Computational Approach To Reveal the Structural Stability and

Feb 25, 2019 - PDF (4 MB) ...... pdf. ao8b03433_si_001.pdf (5.8 MB) ...... 37. Mahadevi, A. S.; Sastry, G. N. Cation−π Interaction: Its Role and Re...
0 downloads 0 Views 4MB Size
Download PDF
This is an open access article published under an ACS AuthorChoice License, which permits copying and redistribution of the article or any adaptations for non-commercial purposes.

Article Cite This: ACS Omega 2019, 4, 4153−4160

http://pubs.acs.org/journal/acsodf

Computational Approach To Reveal the Structural Stability and Electronic Properties of Lithiated M/CNT (M = Si, Ge) Nanocomposites as Anodes for Lithium-Ion Batteries T. K. Bijoy,†,‡ Karthikeyan J,†,‡,§ and P. Murugan*,†,‡ †

Academy of Scientific and Innovative Research (AcSIR) and ‡Functional Materials Division, CSIRCentral Electrochemical Research Institute, Karaikudi, Tamil Nadu 630003, India

ACS Omega 2019.4:4153-4160. Downloaded from pubs.acs.org by 191.101.54.249 on 02/25/19. For personal use only.

S Supporting Information *

ABSTRACT: This work is motivated to explore the structural stability and electronic and electrochemical properties of nanocomposites of M4Lin (M = Si and Ge)−carbon nanotube (CNT) by employing first-principles density functional theory calculations. By analyzing the structural stability of various M4Lin (n = 0−10) clusters, it is revealed that a tetrahedron-shaped M4Li4 Zintl cluster is found to be highly stable. Our study on the interaction between the lithiated clusters and CNT illustrates that the charge transfer from the former to latter plays a pivotal role in stabilizing these nanocomposites. The structural stability of those nanocomposites arises as a consequence of bonding between lithiated clusters and CNT, which is mediated through the cation−π interaction. The strength of the interaction between them is well reflected in electronic structure calculations by shifting the energy levels with respect to the Fermi energy. Further, the electrochemical properties of these nanocomposites are explored by forming an assembly of the cluster-inserted CNT. The calculated average intercalation voltage of the systems is found to be low (maximum ∼1.0 V for M = Si and 1.05 V for M = Ge), which demonstrates their anodic behavior.

■

INTRODUCTION At present, rechargeable lithium-ion batteries (LIBs) emerge as an irrefutable power source for portable electronic and electrical devices.1,2 Because of its various salient features such as high capacity, better operational voltage, and compact size, the LIB outperforms other conventional energy storage devices.3 Even though LIBs are well commercialized long back, further improvements in their components are still considered as an important task to achieve the fast growing demands in large-scale electrical automobiles.4 Especially, the wellcommercialized anode material is graphite; however, the theoretical specific capacity is limited to 372 mAh/g, as the result of the structural restriction of graphite, which leads to form the LiC6 compound.5,6 In this perspective, other group IV elements, such as, silicon, germanium, and tin are well attracted among the researchers because these materials possess higher theoretical capacity than that of graphite.7 However, it is well known that all these materials undergo an enormous volume expansion when they are alloyed with Li ions. This leads to affect the battery performance by losing the contact between electrode and current collector.8−10 In addition, it is observed that the interparticle electronic conductivity of the electrode material is significantly reduced during battery cycling as the consequence of volume expansion.11 Hence, it is unequivocally considered that the practical implementation of bulk phase Si/ Ge/Sn as such in LIBs is not possible. © 2019 American Chemical Society

The most common strategies adopted to nullify the effect of volume change in the anode of the LIB are particle size reduction of electrode materials and/or by making carbon composite of these materials.12−17 Because of the superior electronic and mechanical properties of the carbon nanostructures, the second approach is widely considered. Among various carbon nanostructures, carbon nanotube (CNT) is the one which is considered as an excellent option for making nanocomposite anode materials for LIBs.18,19 The porosity and one-dimensional structure of CNT significantly help to reduce the volume expansion in such composites during battery cycling.20,21 Toward this direction, in the recent past, Yu et al.22 synthesized Si nanoparticle-filled CNT, and this nanocomposite shows an excellent specific capacity of 1750 mA h/ g. In addition, the volume expansion of a Si nanoparticle during alloying with lithium is effectively shielded by the CNT matrix. In a similar fashion, few reports23−25 suggest that the hybrid material consisting of Ge/Sn and carbon nanostructures can be used as anode of LIB. Even though lithium diffusion in Ge is relatively faster as compared with Si, this nanocomposite is not well established, owing to its high cost. To the best of our knowledge, the atomistic understanding of Si/Ge nanoparticle incorporated CNT system is not well studied so far. Received: December 7, 2018 Accepted: February 8, 2019 Published: February 25, 2019 4153

DOI: 10.1021/acsomega.8b03433 ACS Omega 2019, 4, 4153−4160

ACS Omega

Article

structure of Sn4Lin clusters in few cases are fairly different (refer Figure 1). Especially for Sn4Lin clusters, it has been realized that the tetrahedron-shaped geometry is converted into a butterfly-like one (n > 4), by breaking one of its Sn−Sn bonds. This is consistent with our previous study.26 In contrast, for both Si4Li6 and Ge4Li6 clusters, the tetrahedral unit exists up to n = 6. The structure of Si4Li6/Ge4Li6 cluster consists of interconnected Li4 and Si4/Ge4 tetrahedral units, and remaining two Li atoms occupy on two faces of the Si4/ Ge4 cluster as shown in Figure 1. A similar structural difference between Sn4Li10 and Si4Li10/Ge4Li10 clusters is also noticed. Further, to understand the structural stability, we calculated the binding energy (BE) per atom and the highest-occupied molecular orbital−lowest-unoccupied molecular orbital (HOMO−LUMO) energy gap of all clusters. The BE is deduced from

Considering this, here we explored the structural, electronic, and electrochemical properties of M−Li−CNT (M = Si and Ge) composites by employing first-principles density functional theory (DFT) calculations. Our results show that lithiated Si4 and Ge4 clusters are strongly bonded with CNT, through cation−π interaction as observed in our earlier work on Sn−Li−CNT nanocomposites.26 The strength of adsorption between them is well reflected in the electronic structures of those nanocomposites. The calculated average intercalation voltage (AIV) of the nanocomposites manifests that they are suitable for the anode of LIBs.

■

RESULTS AND DISCUSSION For modeling the nanocomposites M4Lin−CNT (M = Si and Ge), we initially considered the cluster having four atoms. The calculated structural and electronic properties of these clusters are compared with that of Sn4Lin clusters, which is studied in our previous work.26 From our preliminary calculations, it is clear that in the ground state, both Si4 and Ge4 clusters prefer to be in rhombus geometry similar to the Sn4 cluster. Our results are well consistent with previous reports.27,28 The deduced electronic density of states (DOS) of these clusters (shown in Figure S1, the Supporting Information) suggest that the Si4 cluster could be the more stable one as the bonding 3p states, next lower to highest-occupied molecular orbital (HOMO) level, is nondegenerated. On the other hand, for Ge4 and Sn4 clusters, respective 4p and 5p states are almost degenerated and they are predominantly located in higher energy region as compared to Si cluster. Therefore, the stability of these two clusters is expected to be lesser than that of Si4 atomic cluster (refer Figure 2). Further, Li atoms are introduced on these stable rhombusshaped clusters to form M4Lin clusters (n = 1−10). Here, for deriving the stable structures of M4Lin clusters, we considered the models of various atomic clusters reported in the previous studies.29−33 The optimized geometries of the stable lithiated clusters are presented in Figures 1 and S2 (Supporting

BE =

[4E(M) + nE(Li)] − E(M4Li n) 4+n

(1)

where E(M) and E(Li) are the atomic energies of Si/Ge/Sn and Li atoms, respectively. E(M4Lin) corresponds to total energies of the lithiated cluster. As expected, the HOMO− LUMO gap of M4Lin clusters with an odd n value is found to be low, owing to the presence of an unpaired electron in the HOMO level; hence, they are magnetic with the spin moment of 1 μB. Here, we also verified the type of magnetism by calculating the energies of the clusters in the antiferromagnetic (AFM) configurations. The results show that the clusters with an odd n value prefer to be in ferromagnetic (FM) configuration rather than respective AFM configuration (refer Figure S9, the Supporting Information). On the other hand, all the clusters having even numbered Li atoms are nonmagnetic and they possess larger gap as compared to nearby odd numbered clusters. Thus, it can be seen in Figure 2a that the HOMO−LUMO gap of the cluster oscillates with respect to n value. It is also observed that Si4Li4, Ge4Li4, and Sn4Li4 clusters have the large HOMO−LUMO gap and high

Figure 1. Optimized structures of stable Si4Lin, Ge4Lin, and Sn4Lin clusters are given three rows (a−d,e−h,i−l), respectively. The n value of cluster varies from 4, 6, 8, and 10 (in column wise).

Information). In addition, the other local minima structures of various lithiated clusters are also provided in Figures S3−S8 (Supporting Information). It is observed that during lithiation, the rhombus-shaped clusters are slowly bent into a closed shape, and they attain perfect tetrahedral geometry when n = 4 (refer Figure 1). Both Si4Lin and Ge4Lin clusters possess almost similar geometry in all the cases. On the other hand, the

Figure 2. (a) BE and (b) HOMO−LUMO gap of various M4Lin clusters (M = Si, Ge and Sn), n varies from 0 to 10. 4154

DOI: 10.1021/acsomega.8b03433 ACS Omega 2019, 4, 4153−4160

ACS Omega

Article

Table 1. Average M−M, M−Li, Li−Li, C−Li Distances for Various M4Lin Cluster-Inserted CNTsa M4Lin/CNT

M−M (Å)

Si4/[8,8] Si4Li2/[8,8] Si4Li4/[8,8] Si4Li6/[8,8] Si4Li8/[8,8] Si4Li10/[8,8] Ge4/[8,8] Ge4Li2/[8,8] Ge4Li4/[8,8] Ge4Li6/[8,8] Ge4Li8/[8,8] Ge4Li10/[8,8] Sn4/[8,8] Sn4Li2/[8,8] Sn4Li4/[8,8] Sn4Li6/[8,8] Sn4Li8/[9,9] Sn4Li10/[9,9]

2.34(2.34) 2.41(2.41) 2.47(2.50) 2.44(2.47) 2.59(2.54) 2.42(2.51) 2.50(2.51) 2.50(2.47) 2.66(2.67) 2.62(2.66) 2.62(2.63) 2.69(2.74) 3.04(3.01) 2.87(2.89) 3.04(3.06) 3.02(3.09) 3.05(3.02) 2.87(3.14)

M−Li (Å) 2.77(2.44) 2.61(2.54) 2.60(2.55) 2.59(2.54) 2.59(2.60) 2.82(2.49) 2.63(2.59) 2.64(2.60) 2.65(2.57) 2.61(2.58) 2.93(2.92) 2.96(2.81) 2.95(2.78) 2.90(2.79) 2.95(2.76)

NLi−Li

Li−Li (Å)

C−Li (Å)

−(2.99) 3.08(3.04) 2.92(2.91)

2.51 2.92 2.47 2.53 2.50

6(0) 8(5) 9(6)

−(2.99) 3.11(2.89) 3.01(2.97)

2.46 2.96 2.47 2.47 2.48

6(0) 8(5) 9(6)

3.25(3.08) 3.12(3.08) 3.11(2.97)

2.42 2.58 2.50 2.50 2.45

1(0) 3(0) 9(5)

δQB (e−) 0 0.43 0.51 2.26 2.34 2.41 0 0.42 0.64 2.23 2.35 2.43 0 0.40 1.00 2.32 2.36 2.62

a For pristine clusters, distances are given in parenthesis. The number of Li−Li bonds (NLi−Li) before and (after) inserting into CNT is also provided. The charge transfer is calculated from the Bader charge analysis and is provided. Size of CNT is also given in square parenthesis.

Ead values are also tabulated in Table 1. From the obtained Ead values, we understood that the [8,8] CNT is most preferred for inserting the Si4Lin/Ge4Lin clusters with n = 0−8. However, for inserting Si4Li10 or Ge4Li10 clusters, a [9,9] CNT is found to be optimum. The size of Sn4Lin clusters are comparatively larger than that of both Si4Lin and Ge4Lin. Thus, for inserting Sn4Lin clusters, [8,8] CNT is preferred only up to n = 6, beyond this, the [9,9] CNT shows maximum Ead. Meanwhile the calculated Ead values indicate that the interaction between all three bare M4 clusters and CNT is weak; the corresponding Ead values are 0.15, 0.12, and 0.03 eV for M = Si, Ge, and Sn, respectively. In fact Figure 3 implies that the adsorption becomes gradually

structural stability in their respective cluster family, as they are stable Zintl clusters.34,35 Figure 2b illustrates the variation of BE of M4Lin clusters with respect to n. The BE of these clusters varies in the order Si4Lin > Ge4Lin > Sn4Lin. This suggests that structurally, the lithiated Si clusters (including pristine) are more stable as compared to Ge/Sn clusters. However, the observed trend shows that in each series BE is reduced when the pristine M4 cluster is lithiated. It is worthy to mention that even though BE reduces gradually as the lithiation proceed, at n = 4, we could observe a rise in BE value as shown in Figure 2b, which confirms the high structural stability of the M4Li4 cluster as compared to any other lithiated cluster we studied. These calculations show that for all M4Lin clusters, the HOMO− LUMO gap and BE are well correlated. Yet, in few cases, the atomic structures of both Si4Lin and Ge4Lin clusters are slightly different from that of Sn4Lin clusters owing to their large BE as compared to Sn4Lin clusters. Next, to understand the interaction between these clusters and CNT, we preferred various stable clusters for depositing them into CNTs. Here, different sized arm chair [m,m] CNTs with m = 7−10 are taken for our calculations. The arm chair CNT is chosen as it is superior to both zigzag and chiral CNTs in terms of mechanical stability and electronic conductivity. The clusters are inserted into the CNT, so that the system will mimic the M−Li−CNT nanocomposite as observed in various experiments.23,24 The optimum size of the CNT for inserting these clusters into CNT is decided by calculating the adsorption energy (Ead), from. Ead = E(M4Li n) + E(CNT) − E(M4Li n + CNT)

Figure 3. Adsorption energy for various M4Lin clusters in their optimum sized arm chair CNT.

stronger once the pristine M4 clusters are lithiated. Moreover, it is also revealed that beyond n = 4, the Ead values are remarkably enhanced, suggesting that the clusters are firmly adsorbed in the inner side of the CNT after the formation of stable Zintl cluster. Because these clusters are strongly adsorbed, beyond n = 4, it is obvious that their geometry is affected. From the relaxed structures of the composite, we identified that the lithium dimer in clusters is becoming weaker when they are inserted into the CNT. For instance, the Li4 tetrahedron in the Si4Li6 cluster has deformed and the respective Li atoms started to interact with the C6 ring of the CNT when this cluster is inserted into the CNT (refer Figure S14, the Supporting Information). Moreover, we

(2)

where E(M4Lin), E(CNT), and E(M4Lin + CNT) correspond to total energies of M4Lin cluster, CNT, and cluster inserted CNT, respectively. Note that, for obtaining the stable configuration of cluster inside CNT, we attempted several models in which the clusters are oriented inside the CNT in various fashion as given in Figures S10−S20 (refer the Supporting Information). Among them, the lowest energy configuration is considered for further studies. The calculated 4155

DOI: 10.1021/acsomega.8b03433 ACS Omega 2019, 4, 4153−4160

ACS Omega

Article

Si4Li6 and Ge4Li6 clusters, the Sn4Li6 moves slightly lesser and it remains partially near to the edges as shown in Figure S23 (refer the Supporting Information). To quantify the decrease in energy during the insertion of the cluster, we calculated the relative energy (ΔE) for various cases with respect to the most stable configuration and the same is plotted against the distance (X) of the cluster from the CNT as shown in Figure 4.

counted the number of Li−Li bonds (NLi−Li) present in the clusters before and after inserting into the CNT, and the values are provided in Table 1, which infers that clusters possess lesser number of Li−Li bonds when they are inside CNT. Hence, it is revealed that the C−Li bond is more preferred over the Li−Li bond, when the clusters are embedded inside the CNT. In a similar fashion, the geometry of the Si4Li8 cluster is significantly modified when it is adsorbed inside the CNT as shown in Figure S15 (refer the Supporting Information). Here, the Si4 unit of the Si4Li8 cluster adopts a square like structure by forming one extra Si−Si bond. Likewise, Ge4Li6 and Ge4Li8 clusters also undergo similar structural changes when they are encapsulated in CNTs. Thus, on the basis of the adsorption study of the cluster, we concluded that the lithiated clusters are strongly adsorbed inside the CNT when compared with that of pristine M4 clusters. Besides this, the reduction in the number of Li−Li bonds within the clusters when they are inside CNT indicates that such a cluster/CNT composite may suppress the growth of Li dendrite in anode during the charging process; hence, the electrochemical performance of these nanocomposites will be better.36 In addition, we also calculated the Ead using the DFTD2 method. Here, we found that the dispersion effect leads to significant enhancement in the Ead and it is more pronounced for Sn4Lin clusters, which is evidenced from the comparatively larger Ead than that of Si4Lin and Ge4Lin clusters when they composite with the CNT (refer Figure S21, the Supporting Information). However, here also we confirmed that the clusters are firmly adsorbed only after the formation of stable M4Li4 clusters. Apart from inserting clusters into the CNT, we also carried out calculations for various cases, in which the clusters are deposited on the outer wall of the CNT and the Ead for these configurations is compared with that of cluster inserted in the inner side of CNT cases. Interestingly, here, we noticed that in most of the cases the clusters are strongly interacting with CNT when they are present inside the CNT owing to the formation of more number of C−Li bonds. The converged structures and corresponding Ead for various Si4Lin cluster anchored on CNT’s wall are provided in Figure S22 (refer the Supporting Information). In all these cases, we have exclusively studied the interaction between clusters with CNT when they are incorporated inside the CNT. Hence, it is important to assess the amount of reduction in energy during the insertion of the cluster into the CNT. Toward this direction, we selected the M4Li6 cluster as the representative one for inserting into a piece of a nonperiodic [8,8] CNT having the length of ≈10.45 Å. Note that the edges of the CNT are terminated with hydrogen atoms owing to the charge compensation. Further to determine energetics of cluster insertion, we modeled various configurations in such a way that the cluster which is initially kept a distance of 4.5 Å from the CNT’s edge is gradually moved (0.5 Å in each step) until it reaches the inner side of the nanotube. In this way, the total energies are calculated for various configurations. Our results suggest that the energy of the composite is reduced, when the cluster approaches near to the edges and then suddenly drops to a minimum value when it enters inside the CNT. It is also revealed that during structural relaxation, the cluster will move itself into the CNT, if it is placed at the close proximity of CNT’s edges ( 4, refer Table 1). This charge-transfer analysis elucidates that for lithiated cluster-loaded CNT system, the excess charge from the cluster to CNTs wall occurs through 4156

DOI: 10.1021/acsomega.8b03433 ACS Omega 2019, 4, 4153−4160

ACS Omega

Article

pristine [8,8] CNT. In addition, the localized Si 3p states are presented near to Ef. However, for lithiated clusters, the total DOS is shifted away from the Ef to the valance band region and this increases with respect to the increase in the n value. However, with increasing lithiation, from n = 2−4, we could not find much accountable shift in total DOS. The reason is that as the Si4Li4 cluster is highly stable, the charge transfer from cluster to CNT in those cases is found to be less. Perceptible deviation in DOS away from the Ef is recognized for n > 4, and at n = 10, it is found to be 0.76 eV. The DOS of Ge4Lin/CNT as well as Sn4Lin/CNT composite also shows similar features (refer Figures S24 and S26, the Supporting Information). Beside DOS, the band structure (BS) of the aforesaid systems is plotted for reaffirming the role of charge transfer on their electronic structure and the same is shown in Figure 7. From BS of Si4/[8,8] CNT composite, it is well clear that the insertion of Si4 cluster does not affect the CNT’s bands significantly owing to the weak interaction existing between them. Also, one discrete band arising from the Si4 cluster is presented near to the Ef. The BS of Ge4/[8,8] CNT composite shows similar features (shown in Figure S25, the Supporting Information), while in the case of Sn4/[8,8]CNT, the energy band of cluster is lying above the Ef level (∼0.24 eV) (refer Figure S27, the Supporting Information). However, for Si4Li2/ [8,8] CNT and Si4Li4/[8,8] CNT cases, the BSs show that the discrete bands arising from the Si atoms are occupied as the result of Li to Si charge transfer. In addition, we noted that the linearly dispersed band (cone-like) arising from the CNT is slightly occupied (refer Figure 7c,d). In contrast, for the remaining cases such as Si4Li6/[8,8], Ge4Li6/[8,8], and Sn4Li6/ [8,8], the aforesaid CNT band is completely occupied as the charge transfer from cluster to CNT is significantly higher in these cases. Beyond n = 6, we could not observe significant shift in CNT bands. Therefore, more bands from cluster appeared in the valence band region for n > 6 cases. From the electronic structure calculations, we concluded that for stabilizing M4Lin/CNT composite the charge transfer plays a vital role. Moreover, CNTs easily accept two electrons from the lithiated cluster. This result is in agreement with the Bader charge analysis. Finally, to unveil the electrochemical properties of these composites, the AIV41 is calculated from

Figure 5. (a,b) Top and side views of charge-transfer diagrams for Si4/[8,8] CNT, (c,d) correspond to Si4Li6/[8,8] CNT composite, respectively. Cyan and yellow colored isosurfaces denote excess and depletion of charge densities, respectively.

cation−π interaction which help to adhere the cluster on the CNT’s wall, and thus stabilizes these composites. Thereon, to inspect the electronic properties of these composites, we deduced electronic DOS. The DOS of Si4Lin/ CNT composite is shown in Figure 6. As discussed earlier, even though the interaction between pristine Si4 with CNT wall is weak, we could observe a narrow shift in total DOS of the composite toward the Fermi energy (Ef) as compared to

AIV = [E(Lix 2@M4Li n) − E(Lix1@M4Li n) − (x1 − x 2)E(Li)] − (x 2 − x1)F (4)

where E(Lix1@M4Lin), E(Lix2@M4Lin), E(Li), and F denote total energies of x1 and x2 Li ions intercalated in M4Lin cluster inserted into CNT, chemical potential of Li−metal, and the Faraday constant, respectively. For intercalating Li atoms, a hexagonally stacked assembly of cluster-inserted CNT composite is modeled, and the Li atoms are allowed to intercalate in the voids which are formed between the adjacent CNTs as shown in Figure S28 (refer the Supporting Information). It is important to mention that in the assembly of CNTs, the interaction between the adjacent nanotubes is weak van der Waals (vdW) force. Hence, we have included the dispersion correction throughout these calculations as mentioned in the computational methodology. Therefore, all the energies used for calculating AIV included vdW correction.

Figure 6. Electronic DOS of (a) Si4/[8,8] CNT, (b) Si4Li2/[8,8] CNT, (c) Si4Li4/[8,8] CNT, (d) Si4Li6/[8,8] CNT, (e) Si4Li8/[8,8] CNT, (f) Si4Li10/[9,9] CNT. The shaded cyan region corresponds to total DOS of the composite and black colored states represent the DOS pristine CNT. The value and direction (arrow) of shift in the energy level of composite with respect to bare CNT are given in each case. 4157

DOI: 10.1021/acsomega.8b03433 ACS Omega 2019, 4, 4153−4160

ACS Omega

Article

Figure 7. Electronic BSs of (a) pristine [8,8] CNT, (b) Si4/[8,8] CNT, (c) Si4Li2/[8,8], (d) Si4Li4/[8,8], (e) Si4Li6/[8,8], (f) Si4Li8/[8,8], and (g) Si4Li10/[9,9] composite.

higher than that of Si4Li8/[8,8] CNT assembly’s AIV. It is important to mention that the calculated AIV values for both composites are lesser than that of the Sn4/CNT (≈1.11 V) composite. For the Sn4Lin/CNT composite, the AIV is lowered from 1.11 V to 0.55 V as we move from Sn4 to Sn4Li8 clusters. Moreover, we compared these voltage with that of pristine CNT; for which the AIV has been found to be approximately 0.90 V. From this, we confirmed that the insertion of lithiated M4 cluster into CNT significantly reduces the intercalation voltage of the composite. A lower value of AIV for the composite indicates that the intercalated Li atoms can easily shuttle between anode and cathode during cell cycling. Therefore, M4Lin/CNT composite can be used as a potential candidate for the anode of LIBs.

The AIV profile for all M4Lin/CNT composites is shown in Figure 8. From the obtained AIV profile of the Si4Lin/CNT

■

CONCLUSIONS To summarize, we have investigated the structural, electronic, and Li intercalation properties of the composite consisting of Si4Lin/Ge4Lin cluster-inserted CNTs for the anode of LIB applications. Initially, we compared the structural and electronic properties of various Si4Lin/Ge4Lin with Sn4Lin clusters. We proved from the BE values that the Si4Lin clusters possess high structural stability and they are less vulnerable toward lithiation. Moreover, the large HOMO−LUMO gap of the M4Lin clusters having even number of Li atoms depicts their higher electronic stability over those with odd number of Li atoms. Further, these stable clusters are selected for studying the interaction with CNT. From this, we understood the lithiated clusters exhibit strong interaction with the wall of CNT as the result of charge transfer from cluster to CNT via cation−π interaction and it plays crucial role for stabilizing the nanocomposites. Moreover, the Li−Li clustering is found to be undermined when the lithiated clusters are present inside CNT; which is expected to be beneficial the electrode performance. Indeed, our calculations on the M4Li6 cluster and hydrogen-terminated CNT demonstrate that the encapsulation of lithiated cluster into the CNT is energetically highly favored. The deduced electronic structures of the composite also suggest that the insertion of lithiated cluster into CNT enhances the stability of the composite. Eventually, the Li intercalation properties of the composite are determined by calculating AIV, which infers that the CNT composites of Si4Lin and Ge4Lin have potential which is comparable to that of a CNT. Moreover, the presence of a lithiated cluster inside the CNT leads to the significant reduction in the AIV of the composite, thus allowing easy shuttling of intercalated Li ions during charge−discharge process. Hence our study proves that these nanocomposites can be used as a potential candidate for the anode of LIBs.

Figure 8. AIV profile for (a) Si4Lin/CNT, (b) Ge4Lin/CNT, and (c) Sn4Lin/CNT composites. The black, red, blue, green, and pink colored lines in each plot represent M4/CNT, M4Li2/CNT, M4Li4/ CNT, M4Li6/CNT, and M4Li8/CNT (M = Si, Ge, and Sn) composites, respectively.

composite, it clears that the Si4/[8,8] CNT assembly has the maximum AIV (≈1.00 V), and this value is slowly reduced for lithiated cluster-inserted composites (refer Figure 8). Here, for Si4Li8/[8,8] CNT assembly, the calculated mean AIV is 0.43 V. However, for Ge4Lin/CNT composite the highest AIV is observed in the case of Ge4/[8,8] CNT system (≈1.05 V). Alike the Si/CNT composite, here also the potential is reduced for lithiated cluster-incorporated CNT composites as shown in the AIV profile. For instance, in the case of Ge4Li8/[8,8] assembly, the mean AIV is found to be 0.49 V, which is slightly 4158

DOI: 10.1021/acsomega.8b03433 ACS Omega 2019, 4, 4153−4160

ACS Omega

■

■

COMPUTATIONAL METHODOLOGY

In this work, first-principles density functional calculations were performed using Vienna Ab initio Simulation Package (VASP)42 to understand the structural and electronic properties of Si−Li and Ge−Li clusters with single-walled CNTs. To understand the structural changes from 2D to 3D geometry, at least four atoms containing cluster is essential; hence, we considered four atoms containing atomic clusters of Si and Ge throughout this work. The projector augmented wave43 pseudopotential method was used to describe the atoms, and electron−electron interactions were correlated by generalized gradient approximations.44 All the atomic clusters were optimized by keeping them in sufficiently large (a = 15 Å) cubic unit cell. Similarly, large a and b (= 21 Å) lattice constants were chosen for isolated and cluster-inserted CNTs. Note that for modeling CNTs, 1 × 1 × 5 supercells were constructed. A plane-wave cutoff energy of 400 eV was used for all the structural relaxation calculations. To study the closed pack assembly of CNTs, the unit cell is completely relaxed without considering any symmetry. To account the weak vdW interaction between the adjacent CNTs in the hexagonal assembly, we performed dispersion-corrected DFT-D2 calculations for the respective cases.45 For sampling the Brillouin zone, Monkhorst−Pack 1 × 1 × 1, 1 × 1 × 4, and 2 × 2 × 4 kmeshes were used for optimizing the clusters, cluster-inserted CNT, and its assembly, respectively. Dense k-mesh of 1 × 1 × 50 was chosen for deducing electronic DOS and BSs for all the cases. Throughout the geometry optimizations, all the ions were relaxed self-consistently without considering any symmetry and the iterative relaxation processes were repeated until absolute force on each ion was converged to the order of 0.01 eV/Å. The convergence for energy was also set to be 10−6 eV in our calculations. We repeated all calculations with aforesaid criteria to find the magnetic solution of the system and this solution is favored for cluster with odd number of Li atoms.

■

Article

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Phone: +91-4565-241443. Fax: +91-4565-227779. ORCID

Karthikeyan J: 0000-0002-1781-8357 P. Murugan: 0000-0003-0062-4828 Present Address §

Department of Applied Physics, School of Science, Aalto University, 00076 Aalto, Finland. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS All authors express their deep gratitude to CSIR−CECRI, CSIR−NCL, and CSIR−CMMACS for sharing cluster computing facilities where all these calculations were carried out. P.M. acknowledges DST-SERB for the financial support through EMR/2016/007112 project. T.K.B. and J.K. are grateful to Dr. Vijay Kumar Foundation, Gurgaon, Haryana, for the kind hospitality at the time of this manuscript preparation.

■

REFERENCES

(1) Deng, D. Li-Ion Batteries: Basics, Progress, and Challenges. Energy Sci. Eng. 2015, 3, 385−418. (2) Bruce, P. G.; Scrosati, B.; Tarascon, J.-M. Nanomaterials for Rechargeable Lithium Batteries. Angew. Chem., Int. Ed. 2008, 47, 2930−2946. (3) Qi, W.; Shapter, J. G.; Wu, Q.; Yin, T.; Gao, G.; Cui, D. Nanostructured Anode Materials for Lithium-Ion Batteries: Principle, Recent Progress and Future Perspectives. J. Mater. Chem. A 2017, 5, 19521−19540. (4) Yang, C.; Deng, S.; Lin, C.; Lin, S.; Chen, Y.; Li, J.; Wu, H. Porous TiNb24O62 microspheres as high-performance anode materials for lithium-ion batteries of electric vehicles. Nanoscale 2016, 8, 18792−18799. (5) Zhao, Y.; Li, J.; Ding, Y.; Guan, L. A nanocomposite of SnO2 and single-walled carbon nanohorns as a long life and high capacity anode material for lithium ion batteries. RSC Adv. 2011, 1, 852−856. (6) Bandyopadhyay, D. Density functional study of the electronic structure and properties of lithium intercalated graphite. Eur. Phys. J. D 2009, 54, 643−655. (7) Mayo, M.; Morris, A. J. Structure Prediction of Li-Sn and Li-Sb Intermetallics for Lithium-Ion Batteries Anodes. Chem. Mater. 2017, 29, 5787−5795. (8) Wang, B.; Luo, B.; Li, X.; Zhi, L. The dimensionality of Sn anodes in Li-ion batteries. Mater. Today 2012, 15, 544−552. (9) Tian, H.; Xin, F.; Wang, X.; He, W.; Han, W. High capacity group-IV elements (Si, Ge, Sn) based anodes for lithium-ion batteries. J. Materiomics 2015, 1, 153−169. (10) Park, C.-M.; Kim, J.-H.; Kim, H.; Sohn, H.-J. Li-alloy based anode materials for Li secondary batteries. Chem. Soc. Rev. 2010, 39, 3115−3141. (11) Lijuan, K.; Ruiyi, L.; Yongqiang, Y.; Zaijun, L. Multi-faceted design of a silicon anode for high performance lithium ion batteries using silicon nanoparticles encapsulated by a multiple graphene aerogel electrode material and a tryptophan-functionalized graphene quantum dot-sodium alginate binder. RSC Adv. 2016, 6, 76344− 76354. (12) Zhao, M.; Zhao, Q.; Qiu, J.; Xue, H.; Pang, H. Tin-based nanomaterials for electrochemical energy storage. RSC Adv. 2016, 6, 95449−95468.

ASSOCIATED CONTENT

* Supporting Information S

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsomega.8b03433. (1) Total and partial DOS of various M4 clusters, (2) optimized structure of various M4Lin clusters, (3−8) ball and stick models of various low lying isomers of lithiated Si4 atomic clusters, (9) relative stability of M4Lin clusters in FM and AFM configuration, (10) optimized structure of M4Lin/CNT composites, (11−20) various configurations of M4Lin cluster inside CNT, (21) dispersion corrected adsorption energy of M4Lin/CNT composite, (22) initial and final structure of M4Li6 cluster kept near to the edge of CNT electronic DOS of Ge4Lin/CNT composites, (23) charge-transfer M4Lin cluster to CNT calculated using PBE functional, (24) electronic DOS of Ge4Lin/CNT composites, (25) electronic BSs of Ge4Lin/ CNT composite, (26) electronic DOS of Sn4Lin/CNT composites, (27) electronic BSs of Sn 4Lin/CNT composites, and (28) Li intercalation in hexagonal assembly of Si4Li6/[8,8] CNT and Ge4Li4/[8,8] CNT systems (PDF) 4159

DOI: 10.1021/acsomega.8b03433 ACS Omega 2019, 4, 4153−4160

ACS Omega

Article

(13) Bogart, T. D.; Lu, X.; Gu, M.; Wang, C.; Korgel, B. A. Enhancing the lithiation rate of silicon nanowires by the inclusion of tin. RSC Adv. 2014, 4, 42022−42028. (14) Wen, Z.; Cui, S.; Kim, H.; Mao, S.; Yu, K.; Lu, G.; Pu, H.; Mao, O.; Chen, J. Binding Sn-based nanoparticles on graphene as the anode of rechargeable lithium-ion batteries. J. Mater. Chem. 2012, 22, 3300− 3306. (15) Cui, G.; Hu, Y.-S.; Zhi, L.; Wu, D.; Lieberwirth, I.; Maier, J.; Müllen, K. A One-Step Approach Towards Carbon-Encapsulated Hollow Tin Nanoparticles and Their Application in Lithium Batteries. Small 2007, 3, 2066−2069. (16) Zeng, L.; Huang, X.; Chen, X.; Zheng, C.; Qian, Q.; Chen, Q.; Wei, M. Ge/GeO2-Ordered Mesoporous Carbon Nanocomposite for Rechargeable Lithium-Ion Batteries with a Long-Term Cycling Performance. ACS Appl. Mater. Interfaces 2016, 8, 232−239. (17) Wang, W.; Favors, Z.; Li, C.; Liu, C.; Ye, R.; Fu, C.; Bozhilov, K.; Guo, J.; Ozkan, M.; Ozkan, C. S. Silicon and Carbon Nanocomposite Spheres with Enhanced Electrochemical Performance for Full Cell Lithium Ion Batteries. Sci. Rep. 2017, 7, 44838. (18) Miao, M. Electrical conductivity of pure carbon nanotube yarns. Carbon 2011, 49, 3755−3761. (19) Ebbesen, T. W.; Lezec, H. J.; Hiura, H.; Bennett, J. W.; Ghaemi, H. F.; Thio, T. Electrical conductivity of individual carbon nanotubes. Nature 1996, 382, 54−56. (20) Fan, Y. J.; Meng, X. S.; Li, H. Y.; Kuang, S. Y.; Zhang, L.; Wu, Y.; Wang, Z. L.; Zhu, G. Stretchable Porous Carbon NanotubeElastomer Hybrid Nanocomposite for Harvesting Mechanical Energy. Adv. Mater. 2016, 29, 1603115. (21) Paul, R.; Etacheri, V.; Pol, V. G.; Hu, J.; Fisher, T. S. Highly porous three-dimensional carbon nanotube foam as a freestanding anode for a lithium-ion battery. RSC Adv. 2016, 6, 79734−79744. (22) Yu, W.-J.; Liu, C.; Hou, P.-X.; Zhang, L.; Shan, X.-Y.; Li, F.; Cheng, H.-M. Lithiation of Silicon Nanoparticles Confined in Carbon Nanotubes. ACS Nano 2015, 9, 5063−5071. (23) Kumar, T. P.; Ramesh, R.; Lin, Y. Y.; Fey, G. T.-K. Tin-filled carbon nanotubes as insertion anode materials for lithium-ion batteries. Electrochem. Commun. 2004, 6, 520−525. (24) Kennedy, T.; Bezuidenhout, M.; Palaniappan, K.; Stokes, K.; Brandon, M.; Ryan, K. M. Nanowire Heterostructures Comprising Germanium Stems and Silicon Branches as High-Capacity Li-Ion Anodes with Tunable Rate Capability. ACS Nano 2015, 9, 7456− 7465. (25) Graetz, J.; Ahn, C. C.; Yazami, R.; Fultz, B. Nanocrystalline and Thin Film Germanium Electrodes with High Lithium Capacity and High Rate Capabilities. J. Electrochem. Soc. 2004, 151, A698−A702. (26) Bijoy, T. K.; Karthikeyan, J.; Murugan, P. Structural stability and bonding nature of Li-Sn-carbon nanocomposites as Li-ion battery anodes: first principles approach. RSC Adv. 2015, 5, 123−129. (27) Dai, D.; Balasubramanian, K. Electronic structure of group IV tetramers (Si4-Pb4). J. Chem. Phys. 1992, 96, 8345−8353. (28) Raghavachari, K. Theoretical study of small silicon clusters: Equilibrium geometries and electronic structures of Sin (n=2-7,10). J. Chem. Phys. 1986, 84, 5672−5686. (29) Bandyopadhyay, D.; Sen, P. Density Functional Investigation of Structure and Stability of Genand GenNi (n= 1−20) Clusters: Validity of the Electron Counting Rule. J. Phys. Chem. A 2010, 114, 1835−1842. (30) Bandyopadhyay, D. Chemisorptions effect of oxygen on the geometries, electronic and magnetic properties of small size Nin (n = 1-6) clusters. J. Mol. Model. 2011, 18, 737−749. (31) Dhaka, K.; Bandyopadhyay, D. Study of the electronic structure, stability and magnetic quenching of CrGen (n = 1-17) clusters: a density functional investigation. RSC Adv. 2015, 5, 83004− 83012. (32) Kumar, M.; Bhattacharyya, N.; Bandyopadhyay, D. Architecture, electronic structure and stability of TM@Ge(n) (TM = Ti, Zr and Hf; n = 1-20) clusters: a density functional modeling. J. Mol. Model. 2011, 18, 405−418.

(33) Trivedi, R.; Bandyopadhyay, D. Hydrogen storage in small size MgnCo clusters: A density functional study. Int. J. Hydrogen Energy 2015, 40, 12727−12735. (34) Corbett, J. D. Polyatomic Zintl anions of the post-transition elements. Chem. Rev. 1985, 85, 383−397. (35) Von Schnering, H. G.; Baitinger, M.; Bolle, U.; CarrilloCabrera, W.; Curda, J.; Grin, Y.; Heinemann, F.; Llanos, J.; Peters, K.; Schmeding, A.; Somer, M. Binary Alkali Metal Compounds with the Zintl Anions [Ge9I4‑ and [Sn9I4‑. Z. Anorg. Allg. Chem. 1997, 623, 1037−1039. (36) Vetter, J.; Novák, P.; Wagner, M. R.; Veit, C.; Möller, K.-C.; Besenhard, J. O.; Winter, M.; Wohlfahrt-Mehrens, M.; Vogler, C.; Hammouche, A. Ageing mechanisms in lithium-ion batteries. J. Power Sources 2005, 147, 269−281. (37) Mahadevi, A. S.; Sastry, G. N. Cation−π Interaction: Its Role and Relevance in Chemistry, Biology, and Material Science. Chem. Rev. 2012, 113, 2100−2138. (38) Bania, K. K.; Guha, A. K.; Bhattacharyya, P. K.; Sinha, S. Effect of substituent and solvent on cation-π interactions in benzene and borazine: a computational study. Dalton Trans. 2014, 43, 1769−1784. (39) Henkelman, G.; Arnaldsson, A.; Jónsson, H. A fast and robust algorithm for Bader decomposition of charge density. Comput. Mater. Sci. 2006, 36, 354−360. (40) Bader, R. F. W. Atoms in MoleculesA Quantum Theory; Oxford University Press: Oxford, UK, 1990. (41) Persson, K.; Hinuma, Y.; Meng, Y. S.; Van der Ven, A.; Ceder, G. Thermodynamic and kinetic properties of the Li-graphite system from first-principles calculations. Phys. Rev. B: Condens. Matter Mater. Phys. 2010, 82, 125416. (42) Kresse, G.; Joubert, D. From ultrasoft pseudopotentials to the projector augmented-wave method. Phys. Rev. B: Condens. Matter Mater. Phys. 1999, 59, 1758−1775. (43) Blöchl, P. E. Projector augmented-wave method. Phys. Rev. B: Condens. Matter Mater. Phys. 1994, 50, 17953−17979. (44) Perdew, J. P.; Chevary, J. A.; Vosko, S. H.; Jackson, K. A.; Pederson, M. R.; Singh, D. J.; Fiolhais, C. Atoms, molecules, solids, and surfaces: Applications of the generalized gradient approximation for exchange and correlation. Phys. Rev. B: Condens. Matter Mater. Phys. 1992, 46, 6671−6687. (45) Grimme, S. Semiempirical GGA-type density functional constructed with a long-range dispersion correction. J. Comput. Chem. 2006, 27, 1787−1799.

4160

DOI: 10.1021/acsomega.8b03433 ACS Omega 2019, 4, 4153−4160