Next Article in Journal
Eugenol Polysiloxane-Polycarbonate/Graphene Nanocomposite: Enhanced in Thermostability and Barrier Property
Previous Article in Journal
Dendrimers as Soft Nanomaterials for Electrochemical Immunosensors
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Pt Cluster Modified h-BN for Gas Sensing and Adsorption of Dissolved Gases in Transformer Oil: A Density Functional Theory Study

1
College of Engineering and Technology, Southwest University, Chongqing 400715, China
2
Key Laboratory of Industrial Internet of Things & Networked Control, Ministry of Education, Chongqing University of Posts and Telecommunications, Chongqing 400065, China
*
Author to whom correspondence should be addressed.
Nanomaterials 2019, 9(12), 1746; https://doi.org/10.3390/nano9121746
Submission received: 14 October 2019 / Revised: 25 November 2019 / Accepted: 6 December 2019 / Published: 8 December 2019

Abstract

:
Hexagonal-Boron nitride nanotubes (h-BN) decorated with transition metals have been widely studied due to their enhanced physicochemical properties. In this paper, Pt cluster-modified h-BN is proposed as a sensitive material for a novel gas sensor for the online malfunction monitoring of oil-immersed transformers. The inner oil is ultimately decomposed to various gases during the long-term use of oil-immersed transformers. Exposure to excessively high temperatures produces the alkanes CH4 and C2H6, whereas different degrees of discharge generate H2 and C2H2. Therefore, the identification of H2, CH4, and C2H2 gas efficiently measures the quality of transformers. Based on the density functional theory, the most stable h-BN doped with 1–4 Pt atoms is employed to simulate its adsorption performance and response behavior to these typical gases. The adsorption energy, charge transfer, total density of states, projected density of states, and orbital theory of these adsorption systems are analyzed and the results show high consistency. The adsorption ability for these decomposition components are ordered as follows: C2H2 > H2 > CH4. Pt cluster-modified h-BN shows good sensitivity to C2H2, H2, with decreasing conductivity in each system, but is insensitive to CH4 due to its weak physical sorption. The conductivity change of Ptn-h-BN is considerably larger upon H2 than that upon C2H2, but is negligible upon CH4. Our calculations suggest that Pt cluster modified h-BN can be employed in transformers to estimate their operation status.

1. Introduction

Oil-immersed transformers are extensively employed in modern power systems. Oil-immersed transformers are recognized as one of the most essential electric equipment because of their outstanding power conversion function and universal application. However, inevitable malfunctions associated with the long-term use of transformers, such as local overheating, partial discharge, and equipment wear out, are likely to cause catastrophic damage to the entire power system. Thus, effective methods for the online monitoring of transformers have been developed by the power sector. Along with different types of faults, various impurity gases, including H2, CH4, C2H6, and C2H2, are generated by transformers [1,2,3]. Exposure to excessive temperature produces the alkanes CH4 and C2H6, whereas different degrees of discharge generate H2 and C2H2. Given that H2, CH4, and C2H2 are the three typical gases in transformer faults, their identification effectively measures the quality of transformers. The condition of the transformer can be speculated by analyzing the concentration of these three typical gases and the variation of conductivity in the oil. Given the good sensitivity and versatility of gas sensors comparing with other testing methods [4], a new type of gas sensor with selectivity and sensitivity can be developed to strengthen the running stability of oil-immersed transformers.
Boron nitride nanotubes (BN), also known as white graphene, are widely employed in various fields. BN exhibits excellent performance, including heat resistance and acid-alkali resistance [5,6]. BN is an insulator with good gas adsorption performance due to its large specific surface area. BN is employed in biological and chemical fields. The mechanism of action between metal-doped BN and ozone, and the adsorption behavior of the drug metformin on BN fullerenes have been studied [7,8]. Hexagonal-BN nanotubes (h-BN) (8,0) are widely applied in gas adsorption and sensing [9,10,11,12]. Nevertheless, intrinsic h-BN exhibits a limited response to some inert gases, such as CH4 and C2H2. Metal cluster modification, such as Pd, Pt, and Ni doping [2,13,14], is commonly employed to strengthen gas sensitive responsiveness. Pt is one of the most widely used modification metals and is considered in this work.
Metal-doped material on gas sensing has developed rapidly, with numerous transition metals considered. We proposed the use of Pt cluster-modified h-BN as a promising sensing material for the detection of dissolved gases in oil-immersed transformers in order to monitor the performance of transformers. Given the activity and catalytic properties of Pt, Pt cluster modification can ameliorate the chemical activity of the h-BN surface so obtaining improved adsorption. This doped structure has rarely been studied. Thus, we analyzed the best adsorption and structural stability of Pt cluster-modified h-BN doped with 1–4 Pt atoms. We then selected the optimal structures to examine the adsorption mechanism of H2, CH4, and C2H2. The mechanisms of how the Pt cluster-modified h-BN reacts with these components must be determined to allow for the preparation of sensors to detect transformer malfunctions.

2. Computational Details

The entire computations in this work were conducted based on the density functional theory (DFT) [15,16]. Spin-polarized calculations were unrestricted. The supercell periodic boundary condition was set as 20 Å × 20 Å × 8.5 Å to prevent interaction between neighboring cells [17,18]. The generalized gradient approximation was utilized for computation by Perdew Burke Ernzerhof for the approximation treatment of exchange-correlation functional, which is widely used in the calculation of extensive materials and their surfaces [19,20,21]. The Brillouin-zone was carried out by the Monkhorst-Pack scheme with the k-point set to 3 × 3 × 1, which presents good approximation for h-BN [22,23]. The double numerical plus polarization was selected as the atomic orbital basis set, whereas the DFT semi-core pseudopotential method was applied considering the relativistic effect of transition elements [24]. The energy tolerance accuracy, maximum force, and displacement were selected as 10−5 Ha, 2 × 10−3 Ha/Å, and 5 × 10−3 Å, respectively [25,26]. For the stationary electronic structure, a precise convergence criterion of 1.0 × 10−6 Ha for self-consistent field tolerance was employed [27].
We optimized the geometry of the H2, C2H2, and CH4 in oil, as well as the majorization treatment of the intrinsic h-BN structure. Pt cluster modification was considered to obtain improved adsorption performance. In this work, all possible Pt cluster-modified h-BNs doped with 1–4 Pt atoms were optimized. The best performance and the most stable structure with different Pt cluster decorations was obtained by calculating the adsorption energy (Ed) of doped fabric, and is defined in Equation (1), combined with the specific bond length and collection shape.
Ed = EPtn/h-BNEh-BNEPtn
where n represents the number of doping atoms and Ed expresses the energy released during binding. To analyze the mechanism and effect of the adsorption process, we calculated the charge transfer (Qt) and adsorption energy (Eads) based on Equations (2) and (3).
Qt = Q1Q2
Eads = Egas/sufEgasEsuf
where Qt indicates the charge transfer amount from the gas molecules to the surface of Pt cluster modified h-BN though Mulliken population analysis while Q1 and Q2 represent the total net charge after and before gas adsorption, respectively. In Equation (3), Eads represents the total energy of the gas molecules before adsorption. A negative Eads indicates that the adsorption process is exothermic and spontaneous. The adsorption mechanism can greatly affect the conductivity and gas concentration of the entire system, which is beneficial to fabricate a gas sensor with selectivity and sensitivity. The electrical conductivity is related to energy gap and the change of energy gap will affect conductivity exponentially [28]. To further obtain the conductivity of the entire system, we determined the energy gap (Eg) of the molecular orbital between the highest occupied molecular orbital (HOMO) and the lowest occupied molecular orbital (LUMO), as defined in Equation (4) [29].
Eg = |ELUMOEHOMO|
Total density of states (TDOS) and projected density of states (PDOS) in the adsorption are analyzed in detail for comprehension of the mechanism of the adsorption process.

3. Results and Discussions

3.1. Optimized Structure of Gas Molecules and Intrinsic h-BN

As shown in Figure 1, the optimized structure of gas molecules and intrinsic h-BN are presented. Both H2 and C2H2 are linear gas molecules, whereas CH4 is a three-dimensional tetrahedral structure, the H–H bond length in the H2 molecule is 0.75 Å, given the weak atomic force and short atomic radius of the H atom. C–H bond lengths are 1.10 and 1.07 Å in CH4 and C2H2 molecules. Different orbital hybridization causes slight difference in the C-H bond lengths. In the CH4 molecule, the H–C–H bond angle is 103.39°. The symmetrical structure and four C-H bonds of the CH4 molecule imparts its chemical stability. An h-BN with a perfect crystal structure is shown in Figure 1b. The mesh-like hexagonal structure is beneficial to the adsorption performance. The B-N bond length of 1.44 Å is appropriate for structural stability, both axial and circumferential distances are around this value. When all aspects are considered, h-BN is a suitable material for gas adsorption.

3.2. Pt Cluster Modified h-BN

Figure 2 shows the different doping structures with 1–4 Pt atoms. The adsorption distance, charge transfer, and adsorption energy are calculated to find the most stable structures with different doping Pt atoms.
Figure 2a shows the most stable structure doped with one Pt atom, which is also the only possible structure at this situation. The atomic arrangement of h-BN exhibits central symmetry and requires the location of the Pt atom to be centered. We considered different situations in the experiments. However, only the bridge site for Pt atom doping can be optimized successfully. The Pt–N bond length is 2.00 Å, which is slightly shorter than the Pt-B bond length (2.19 Å). This phenomenon illustrates the strong force between Pt and N atoms. The adsorption energy (Ed) of −1.98 eV demonstrates fine stability from the perspective of energy.
Two Pt-atom doping is also considered in this work. Different systems have been computed, as presented in Figure 2b1,b2. The two structures are the adjacent and opposite sides of the hexagon where Pt atom doping occurs. The Pt–Pt bond lengths are 2.51 and 2.47 Å, meeting the bond length between heavy metals. The Ed values of the adjacent and opposite sides are −4.83 and −4.84 eV, as shown in Figure 2b1. Dispersed Pt atoms may cause the two- and single-atom-doped structures to be similar.
As shown in Figure 2c1–c5, five possible models were generated. The clustering of the three metal atoms has been widely employed in the field of materials. Thus, the Pt3 cluster is treated emphatically in this work to enhance gas sensitivity response. The surface structure of the intrinsic h-BN is greatly altered after Pt3 modification. The atom tends to protrude outward. Ed values from (c1) to (c5) are −9.02, −8.16, −8.31, −8.30, and −8.30 eV, respectively. A high Ed proves that the structure shown in Figure 2c1 exhibits stronger stability compared with the other structures.
As shown in Figure 2d1–d4, the Pt4 cluster is employed for analysis. Many possible structures are also obtained. We finally selected the most stable structure according to the Ed and geometry structure, as shown in Figure 2d1. The adsorption energy of this structure is −12.59 eV, which is higher than that of any other structure. The three-dimensional feature also imparts stability to this structure.
To gain insight into the diversification in conductivity, we studied the band structure of systems in Figure 2a,b2,c1,d1. This analysis allows for the advanced understanding of the gas adsorption mechanism. The band structure of the systems is drawn as follows:
As shown in Figure 3, the band gaps from Figure 3a–e are 3.73, 1.50, 0.30, 0.0, and 0.38 eV, respectively. The intrinsic h-BN exhibits a huge band gap of 3.73 eV, which explains the difficulty for electrons to jump from the top of the valence band to the bottom of the conduction band. This result is also consistent with h-BN being an insulator. After Pt cluster modification, the band gap in each case is greatly decreased, enhancing electrical conductivity throughout the system. After Pt3 doping, the entire band structure is almost continuous, which may be largely due to the doping position.

3.3. Adsorption of H2, CH4, C2H2 on the Surface of Intrinsic h-BN

The optimization results for the adsorption of H2, CH4, and C2H2 on intrinsic h-BN are presented in Figure 4. The geometric optimization results were calculated as the lowest energy structure according to the DFT, that is, relatively thermodynamic stability, so as to determine the adsorption energy and adsorption distance. According to the final analysis results, intrinsic h-BN exhibits low gas sensitivity response to these typical gases. The H2, C2H2, and CH4 molecular structures remain unchanged during the adsorption process, and the long adsorption distance and weak adsorption energy suggest a weak physical mechanism between gas molecules and the intrinsic h-BN. The specific parameters in the adsorption process are provided in Table 1 to further explain the adsorption performance.
The distance of the gas molecules to h-BN is very far, ranging from 3.23 Å to 3.63 Å, explaining the weak interaction between the molecules. The adsorption energies of H2, CH4, and C2H2 are −0.08, −0.06, and −0.10 eV, respectively. Such small adsorption energy values also correspond to the long adsorption distance. The final simulation data show that the intrinsic h-BN response to these three gases is ordered as follows: C2H2 > H2 > CH4. This order seems to be related to the characteristics of the gas molecules.

3.4. Adsorption of Gas Molecules on Pt, Pt2 Doped h-BN

Figure 5a–f depicts the gas adsorption system on Pt- and the Pt2-doped h-BN. The calculation parameters are shown in Table 2. Based on the optimized adsorption structure, Pt doping can improve the gas sensitivity of the entire system, especially for H2 and C2H2, where strong chemical adsorption process occurs after Pt and Pt2 doping. However, the adsorption capacity for CH4 is insufficient, with merely weak physical and catalytic effects between the molecules. The adsorption capacity based on the simulation results is as follows: C2H2 > H2 > CH4, and the effects in all occasions are strengthened.
Figure 5a–d shows that the H–H bond is broken during the adsorption process. Two H atoms are bonded to the Pt atom at a bond length of 1.55 Å, and the adsorption energy in both cases are 1.94 and 1.97 eV. No difference is found between single- and two-atom doping due to the far distance between the H2 molecule and the second Pt atom. After Pt atom doping, Eads is increased from −0.08 eV to −1.94 eV. The rapid increase of Eads with a large Qt (0.26 e) indicates that the Pt atom can enhance the adsorption activity of the system, thereby producing a strong chemical action with the H2 molecule.
The optimization results for CH4 adsorption are shown in Figure 5b–e. Compared with intrinsic h-BN, the Pt- and Pt2-doped h-BN can considerably improve the adsorption capacity, as proven by the large Eads and short adsorption process distance. The bond lengths of the H atom in the CH4 molecules to the Pt atom are 1.76 and 2.53 Å, and the molecular structure of CH4 remains substantially unchanged. We speculate that the role between the CH4 and Pt is biased toward physical adsorption. The structure in Figure 5e shows a weaker effect, and the stability of the doped structure causes the lower adsorption of diatomic doping compared with that of monoatomic doping.
Figure 5c–e demonstrates the large influence between C2H2 molecule and doped h-BN. In the single Pt atom-doped system, the Pt–B bond is fractured because of the great force, and the structure of C2H2 is changed from a straight line to a distorted planar structure. In the two Pt atom-doped system, the degree of distortion is evident. Eads with magnitudes of 2.60 and 3.4 eV at different situations suggest a complicated process that occurs, and the adsorption effect is improved by a leap.
For improved understanding of the adsorption for manufacturing suitable gas sensors, the total density of states (TDOS), partial density of states (PDOS), and molecular orbitals for different gas adsorption scenarios were analyzed. An accurate sensor applied to the online monitoring of transformers can be implemented by identifying different parameters, including conductivity and Eg.
As shown in Figure 6a–f, the TDOS is changed after gas adsorption, and Figure 6a1–f1 presents the PDOS of corresponding situations. The adsorption of different gases shows varying effects on the TDOS. The conductivity of the entire system can be precisely modified through TDOS analysis. Finally, a sensitive gas sensor can be fabricated by analyzing the resistance value of the system. PDOS was calculated to explore advanced adsorption mechanisms, particularly the mechanism of intermolecular chemical bonding. A combination analysis can provide accurate simulation data applied to gas sensor manufacturing.
In TDOS, both H2 and C2H2 show remarkable changes in TDOS, but the change in CH4 is not evident. In a single-atom-doped system, TDOS is approximately equal to 0 eV between the conduction band and the valence band, reflecting the decline in conductivity. However, the occurrence of this phenomenon is minimal in the two-atom-doped system, because two-Pt-atoms doping increases the conductivity of the system. As shown in Figure 1a, TDOS moves to the right as a whole and then drops slightly at the right side of the Fermi level. This result indicates the reduced number of electronic fillings of the conduction band. Furthermore, the conductive property of the structure is decreased, the reason for which can be analyzed from the distribution of PDOS. Considerable orbital hybridization exists in the 5d orbital of Pt and the 1s orbital of H, which is related to the formation of Pt–H bond in the adsorption structure. Strong chemical action decreases conductivity, but this effect is minimal in the two Pt atoms doping system.
In the TDOS and PDOS of CH4 adsorption, no considerable change is found, as shown in Figure 7b–e. Therefore, the electrical conductivity of the entire system remains unchanged. The 2p orbital of C is mainly distributed between −7.5 and −5 eV. Therefore, the effect on the entire energy band structure is minimal, as demonstrated by the weak physical function and small Eads (0.97 and 0.1 eV). Such feature can be used to study new types of sensors with selectivity.
The same analytical method was used to analyze the adsorption of C2H2. The Pt 5d and C 2p orbitals show simultaneous peaks at the same energy level. The orbital hybridization illustrates the strong interaction between the Pt and C atoms, and the high effect between these atoms also reduces the distribution of TDOS around the Fermi level. Thus, the conductivity of the entire system will also decrease.
Frontier molecular orbital theory was analyzed for different systems, and the effect can be obtained based on the electronic behavior of Pt- and Pt2-doped h-BN in the presence of gas molecules. Determining the features that can be practically modified will be helpful for the exploitation of gas sensors. Based on molecular orbital theory, we calculated the HOMO and LUMO distributions of the adsorption system, as shown in Figure 7. We also calculated Eg to evaluate conductivity changes, as presented in Table 3. HOMO and LUMO are mostly located near the Pt atom doping site, which is associated with the good conductivity and insulation properties of h-BN. After gas adsorption, the LUMO position is drastically changed, whereas the HOMO is slightly altered. As shown in Figure 7a–d, the 1.51 eV Eg of Pt-h-BN reflects good conductivity. When the gas molecules are close to h-BN, Eg is rapidly increased to >2.8 eV. As gas molecules transfer charge to the h-BN system, many LUMOs are on the surface of the gas molecules. In the H2 molecule model, Eg can be remarkably improved, and a strong chemical adsorption process occurs, greatly reducing the conductivity of the entire system. Based on the comparison of different situations, almost no HOMO and LUMO distribution occurs on the CH4 molecules, and weak physical effects cannot largely change the conductivity. The contribution of C2H2 to the conductivity is slightly lesser than that of H2. Thus, the conductivity of the entire system is slightly higher than that of the H2 adsorption system. Figure 7e–h show that the HOMO and LUMO distribution of the gas molecule/Pt2-h-BN system presents a similar situation. CH4 exhibits a minor role, and Eg remains unchanged. In general, the molecular adsorption of H2 shows the largest influence on conductivity, whereas that of CH4 exhibits nearly no effect on conductivity.
In summary, for different adsorption scenarios, conductivity decreases at varying degrees. H2 molecular adsorption exhibits the greatest influence on conductivity, followed by C2H2 adsorption. CH4 molecule adsorption is very weak, minimally contributing to conductivity. The estimated final conductivity is arranged as follows: CH4 adsorption system >C2H2 adsorption system >H2 adsorption system.

3.5. Adsorption of Gas Molecules on Pt3, Pt4 Doped h-BN

The gas adsorption system on Pt3- and Pt4-doped h-BN is depicted in Figure 8a–f. Table 4 shows very similar results between Pt- and Pt2-doped h-BN systems. Doping structures with more modified Pt atoms are more responsive to H2 and C2H2 than to CH4, and the adsorption process with the CH4 molecule is very weak. Drastic changes in the charge transfer amount occur at different situations. Different doping structures were compared, and the results are presented as follows.
In the H2 adsorption system, a strong chemical adsorption process exists between Pt3- or Pt4-doped structure and the H2 molecule, the adsorption energy slightly decreasing from −1.97 eV to −1.67 eV. Different degrees of adsorption process mechanism cause large changes in Qt. In the H2/Pt3-h-BN system, such a large Qt indicates that the H atom acquires numerous electrons. Therefore, great changes in conductivity exist.
In the CH4 adsorption model, the adsorption distance of approximately 2.30 Å and adsorption energy of <1 eV greatly reflect the weak physical effects, which is consistent with previous conclusions. In the system of C2H2/Pt3- and Pt4-doped h-BN, Eads of 2.16 and 2.88 eV are approaching that of the C2H2/Pt- and Pt2-doped h-BN systems. This result illustrates that Eads remains unchanged, and only slight differences in conductivity occurs.
The TDOS and PDOS of gas molecule adsorption on Pt3- and Pt4-doped h-BN are presented in Figure 9. When three Pt atoms are doping, TDOS is decreased rapidly at the Fermi level, and the TDOS to the right of the Fermi level is increased, indicating increased electronic filling conduction band. The band gap of Pt3-h-BN is 0.01 eV, which reflects very good electrical conductivity. Additional impurity bands and chemical action worsens the conductivity. Figure 9b,c shows a spike at −5 eV, which is caused by the hybridization of the C-2p orbit and Pt-5d orbit from the PDOS. In C2H2 adsorption, the C-2p and Pt-5d orbits show strong hybridization at many energy levels. Strong molecular forces exhibit a great effect between the two atoms. Thus, the C2H2 molecule undergoes a great bend.
The overall change of TDOS in the Pt4-h-BN adsorption system is relatively weak, as shown in Figure 9d–f. Only the adsorption of the H2 gas molecule changes the TDOS at the Fermi level, and a low TDOS decreases the conductivity in this situation. As shown in Figure 9e, almost no change occurs in TDOS around the Fermi level, indicating that the CH4 molecule contributes slightly to conductivity. The PDOS from −2.5 eV to 2.5 eV in this case is also a good illustration of this phenomenon. Almost no electronic filling is expected in the Pt atom. Only the Pt doping changes the band structure. Therefore, the conductivity does not change substantially after CH4 adsorption. The results in Figure 9f–f1 are almost identical to the previous analysis of C2H2. Extraordinary orbit hybridizations occur between the C-2p and Pt-5d orbits, but the change in conductivity may be small due to the weak change around the Fermi level.
Based on the molecular orbital theory, the HOMO and LUMO distributions of the adsorbed structures were determined using the same formula. Figure 10 shows an intuitive graphical distribution, and Table 5 lists the arranged data.
As shown in Figure 10, Eg is decreased between HOMO and LUMO, indicating that increased metal atom doping can reduce Eg, thereby increasing the conductivity. In other respects, the contribution of gas molecules to electrical conductivity is not as remarkable as previously observed, but the degree of discrimination is still quite valuable. In general, gas adsorption can decrease the conductivity of the system, and conductivity caused by Pt3 doping is more intense than that by Pt4 doping.
In the Pt3 doping system, the Eg of 0.1 eV reflects good electrical conductivity of the system. When accompanied by the adsorption of different gases, the conductivity is decreased, especially when H2 is close to the doped structure. Eg is changed from 0.1 eV to 1.15 eV, indicating that strong adsorption with chemical bond formation decreases electrical conductivity. The role of CH4 is non-negligible under this model, and the relatively large Eg (0.63 eV) also remarkably changes the conductivity of the system. The same effect is observed in the C2H2 adsorption model. Finally, we analyzed the gas/Pt4-h-BN. As shown in Figure 10e–h, the effect of gas molecules on conductivity is slightly decreased, and no considerable changes occur after gas adsorption.
In summary. The adsorption of different molecules decreases conductivity at varying degrees. The adsorption of H2 and CH4 shows the greatest and weakest influence on conductivity, respectively. The final conductivity is ordered as follows: CH4 adsorption system >C2H2 adsorption system >H2 adsorption system, which is the same as that in Pt- and Pt2-doped h-BN systems.

3.6. Comparison of Charge Transfer, Adsorption Energy, Energy Gap at Different Systems Through Table and Histogram

A comparative analysis of adsorption parameters (Qt, Eads, Eg) are carried out to present more intuitive comparison as shown in Table 6 and Figure 11. After gas molecules adsorption, Qt is ranging from −0.05–0.28 e. It is obvious that the charge transfer amount of CH4 is smaller than H2 and C2H2. By comparing the Qt and Eads of each adsorption systems, chemical effect occurs in in H2, C2H2, while physical effect occurs in CH4 during the adsorption process. The adsorption ability for these decomposition components are ordered as follows: C2H2 > H2 > CH4. As shown in Figure 11c, Eg has different degrees of reduction after different Pt atoms modification, meaning that the conductivity of the entire system increases. When gas molecules are close to the Ptn-h-BN, the rising Eg indicates that the conductivity of the system decrease with different degrees, H2 and C2H2 molecules drastically decrease the conductivity, while CH4 nearly does not affect electronic distribution. The final conductivity is ordered as follows: CH4 adsorption system >C2H2 adsorption system >H2 adsorption system. This work can be applied to develop new gas sensors based on the Pt cluster modified h-BN using online monitoring of the dissolved gases in transformer oil.

4. Conclusions

In this study, the most stable structure of h-BN under various doping strategies is analyzed, and the Pt cluster-modified h-BN is employed to investigate its adsorbing performance on oil-decomposed components. DOS analysis and frontier molecular orbital theory are considered to comprehensively understand the interactions between gas molecules and Pt cluster-modified h-BN. A new gas sensor with selectivity and sensitivity can be developed by analyzing the responsiveness to typical gases. The specific response mechanism and conclusions are as follows:
  • Pt cluster-modified h-BN exhibits good sensitivity to C2H2 and H2 due to chemical adsorption process but is insensitive to CH4, with only weak physical adsorption process between them. The adsorption ability for these decomposition components occurs in the following order: C2H2 > H2 > CH4.
  • In all doping systems, the adsorption of each gas molecule decreases the conductivity of the entire system. H2 and C2H2 molecules can drastically change the conductivity, and the decreased conductivity values for C2H2 and H2 are ordered as follows: H2 > C2H2. CH4 does not affect electronic distribution.
  • In doping by different Pt atoms, the adsorption process mechanisms and adsorption are slightly different, but the good sensitivity to H2 and C2H2 is consistent. On the basis of the large difference in characteristics after adsorption, Pt cluster-modified h-BN is a suitable gas sensor.
Therefore, our calculations suggest that Pt cluster-modified h-BN prepared sensors provide a facile way to practically detect transformer oil-dissolved components and can effectively estimate the operation status of transformers. Pt cluster-modified h-BN is a promising material as a gas sensor and for employment in power systems.

Author Contributions

Y.G. conceived and designed the experiments; T.L. and Y.G. performed the experiments; T.L. and P.Y. analyzed the data; Z.D. contributed analysis tools, supervision; Y.G. and Z.D. contributed funding; T.L. and X.H. proposed the methodology; P.Y. wrote the original draft; X.H. provides guidance for revise.

Funding

This work is supported by the National Natural Science Foundation of China (Grant No. 51907165), Key Laboratory of Industrial Internet of Things & Networked Control (Grant No. 2018FF04), Chongqing Research Program of Basic Research and Frontier Technology (Grant No. cstc2018jcyjAX0068).

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Lu, J.; Zhang, X.; Wu, X.; Dai, Z.; Zhang, J. A Ni-Doped Carbon Nanotube Sensor for Detecting Oil-Dissolved Gases in Transformers. Sensors 2015, 15, 13522–13532. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  2. Zhang, X.; Hao, C.; Zhang, J.; Ju, T. Adsorption characteristic of Pd-4 cluster carbon nanotube towards transformer oil dissolved components: A simulation. Appl. Surf. Sci. 2017, 419, 802–810. [Google Scholar] [CrossRef]
  3. Sun, H.C.; Huang, Y.C.; Huang, C.M. A Review of Dissolved Gas Analysis in Power Transformers. Energy Procedia 2012, 14, 1220–1225. [Google Scholar] [CrossRef] [Green Version]
  4. Lee, J.H. Gas sensors using hierarchical and hollow oxide nanostructures: Overview. Sens. Actuators B Chem. 2009, 140, 319–336. [Google Scholar] [CrossRef]
  5. Jia, J.F.; Wu, H.S.; Jiao, H. The structure and electronic property of BN nanotube. Phys. B Condens. Matter 2006, 381, 90–95. [Google Scholar] [CrossRef]
  6. Bo, Z.; Huang, X.; Wen, G.; Long, X.; Yu, H.; Bai, H. Preparation and Ultraviolet-Visible Luminescence Property of Novel BN Whiskers with a Cap-Stacked Structure. J. Phys. Chem. C 2010, 114, 21165–21172. [Google Scholar]
  7. Farmanzadeh, D.; Askari, N. Theoretical study of ozone adsorption on the surface of Fe, Co and Ni doped boron nitride nanosheets. Appl. Surf. Sci. 2018, 444. [Google Scholar] [CrossRef]
  8. Anota, E.C.; Cocoletzi, H.G. GGA-based analysis of the metformin adsorption on BN nanotubes. Phys. E Low-Dimens. Syst. Nanostruct. 2014, 56, 134–140. [Google Scholar] [CrossRef]
  9. Xuan, Z.; Wei, C.; Zhou, Y.; Sun, W.; Ying, X. DFT simulation on H 2 adsorption over Ni-decorated defective h -BN nanosheets. Appl. Surf. Sci. 2018, 439, 246–253. [Google Scholar]
  10. Min, G.; Lyalin, A.; Taketsugu, T. Catalytic Activity of Au and Au2 on the h-BN Surface: Adsorption and Activation of O2. J. Phys. Chem. C 2012, 116, 9054–9062. [Google Scholar]
  11. Koch, H.P.; Laskowski, R.; Blaha, P.; Schwarz, K. Adsorption of gold atoms on the h-BN/Rh (111) nanomesh. Phys. Rev. B Condens. Matter 2011, 84, 2335–2341. [Google Scholar] [CrossRef] [Green Version]
  12. Gui, Y.; Wang, Y.; Duan, S.; Tang, C.; Zhou, Q.; Xu, L.; Zhang, X. Ab Initio Study of SOF2 and SO2F2 Adsorption on Co-MoS2. ACS Omega 2019, 4, 2517–2522. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  13. Wang, Y.; Gui, Y.; Ji, C.; Tang, C.; Zhou, Q.; Li, J.; Zhang, X. Adsorption of SF6 decomposition components on Pt3-TiO2(1 0 1) surface: A DFT study. Appl. Surf. Sci. 2018, 459, 242–248. [Google Scholar] [CrossRef]
  14. Wang, Q.; Lim, K.H.; Yang, S.W.; Yang, Y.; Chen, Y. Atomic carbon adsorption on Ni nanoclusters: A DFT study. Theor. Chem. Acc. 2011, 128, 17–24. [Google Scholar] [CrossRef]
  15. Russo, T.V.; Martin, R.L.; Hay, P.J. Effective Core Potentials for DFT Calculations. J. Phys. Chem. 1995, 99, 17085–17087. [Google Scholar] [CrossRef]
  16. Hensley, A.J.R.; Ghale, K.; Rieg, C.; Dang, T.; Ye, X. DFT-Based Method for More Accurate Adsorption Energies: An Adaptive Sum of Energies from RPBE and vdW Density Functionals. J. Phys. Chem. C 2017, 121, 4937–4945. [Google Scholar] [CrossRef]
  17. Schultz, P.A. Local electrostatic moments and periodic boundary conditions. Phys. Rev. B 1998, 60, 1551–1554. [Google Scholar] [CrossRef]
  18. Makov, G.; Payne, M.C. Periodic boundary conditions in ab initio calculations. Phys. Rev. B Condens. Matter 1995, 51, 4014. [Google Scholar] [CrossRef] [Green Version]
  19. Li, J.; Gui, Y.; Ji, C.; Tang, C.; Zhou, Q.; Wang, Y.; Zhang, X. Theoretical study of the adsorption of SF6 decomposition components on Ni(1 1 1) surface. Comput. Mater. Sci. 2018, 152, 248–255. [Google Scholar] [CrossRef]
  20. Javid, M.A.; Khan, Z.U.; Mehmood, Z.; Nabi, A.; Hussain, F.; Imran, M.; Nadeem, M.; Anjum, N. Structural, electronic and optical properties of LiNbO3 using GGA-PBE and TB-mBJ functionals: A DFT study. Int. J. Mod. Phys. B 2018, 32, 1850168. [Google Scholar] [CrossRef]
  21. Amaya-Roncancio, S.; Linares, D.H.; Duarte, H.A.; Sapag, K. A DFT Study of Hydrogen Assisted Dissociation of CO by HCO, COH and HCOH Formation on Fe (100). J. Phys. Chem. C 2016, 120. [Google Scholar] [CrossRef]
  22. Esrafili, M.D. BN co-doped graphene monolayers as promising metal-free catalysts for N 2 O reduction: A DFT study. Chem. Phys. Lett. 2018, 705, 44–49. [Google Scholar] [CrossRef]
  23. Froyen, S. Brillouin-zone integration by Fourier quadrature: Special points for superlattice and supercell calculations. Phys. Rev. B Condens. Matter 1989, 39, 3168–3172. [Google Scholar] [CrossRef] [PubMed]
  24. Yang, Q.; Zhong, C. Understanding hydrogen adsorption in metal-organic frameworks with open metal sites: A computational study. J. Phys. Chem. B 2006, 110, 655–658. [Google Scholar] [CrossRef]
  25. Gui, Y.; Chao, T.; Qu, Z.; Xu, L.; Zhao, Z.; Zhang, X. The sensing mechanism of N-doped SWCNTs toward SF 6 decomposition products: A first-principle study. Appl. Surf. Sci. 2018, 440, 846–852. [Google Scholar] [CrossRef]
  26. Huo, C.F.; Ren, J.; Li, Y.W.; Wang, J.G.; Jiao, H.J. CO dissociation on clean and hydrogen precovered Fe(111) surfaces. J. Catal. 2007, 249, 174–184. [Google Scholar] [CrossRef]
  27. Motl, P.M.; Tohline, J.E.; Frank, J. Numerical Methods for the Simulation of Dynamical Mass Transfer in Binaries. Astrophys. J. Suppl 2001, 138, 121. [Google Scholar] [CrossRef] [Green Version]
  28. He, X.; Gui, Y.; Xie, J.; Liu, X.; Wang, Q.; Tang, C. A DFT study of dissolved gas (C2H2, H2, CH4) detection in oil on CuO-modified BNNT. Appl. Surf. Sci. 2020, 500. [Google Scholar] [CrossRef]
  29. Hariharan, P.C.; Pople, J.A. The influence of polarization functions on molecular orbital hydrogenation energies. Theor. Chim. Acta 1973, 28, 213–222. [Google Scholar] [CrossRef]
Figure 1. (a) H2, CH4, C2H2 molecules (b) Intrinsic hexagonal boron nitride nanotubes (h-BN).
Figure 1. (a) H2, CH4, C2H2 molecules (b) Intrinsic hexagonal boron nitride nanotubes (h-BN).
Nanomaterials 09 01746 g001
Figure 2. The structures of Pt cluster modified h-BN. The h-BN is modified by a cluster containing a Pt atom (a), two Pt atom (b1,b2), three Pt atom (c1c5), four Pt atom (d1d4).
Figure 2. The structures of Pt cluster modified h-BN. The h-BN is modified by a cluster containing a Pt atom (a), two Pt atom (b1,b2), three Pt atom (c1c5), four Pt atom (d1d4).
Nanomaterials 09 01746 g002
Figure 3. Energy band diagram of different structures, (a) intrinsic h-BN, (b) Pt-h-BN, (c) Pt2-h-BN, (d) Pt3-h-BN, (e) Pt4-h-BN.
Figure 3. Energy band diagram of different structures, (a) intrinsic h-BN, (b) Pt-h-BN, (c) Pt2-h-BN, (d) Pt3-h-BN, (e) Pt4-h-BN.
Nanomaterials 09 01746 g003
Figure 4. The structures of gas molecular adsorption on intrinsic h-BN. (a) H2, (b) CH4, (c) C2H2.
Figure 4. The structures of gas molecular adsorption on intrinsic h-BN. (a) H2, (b) CH4, (c) C2H2.
Nanomaterials 09 01746 g004
Figure 5. Adsorption of gas molecules on Pt, Pt2 modified h-BN. (a) H2 molecule on Pt modified h-BN, (b) CH4 molecule on Pt modified h-BN, (c) C2H2 molecule on Pt modified h-BN, (d) H2 molecule on Pt2 modified h-BN, (e) CH4 molecule on Pt2 modified h-BN, (f) C2H2 molecule on Pt2 modified h-BN.
Figure 5. Adsorption of gas molecules on Pt, Pt2 modified h-BN. (a) H2 molecule on Pt modified h-BN, (b) CH4 molecule on Pt modified h-BN, (c) C2H2 molecule on Pt modified h-BN, (d) H2 molecule on Pt2 modified h-BN, (e) CH4 molecule on Pt2 modified h-BN, (f) C2H2 molecule on Pt2 modified h-BN.
Nanomaterials 09 01746 g005
Figure 6. Total density of states (TDOS) and partial density of states (PDOS) of adsorption of gas molecules on Pt, Pt2 modified h-BN. (a) TDOS of H2 on Pt modified h-BN, (b) TDOS of CH4 on Pt modified h-BN, (c) TDOS of C2H2 on Pt modified h-BN, (d) TDOS of H2 on Pt2 modified h-BN, (e) TDOS of CH4 on Pt2 modified h-BN, (f) TDOS of C2H2 on Pt2 modified h-BN. (a1) to (f1) are PDOS of the corresponding subgraph.
Figure 6. Total density of states (TDOS) and partial density of states (PDOS) of adsorption of gas molecules on Pt, Pt2 modified h-BN. (a) TDOS of H2 on Pt modified h-BN, (b) TDOS of CH4 on Pt modified h-BN, (c) TDOS of C2H2 on Pt modified h-BN, (d) TDOS of H2 on Pt2 modified h-BN, (e) TDOS of CH4 on Pt2 modified h-BN, (f) TDOS of C2H2 on Pt2 modified h-BN. (a1) to (f1) are PDOS of the corresponding subgraph.
Nanomaterials 09 01746 g006
Figure 7. Molecular orbital before and after gas adsorption. (a) Pt-h-BN system, (b) Pt2-h-BN system.
Figure 7. Molecular orbital before and after gas adsorption. (a) Pt-h-BN system, (b) Pt2-h-BN system.
Nanomaterials 09 01746 g007
Figure 8. Adsorption of gas molecules on Pt3, Pt4 doped h-BN. (a) H2 molecule on Pt3 modified h-BN, (b) CH4 molecule on Pt3 modified h-BN, (c) C2H2 molecule on Pt3 modified h-BN, (d) H2 molecule on Pt4 modified h-BN, (e) CH4 molecule on Pt4 modified h-BN, (f) C2H2 molecule on Pt4 modified h-BN.
Figure 8. Adsorption of gas molecules on Pt3, Pt4 doped h-BN. (a) H2 molecule on Pt3 modified h-BN, (b) CH4 molecule on Pt3 modified h-BN, (c) C2H2 molecule on Pt3 modified h-BN, (d) H2 molecule on Pt4 modified h-BN, (e) CH4 molecule on Pt4 modified h-BN, (f) C2H2 molecule on Pt4 modified h-BN.
Nanomaterials 09 01746 g008
Figure 9. TDOS and PDOS of adsorption of gas molecules on Pt3, Pt4 modified h-BN. (a) TDOS of H2 on Pt3 modified h-BN, (b) TDOS of CH4 on Pt3 modified h-BN, (c) TDOS of C2H2 on Pt3 modified h-BN, (d) TDOS of H2 on Pt4 modified h-BN, (e) TDOS of CH4 on Pt4 modified h-BN, (f) TDOS of C2H2 on Pt4 modified h-BN. (a1) to (f1) are PDOS of the corresponding subgraph.
Figure 9. TDOS and PDOS of adsorption of gas molecules on Pt3, Pt4 modified h-BN. (a) TDOS of H2 on Pt3 modified h-BN, (b) TDOS of CH4 on Pt3 modified h-BN, (c) TDOS of C2H2 on Pt3 modified h-BN, (d) TDOS of H2 on Pt4 modified h-BN, (e) TDOS of CH4 on Pt4 modified h-BN, (f) TDOS of C2H2 on Pt4 modified h-BN. (a1) to (f1) are PDOS of the corresponding subgraph.
Nanomaterials 09 01746 g009
Figure 10. Molecular orbital before and after gas adsorption. (a) Pt3-h-BN system, (b) Pt4-h-BN system.
Figure 10. Molecular orbital before and after gas adsorption. (a) Pt3-h-BN system, (b) Pt4-h-BN system.
Nanomaterials 09 01746 g010
Figure 11. Contrast histogram analysis of adsorption parameters. (a) Charge transfer of all adsorption system, (b) adsorption of all adsorption system, (c) energy gap of all adsorption system.
Figure 11. Contrast histogram analysis of adsorption parameters. (a) Charge transfer of all adsorption system, (b) adsorption of all adsorption system, (c) energy gap of all adsorption system.
Nanomaterials 09 01746 g011
Table 1. Adsorption process parameters of adsorption of H2, CH4, and C2H2 on the surface of intrinsic h-BN.
Table 1. Adsorption process parameters of adsorption of H2, CH4, and C2H2 on the surface of intrinsic h-BN.
SystemStructureAdsorption Distance (d)Adsorption Energy (Eads)
H2 adsorptionFigure 4a3.23 Å−0.08 eV
CH4 adsorptionFigure 4b3.45 Å−0.06 eV
C2H2 adsorptionFigure 4c3.63 Å−0.10 eV
Table 2. Adsorption parameters of gas molecules on Pt, Pt2 doped h-BN: charge transfer (Qt), adsorption distance (d), adsorption energy (Eads).
Table 2. Adsorption parameters of gas molecules on Pt, Pt2 doped h-BN: charge transfer (Qt), adsorption distance (d), adsorption energy (Eads).
SystemStructureAdsorption Distance (6)Adsorption Energy (Eads)Charge Transfer (Qt)
H2/Pt-h-BNFigure 5a1.56 Å−1.94 eV0.26 e
CH4/Pt-h-BNFigure 5b1.76 Å−0.97 eV0.10 e
C2H2/Pt-h-BNFigure 5c2.03 Å−2.60 eV0.18 e
H2/Pt2-h-BNFigure 5d1.54 Å−1.97 eV0.30 e
CH4/Pt2-h-BNFigure 5e2.53 Å−0.10 eV0.06 e
C2H2/Pt2-h-BNFigure 5f1.97 Å−3.40 eV0.14 e
Table 3. Molecular orbitals and energy gaps of individual molecules before and after adsorption.
Table 3. Molecular orbitals and energy gaps of individual molecules before and after adsorption.
Adsorption systemStructureEHOMO(eV)ELUMO(eV)Eg(eV)
Pt-h-BNFigure 7a−5.34−3.851.51
H2/Pt-h-BNFigure 7b−5.97−2.843.31
CH4/Pt-h-BNFigure 7c−5.33−2.482.85
C2H2/Pt-h-BNFigure 7d−5.60−2.583.02
Pt2-h-BNFigure 7e−5.19−4.820.37
H2/Pt2-h-BNFigure 7f−5.16−3.741.42
CH4/Pt2-h-BNFigure 7g−5.16−4.790.37
C2H2/Pt2-h-BNFigure 7h−5.11−4.091.02
Table 4. Adsorption parameters of gas molecules on Pt3, Pt4 doped h-BN: charge transfer (Qt), adsorption distance (d), adsorption energy (Eads).
Table 4. Adsorption parameters of gas molecules on Pt3, Pt4 doped h-BN: charge transfer (Qt), adsorption distance (d), adsorption energy (Eads).
SystemStructureAdsorption Distance (d)Adsorption Energy (Eads)Charge Transfer (Qt)
H2/Pt3-h-BNFigure 8a1.53 Å1.67 eV0.20 e
CH4/Pt3-h-BNFigure 8b2.32 Å0.56 eV0.21 e
C2H2/Pt3-h-BNFigure 8c1.97 Å2.16 eV0.25 e
H2/Pt4-h-BNFigure 8d1.55 Å1.58 eV0.22 e
CH4/Pt4-h-BNFigure 8e2.31 Å0.10 eV0.05 e
C2H2/Pt4-h-BNFigure 8f1.98 Å2.88 eV0.28 e
Table 5. Molecular orbitals and energy gaps of individual molecules before and after adsorption.
Table 5. Molecular orbitals and energy gaps of individual molecules before and after adsorption.
Adsorption SystemStructureEHOMO (eV)ELUMO (eV)Eg (eV)
Pt3-h-BNFigure 10a−4.62−4.520.10
H2/Pt3-h-BNFigure 10b−4.81−3.661.15
CH4/Pt3-h-BNFigure 10c−5.08−4.450.63
C2H2/Pt3-h-BNFigure 10d−4.60−4.030.57
Pt4-h-BNFigure 10e−4.61−4.210.40
H2/Pt4-h-BNFigure 10f−4.81−4.280.53
CH4/Pt4-h-BNFigure 10g−4.57−4.170.40
C2H2/Pt4-h-BNFigure 10h−4.47−3.990.48
Table 6. Adsorption parameters summary of different systems.
Table 6. Adsorption parameters summary of different systems.
Adsorption TypeQt (e)Eads (eV)Eg (eV)
Pt-h-BN//1.51
Pt2-H-BN//0.37
Pt3-h-BN//0.10
Pt4-h-BN//0.40
H2/Pt-h-BN0.261.943.31
H2/Pt2-h-BN0.301.971.42
H2/Pt3-h-BN0.201.671.15
H2/Pt4-h-BN0.221.580.53
CH4/Pt-h-BN0.100.972.85
CH4/Pt2-h-BN0.060.100.37
CH4/Pt3-h-BN0.210.560.63
CH4/Pt4-h-BN0.050.100.40
C2H2/Pt-h-BN0.182.603.02
C2H2/Pt2-h-BN0.143.401.02
C2H2/Pt3-h-BN0.252.160.57
C2H2/Pt4-h-BN0.282.880.48

Share and Cite

MDPI and ACS Style

Gui, Y.; Li, T.; He, X.; Ding, Z.; Yang, P. Pt Cluster Modified h-BN for Gas Sensing and Adsorption of Dissolved Gases in Transformer Oil: A Density Functional Theory Study. Nanomaterials 2019, 9, 1746. https://doi.org/10.3390/nano9121746

AMA Style

Gui Y, Li T, He X, Ding Z, Yang P. Pt Cluster Modified h-BN for Gas Sensing and Adsorption of Dissolved Gases in Transformer Oil: A Density Functional Theory Study. Nanomaterials. 2019; 9(12):1746. https://doi.org/10.3390/nano9121746

Chicago/Turabian Style

Gui, Yingang, Tao Li, Xin He, Zhuyu Ding, and Pingan Yang. 2019. "Pt Cluster Modified h-BN for Gas Sensing and Adsorption of Dissolved Gases in Transformer Oil: A Density Functional Theory Study" Nanomaterials 9, no. 12: 1746. https://doi.org/10.3390/nano9121746

APA Style

Gui, Y., Li, T., He, X., Ding, Z., & Yang, P. (2019). Pt Cluster Modified h-BN for Gas Sensing and Adsorption of Dissolved Gases in Transformer Oil: A Density Functional Theory Study. Nanomaterials, 9(12), 1746. https://doi.org/10.3390/nano9121746

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop