以第一原理計算探討石墨烯中硼、氮摻雜對多硫化合物穩定機制之影響

Abstract

本研究利用第一原理探討硼、氮摻雜對於鋰硫電池中石墨烯正極材料穩定多硫化合物的機制與影響。多硫化合物為鋰硫電池放電過程中會溶於電解液的中間產物，若能有效將其吸附於石墨烯表面，則能有效提升鋰硫電池的循環能力。本研究第一部分討論並比較純硼、純氮摻雜與缺陷對石墨烯穩定多硫化合物的能力之影響。第二部分則討論硼、氮共同摻雜石墨烯吸附多硫化合物的行為與機制。 第一部份的研究為比較純硼與純氮在吸附多硫化合物的行為與機制差異，解釋為何實驗上觀察到少量硼摻雜的循環能力優於大量氮摻雜。首先研究的是硼、氮摻雜在石墨烯中的排列結構。結果顯示在沒有空缺時，氮原子傾向於遠離彼此，而硼則會傾向於聚集成一區域，因此只要少量的硼摻雜就可以在局域生成高硼濃度的區域。接下來使用前面所得的結構探討其對於多硫化合物的吸附能力，結果顯示氮摻雜只有在與兩顆碳鍵結的位置中有較強的吸附能，相反的硼摻雜只要濃度夠高就可以提供相當的吸附能，因此硼摻雜中能提升吸附能的比例較氮大上許多。此外在氮摻雜吸附多硫化合物的結構中觀察到鋰會脫離多硫化合物並直接吸附到石墨烯上，這顯示在實際放電過程中會有鋰先與石墨烯吸附後再與硫反應生成多硫化合物，因此實際情形應是多硫化合物吸附在含鋰的石墨烯上，結果顯示當石墨烯上已經有鋰時會使多硫化合物的吸附能下降，表示前面計算得到氮摻雜的吸附能應是被高估的，而這種情況並不會發生在硼摻雜的結構中。此外，本文也比較硼、氮摻雜結構吸附多顆多硫化合物的能力，結果顯示在相同摻雜濃度下，硼摻雜結構能吸附多硫化合物的數目上遠大於氮摻雜的結構。總結來說，硼摻雜結構可以較氮摻雜有較好的循環能力有四個原因，分別為摻雜原子傾向於聚集、較多的可吸附結構、不會有鋰先與石墨烯吸附與較強吸附多顆多硫化合物的能力。 第二部份主要討論硼、氮共同摻雜對於穩定多硫化合物之機制。首先同樣是找尋較穩定的摻雜結構，結果顯示在沒有缺陷時，硼、氮之間形成鍵結聚集成一區為最穩定的結構。但是即使不形成硼-氮鍵結，兩種摻雜元素之間仍可以使彼此摻雜所需能量降低，這顯示硼、氮共同摻雜較純硼或氮摻雜更容易達到高摻雜濃度。另外在摻雜結構方面有一個有趣的發現，在純氮或硼的系統中，氮摻雜的空缺都較硼摻雜空缺的形成能低上許多。但是在共同摻雜的空缺中，反而是硼原子在空缺周圍較穩定。接下來同樣使用前面所得的摻雜結構討論其吸附多硫化合物的能力，結果顯示在沒有空缺的情況下，過往實驗中認為可以吸附多硫化合物的硼、氮鍵結結構實際上無法吸附多硫化合物，這是由於氮摻雜抵消硼摻雜產生的p型摻雜效果。因此在沒有空缺的情況下，只有在硼、氮之間沒有形成鍵結且硼聚集的結構有辦法吸附多硫化合物。至於在有空缺的結構部份分為在氮摻雜空缺旁加入硼與硼摻雜空缺旁加入氮兩種情況作討論。在氮摻雜空缺旁加入硼的結果顯示硼加入可以有效提升該缺陷對於多硫化合物的吸附能，且即使在第一部份提到預先有鋰吸附的結構仍可提供一定的吸附能，但硼的加入同時會使該結構的形成能上升，使其不易形成，甚至更傾向於生成沒有空缺的結構。而在硼摻雜的結構中，發現隨著氮的加入該結構對多硫化合物的吸附能會逐漸下降，唯有在硼濃度高於氮時才能提供足夠的吸附能，但氮摻雜的加入同時會使該結構形成能下降，使其越容易形成，甚至較形成沒有空缺的結構來的容易。總結來說，以材料設計的角度出發在製備硼、氮共同摻雜的石墨烯做為鋰硫電池正極材料時，應適當控制條件使硼濃度大於氮，使其可以在達到高摻雜濃度的同時不會生成無法吸附多硫化合物的缺陷結構。

In this thesis, first principles calculations are employed to explore stabilization behaviors of lithium polysulfides on B, N doped graphene. Lithium polysulfides are soluble intermediates produced by discharging process of Li-S batteries. If the substrates can adsorb lithium polysulfides, the cycle performance of Li-S batteries can be improved. Therefore, it would be of great interest to develop detailed atomistic scale understanding of this materials system in many fundamental aspects, especially for the interactions between lithium polysulfides and substrate. In the first part of this thesis, we investigated the adsorption behaviors of lithium polysulfides on pure B and N doped graphene. The doping configurations of pure B and N doped graphene were first studied. The graphitic N dopants prefer to stay away from each other. In contrast, graphitic B prefer to stay together, which makes it easier to reach high dopant contents in local area. The adsorption energies of lithium polysulfides on substrates with low formation energies were calculated. The result shows that N-doped graphene can afford strong binding energy only if N dopants locate at two-fold bonded sites with vacancies. However, B-doped graphene can adsorb lithium polysulfides no matter what sites B are located. In addition, we found that one of Li would be grabbed from lithium polysulfides to N-doped substrates with vacancy. Thus, in the real discharging process, Li will first bind to substrate then interact with sulfur. So the lithium polysulfides in that kind of system will adsorb on substrate with trapped Li. Our result indicates that Li-trapping effect will lower the binding energy of lithium polysulfides, revealing that the adsorption energies calculated in previous section of N-doped graphene are overestimated. We also studied the ability of substrates to adsorb multiple lithium polysulfides. The result shows that B-doped graphene can bind more lithium polysulfides than N-doped graphene, which is one of the advantage to B-doped substrates. In summary, B is a better dopant than N due to four reasons: preference of dopant clustering, more dopant configurations capable for adsorbing lithium polysulfides, no Li-trapping effect and stronger ability to adsorb multiple lithium polysulfides. In the second part of this thesis, we studied the anchoring mechanism of lithium polysulfides on B, N co-doped graphene. The dopant configurations are first investigated. Our result indicates that B and N can synergically reduce the formation energies of substrates, showing that high dopant contents can be easily reached. As for adsorption energy calculations, our result show that structures with B-N pairs, which thought to be able to adsorb lithium polysulfides by previous experimental studies, cannot afford strong adsorption energy due to the cancellation of p-type doping effect by N dopants. Instead, the structures without B-N pair and having more graphitic B dopants than N can anchor the lithium polysulfides. We also explored the effect of B to N-doped vacancy and the effect of N to B-doped vacancy. Adsorption energies of N-doped vacancy significantly increase with the increasing B contents. Even after Li-trapped, the substrates could still adsorb lithium polysulfides. However, the formation energies of those structures become higher in the meantime. On the other hand, with the increasing N contents, adsorption energies of B-doped vacancy decrease and the formation energies of those structures also decrease. In the view of materials design, graphene with more boron than nitrogen dopants are better cathode materials for Li-S batteries, which could not only reach high dopant concentration but also creating structures that can strongly anchor lithium polysulfides.