𝜃) = ∑𝑛 Figure 3. a) Schematic representation of chemical shift tensor orientation for POPy2. The relationship between the P=O orientation and the chemical shift is also shown in calculated CSA spectrum with completely random orientations. b and c) Experimental 31P CSA spectra of POPy2 b) vacuum-deposited and c) drop-cast on SiO2 with DNP enhancement (black line). Simulated best-fit spectra are shown by the red line. d and e) Orientation distributions p(θ) (thick red lines) and contributions of respective terms (thin lines) of Legendre polynomials Pn(cosθ) for d) the vacuum-deposited film and e) the drop-cast film and on SiO2. Owing to the symmetry, values of p(θ) with θ > 90° are folded in between 0 and 90°.
The CSA patterns of the vacuum-deposited and drop-cast POPy2 were spread over a wide chemical shift range from 100 to −100 ppm depending on the P=O direction. The experimentallyobtained CSA spectra were axially-symmetric (Figures 2 and S5). From the structural symmetry of POPy2, the principal axes can be assigned as shown in Figure 3a. This assignment is confirmed by gauge-including atomic orbital (GIAO) calculations. The signal at −100 ppm, labeled 𝜎⫽ , corresponds to the parallel alignment of P=O axis of POPy2 relative to B0 (the angle between the unique principal axis of the uniaxial chemical shift tensor, 𝜎⫽ , and the P=O axis was calculated to be 3.2°). The signal at 100 ppm, labeled 𝜎⊥ , corresponds to the perpendicular alignment of the P=O axis relative to B0. The CSA spectrum of the vacuum-deposited POPy2 (Figure 3b) showed an enhanced signal intensity around −100 ppm compared with the drop-cast sample (Figure 3c), indicating the greater contribution of the P=O axis parallel to the B0.
2𝑛+1 2
〈𝑃𝑛 〉𝑃𝑛 (cos 𝜃) , 𝑛 = 0, 2, 4, 6, ⋯
(1)
𝜋
∫0 𝑝(𝜃) sin 𝜃 𝑑𝜃 = 2〈𝑃0 〉 = 1 〈𝑃𝑛 〉 =
(2)
𝜋
∫0 𝑝(𝜃)𝑃𝑛 (cos 𝜃) sin 𝜃𝑑𝜃
(3)
𝜋
∫0 𝑝(𝜃) sin 𝜃𝑑𝜃
Here, θ is the angle between the P=O axis and B0, and B0 is normal to the SiO2 substrate (Figure 3a). The function p(θ) gives the probability of finding a molecule at a particular orientation of θ. The experimental data were fitted by least squares minimization. The fitting parameters were the principal values of 𝜎⫽ and 𝜎⊥ , n-th order parameters of 〈𝑃𝑛 〉 (n = 2, 4, 6), and a Gaussian line broadening factor (BF). Higher order parameters (n ≥ 8) were ignored because the contribution of the 6th order parameter was negligibly small, as shown in Table 1. The vacuum-deposited and drop-cast 31P CSA spectra were well reproduced by the best-fit simulated spectra, shown
Table 1. Chemical shift tensor principal values (𝜎⫽ and 𝜎⊥ ), order parameters (〈𝑃2 〉, 〈𝑃4〉, 〈𝑃6〉), and Gaussian line broadening factor (BF) for the vacuumdeposited and drop-cast POPy2 films determined from fitting to the 31P CSA spectra in Figure 3b and c.
Sample
𝜎⫽ (ppm)
𝜎⊥ (ppm)
〈𝑃2〉
〈𝑃4〉
〈𝑃6 〉
BF (ppm)
Vacuumdeposited
114.3
-93.4
0.15
0.031
-0.0010
8.4
Dropcast
110.1
-88.4
-0.019
0.0059
-0.0014
7.2
This article is protected by copyright. All rights reserved.
Accepted Manuscript
COMMUNICATION
10.1002/anie.201707208
Angewandte Chemie International Edition
by the red curves in Figure 3b and c. The best-fit parameters are shown in Table 1. The thick red curves in Figure 3d and e represent p(θ) for the vacuum-deposited and drop-cast thin film samples. The respective contributions of Pn(cosθ) are also shown in the figures. For the vacuum-deposited films, the probability of finding a P=O orientation at θ = 0 was highest and decreased with increasing θ. This originates from the large contribution of the order parameter 〈𝑃2 〉 of 0.15 in addition to 〈𝑃0 〉. We detected some contribution from 〈𝑃4 〉 , but that of 〈𝑃6 〉 is negligible. For the drop-cast film, the contributions of 〈𝑃𝑛 〉, where n is higher than two are quite small (Figure 3e) and the CSA spectrum is mostly determined by 〈𝑃0 〉, indicating that the P=O axis of POPy2 is isotropically distributed in the drop-cast film. Differences in molecular orientation can be expected to affect the photocurrent transients of POPy2. Indeed, the transient photocurrent of the vacuum-deposited and drop-cast films of POPy2, measured by the TOF technique, exhibited different behavior, as shown in Figure 4 (see Supporting Information for experimental details). The vacuum-deposited POPy2 showed a “non-dispersive” photocurrent with a clear bending point, while the drop-cast POPy2 exhibited a featureless “dispersive” photocurrent. The random molecular orientation in drop-casted POPy2 films is considered to generate larger disorders (energetic disorder and structural disorder) compared to vacuum-deposited POPy2, resulting in disordered dispersive photocurrent. To gain further insight about the origin of different photocurrent transients, we will further develop our currently performing multiscale charge transport simulations[13] to link molecular-level structures and charge mobilities. In summary, we performed static DNP-ssNMR measurements of vacuum-deposited and drop-cast POPy2 thin films. Under the DNP-enhanced condition, we could obtain 31P CSA spectra with a S/N ratio sufficient for orientation analysis. The CSA analysis based on the Legendre moment expansion approach revealed a quantitative orientational distribution of the vacuum-deposited and drop-cast amorphous POPy2 thin films. The vacuum-deposited POPy2 favors a perpendicular orientation of the P=O axes to the SiO2 substrates, while drop-cast POPy2 exhibits an isotropic random orientation. This difference in orientation affects the photocurrent transient behavior.
[1]
[2]
[3]
[4]
[5]
[6] [7]
Acknowledgements This work was supported by Japan Society and the Promotion of Science (JSPS) Grant-in-Aid for Scientific Research (A) (No. 17H01231) and Grant-in-Aid for JSPS Fellows (No. 14J04794), and also supported by Kyoto University for Supporting Program for Interaction-Based Initiative Team Studies (SPIRITS). FA and FE are grateful to Christian Reiter, Hiba Sarrouj, and Armin Purea, Bruker Biospin, for the extensive engineering work on the static DNP probe. We thank Andrew Jackson, PhD, from Edanz Group (www.edanzediting.com/ac) for editing a draft of this manuscript. Keywords: Amorphous materials • Dynamic nuclear polarization • NMR spectroscopy • Molecular orientation • Organic semiconductor
[8]
[9] [10]
[11] [12]
a) L. Dou, Y. Liu, Z. Hong, G. Li, Y. Yang, Chem. Rev. 2015, 115, 12633-12665; b) O. Ostroverkhova, Chem. Rev. 2016, 116, 1327913412; c) Z. Yang, Z. Mao, Z. Xie, Y. Zhang, S. Liu, J. Zhao, J. Xu, Z. Chi, M. P. Aldred, Chem. Soc. Rev. 2017, 46, 915-1016; d) H. Kaji, H. Suzuki, T. Fukushima, K. Shizu, K. Suzuki, S. Kubo, T. Komino, H. Oiwa, F. Suzuki, A. Wakamiya, Y. Murata, C. Adachi, Nat. Commun., 2015, 6, 8476; e) T. Miwa, S. Kubo, K. Shizu, T. Komino, C. Adachi, H. Kaji, Sci. Rep. 2017, 7, 284. a) J. Frischeisen, D. Yokoyama, C. Adachi, W. Brütting, Appl. Phys. Lett. 2010, 96, 073302; b) J. W. Sun, J.-H. Lee, C.-K. Moon, K.-H. Kim, H. Shin, J.-J. Kim, Adv. Mater. 2014, 26, 5684-5688. a) D. Yokoyama, Y. Setoguchi, A. Sakaguchi, M. Suzuki, C. Adachi, Adv. Funct. Mater. 2010, 20, 386-391; b) M. Shibata, Y. Sakai, D. Yokoyama, J. Mater. Chem. C 2015, 3, 11178-11191. a) K. Schmidt-Rohr, H. W. Spiess, Multidimensional solid-state NMR and polymers, Academic Press, 1994; b) H. Kaji, N. Miura, K. SchmidtRohr, Macromolecules 2003, 36, 6100-6113; c) M. Sharma, M. Yi, H. Dong, H. Qin, E. Peterson, D. D. Busath, H.-X. Zhou, T. A. Cross, Science 2010, 330, 509-512; d) R. Bhattacharyya, B. Key, H. Chen, A. S. Best, A. F. Hollenkamp, C. P. Grey, Nat. Mater. 2010, 9, 504-510; e) J. K. Williams, D. Tietze, M. Lee, J. Wang, M. Hong, J. Am. Chem. Soc. 2016, 138, 8143-8155. a) M. Utz, M. Nandagopal, M. Mathai, F. Papadimitrakopoulos, Appl. Phys. Lett. 2003, 83, 4023-4025; b) H. Kaji, Y. Kusaka, G. Onoyama, F. Horii, J. Am. Chem. Soc. 2006, 128, 4292-4297; c) Y. Nishiyama, T. Fukushima, K. Takami, Y. Kusaka, T. Yamazaki, H. Kaji, Chem. Phys. Lett. 2009, 471, 80-84; d) T. Fukushima, H. Kimura, Y. Shimahara, H. Kaji, Appl. Phys. Lett. 2011, 99, 223301; e) F. Suzuki, T. Fukushima, M. Fukuchi, H. Kaji, J. Phys. Chem. C 2013, 117, 18809-18817; f) F. Suzuki, Y. Nishiyama, H. Kaji, Chem. Phys. Lett. 2014, 605–606, 1-4; g) S. R. Chaudhari, J. M. Griffin, K. Broch, A. Lesage, V. Lemaur, D. Dudenko, Y. Olivier, H. Sirringhaus, L. Emsley, C. P. Grey, Chem. Sci. 2017, 8, 3126-3136. Y. Nishiyama, T. Fukushima, M. Fukuchi, S. Fujimura, H. Kaji, Chem. Phys. Lett. 2013, 556, 195-199. a) D. A. Hall, D. C. Maus, G. J. Gerfen, S. J. Inati, L. R. Becerra, F. W. Dahlquist, R. G. Griffin, Science 1997, 276, 930-932; b) H. Takahashi, D. Lee, L. Dubois, M. Bardet, S. Hediger, G. De Paëpe, Angew. Chem. Int. Ed. 2012, 51, 11766-11769; Angew. Chem. 2012, 124, 1193611939; c) A. J. Rossini, C. M. Widdifield, A. Zagdoun, M. Lelli, M. Schwarzwälder, C. Copéret, A. Lesage, L. Emsley, J. Am. Chem. Soc. 2014, 136, 2324-2334; d) D. Lee, G. Monin, N. T. Duong, I. Z. Lopez, M. Bardet, V. Mareau, L. Gonon, G. De Paëpe, J. Am. Chem. Soc. 2014, 136, 13781-13788; e) A. N. Smith, J. R. Long, Anal. Chem. 2016, 88, 122-132; f) D. Stöppler, C. Song, B.-J. van Rossum, M.-A. Geiger, C. Lang, M.-A. Mroginski, A. P. Jagtap, S. T. Sigurdsson, J. Matysik, J. Hughes, H. Oschkinat, Angew. Chem. Int. Ed. 2016, 55, 16017-16020; Angew. Chem. 2016, 128, 16251-16254; g) M. T. Colvin, R. Silvers, Q. Z. Ni, T. V. Can, I. Sergeyev, M. Rosay, K. J. Donovan, B. Michael, J. Wall, S. Linse, R. G. Griffin, J. Am. Chem. Soc. 2016, 138, 9663-9674. a) E. S. Salnikov, H. Sarrouj, C. Reiter, C. Aisenbrey, A. Purea, F. Aussenac, O. Ouari, P. Tordo, I. Fedotenko, F. Engelke, B. Bechinger, J. Phys. Chem. B 2015, 119, 14574-14583; b) E. S. Salnikov, C. Aisenbrey, F. Aussenac, O. Ouari, H. Sarrouj, C. Reiter, P. Tordo, F. Engelke, B. Bechinger, Sci. Rep. 2016, 6, 20895. T. Matsushima, C. Adachi, Appl. Phys. Lett. 2006, 89, 253506. Y. Matsuki, T. Maly, O. Ouari, H. Karoui, F. Le Moigne, E. Rizzato, S. Lyubenova, J. Herzfeld, T. Prisner, P. Tordo, R. G. Griffin, Angew. Chem. Int. Ed. 2009, 48, 4996-5000; Angew. Chem. 2009, 121, 50965100. B. Corzilius, L. B. Andreas, A. A. Smith, Q. Z. Ni, R. G. Griffin, J. Magn. Reson. 2014, 240, 113-123. a) R. Hentschel, J. Schlitter, H. Sillescu, H. W. Spiess, J. Chem. Phys. 1978, 68, 56-66; b) W. Gabriëlse, H. Angad Gaur, W. S. Veeman, Macromolecules 1996, 29, 4125-4133.
This article is protected by copyright. All rights reserved.
Accepted Manuscript
COMMUNICATION
10.1002/anie.201707208
Angewandte Chemie International Edition
COMMUNICATION H. Uratani, S. Kubo, K. Shizu, F. Suzuki, T. Fukushima, H. Kaji, Sci. Rep. 2016, 6, 39128.
Accepted Manuscript
[13]
This article is protected by copyright. All rights reserved.
10.1002/anie.201707208
Angewandte Chemie International Edition
COMMUNICATION Entry for the Table of Contents
COMMUNICATION The orientation of organic semiconducting molecules in an amorphous thin film was revealed by static dynamic nuclear polarization enhanced solid-state NMR. The P=O axis of an electron transport material, called POPy2, in a vacuum-deposited film tends to orient perpendicular to the substrates. We find a close relationship between the orientational distribution and charge transport behavior.
Katsuaki Suzuki, Shosei Kubo, Fabien Aussenac, Frank Engelke, Tatsuya Fukushima, and Hironori Kaji*
Analysis of Molecular Orientation in Organic Semiconducting Thin Films Using Static Dynamic Nuclear Polarization Enhanced Solid-State NMR
This article is protected by copyright. All rights reserved.
Accepted Manuscript
Page No. – Page No.
International Edition
Chemie www.angewandte.org
Accepted Article Title: Analysis of Molecular Orientation in Organic Semiconducting Thin Films Using Static Dynamic Nuclear Polarization Enhanced SolidState NMR Authors: Katsuaki Suzuki, Shosei Kubo, Fabien Aussenac, Frank Engelke, Tatsuya Fukushima, and Hironori Kaji This manuscript has been accepted after peer review and appears as an Accepted Article online prior to editing, proofing, and formal publication of the final Version of Record (VoR). This work is currently citable by using the Digital Object Identifier (DOI) given below. The VoR will be published online in Early View as soon as possible and may be different to this Accepted Article as a result of editing. Readers should obtain the VoR from the journal website shown below when it is published to ensure accuracy of information. The authors are responsible for the content of this Accepted Article. To be cited as: Angew. Chem. Int. Ed. 10.1002/anie.201707208 Angew. Chem. 10.1002/ange.201707208 Link to VoR: http://dx.doi.org/10.1002/anie.201707208 http://dx.doi.org/10.1002/ange.201707208
10.1002/anie.201707208
Angewandte Chemie International Edition
COMMUNICATION Analysis of Molecular Orientation in Organic Semiconducting Thin Films Using Static Dynamic Nuclear Polarization Enhanced Solid-State NMR
Abstract: Molecular orientation in amorphous organic semiconducting thin film devices is an important issue affecting device performances. However, to date it has not been possible to analyze the “distribution” of the orientations. Although solid-state NMR (ssNMR) can provide information on the distribution of molecular orientations, the technique is limited because of the small amounts of sample in the devices and the low sensitivity of ssNMR. Here, we report the first application of dynamic nuclear polarization enhanced ssNMR (DNP-ssNMR) to orientational analysis of amorphous phenyldi(pyren-1-yl)phosphine oxide (POPy2). The 31P DNP-ssNMR spectra exhibited a sufficient signal-to-noise ratio to quantify the distribution of molecular orientations in amorphous films: the P=O axis of the vacuum-deposited and drop-cast POPy2 shows anisotropic and isotropic distribution, respectively. The different molecular orientations reflect the molecular origin of the different charge transport behaviors.
Organic thin-film semiconducting devices, such as organic lightemitting diodes (OLEDs), organic solar cells (OSCs), and organic thin-film transistors, are expected to be the next generation devices owing to their light-weight, cost effectiveness, and flexible properties. Thus, organic semiconductors have been extensively studied over the last few decades.[1] The properties inherent to these devices, such as charge carrier mobility, light emission, and light out-coupling, depend on the intra- and intermolecular structures, including the orientation of organic molecules in the devices.[1d, e] Organic molecules in OLEDs and OSCs are often in an amorphous state, which limits the application of diffraction techniques based on X-rays and neutrons for detailed analysis of the orientation of organic amorphous films. The molecular orientation can be determined by angular-dependent photoluminescence measurements,[2] and variable angle spectroscopic ellipsometry[3]; however, these methods provide only average values, such as the order parameters. There are currently no appropriate methods for revealing the “distribution” of molecular orientations in an amorphous aggregate.
[a]
[b]
[c]
Dr. Katsuaki Suzuki, Shosei Kubo, Dr. Tatsuya Fukushima, and Prof. Dr. Hironori Kaji* Institute for Chemical Research Kyoto University Uji, Kyoto 611-0011, Japan E-mail: [email protected] Dr. Fabien Aussenac Bruker BioSpin 34, rue de l’Industrie, 67166 Wissembourg, France Dr. Frank Engelke Bruker BioSpin Silberstreifen, 76287 Rheinstetten, Germany Supporting information for this article is given via a link at the end of the document.
Figure 1. a) Molecular structures of POPy2 and bTbK. b) Schematic representation of sample preparation. The samples on glass (SiO2) or polytetrafluoroethylene (PTFE) are inserted into a 5-mmø quartz tube.
Solid-state nuclear magnetic resonance (ssNMR) is a powerful technique for analyzing the structure and dynamics of materials with atomic resolution, with both crystalline and amorphous morphologies.[4] Thus, it may be possible to perform a detailed analysis of the structure of semiconducting materials in organic devices. However, most ssNMR studies are based on bulk samples[5] and reports on thin film samples are rare.[6] One reason for this lack of previous studies is that the limited amount of organic material in the devices results in a low signal-to-noise (S/N) ratio. To enhance the sensitivity of the NMR signal, dynamic nuclear polarization (DNP) enhanced ssNMR (DNP-ssNMR) has recently attracted considerable attention. [7] In DNP-ssNMR experiments, radicals dispersed in a sample are polarized by microwave irradiation leading to a high electron polarization. The resulting electron polarization is transferred to the 1H population in the sample. In most cases, the enhanced 1H polarization is further transferred to other nuclei, typically by cross polarization (CP). Although most recent DNP-ssNMR measurements have been performed under magic angle spinning conditions,[5g, 7] Bechinger’s group reported static DNP-ssNMR measurements, which enabled insight into the structure of membrane polyproteins in lipid bilayers.[8] Here, we performed static DNP-ssNMR measurements for a semiconducting organic material, phenyldi(pyren-1yl)phosphine oxide (POPy2) (Figure 1a). POPy2 have electron transport properties,[9] and is thus frequently used in OLEDs. As shown later, we have detected different electron transport behaviors depending on different preparation methods, vacuum-
This article is protected by copyright. All rights reserved.
Accepted Manuscript
Katsuaki Suzuki[a], Shosei Kubo[a], Fabien Aussenac[b], Frank Engelke[c], Tatsuya Fukushima[a], and Hironori Kaji*[a]
10.1002/anie.201707208
Angewandte Chemie International Edition
deposition and drop-casting methods (Figure 1b). We selected a bisnitroxide radical, bTbK[10] as the polarizing agent, (Figure 1a), because our thermogravimetric measurements (Figure S1) revealed that bTbK can be sublimed before decomposition. The concentration of radicals is the main factor affecting the signal enhancement, because strong electron-electron exchange couplings among radicals are reported to decrease the DNP efficiency.[11] Thus, we used a radical doping concentration of 0.25 wt%. Electron paramagnetic resonance (EPR) spectra of the sublimed bTbK doped in POPy2 showed essentially the same spectra as the drop-cast spectra, as shown in Figure S2. The EPR signals were well-resolved, indicating that the radicals in the sample films did not feature any notable electron-electron exchange couplings. Amorphous POPy2 and bTbK (0.25 wt%) layers were formed on glass (SiO2) substrates. The thickness of the SiO2 substrate was 30 μm. The thickness of the vacuum-deposited POPy2 layer, 1.5 μm (ca. 52 μg for each SiO2 substrate), was thinner than the typical sample thickness required for time-offlight (TOF) experiments (normally, thicker than several μm), which is a most frequently used method to measure photocurrent charge transients. As shown in Figure S3, no distinct peaks were found in the XRD profiles for vacuumdeposited and drop-cast films, confirming that these layers are amorphous states. The Halo patterns for both the films are similar, indicating that XRD measurements do not provide information on the molecular orientation in amorphous films, even qualitatively. Details of sample preparation and measurements are provided in the Supporting Information. The POPy2 films on the substrates were set perpendicular to the external magnetic field, B0 (Figure 1b). We performed 31P CP DNP-ssNMR experiments under static conditions to obtain chemical shift anisotropy (CSA) spectra, which provide information on 31P=O orientations. Figure 2 shows the 31P CSA spectra of POPy2 with and without DNP enhancement. According to reports from Bechinger’s group,[8] we calculated the DNP enhancement factor (εon/off) to be the ratio of the integral signal intensity of the CSA spectra with and without DNP enhancements. The values of εon/off for the vacuum-deposited and drop-cast films on SiO2 (12 sheets for the vacuum-deposited films; 15 sheets for the dropcast films) were 3.0 and 2.0, respectively (Figure 2a and b). Figure 2c shows the 31P CSA spectrum of a single sheet of a POPy2 thin film with a measurement time of ~16 h. Compared with the stacked samples in Figure 2a and b, the εon/off value increased to 6.0. The diferrence in εon/off for stacked samples and one sheet sample is probably due to the difference of cooling efficiency of thin-film. The one sheet sample is more effectively cooled, leading to a higher εon/off value. Through the DNP enhancement, we successfully obtained a CSA spectrum with a sufficient S/N ratio from a sample of even a single sheet. Note that the POPy2 films on polytetrafluoroethylene (PTFE) substrates (10 sheets for the
Figure 2. 31P CSA spectra of POPy2 with (black) and without (gray) DNP enhancement. a) Vacuum-deposited on SiO2 (12 sheets), b) drop-cast on SiO2 (15 sheets), and c) vacuum-deposited on SiO2 (1 sheet). The normalized CSA lineshapes with and without microwave irradiations were the same as shown in Figure S4.
vacuum-deposited films; 15 sheets for the drop-cast films) exhibited εon/off values of 10 and 9.3, respectively (Figure S5a and b). These values are much higher than those on SiO 2. This difference in the values of εon/off for the samples on SiO2 and PTFE is likely caused by the different degrees of sample heating under microwave irradiation of SiO2 and PTFE substrates.[8a]
This article is protected by copyright. All rights reserved.
Accepted Manuscript
COMMUNICATION
10.1002/anie.201707208
Angewandte Chemie International Edition
Figure 4. Photocurrent transient curves of a) the vacuum-deposited POPy2 and b) drop-cast POPy2.
To quantify the molecular orientation distributions, the experimental CSA spectra were fitted by the Legendre moment expansion approach.[4a, 12] The spectra simulated by assuming a Gaussian-type orientation distribution function showed slightly poor reproducibility (Figure S6). Because the CSA spectra of POPy2 have an axially symmetric pattern, the orientational distribution function p(θ) was defined as an expansion of Legendre polynomials Pn(cosθ) with the following equations (1)– (3). 𝑝(𝜃) = ∑𝑛 Figure 3. a) Schematic representation of chemical shift tensor orientation for POPy2. The relationship between the P=O orientation and the chemical shift is also shown in calculated CSA spectrum with completely random orientations. b and c) Experimental 31P CSA spectra of POPy2 b) vacuum-deposited and c) drop-cast on SiO2 with DNP enhancement (black line). Simulated best-fit spectra are shown by the red line. d and e) Orientation distributions p(θ) (thick red lines) and contributions of respective terms (thin lines) of Legendre polynomials Pn(cosθ) for d) the vacuum-deposited film and e) the drop-cast film and on SiO2. Owing to the symmetry, values of p(θ) with θ > 90° are folded in between 0 and 90°.
The CSA patterns of the vacuum-deposited and drop-cast POPy2 were spread over a wide chemical shift range from 100 to −100 ppm depending on the P=O direction. The experimentallyobtained CSA spectra were axially-symmetric (Figures 2 and S5). From the structural symmetry of POPy2, the principal axes can be assigned as shown in Figure 3a. This assignment is confirmed by gauge-including atomic orbital (GIAO) calculations. The signal at −100 ppm, labeled 𝜎⫽ , corresponds to the parallel alignment of P=O axis of POPy2 relative to B0 (the angle between the unique principal axis of the uniaxial chemical shift tensor, 𝜎⫽ , and the P=O axis was calculated to be 3.2°). The signal at 100 ppm, labeled 𝜎⊥ , corresponds to the perpendicular alignment of the P=O axis relative to B0. The CSA spectrum of the vacuum-deposited POPy2 (Figure 3b) showed an enhanced signal intensity around −100 ppm compared with the drop-cast sample (Figure 3c), indicating the greater contribution of the P=O axis parallel to the B0.
2𝑛+1 2
〈𝑃𝑛 〉𝑃𝑛 (cos 𝜃) , 𝑛 = 0, 2, 4, 6, ⋯
(1)
𝜋
∫0 𝑝(𝜃) sin 𝜃 𝑑𝜃 = 2〈𝑃0 〉 = 1 〈𝑃𝑛 〉 =
(2)
𝜋
∫0 𝑝(𝜃)𝑃𝑛 (cos 𝜃) sin 𝜃𝑑𝜃
(3)
𝜋
∫0 𝑝(𝜃) sin 𝜃𝑑𝜃
Here, θ is the angle between the P=O axis and B0, and B0 is normal to the SiO2 substrate (Figure 3a). The function p(θ) gives the probability of finding a molecule at a particular orientation of θ. The experimental data were fitted by least squares minimization. The fitting parameters were the principal values of 𝜎⫽ and 𝜎⊥ , n-th order parameters of 〈𝑃𝑛 〉 (n = 2, 4, 6), and a Gaussian line broadening factor (BF). Higher order parameters (n ≥ 8) were ignored because the contribution of the 6th order parameter was negligibly small, as shown in Table 1. The vacuum-deposited and drop-cast 31P CSA spectra were well reproduced by the best-fit simulated spectra, shown
Table 1. Chemical shift tensor principal values (𝜎⫽ and 𝜎⊥ ), order parameters (〈𝑃2 〉, 〈𝑃4〉, 〈𝑃6〉), and Gaussian line broadening factor (BF) for the vacuumdeposited and drop-cast POPy2 films determined from fitting to the 31P CSA spectra in Figure 3b and c.
Sample
𝜎⫽ (ppm)
𝜎⊥ (ppm)
〈𝑃2〉
〈𝑃4〉
〈𝑃6 〉
BF (ppm)
Vacuumdeposited
114.3
-93.4
0.15
0.031
-0.0010
8.4
Dropcast
110.1
-88.4
-0.019
0.0059
-0.0014
7.2
This article is protected by copyright. All rights reserved.
Accepted Manuscript
COMMUNICATION
10.1002/anie.201707208
Angewandte Chemie International Edition
by the red curves in Figure 3b and c. The best-fit parameters are shown in Table 1. The thick red curves in Figure 3d and e represent p(θ) for the vacuum-deposited and drop-cast thin film samples. The respective contributions of Pn(cosθ) are also shown in the figures. For the vacuum-deposited films, the probability of finding a P=O orientation at θ = 0 was highest and decreased with increasing θ. This originates from the large contribution of the order parameter 〈𝑃2 〉 of 0.15 in addition to 〈𝑃0 〉. We detected some contribution from 〈𝑃4 〉 , but that of 〈𝑃6 〉 is negligible. For the drop-cast film, the contributions of 〈𝑃𝑛 〉, where n is higher than two are quite small (Figure 3e) and the CSA spectrum is mostly determined by 〈𝑃0 〉, indicating that the P=O axis of POPy2 is isotropically distributed in the drop-cast film. Differences in molecular orientation can be expected to affect the photocurrent transients of POPy2. Indeed, the transient photocurrent of the vacuum-deposited and drop-cast films of POPy2, measured by the TOF technique, exhibited different behavior, as shown in Figure 4 (see Supporting Information for experimental details). The vacuum-deposited POPy2 showed a “non-dispersive” photocurrent with a clear bending point, while the drop-cast POPy2 exhibited a featureless “dispersive” photocurrent. The random molecular orientation in drop-casted POPy2 films is considered to generate larger disorders (energetic disorder and structural disorder) compared to vacuum-deposited POPy2, resulting in disordered dispersive photocurrent. To gain further insight about the origin of different photocurrent transients, we will further develop our currently performing multiscale charge transport simulations[13] to link molecular-level structures and charge mobilities. In summary, we performed static DNP-ssNMR measurements of vacuum-deposited and drop-cast POPy2 thin films. Under the DNP-enhanced condition, we could obtain 31P CSA spectra with a S/N ratio sufficient for orientation analysis. The CSA analysis based on the Legendre moment expansion approach revealed a quantitative orientational distribution of the vacuum-deposited and drop-cast amorphous POPy2 thin films. The vacuum-deposited POPy2 favors a perpendicular orientation of the P=O axes to the SiO2 substrates, while drop-cast POPy2 exhibits an isotropic random orientation. This difference in orientation affects the photocurrent transient behavior.
[1]
[2]
[3]
[4]
[5]
[6] [7]
Acknowledgements This work was supported by Japan Society and the Promotion of Science (JSPS) Grant-in-Aid for Scientific Research (A) (No. 17H01231) and Grant-in-Aid for JSPS Fellows (No. 14J04794), and also supported by Kyoto University for Supporting Program for Interaction-Based Initiative Team Studies (SPIRITS). FA and FE are grateful to Christian Reiter, Hiba Sarrouj, and Armin Purea, Bruker Biospin, for the extensive engineering work on the static DNP probe. We thank Andrew Jackson, PhD, from Edanz Group (www.edanzediting.com/ac) for editing a draft of this manuscript. Keywords: Amorphous materials • Dynamic nuclear polarization • NMR spectroscopy • Molecular orientation • Organic semiconductor
[8]
[9] [10]
[11] [12]
a) L. Dou, Y. Liu, Z. Hong, G. Li, Y. Yang, Chem. Rev. 2015, 115, 12633-12665; b) O. Ostroverkhova, Chem. Rev. 2016, 116, 1327913412; c) Z. Yang, Z. Mao, Z. Xie, Y. Zhang, S. Liu, J. Zhao, J. Xu, Z. Chi, M. P. Aldred, Chem. Soc. Rev. 2017, 46, 915-1016; d) H. Kaji, H. Suzuki, T. Fukushima, K. Shizu, K. Suzuki, S. Kubo, T. Komino, H. Oiwa, F. Suzuki, A. Wakamiya, Y. Murata, C. Adachi, Nat. Commun., 2015, 6, 8476; e) T. Miwa, S. Kubo, K. Shizu, T. Komino, C. Adachi, H. Kaji, Sci. Rep. 2017, 7, 284. a) J. Frischeisen, D. Yokoyama, C. Adachi, W. Brütting, Appl. Phys. Lett. 2010, 96, 073302; b) J. W. Sun, J.-H. Lee, C.-K. Moon, K.-H. Kim, H. Shin, J.-J. Kim, Adv. Mater. 2014, 26, 5684-5688. a) D. Yokoyama, Y. Setoguchi, A. Sakaguchi, M. Suzuki, C. Adachi, Adv. Funct. Mater. 2010, 20, 386-391; b) M. Shibata, Y. Sakai, D. Yokoyama, J. Mater. Chem. C 2015, 3, 11178-11191. a) K. Schmidt-Rohr, H. W. Spiess, Multidimensional solid-state NMR and polymers, Academic Press, 1994; b) H. Kaji, N. Miura, K. SchmidtRohr, Macromolecules 2003, 36, 6100-6113; c) M. Sharma, M. Yi, H. Dong, H. Qin, E. Peterson, D. D. Busath, H.-X. Zhou, T. A. Cross, Science 2010, 330, 509-512; d) R. Bhattacharyya, B. Key, H. Chen, A. S. Best, A. F. Hollenkamp, C. P. Grey, Nat. Mater. 2010, 9, 504-510; e) J. K. Williams, D. Tietze, M. Lee, J. Wang, M. Hong, J. Am. Chem. Soc. 2016, 138, 8143-8155. a) M. Utz, M. Nandagopal, M. Mathai, F. Papadimitrakopoulos, Appl. Phys. Lett. 2003, 83, 4023-4025; b) H. Kaji, Y. Kusaka, G. Onoyama, F. Horii, J. Am. Chem. Soc. 2006, 128, 4292-4297; c) Y. Nishiyama, T. Fukushima, K. Takami, Y. Kusaka, T. Yamazaki, H. Kaji, Chem. Phys. Lett. 2009, 471, 80-84; d) T. Fukushima, H. Kimura, Y. Shimahara, H. Kaji, Appl. Phys. Lett. 2011, 99, 223301; e) F. Suzuki, T. Fukushima, M. Fukuchi, H. Kaji, J. Phys. Chem. C 2013, 117, 18809-18817; f) F. Suzuki, Y. Nishiyama, H. Kaji, Chem. Phys. Lett. 2014, 605–606, 1-4; g) S. R. Chaudhari, J. M. Griffin, K. Broch, A. Lesage, V. Lemaur, D. Dudenko, Y. Olivier, H. Sirringhaus, L. Emsley, C. P. Grey, Chem. Sci. 2017, 8, 3126-3136. Y. Nishiyama, T. Fukushima, M. Fukuchi, S. Fujimura, H. Kaji, Chem. Phys. Lett. 2013, 556, 195-199. a) D. A. Hall, D. C. Maus, G. J. Gerfen, S. J. Inati, L. R. Becerra, F. W. Dahlquist, R. G. Griffin, Science 1997, 276, 930-932; b) H. Takahashi, D. Lee, L. Dubois, M. Bardet, S. Hediger, G. De Paëpe, Angew. Chem. Int. Ed. 2012, 51, 11766-11769; Angew. Chem. 2012, 124, 1193611939; c) A. J. Rossini, C. M. Widdifield, A. Zagdoun, M. Lelli, M. Schwarzwälder, C. Copéret, A. Lesage, L. Emsley, J. Am. Chem. Soc. 2014, 136, 2324-2334; d) D. Lee, G. Monin, N. T. Duong, I. Z. Lopez, M. Bardet, V. Mareau, L. Gonon, G. De Paëpe, J. Am. Chem. Soc. 2014, 136, 13781-13788; e) A. N. Smith, J. R. Long, Anal. Chem. 2016, 88, 122-132; f) D. Stöppler, C. Song, B.-J. van Rossum, M.-A. Geiger, C. Lang, M.-A. Mroginski, A. P. Jagtap, S. T. Sigurdsson, J. Matysik, J. Hughes, H. Oschkinat, Angew. Chem. Int. Ed. 2016, 55, 16017-16020; Angew. Chem. 2016, 128, 16251-16254; g) M. T. Colvin, R. Silvers, Q. Z. Ni, T. V. Can, I. Sergeyev, M. Rosay, K. J. Donovan, B. Michael, J. Wall, S. Linse, R. G. Griffin, J. Am. Chem. Soc. 2016, 138, 9663-9674. a) E. S. Salnikov, H. Sarrouj, C. Reiter, C. Aisenbrey, A. Purea, F. Aussenac, O. Ouari, P. Tordo, I. Fedotenko, F. Engelke, B. Bechinger, J. Phys. Chem. B 2015, 119, 14574-14583; b) E. S. Salnikov, C. Aisenbrey, F. Aussenac, O. Ouari, H. Sarrouj, C. Reiter, P. Tordo, F. Engelke, B. Bechinger, Sci. Rep. 2016, 6, 20895. T. Matsushima, C. Adachi, Appl. Phys. Lett. 2006, 89, 253506. Y. Matsuki, T. Maly, O. Ouari, H. Karoui, F. Le Moigne, E. Rizzato, S. Lyubenova, J. Herzfeld, T. Prisner, P. Tordo, R. G. Griffin, Angew. Chem. Int. Ed. 2009, 48, 4996-5000; Angew. Chem. 2009, 121, 50965100. B. Corzilius, L. B. Andreas, A. A. Smith, Q. Z. Ni, R. G. Griffin, J. Magn. Reson. 2014, 240, 113-123. a) R. Hentschel, J. Schlitter, H. Sillescu, H. W. Spiess, J. Chem. Phys. 1978, 68, 56-66; b) W. Gabriëlse, H. Angad Gaur, W. S. Veeman, Macromolecules 1996, 29, 4125-4133.
This article is protected by copyright. All rights reserved.
Accepted Manuscript
COMMUNICATION
10.1002/anie.201707208
Angewandte Chemie International Edition
COMMUNICATION H. Uratani, S. Kubo, K. Shizu, F. Suzuki, T. Fukushima, H. Kaji, Sci. Rep. 2016, 6, 39128.
Accepted Manuscript
[13]
This article is protected by copyright. All rights reserved.
10.1002/anie.201707208
Angewandte Chemie International Edition
COMMUNICATION Entry for the Table of Contents
COMMUNICATION The orientation of organic semiconducting molecules in an amorphous thin film was revealed by static dynamic nuclear polarization enhanced solid-state NMR. The P=O axis of an electron transport material, called POPy2, in a vacuum-deposited film tends to orient perpendicular to the substrates. We find a close relationship between the orientational distribution and charge transport behavior.
Katsuaki Suzuki, Shosei Kubo, Fabien Aussenac, Frank Engelke, Tatsuya Fukushima, and Hironori Kaji*
Analysis of Molecular Orientation in Organic Semiconducting Thin Films Using Static Dynamic Nuclear Polarization Enhanced Solid-State NMR
This article is protected by copyright. All rights reserved.
Accepted Manuscript
Page No. – Page No.
