Theoretical Studies On The Structures And Stability Of Silicon, Phosphorus Or Sulfur-related Molecules

Silicon, carbon, nitrogen, and oxygen-containing chemistries haveattracted continuing interest from various fields. One particular interest is theirpossible role in astrophysical chemistry. Up to now, some silicon-, carbon-,nitrogen-, or oxygen-containing molecules, such as SiC_n (n=1-4), SiCN, SiN,SiO, SiS, C₂, CO, CN, NO and NP, have been detected in interstellar space.Their direct addition may possibly lead to the formation of several NC2O, [Si,C, N, O] or Si2NO isomers, such as NC+CO→NCCO, SiN+CO→SiNCO, andSiN+SiO→SiNSiO. Therefore, the 19-valence-electron series XX'NO (X,X'=C, Si) radicals should be of interstellar interest. It is well known thatsilicon and carbon have iso-valence electrons, however, their chemicalproperties are rather different from each other. Silicon is usually reluctant toform multiple-bonding whereas carbon prefers. Therefore, the structures,stabilities, and bonding properties of XX'NO (X, X'=C, Si) isomers shouldhave distinct features. A detailed theoretical study of XX'NO (X, X'=C, Si)radicals will promote our understanding of the structural, energetic, andbonding changes among the series of molecules. On the other hand, the silicon,carbon, nitrogen, and oxygen-related chemistries have also been believed toplay important roles in microelectronic materials. The knowledge about thestructures, energies, and bonding natures of various [Si, C, N, O] and Si2NOisomers may be helpful for understanding the initial step of the growingmechanism during the NO-doped SiC or Sin vaporization process.Phosphorus, nitrogen, and sulfur chemistries have received considerableattention from various aspects. One interesting field is interstellar chemistry.Up to now, many phosphorus-, nitrogen-, or sulfur-containing molecules, suchas CnN (n=1, 3, 5), CnS (n=1-3), CP, NP, SiN, SiS, NO, and NS, have beendetected in interstellar space. Furthermore, several phosphorus-containingmolecules, such as C2P, C3P, and HC2P, might be detectable in space if theyare unreactive with oxygen atoms. It is well known that the larger CnN, CnP,and CnS species have been extensively studied and are expected to be carriersof some interstellar bands. Also, the mixed SCnX (X = N or P) species maypresent a bridge between the CnS and CnX (X = N or P) clusters and may be ofastronomical interest. Therefore, understanding the structural, bonding, andstability properties of CnX or SCnX (X = N or P) clusters may be helpful forfuture identification of new N, P, or S-containing species either in laboratoryor in space and also for elucidation of the formation mechanism of the N orP-doped Cn or CnS clusters or S-doped CnN or CnP clusters.The structures, stability and bonding natures of XX'NO (X, X'=C, Si)radicals, SCnX (X=N, P;n=2, 3) molecules and C4P radical have been studiedby quantum chemistry computation. The main conclusions are summarized asfollows:1. At the B3LYP, QCISD, CASSCF, CCSD(T) and CASPT2 levels, thedoublet potential energy surface (PES) studies are performed on the19-valence-electron series XX'NO (X, X'=C, Si) radicals. The electronicstructures and bonding nature of the important isomers are analyzed. Possibleformation strategies of the relevant isomers in laboratory and space arediscussed in detail. Moreover, the NO-doped SiC or Sin vaporization processesare also discussed.For NC2O, at the CCSD(T)/6-311+G(2df)//QCISD/6-311G(d)+ZPVElevel, the lowest-lying isomer is experimentally known NCCO 1 with bentstructure followed by the bent isomer CNCO 2. The potential energy surfaceindicates two isomers 1, 2 and another high-lying species CCNO 4 with bentform are kinetically stable with a barrier of at least 20 kcal/mol. Therefore,they could be observable in laboratory and in interstellar space. This result isin agreement with the indication of neutralization-reionization massspectrometry experiments.For [Si, C, N, O], at the CCSD(T)/6-311+G(2df)//QCISD/6-311G(d)+ZPVE level, the lowest-lying isomer is found to be linear SiNCO 1. Two bentisomers OSiCN 2 and OSiNC 3 are energetically close to the species 1. All thethree isomers 1, 2, 3, and another high-lying species, linear SiCNO 5, arecalculated to possess considerable kinetic stability and may be observable inthe laboratory and in interstellar space. The predicted results of the isomers 1and 2 are in agreement with the indication from the known mass spectrometryexperiments. Moreover, the fourth low-lying species SiOCN 4 with bent formmay be observable in low-temperature environments.For Si2NO, at the CCSD(T)/6-311+G(2df)//QCISD/6-311G(d)+ZPVElevel, three low-lying isomers are located, i.e., one bent SiNSiO 3 and twocyclic isomers cSiNSiO 1 and cSiNSiO 1'. In addition to experimentallyknown species 3, two new low-lying isomers 1 and 1' have reasonable kineticstabilities and could be observable in laboratory and in interstellar space. Thepredicted possible formation process of 3 is consistent with the recentexperimental indication.By comparison, we find that the PES of [Si, C, N, O] radical is similar tothat of C2NO, on their PESs, the linear or chainlike species arethermodynamically and kinetically more stable and no cyclic isomers are ofinterest. On the Si2NO PES, besides chainlike species, four-membered ringisomers are also of interest. Obviously, for the 19-valence-electron seriesXX'NO (X, X'=C, Si) radicals, with the increasing number of the Si atom, thecyclic structures become increasingly important, which can be attributed to thefact that the second-row element Si has higher preference to form σ-bondingthan the first-row element C. Being aware that up to now, no cyclicnitrogen-containing species have been detected in space, two cyclic Si2NOisomers 1 and 1' could be promising candidates. Finally, the calculated resultsare expected to be useful for future identification of various XX'NO (X, X'=C,Si) isomers either in laboratory or in interstellar space and also understandingthe initial growing step for the NO-doped SiC and Sin vaporization processes.2. At the B3LYP, QCISD and CCSD(T) levels, the structures, energetics,and spectroscopies of SCnX (X=N, P;n=2, 3) isomeric species are explored.The doublet potential energy surfaces for them are established, respectively.The electronic structures and bonding nature of the relevant isomers areanalyzed.For NC2S, at the CCSD(T)/6-311+G(2df)//QCISD/6-311G(d)+ZPVElevel, in addition to the experimentally known isomers bent NCCS 1(I) andquasi-linear NCCS 1(II), three new isomers bent CNCS 2(I), quasi-linearCNCS 2(II), and linear CCNS 4 are predicted to possess very high kineticstability of more than 25 kcal/mol. They could be produced under certainlaboratory and interstellar conditions. The isomers NCCS 1(II) and CNCS 2(II)are the Renner-Teller pair states for NCCS 1(I) and CNCS 2(I), respectively.In addition, the three metastable isomers with slightly lower kinetic stabilityare possibly observed in laboratory and in space, i.e., bent CSCN 3, bentCCSN 6 and cyclic S-cCCN 7. The possible formation strategies of therelevant isomers in laboratory and in space are discussed in detail.For NC3S, at the CCSD(T)/6-311G(2df)//QCISD/6-311G(d)+ZPVE level,in addition to the experimentally known isomer NCCCS 1, eight new isomersCNCCS 2, CCNCS 3, CCSCN 4, CCSNC 5, CCCNS 6, NC-cCCS 9, 9' andCN-cCCS 10 are predicted to possess very high kinetic stability of around ormore than 20 kcal/mol. They could be produced under certain laboratory andinterstellar conditions. All the nine NC3S isomers can mainly be considered asthe different combined forms between the C3N radical and S-atom. Therefore,it's better to consider NC3S as S-doped C3N molecules rather than theN-doped C3S molecules. Similarly, we expect that the isomers (S-addition atterminal atoms of CnN, S-insertion to the C-C single bonding of CnN, andS-addition to the terminal CC Ï€-bonding of CnN) conceived from thecombination between CnN and S-atom can be kinetically stable for largerNCnS molecules for which detailed potential energy surveys are even morecost expensive.For PC2S, at the CCSD(T)/6-311+G(2df)//QCISD/6-311G(d)+ZPVElevel, the lowest-lying isomer is a linear form PCCS 1 that is predicted topossess great kinetic stability about 50 kcal/mol. Additionally, two high-lyingisomers bent CCPS 2 and bent CCPS 2' also have considerable kineticstability. Interestingly, two isomers 2 and 2' can be easily converted to eachother, and both could coexist. All the three isomers are very promisingcandidates for future laboratory and astrophysical detection. Possibleformation strategies of the isomers 1, 2, and 2' in laboratory and in space arealso discussed. The present PC2S work represents the first report on a detailedpotential energy surface of the PCnS (n>1) series.The calculated results are expected to be useful for future identification ofvarious SCnX (X=N, P;n=2, 3) isomers either in laboratory or in interstellarspace and also understanding the isomerism of larger NCnS radicals as well asthe formation process of S-doped CnN materials.3. At the B3LYP, QCISD and CCSD(T) levels, a detailed potential energysurface of C4P is theoretically established. At the CCSD(T)/6-311G(2df)//QCISD/6-311G(d)+ZPVE level, the lowest-lying isomer is the quasi-linearCCCCP 1. Previous B3LYP calculations predict a linear form and largerbending mode. The second low-lying isomer is the three-membered ringPC-cCCC 3. Both species 1 and 3 have considerable kinetic stability and couldbe observed in laboratory and in interstellar space. Moreover, four high-energyisomers, i.e., bent CCPCC 2, bent CCPCC 2', cage-like 10 and cage-like 11,also possess considerable kinetic stability. Together with the lower-energyones 1 and 3, they can be produced during the thermal vaporization process ofphosphorus-doped carbon clusters. They can finally isomerize to 3 and 1. Inaddition, significant discrepancies are found in the bonding nature and dipolemoments between C4P and its nitrogen-analogue C4N, which are be accountedfor by the different preference to form multiple-bonding or single-bonding. Wehope that the calculated results may stimulate future identification of the C4Pradical either in laboratory or in interstellar space. The calculated results arealso expected to provide useful information for future investigation of largerCnP (n>4) radicals and for understanding the mechanism of P-doped carbonclusters.