Journal of Thermal Analysis and Calorimetry

, Volume 133, Issue 1, pp 797–803 | Cite as

Analysis of the thermal hazards of 1-butyl-3-methylimidazolium chloride mixtures with cellulose and various metals

  • Nana Yamaki
  • Kento Shiota
  • Yu-ichiro Izato
  • Atsumi Miyake


Ionic liquids (ILs) are a relatively new class of environmentally benign and comparatively safe solvents and are expected to have numerous applications in chemical processes. Although pure ILs are thermally stable, the presence of impurities can affect their thermal stability and decomposition behavior. In addition, ILs decomposition products include flammable gases that may present a fire hazard. When designing safer processes and operating conditions, it is therefore important to investigate IL thermal properties and decomposition products in combination with additives. The present work focused on cellulose dissolution which is promising application of ILs to obtain better understanding of thermal hazards. Mixtures of cellulose, iron (III) oxide (Fe2O3), copper(II) oxide (CuO), chromium, and nickel with 1-butyl-3-methylimidazolium chloride (BmimCl) were examined, using differential scanning calorimetry, high-sensitivity calorimetry, and thermogravimetry–differential thermal analysis–mass spectrometry. The addition of CuO was found to generate an exothermic reaction below the decomposition temperature of BmimCl and also to lower the decomposition temperature. BmimCl/CuO mixtures also produced extremely flammable gases below the decomposition temperature of pure BmimCl.


Ionic liquids Thermal hazards Cellulose dissolving 1-Butyl-3-methylimidazolium chloride Evolved gas analysis 


Ionic liquids (ILs), generally defined as salts with a melting point below 100 °C, have been proposed as environmentally benign, safer solvents [1, 2, 3, 4]. ILs have many attractive properties, including low volatility and flammability, good thermal stability, wide electrochemical windows, and high ion conductivities. Furthermore, these properties can be finely tuned by varying the combination of ILs anion and cation as well as the structures of this component [1, 2, 3, 4, 5, 6, 7, 8, 9, 10]. Hence, ILs are sometimes referred to as “designer solvents” and the applications of IL to various chemical processes are continually increasing. The design of safe, effective chemical processes requires an understanding of the thermal hazards associated with the procedure and the associated materials [11, 12]. Although the majority of pure ILs are nonflammable, the decomposition products of ILs may include flammable gases, so the thermal decomposition of ILs can be a potential hazard [13]. The decomposition reactions of ILs are also accelerated by the presence of impurities. As an example, the presence of quartz or amorphous silica lowers the decomposition temperatures of 1-alkyl-3-methylimidazolium phosphate and triflate in ambient air [14] and aluminum both reduces the decomposition temperature of Imidazolium-based ILs and changes the decomposition from endothermic to exothermic [15]. Nucleophilic anions have been shown to lower the decomposition temperature of tetraalkylimidazolium ILs by more than 150 °C [16, 17, 18]. ILs are typically used with various additives during chemical processing; ILs also have higher viscosities than conventional solvents [19] and thus may store heat of reaction and can be difficult to cool. Therefore, if exothermic reactions occur in ILs at or below the processing temperature, the heats of reaction may accumulate in the ILs, leading to overheat and decomposition to generate flammable gases. Thus, it is important to investigate the thermal properties and gas evolution of ILs in conjunction with additives or impurities. The present work focused on systems in which cellulose was dissolved in ILs, since this is one of the expected applications of ILs.

Cellulose is a highly abundant renewable material and can be a valuable material for biofuel manufacturing and the textile sector [20]. Cellulose requires chemical derivatization because it is insoluble in common solvents due to its highly crystalline structure and strong network of hydrogen bonds between fibrils [21]. Conventional chemical derivatization systems for biopolymers are associated with environmental and safety concerns because of the use of large amounts of acidic or thermally unstable solvents [21, 22]. However, in 2002, Richard et al. [20, 23] reported that cellulose can be dissolved in imidazolium-based ILs. Since then, cellulose dissolution using ILs has been an important research topic and many studies investigated the capacity and efficiency of cellulose dissolution in ILs [20, 21, 22, 23, 24, 25, 26]. Despite this, thermal hazards, especially in the case of ILs with additives, are not well understood. Frank et al. [20] measured the onset temperature, heat of reaction, and pressure rise of a cellulose solution in 1-ethyl-3-methylimidazolium acetate (EmimAc). Their work revealed that the addition of cellulose to EmimAc lowered the decomposition temperature of IL while increasing its heat of reaction and pressure rise. Yamamoto et al. [27] revealed that co-solvent composed of 1-butyl-3-methylimidazolium acetate (BmimAc) and dimethyl sulfoxide (DMSO) could undergo a violent exothermal reaction. These results demonstrated that ILs mixtures can be more hazardous than pure ILs. Therefore, thermal hazard analysis for ILs with additives is needed.

The purpose of this study was to obtain a better understanding of the thermal hazards of ILs mixtures. The compound 1-butyl-3-methylimidazolium chloride (BmimCl), which is an excellent solvent for cellulose, was selected as a model IL. Cellulose, iron (III) oxide (Fe2O3), copper(II) oxide (CuO), chromium (Cr), and nickel (Ni) were used as the additives. Cellulose is an ingredient of biofuel and solute of BmimCl. Fe2O3 is a main component of rust which may be present on processing equipment. CuO is the main component of wood preservative and would be included in construction waste. Cr and Ni are components of stainless steel. The chloride ion in BmimCl could potentially corrode stainless steel equipment, such as pipes and reactors, releasing their chemical component. These additives may contaminate with BmimCl in the cellulose dissolving process. The thermal behavior of BmimCl containing additives was assessed using differential scanning calorimetry (DSC). Following these trials, high-sensitivity calorimetry was conducted to examine samples that showed abnormal reactions by detecting small exothermic reactions and heats of reaction. Evolved gas analysis was also conducted for samples that exhibited exotherms below the decomposition temperature of BmimCl.



In the present work, 1-butyl-3-methylimidazolium chloride (BmimCl) was employed as the solvent. The structure of BmimCl is shown in Fig. 1. BmimCl, Fe2O3, Cr, and Ni are obtained from Wako Pure Chemical Industries, Ltd., and CuO was purchased from the Kanto Chemical Co., Inc. The physical properties of BmimCl and the various additives are shown in Table 1.
Fig. 1

Chemical structure of BmimCl

Table 1

Physical properties of the test materials









98% ≧

95% ≧

99.3% ≧

99% ≧

95% ≧

Molecular weight/g mol−1






State of materials







Particle size

< 38 μm

27.2–95.3 nm

< 75 μm

< 150 μm

Thermal analysis for screening

The thermal behavior and reaction onset temperature of each sample were investigated using differential scanning calorimetry (DSC) (HP-DSC827e, Mettler Toledo). In the case that mixtures of the BmimCl with various additives show the same endo-/exothermic behavior and approximately the same onset temperature as the pure BmimCl, the additives were assumed not to affect the BmimCl and the thermal hazard of the mixture was assessed as very low. In contrast, reduction in the decomposition temperature or exothermic reactions suggested that additives lowered the thermal stability of the BmimCl. In these DSC trials, samples of approximately 4 mg were loaded into an open alumina pan and heated from 30 to 400 °C at a heating rate of 10 K min−1. During DSC measurements, the apparatus was continually purged with a flow of either air or argon (Ar) at 100 mL min−1. Each BmimCl/additive mixture had a mass ratio of 9:1. The DSC apparatus was calibrated using high-purity indium at a heating rate of 10 K min−1.

High-sensitivity calorimetry

A high-sensitivity calorimeter (C80, Setaram) was used to further assess samples that exhibited abnormal thermal behavior. The aim of this test was to detect small exotherms and to determine its heat of reaction below the decomposition temperature of BmimCl. In these trials, samples of approximately 140 mg were loaded into a glass tube within a sealed high-pressure stainless steel vessel and the vessel was purged with Ar. The apparatus was heated to 30 °C and held at that temperature for 5 h to stabilize the heat flow and then heated to 230 °C at 0.1 K min−1.

Evolved gas analysis

Evolved gas and thermal analyses were conducted using thermogravimetry–differential thermal analysis–mass spectrometry (TG–DTA–MS) simultaneously. The aim of this test is to investigate the decomposition products in conjunction with thermal and mass loss behaviors. The instrumentation was consisted of a simultaneous thermal analyzer (STA2500-YKD26 Regulus, Netzsch) and a mass spectrometer (QMS 403D, Netzsch). Samples of approximately 4 mg were loaded into open alumina pan and heated from 30 to 400 °C at 10 K min−1 with a constant 100 mL min−1 flow of helium (He) gas as the carrier gas. Gaseous decomposition products were transported to the mass spectrometer through a quartz capillary tube held at 230 °C. The mass spectrometer was operated in the electro-ionization mode with an ion source temperature of 300 °C, and the electrons are accelerated to 70 eV.

Results and discussion

Thermal analysis for screening

Figures 2 and 3 present the DSC data obtained from mixture of BmimCl with additives under Ar and air, respectively. The onset temperatures are temperatures determined for each baseline shown in Figs. 2 and 3. Onset temperatures for these samples are summarized in Table 2. In this study, the onset temperature was defined as the intersection of the inflectional tangent through the peak slope with the baseline. In the Ar data, each sample exhibits a slight endothermic peak at approximately 100 °C, followed by a more prominent endotherm at a higher temperature. The first endotherm is to the latent heat of vaporization of water, since ILs are hygroscopic and can absorb large amounts of water from the ambient air [28]. The second endotherm, beginning at approximately 235 °C, results from the decomposition of the BmimCl [29]. The addition of Fe2O3, CuO, Cr, or Ni did not lower the decomposition temperature or result in exothermic reactions. Thus, these additives appear not to affect thermal stability of BmimCl under Ar.
Fig. 2

DSC curves obtained from BmimCl mixtures under Ar at a heating rate of 10 K min−1

Fig. 3

DSC curves obtained from BmimCl mixtures under air at a heating rate of 10 K min−1

Table 2

Onset temperatures for reactions of samples


Onset temperature TDSC in argon/°C

Onset temperature TDSC in air/°C






210 (exo.) 251 (endo.)



237 (endo.) 298 (exo.)



242 (endo.) 299 (exo.)







TDSC: the onset temperature of each sample

In air, only the BmimCl/cellulose shows a small exotherm at approximately 210 °C, while the Fe2O3 and CuO mixtures exhibit exothermic reactions that appear in the middle of the BmimCl decomposition. The Ni and Cr mixtures do not show any abnormal reactions. The onset temperatures of the exothermic reactions of the BmimCl/CuO and BmimCl/Fe2O3 were 298 and 299 °C, respectively. These exotherms are assumed to involve the oxidation of BmimCl decomposition products because they occurred only under air. The data show that BmimCl/CuO or BmimCl/Fe2O3 above 300 °C may induce exothermic reactions that raise the system temperature.

High-sensitivity calorimetry

High-sensitivity calorimetry was employed to detect exothermic reactions occurring below the decomposition temperature of the BmimCl and to obtain the heat of reaction. An exothermic reaction below the BmimCl decomposition temperature can result in decomposition of the BmimCl because the resulting heat can accumulate and increase the temperature. Figure 4 presents the resulting data for the pure BmimCl as well as for mixtures with cellulose, Fe2O3, and CuO, either under Ar or air. Neither the BmimCl nor the BmimCl/Fe2O3 exhibits significant exotherms below the decomposition temperature of BmimCl. In contrast, the BmimCl/CuO sample underwent an exothermic reaction at 135 °C with as heat of reaction of 35 J g−1. Under air, this same sample showed a heat of reaction of 91 J g−1. These results demonstrate that oxygen affects exothermic reaction of the BmimCl/CuO combination and increases the associated heat of reaction. The BmimCl/cellulose mixture showed an exothermic reaction at 190 °C, but it was not possible to obtain the heat of reaction because this exotherm did not finish until 230 °C.
Fig. 4

C80 curves obtained from pure BmimCl and mixtures with CuO, cellulose, and Fe2O3 at the heating rate of 1 K min−1

Because exothermic reactions occurring under operation temperature can be hazardous, we focused on the BmimCl/CuO combination, which generated an exotherm under the conditions applied in this study. The Globally Harmonized System of classification and Labelling of Chemicals (GHS) provides criteria for determining self-reactive substances and mixtures, in which the cutoff value for the heat of reaction is greater than 300 J g−1 [30]. Based on this standard, we concluded that the heat of reaction of BmimCl/CuO is not catastrophic, and it presents only a minor thermal hazard.

Evolved gas analysis

Figure 5 provides the TG–DTA–MS data for BmimCl and BmimCl/CuO under He. In these trials, the mass–charge ratio (m/z) range of 10–200 was monitored. The peak top of m/z = 50, which shows the largest intensity among decomposition products of BmimCl, was set at 100. Then, relative intensity for other mass peaks was determined. Considering the TG–DTA results, an endotherm with mass loss resulting from the decomposition reaction of the BmimCl began at 230 °C. In contrast, the mass loss of the BmimCl/CuO sample can be divided into two steps, starting at 168 and 230 °C. The mass spectral data for both the BmimCl and the BmimCl/CuO mixture indicate that products having m/z = 18 (H2O), 28(C2H4, N2, CO), 41(C3H6, C2H3N), 44(CO2, N2O), and 50(CH3Cl) were mainly generated. BmimCl is known to decompose via an SN2 reaction to form 1-butyl imidazole and CH3Cl as the primary decomposition products [31, 32], while the decomposition products of 1-butyl imidazole are 1-methylimidazole, propylene, 1-ethyl imidazole, and ethylene [33]. The most intensive peak in a spectrum shown in NIST chemistry web book is determined as a base peak of the compound, whose relative intensity is 100. From the database in Ref. [34], the peaks at m/z values of 50, 41, and 28 can be attributed to CH3Cl, C3H6, and C2H4 [34]. In addition, because the ring opening of imidazolium-based ILs is difficult, including nitrogen is rarely generated [35]. Thus, the products at m/z = 41, 44, and 50 are identified as propylene (CH2=CH–CH3), carbon dioxide (CO2), and methyl chloride (CH3Cl). Mass spectra including CO2 typically also exhibit a peak at an m/z = 28 due to CO, with a CO2 peak-to-CO peak intensity ratio of 100:10. However, the MS data in Fig. 5 indicate that the m/z = 28 peak is more intense than the m/z = 44, strongly suggesting that ethylene (CH2=CH2) was also produced. Both ethylene and propylene are extremely flammable, and their mixture with air is explosive. Methyl chloride is flammable and also attacks many metals in the presence of moisture. The mixing of CuO with BmimCl generates these hazardous decomposition gases beginning at approximately 168 °C.
Fig. 5

TG–DTA–MS profiles of pure BmimCl and BmimCl/CuO under He condition

A schematic reaction mechanism for a mixture of BmimCl and CuO

Figure 6 summarizes the proposed decomposition mechanism for BmimCl/CuO mixture. Figures 710 show photographic images of samples in alumina pans. At ambient temperature approximately 25 °C, the black powdered CuO can be seen on the solid BmimCl (Fig. 7). Heating the mixture in air changes the sample to red, suggesting that the CuO oxidizes the BmimCl and/or its decomposition products and is reduced to Cu2O (Fig. 8). As the sample is further heated in air, the color gradually changes again, from red to black, indicating that the Cu2O may oxidized back to CuO by oxygen in the air (Fig. 10). From Figs. 710 and the TG–DTA–MS profiles, it can be concluded that the BmimCl was gradually decomposed by oxidative reactions and that CuO promoted this oxidative decomposition.
Fig. 6

A mechanism for the reaction of a BmimCl and CuO mixture

Fig. 7

A picture of BmimCl/CuO at 23 °C

Fig. 8

A picture of BmimCl/CuO at 170 °C

Fig. 9

A picture of BmimCl/CuO at 400 °C in Ar

Fig. 10

A picture of BmimCl/CuO at 400 °C in air


An analysis of the thermal hazards presented by BmimCl in combination with various metals was performed to investigate of the thermal properties and evolved gases of these mixtures. The DSC data show that BmimCl/cellulose mixtures exhibit an exothermic reaction below the decomposition temperature of BmimCl under ambient air. The addition of Fe2O3 or CuO results in an exothermic reaction during the decomposition in air. Cr and Ni do not affect the thermal stability of BmimCl in either argon or air. High-sensitivity calorimetry measurements detected an exothermic reaction in the case of the BmimCl/CuO mixture, but with a relatively low heat of reaction of 35 J g−1, suggesting that this is not hazardous reaction. TG–DTA–MS analysis revealed that CuO lowers the decomposition temperature of BmimCl from 230 to 168 °C. Ethylene, propylene, and methyl chloride, all of which are extremely flammable gases, were detected as decomposition products of both BmimCl and BmimCl/CuO. The BmimCl/CuO mixture also exhibited a color change during heating, suggesting the reaction of CuO with the BmimCl to promote the decomposition of the BmimCl.


  1. 1.
    Zhimin X, Yuwei Z, Xiao-qin Z, Yuanyuan C, Tiancheng M. Thermal stabilities and decomposition mechanism of amino- and hydroxyl-functionalized ionic liquids. Thermochim Acta. 2014;578:59–67.CrossRefGoogle Scholar
  2. 2.
    Martyn JE, Kenneth RS. Ionic liquids. Green solvents for the future. Pure Appl Chem. 2000;72:1391–8.CrossRefGoogle Scholar
  3. 3.
    Ngoc LM, Kihun A, Yoon-Mo K. Methods for recovery of ionic liquids—a review. Process Biochem. 2014;49:872–81.CrossRefGoogle Scholar
  4. 4.
    Pengfei Z, Yutong G, Yiqi L, Yan G, Yong W, Cong W. Ionic liquids with metal chelate anions. Chem Commun. 2012;48:2334–6.CrossRefGoogle Scholar
  5. 5.
    John SW. A short history of ionic liquids—from molten salts to neoteric solvents. Green Chem. 2002;4:73–80.CrossRefGoogle Scholar
  6. 6.
    Fabio B, Cinzia C. The heck reaction in ionic liquids: progress and challenges. Molecules. 2010;15:2211–45.CrossRefGoogle Scholar
  7. 7.
    Christian PM. Supported ionic liquid catalysis. Chem Eur. 2005;11:50–6.CrossRefGoogle Scholar
  8. 8.
    Roberto IC, Joan FB. Comparison of Ionic liquids to conventional organic solvents for extraction of aromatics from aliphatics. J Chem Eng Data. 2016;61:1685–99.CrossRefGoogle Scholar
  9. 9.
    Samir IAE. Ionic liquids recycling and reuse. In: Scott TH, editor. Ionic liquids—classes and properties. London: Intech; 2011. p. 239–72.Google Scholar
  10. 10.
    Barbara R, Martin S, Andrea B, Martin W, Wolfgang K. Polymer electrolyte for lithium batteries based on photochemically crosslinked poly(ethylene oxide) and ionic liquid. Eur Polym J. 2008;44:2986–90.CrossRefGoogle Scholar
  11. 11.
    Zhe W, Steen MR, Bradley DG, Timothy AG. Safety concerns in a pharmaceutical manufacturing process using dimethyl sulfoxide (DMSO) as a solvent. Org Process Res Dev. 2012;16:1994–2000.CrossRefGoogle Scholar
  12. 12.
    Francis S. Atsumi Miyake translator. Thermal safety of chemical process: risk assessment and process design. Tokyo: Maruzen; 2011.Google Scholar
  13. 13.
    Horng-Jang L, Shih-Kai H, Hao-Ying C, Sheng-Nan L. 2012 International symposium on safety science and technology reason for ionic liquids to be combustible. Procedia Engineering, vol. 45; 2012. p. 502–6.Google Scholar
  14. 14.
    Marek K, Jan G, Jarl BR. Thermal stability of low temperature ionic liquids revisited. Thermochim Acta. 2004;412:47–53.CrossRefGoogle Scholar
  15. 15.
    Helen LN, Karen L, Liesl H, Alan BM. Thermal properties of imidazolium ionic liquids. Thermochim Acta. 2000;357–8:97–102.Google Scholar
  16. 16.
    Douglas MF, Jeffrey WG, Hugh CDL, Paul CT. TGA decomposition kinetics of 1-butyl-2,3-dimethylimidazolium tetrafluoroborate and the thermal effects of contaminants. J Chem Thermodyn. 2005;37:900–5.CrossRefGoogle Scholar
  17. 17.
    Walid HA, Jeffrey W, Marc N, Richard HH, Thomas E, John C, Paul CT, Hugh CD, Douglas MF. Thermal degradation studies of alkyl-imidazolium salts and their application in nanocomposites. Thermochim Acta. 2004;409:3–11.CrossRefGoogle Scholar
  18. 18.
    Douglas MF, Walid HA, Jeffrey WG, Paul HM, Hugh CDL, Paul CT. Flammability, thermal stability, and phase change characteristics of several trialkylimidazolium salts. Green Chem. 2003;5:724–7.CrossRefGoogle Scholar
  19. 19.
    Qiwei Y, Kun Y, Huabin X, Baogen S, Zongbi B, Yiwen Y, Qilong R. The effect of molecular solvents on the viscosity, conductivity and ionicity of mixtures containing chloride anion-based ionic liquid. J Ind Eng Chem. 2013;19:1708–14.CrossRefGoogle Scholar
  20. 20.
    Frank W, Loredana NT, Frank M. Thermostability of imidazolium ionic liquids as direct solvents for cellulose. Thermochim Acta. 2012;528:76–84.CrossRefGoogle Scholar
  21. 21.
    Yujin C, Rubing Z, Tao C, Jing G, Mo X, Huizhou L. Imidazolium-based ionic liquids for cellulose pretreatment: recent progresses and future perspectives. Appl Microbiol Biotechnol. 2017;101:521–32.CrossRefGoogle Scholar
  22. 22.
    Martin G, Pedro F, Thomas H. Ionic liquids—promising but challenging solvents for homogeneous derivatization of cellulose. Molecular. 2012;17:7458–502.CrossRefGoogle Scholar
  23. 23.
    Richard PS, Scott KS, John DH, Robin DR. Dissolution of cellulose with ionic liquids. J Am Chem Soc. 2002;124:4974–5.CrossRefGoogle Scholar
  24. 24.
    Hyungsup K, Yongjun A, Seung Y. Comparing the influence of acetate and chloride anions on the structure of ionic liquid pretreated lignocellulosic biomass. Biomass Bioenergy. 2016;93:234–53.Google Scholar
  25. 25.
    Yan C, Jin W, Jun Z, Huiquan L, Yi Z, Jiasong H. Room temperature ionic liquids (RTILs): a new and versatile platform for cellulose processing and derivatization. Chem Eng J. 2009;147:13–21.CrossRefGoogle Scholar
  26. 26.
    Samira VF, Yong-Wah K, Constance AS. A coupled low temperature oxidative and ionic liquid pretreatment of lignocellulosic biomass. Catal Today. 2016;269:2–8.CrossRefGoogle Scholar
  27. 27.
    Yamamoto Y, Miyake A. Influence of a mixed solvent containing ionic liquids on the thermal hazard of the cellulose dissolution process. J Therm Anal Calorim. 2017;127:743–8.CrossRefGoogle Scholar
  28. 28.
    Chieu DT, Silvia L, Daniel O. Absorption of water by room-temperature ionic liquids: effect of anions on concentration and state of water. Appl Spectrosc. 2003;57:152–7.CrossRefGoogle Scholar
  29. 29.
    Anastasia E, Grit H, Peer S. Thermal stability and crystallization behavior of imidazolium halide ionic liquids. Thermochim Acta. 2013;573:162–9.CrossRefGoogle Scholar
  30. 30.
    United Nations. Globally harmonized system of classification and labeling of chemicals (GHS) fourth revised edition. 2011. Accessed 25 June 2017.
  31. 31.
    Maaike CK, Wim B, Cor JP, Geert-Jan W. Quantum chemical aided prediction of the thermal decomposition mechanisms and temperatures of ionic liquids. Thermochim Acta. 2007;465:40–7.CrossRefGoogle Scholar
  32. 32.
    Arindrajit C, Stefan TT. Confined rapid thermolysis/FTIR/ToF studies of imidazolium-based ionic liquids. Thermochim Acta. 2006;443:159–72.CrossRefGoogle Scholar
  33. 33.
    Yan H, Jing P, Shaowen H, Jiuqiang L, Maolin Z. Thermal decomposition of allyl-imidazolium-based ionic liquid studied by TGA–MS analysis and DFT calculations. Thermochim Acta. 2010;501:78–83.CrossRefGoogle Scholar
  34. 34.
    NIST Mass Spec Data Center. Mass Spectra in NIST Chemistry WebBook, NIST Standard Reference Database Number 69. Eds. Accessed 21 July 2017.
  35. 35.
    Siedlecka EM, Czerwicka M, Neumann J, Stepnowski P, Fernández JF, Thöming J. Ionic liquids: methods of degradation and recovery. In: Ionic liquids: theory, properties, new approaches. 2011. Accessed 18 July 2017.

Copyright information

© Akadémiai Kiadó, Budapest, Hungary 2018

Authors and Affiliations

  • Nana Yamaki
    • 1
  • Kento Shiota
    • 1
  • Yu-ichiro Izato
    • 1
  • Atsumi Miyake
    • 1
    • 2
  1. 1.Graduate School of Environment and Information SciencesYokohama National UniversityYokohamaJapan
  2. 2.Institute of Advanced SciencesYokohama National UniversityYokohamaJapan

Personalised recommendations