Surface modification effects of graphite for selective hydrogen absorption by titanium at room temperature

Keita Shinzato a, Yuki Nakagawa b, So Hamamoto a, Yuya Hayashi c, Hiroki Miyaoka de, Shigehito Isobe b, Tamaki Shibayama b, Norio Ogita e and Takayuki Ichikawa *ade
aGraduate School of Engineering, Hiroshima University, 1-4-1 Kagamiyama, Higashi-Hiroshima, 739-8527, Japan. E-mail:
bFaculty of Engineering, Hokkaido University, N-13, W-8, Sapporo, 060-8628, Japan
cGraduate School of Engineering, Hokkaido University, N-13, W-8, Sapporo, 060-8628, Japan
dNatural Science Center for Basic Research and Development, Hiroshima University, 1-3-1 Kagamiyama, Higashi-Hiroshima, 739-8530, Japan
eGraduate School of Advanced Science and Engineering, Hiroshima University, 1-4-1 Kagamiyama, Higashi-Hiroshima, 739-8527, Japan

Received 26th April 2020 , Accepted 25th May 2020

First published on 26th May 2020

Surface modification effects of graphite and organic solvents on Ti were investigated by thermogravimetry (TG), Raman spectroscopy, and transmission electron microscopy (TEM) observations to improve its hydrogen absorption properties. As a result, Ti ball-milled with graphite showed high reactivity and selectivity for hydrogen with high durability.

Hydrogen (H2) is considered as a secondary energy source because it can chemically store energy with high gravimetric energy density. The applications of metal–hydrogen systems have been developing to effectively utilize H2 as an energy carrier.1–7 To establish H2 utilization systems, the development of related technologies is also necessary. Accidents of H2 leakage in closed areas are especially serious due to the low ignition energy and wide explosion range of H2 (4 to 75 vol% in air). Thus, H2 capture materials are required to prevent H2 explosion and utilize H2 safely. Titanium (Ti) is one of the promising materials for H2 capture because of its quite low H2 equilibrium pressure at room temperature. Metals with a higher value of H2 absorption enthalpy change (|ΔHabs|) have low equilibrium pressure at a fixed reaction temperature (thermodynamically favourable). The value of ΔHabs of Ti (−142 kJ mol−1 H28) is higher than that of other practical materials for H2 storage (−30 to −70 kJ mol−1 H29,10). When H2 equilibrium pressure was calculated using thermodynamic parameters reported in the previous literature,8 it was found that Ti absorbs H2 below the high vacuum level (10−17 Pa) at room temperature. For the utilization of Ti as a H2 capture material, the reaction kinetics is an important factor. Various research studies on the kinetics of H2 absorbing reactions have been conducted. In particular, the effects of the surface oxide layer on the H2 absorption properties of Ti have been investigated so far.11–13 However, selective H2 absorption by Ti around room temperature has not been achieved yet. In our previous report, the H2 absorption properties of Ti were drastically improved by surface modification with acetone.13 On the other hand, the reactivity to H2 was lost after 7 days due to surface oxidation, however the surface modification mechanism is still unclear. Here, it was indicated in a previous report that the TiC precursor formed on the Ti surface plays an important role in the reaction selectivity to H2.

In this work, the effects of graphite and several kinds of organic solvents as additives for surface modification of Ti are investigated to understand the effects of carbon on the H2 absorption properties of Ti. Fig. 1 shows the TG curves of Ti ball-milled with graphite (TiGraphite), acetone (TiAcetone), and xylene (TiXylene) obtained under 0.1 MPa H2 flow at around 40 °C. cis-1,2-Dimethylcyclohexane (Ticis-1,2-Dimethylcyclohexane) and 2-methyl-4-pentanone (Ti2-Methyl-4-pentanone) do not show significant surface modification effects (see Fig. S2, ESI). These measurements were carried out every 24 h after sample preparation in order to investigate the degradation of H2 absorption properties with time in a glovebox. The weight gain during TG measurement can be ascribed to H2 absorption by Ti as the phase change from Ti to TiH2 can be confirmed by XRD measurements (Fig. 2 and Fig. S3, ESI). Weight gain of approximately 3 wt% occurred in all the samples even around 40 °C during TG measurement performed immediately after sample preparation. The results indicate that all the as-prepared samples are active in the hydrogenation reaction. TiGraphite, TiAcetone, and TiXylene can absorb H2 after several days, indicating that these additives modify the Ti surface. It is noteworthy that the reactivity of TiGraphite was retained for at least 7 days, suggesting that graphite has superior effects on Ti for selective H2 absorption. Furthermore, even though the H2 absorption temperature increased after TiGraphite was exposed to air for 1 day, it shows a lower reaction temperature compared to Ti without any surface modification, and high stability without burning in air (Fig. S4, ESI). The degradation of H2 absorption properties with time in a glovebox for TiAcetone and TiXylene were different. The H2 absorption capacity of TiAcetone decreased gradually, while that of TiXylene decreased suddenly. These results indicate that the inactivation mechanism is different in both the samples. Several mechanisms for the reaction selectivity have to be considered. One possibility is that the modified surface functions as a metal membrane. Palladium (Pd) thin film, which is known as a metal membrane, dissociates H2 molecules on its surface, and then the H atoms diffuse into the Pd film.14 The metal membrane can separate H2 with high selectivity from mixed gas via a solid-diffusion mechanism. If a functional surface with the metal membrane properties was formed on the Ti sample, the reactivity of Ti could be maintained without inactivation because oxygen and water will not reach Ti. However, the Ti samples were inactivated even in the case of TiGraphite. As a second possibility, the modified surface functions as a porous inorganic membrane such as silica and gases can permeate through the membrane with different permeation rates.15,16 In this case, oxygen also reaches the Ti bulk along with H2, indicating that Ti would be gradually oxidized. The inactivation behavior of TiAcetone is similar to this phenomenon. Third one is that the modified surface functions as an adsorbent for oxygen and water similar to activated carbon.17 When the adsorption sites are fully occupied, the gases reach the Ti surface, leading to its inactivation at once. The inactivation tendency of TiXylene is similar to the adsorbent model. TiGraphite shows high reaction selectivity with the highest H2 capacity of 3.4 wt% in this work. This value is close to the one estimated for the mixture of Ti and graphite without any reaction between both materials (3.7 wt%), indicating that excess graphite remains in the sample. Wei Ye et al. reported on the H2 absorption properties of a Ti/graphene composite with Kubas-type interactions between 100 and 300 °C.18 Although the theoretical H2 capacity of pure Ti is 4.2 wt%, the Ti/graphene composite (20 wt% graphene) can also absorb 4.3 wt% H2 due to the Kubas-type interaction at 300 °C. On the other hand, the H2 capacity of TiGraphite is close to that of the mixture mentioned above, indicating that TiGraphite does not show the Kubas-type interaction.

image file: d0cc03023f-f1.tif
Fig. 1 TG results of TiGraphite, TiAcetone, and TiXylene, performed by apparatus under 0.1 MPa H2 at around 40 °C. TG measurements were performed immediately after sample preparation (immediately), and after several days in a glovebox.

image file: d0cc03023f-f2.tif
Fig. 2 XRD results of TiGraphite, TiAcetone, and TiXylene before and after TG measurements. TG measurements were performed under a H2 and Ar atmosphere up to 400 and 650 °C, respectively.

To indirectly estimate the amount of carbon on the modified surface, TG–DTA measurement was carried out up to 650 °C under a 0.1 MPa Ar flow. After this measurement, TiC was observed by XRD measurement in all the samples (Fig. 2 and Fig. S3, ESI). The TiC/Ti ratio was estimated using the (101) plane of Ti and (200) plane of TiC by XRD measurement, as shown in Fig. 2, without considering the mass absorption coefficient as shown in Table 1. If it is assumed that carbon on the modified surface was changed to TiC by heating, the amount of carbon on the TiGraphite surface is much larger than that on other samples. In addition, the amount of TiC formed in Ticis-1,2-Dimethylcyclohexane is quite small (Table 1), indicating that cis-1,2-dimethylcyclohexane does not show any surface modification effects. Based on these results it is considered that the TiC precursor would affect the H2 absorption properties. However, the H2 absorption properties of TiAcetone, TiXylene, and Ti4-Methyl-2-pentanone are clearly different although the TiC/Ti ratio of each sample is close. Therefore, it is indicated that the amount of TiC precursor does not directly affect the H2 absorption properties.

Table 1 TiC/Ti ratio of each sample heated up to 650 °C under an Ar flow. Ti[thin space (1/6-em)]:[thin space (1/6-em)]C (carbon) is the initial molar ratio of ball-milling. The moles of C are estimated by 20 wt% solvents. The peak areas of Ti and TiC are calculated using the (101) plane for Ti and (200) plane of TiC without considering the mass absorption coefficient, respectively
Sample name Ti[thin space (1/6-em)]:[thin space (1/6-em)]C Area ratio of TiC/Ti
TiGraphite 1[thin space (1/6-em)]:[thin space (1/6-em)]0.5 0.82
TiAcetone 1[thin space (1/6-em)]:[thin space (1/6-em)]0.5 0.06
TiXylene 1[thin space (1/6-em)]:[thin space (1/6-em)]0.7 0.07
Ticis-1,2-Dimethylcyclohexane 1[thin space (1/6-em)]:[thin space (1/6-em)]0.7 0.03
Ti4-Methyl-2-pentanone 1[thin space (1/6-em)]:[thin space (1/6-em)]0.6 0.08

To characterize the chemical state of the TiC precursor, Raman scattering spectrometry was carried out immediately after sample preparation and after several days to investigate the degradation with time (Fig. 3 and Fig. S5, ESI). Here, noise peaks from the apparatus at 1300 and 1550 cm−1 are observed. Two characteristic peaks were clearly observed at 1300 to 1400 cm−1 and 1500 to 1675 cm−1 for TiGraphite. These peaks are assigned to the D and G bands, respectively.19,20 The peaks corresponding to the sp3-carbon and defects of hexatomic rings would appear in the D band area, while those in the G band correspond to local sp2-bonded carbon. TiAcetone also shows two broad peaks at the same positions. However, it is surprising that the D band and G band appeared in the spectrum of TiAcetone because acetone does not have hexatomic rings. From this result, it is believed that acetone was reformed upon reaction with Ti, suggesting that carbon was dissolved on the Ti surface. Namely, TiCx was partially generated on the TiAcetone surface as a TiC precursor. Titanium carbohydride (TiCxHy) is one of the possible precursors, and is observed by high resolution TEM (HRTEM) (see Fig. S6, ESI). The peak intensity of the Raman spectra for TiAcetone decreased day-by-day. This result indicates that TiCx is in the metastable state, and the surface is degraded with time. The degradation properties are consistent with the H2 absorption properties. Thus, it was clarified that TiCx (or TiCxHy) is important for the reaction selectivity. Although the Raman spectrum of TiXylene is similar to that of TiAcetone, the peak intensity does not decrease with time even after the loss of reactivity with H2. This result suggests that the state of carbon observed in TiXylene surface is different from TiCx observed in TiAcetone. The stable G band peak originated from the molecular structure of xylene with a hexatomic ring. Considering the H2 absorption properties, the surface of TiXylene has adsorbent-like properties. If a xylene-based product with a large surface area similar to graphite is formed, the adsorbent effects can be understood. The spectrum of TiXylene is similar to that of amorphous carbons with a mixture of sp2 and sp3 hybrid orbitals.19 Therefore, it is possible that the amorphous carbons formed on the TiXylene surface acts as an adsorbent for water and oxygen. Thus, it is clear that the mechanism of selectivity of the reaction is different for TiAcetone and TiXylene. However, the details of the TiC precursor such as composition and structure are not identified. Thus, further research on the reaction between Ti and organic solvents are required to understand the properties of the TiC precursor for future works. The Raman spectrum of TiGraphite shows no change with time. Although there is a possibility that the Raman spectrum of TiGraphite includes residual graphite, the spectrum is similar to that of turbostratic carbon.21 This result indicates that graphite was partially broken during ball-milling. There is a report on H2 separation using turbostratic carbon.21 In the case of gas separation using turbostratic carbon, the gaseous molecules permeate into carbon through its pores or cracks. The permeability of molecules into carbon depends on their kinetic diameter. Therefore, the permeation of oxygen is inhibited due to its larger size than H2. In addition, A. Wollbrink et al. reported that the permeation of water is prevented due to hydrophobic characteristics of carbon.21 Therefore, it is indicated that turbostratic carbon can prevent the permeation of water as well as oxygen. Although it is difficult to know the existence of the TiCx layer in TiGraphite, graphite exhibits superior surface modification effects on Ti to improve its H2 absorption properties. Analysis of the local area of TiGraphite is necessary to characterize the surface state because the Raman spectrum is mainly contributed by residual graphite.

image file: d0cc03023f-f3.tif
Fig. 3 Raman spectra of TiGraphite, TiAcetone, and TiXylene. The measurements were performed immediately after sample preparation and after several days.

To further characterize TiGraphite, which shows the highest performance, TEM observation was carried out as shown in Fig. 4. Fig. 4a and b show the high-angle annular dark field scanning TEM (HAADF-STEM) image of TiGraphite and the energy-dispersive X-ray spectrometry (EDS) mapping of Ti, C, and O elements. In the HAADF-STEM image, electron-rich areas are observed as the bright part. Because the Ti element was detected by EDS mapping, the bright part of the HAADF-STEM image is mainly corresponding to Ti. EDS mapping of C shows carbon surround the Ti particle. The results of X-ray photoelectron spectrometry showed that Ti was covered with carbon (see Fig. S7, ESI). It was also confirmed by HRTEM observation that turbostratic graphite, which is shown in the inserted figure, formed on the Ti surface (Fig. 4c). This result is consistent with the results of Raman spectra. Energy-loss spectroscopy (EELS) line analyses are performed across the boundary between Ti and C, where measurement areas are expressed as squares, as shown in Fig. 4d. The peaks observed at 285 eV and from 287 to 310 eV (Fig. 4e) were assigned to the π and σ bonding of carbon, respectively.22 These peak positions and shapes are similar to those of amorphous carbon. Therefore, TiGraphite would have adsorption effect on oxygen similar to TiXylene and hydrophobic effects on water similar to turbostratic carbon. In the spectra of Ti L2,3-edge in the Ti bulk (areas 1–4), two peaks were observed at 457 and 463 eV. These peaks were attributed to metallic Ti.23 The peaks of Ti L2,3-edge obtained at the interface between Ti and graphite is shifted to ∼1 eV higher energy than metallic Ti (Fig. 4f). This spectrum is similar to that of TiCx reported before.24 It has been reported that the peaks are gradually shifted to 1.4 eV higher energy with increasing N atoms in TiNx (x = 0 to 1).23 The presence of TiCx is also confirmed by XPS measurements (see Fig. S7, ESI). These results indicate that TiCx is formed at the interface between Ti and graphite. Therefore, it is considered that TiGraphite should have inorganic porous membrane effects due to the TiCx layer, whose function is similar to that of TiAcetone. Thus, the inactivation of TiGraphite was prevented by the mixture of turbostratic and amorphous carbons with “hydrophobic adsorption effects” and the TiCx layer with inorganic porous membrane effects.

image file: d0cc03023f-f4.tif
Fig. 4 TEM observation results of TiGraphite. (a) HAADF-STEM image and (b) EDS mapping of Ti, C, O elements. (c) HRTEM image of the interface between Ti and graphite. Inserted figure is the high magnification image of the graphite area. (d) STEM–EELS mapping of Ti, and spectra of the (e) C K-edge and (f) Ti L2,3-edge at each area.

In this work, it is revealed that graphite is the most effective additive for surface modification of Ti to improve its H2 absorption properties. TiGraphite shows high reaction selectivity to H2. These properties are originated from the mixed turbostratic and amorphous carbons, and the TiCx layer. The Ti surface was also modified with organic solvents. The mechanism of reaction selectivity depends on the properties of each solvent. Finally, the effective factors for selective H2 absorption reaction of Ti were presented in order to utilize Ti as a H2 capture material.

This work was supported by the Satake Fund on Hiroshima University. Part of this work was conducted at the joint-use facilities in Hokkaido University, supported by the “Nanotechnology Platform” and the “Project for promoting public utilization of advanced research infrastructure (Program for supporting introduction of the new sharing system) Grant Number JPMXS0420100519”. The authors are thankful to Mr Keita Suzuki for his help in XPS measurement and Mr Ryo Ota for his help in TEM observation.

Conflicts of interest

There are no conflicts to declare.

Notes and references

  1. L. Schlapbach and A. Züttel, Nature, 2001, 414, 353–358 CrossRef CAS PubMed .
  2. Y. Oumellal, A. Rougier, G. A. Nazri, J. M. Tarascon and L. Aymard, Nat. Mater., 2008, 7, 916–921 CrossRef CAS PubMed .
  3. K. Kawahito, L. Zeng, T. Ichikawa, H. Miyaoka and Y. Kojima, Mater. Trans., 2016, 57, 755–757 CrossRef CAS .
  4. V. V. Solovey, A. I. Ivanovsky, V. I. Kolosov and Y. F. Shmal'ko, J. Alloys Compd., 1995, 231, 903–906 CrossRef CAS .
  5. M. V. Lototskyy, V. A. Yartys, B. G. Pollet and R. C. Bowman, Int. J. Hydrogen Energy, 2014, 39, 5818–5851 CrossRef CAS .
  6. C. Corgnale, B. Hardy, T. Motyka, R. Zidan, J. Teprovich and B. Peters, Renewable Sustainable Energy Rev., 2014, 38, 821–833 CrossRef CAS .
  7. M. Paskevicius, D. A. Sheppard, K. Williamson and C. E. Buckley, Energy, 2016, 88, 469–477 CrossRef .
  8. W.-E. Wang, J. Alloys Compd., 1996, 238, 6–12 CrossRef CAS .
  9. H. H. Van Mal, K. H. J. Buschow and A. R. Miedema, J. Less-Common Met., 1974, 35, 65–76 CrossRef CAS .
  10. T. Kimura, H. Miyaoka, T. Ichikawa and Y. Kojima, Int. J. Hydrogen Energy, 2013, 38, 13728–13733 CrossRef CAS .
  11. Y. Hirooka, M. Miyake and T. Sano, J. Nucl. Mater., 1981, 96, 227–232 CrossRef CAS .
  12. E. Hadjixenophontos, L. Michalek, M. Roussel, M. Hirscher and G. Schmitz, Appl. Surf. Sci., 2018, 441, 324–330 CrossRef CAS .
  13. K. Shinzato, S. Hamamoto, H. Miyaoka and T. Ichikawa, J. Phys. Chem. C, 2019, 123, 19269–19274 CrossRef CAS .
  14. S. Uemiya, N. Sato, H. Ando, Y. Kude, T. Matsuda and E. Kikuchi, J. Membr. Sci., 1991, 56, 303–313 CrossRef CAS .
  15. R. M. de Vos, W. F. Maier and H. Verweij, J. Membr. Sci., 1999, 158, 277–288 CrossRef CAS .
  16. H. R. Lee, M. Kanezashi, Y. Shimomura, T. Yoshioka and T. Tsuru, AIChE J., 2011, 57, 2755–2765 CrossRef CAS .
  17. S.-I. Yang, D.-Y. Choi, S.-C. Jang, S.-H. Kim and D.-K. Choi, Adsorption, 2008, 14, 583–590 CrossRef CAS .
  18. W. Ye, Q. Chi, H. Zhou and P. Gao, Int. J. Hydrogen Energy, 2018, 43, 19164–19173 CrossRef CAS .
  19. L. Zhang and R. V. Koka, Mater. Chem. Phys., 1998, 57, 23–32 CrossRef CAS .
  20. M. Mermoux, S. Chang, H. A. Girard and J.-C. Arnault, Diamond Relat. Mater., 2018, 87, 248–260 CrossRef CAS .
  21. A. Wollbrink, K. Volgmann, J. Koch, K. Kanthasamy, C. Tegenkamp, Y. Li, H. Richter, S. Kämnitz, F. Steinbach, A. Feldhoff and J. Caro, Carbon, 2016, 106, 93–105 CrossRef CAS .
  22. T. Jin, M. Ma, B. Li, Y. Gao, Q. Zhao, Z. Zhao, J. Chen and Y. Tian, Carbon, 2020, 161, 1–6 CrossRef CAS .
  23. I. Abdallah, C. Dupressoire, L. Laffont, D. Monceau and A. V. Put, Corros. Sci., 2019, 153, 191–199 CrossRef CAS .
  24. J. Woltersdorf, A. Feldhoff and O. Lichtenberger, Cryst. Res. Technol., 2000, 35, 653–661 CrossRef CAS .


Electronic supplementary information (ESI) available: Detailed experimental procedures, TEM-EDS results of as-received Ti, the results of TG curves, XRD, Raman spectra, TEM observation, and XPS. See DOI: 10.1039/d0cc03023f

This journal is © The Royal Society of Chemistry 2020