MNRAS 493, 3054–3059 (2020) doi:10.1093/mnras/staa467 Advance Access publication 2020 February 17 On carbon nanotubes in the interstellar medium Qi Li,1,2 Aigen Li,2‹ B. W. Jiang1‹ and Tao Chen3 1Department of Astronomy, Beijing Normal University, Beijing 100875, China 2Department of Physics and Astronomy, University of Missouri, Columbia, MO 65211, USA 3Department of Theoretical Chemistry and Biology, Royal Institute of Technology, SE-10691 Stockholm, Sweden Accepted 2020 February 13. Received 2020 February 12; in original form 2020 January 15 Downloaded from https://academic.oup.com/mnras/article/493/2/3054/5737582 by guest on 01 October 2021 ABSTRACT Since their discovery in 1991, carbon nanotubes (CNTs) – a novel one-dimensional carbon allotrope – have attracted considerable interest worldwide because of their potential technolog- ical applications such as electric and optical devices. In the astrophysical context, CNTs may be present in the interstellar space since many of the other allotropes of carbon (e.g. amorphous carbon, fullerenes, nanodiamonds, graphite, polycyclic aromatic hydrocarbons, and possibly graphene as well) are known to be widespread in the Universe, as revealed by pre-solar grains in carbonaceous primitive meteorites and/or by their fingerprint spectral features in astronomical spectra. In addition, there are also experimental and theoretical pathways to the formation of CNTs in the interstellar medium (ISM). In this work, we examine their possible presence in the ISM by comparing the observed interstellar extinction curve with the ultraviolet/optical absorption spectra experimentally obtained for single-walled CNTs of a wide range of diameters and chiralities. Based on the absence in the interstellar extinction curve of the ∼4.5 and 5.25 eV π-plasmon absorption bands that are pronounced in the experimental spectra of CNTs, we place an upper limit of ∼10 ppm of C/H (i.e. ∼4 per cent of the total interstellar C) on the interstellar CNT abundance. Key words: dust, extinction – ISM: lines and bands – ISM: molecules – infrared: ISM. similartothatusedbyKrotoetal.(1985) for fullerene synthesis. In 1 INTRODUCTION 2004, the first isolation of a single graphene sheet has been achieved As one of the most abundant elements in the Universe only exceeded by Andre Geim and Kostya Novoselov who extracted single-atom- by hydrogen, helium, and oxygen, carbon (C) is a major player in thick layers from bulk graphite (see Novoselov et al. 2004). Because the evolutionary scheme of the Universe (Henning & Salama 1998). of this, they were awarded the 2010 Nobel Prize in Physics. Since C atoms can form strong and stable single, double, and triple The presence of many of these C allotropes in the interstellar covalent bonds, they facilitate the generation of a large variety medium (ISM) has been explicitly revealed or implicitly indi- of allotropic forms, including amorphous carbon, carbon chains, cated from pre-solar grains isolated from carbonaceous primitive carbon nanotubes (CNTs), carbon onions, diamond, fullerenes (e.g. meteorites (e.g. nanodiamonds, graphite; see Lewis et al. 1987; C60,C70), graphene, graphite, and polycyclic aromatic hydrocar- Amarietal.1990), and/or from the observations of molecular bons (PAHs). While diamond and graphite have been known for and solid-state features in astronomical spectra and the realization at least a few thousand years, the laboratory discoveries of such that these features are linked to certain carbonaceous materials low-dimensional carbon nanostructures as zero-dimensional (0D) (e.g. amorphous carbon, graphite, and PAHs; see Stecher & Donn fullerenes, one-dimensional (1D) CNTs, and two-dimensional (2D) 1965;Leger´ & Puget 1984; Allamandola, Tielens & Barker 1985; graphene were far more recent (see Dinadayalane & Leszczynski Pendleton & Allamandola 2002;Qietal.2018). While fullerenes, 2010). CNTs, and graphene are of paramount importance in modern 35 yr ago, C60 and C70 were serendipitously discovered by Kroto science and technology since these carbon nanomaterials provide et al. (1985)–R.F.Curl,H.W.Kroto,andR.E.Smalleyreceivedthe exciting challenges and opportunities for physicists, chemists, 1996 Nobel Prize in Chemistry because of this spectacular discovery biologists, engineers, and material scientists, in recent years, the – in the sooty residue experimentally generated by vapourising astronomical community has also been passionate about their graphite in a (hydrogen-lacking) helium atmosphere. 6 yr later, possible presence and the role they could have played in the CNTs were discovered and synthesized by Iijima (1991) as spin-off interstellar and circumstellar space. It is worth noting that the initial products of fullerenes, using an arc-discharge evaporation method synthesis of fullerenes was actually astrophysically motivated – the experiments that resulted in the discovery of fullerenes were originally aimed at understanding the mechanisms by which long- E-mail: [email protected] (AL); [email protected] (BWJ) chain carbon molecules are formed in the interstellar space and C 2020 The Author(s) Published by Oxford University Press on behalf of the Royal Astronomical Society Interstellar carbon nanotubes 3055 circumstellar shells (Kroto et al. 1985). A quarter of a century later, multishell tubes on the carbon cathode, Iijima & Ichihashi (1993) the detection of C60 and C70 and their ions in the interstellar and found that abundant single-walled carbon nanotubes (SWNTs) with circumstellar space has been reported based on their characteristic diameters of about 1 nm grow in the gas phase. As will be discussed vibrational spectral bands in the infrared (IR; Cami et al. 2010; in Section 4, the possible pathways for the formation of CNTs in Sellgren et al. 2010;Berne,´ Mulas & Joblin 2013;Strelnikov, the ISM would result in SWNTs, more favourably than multiwalled Kern&Kappes2015). Webster (1992) suggested that fullerenes nanotubes. In this work, we will therefore focus on SWNTs. and hydrogenated fullerenes (i.e. fulleranes) could be responsible SWNTs can be considered as a layer of graphene sheet rolled for an appreciable component of the optical and ultraviolet (UV) up into a cylinder. The ends of each nanotube are closed by hemi- interstellar extinction. It has also been proposed that single- and fullerene caps, each containing six five-membered rings. In the limit multishell fullerenes may be the carrier of the still unidentified of large aspect (i.e. length-to-diameter) ratios, however, the ends 2175 Å interstellar extinction feature (e.g. see Iglesias-Groth 2008; have negligible effects on nanotube electronic structure. SWNTs Li et al. 2008). Cataldo (2002) measured the absorption spectra of are novel 1D materials made of an sp2-bonded wall one atom thick fullerite (i.e. carbon soot containing fullerenes) and C60 fullerene and constitute a rich family of structures, with each SWNT structure Downloaded from https://academic.oup.com/mnras/article/493/2/3054/5737582 by guest on 01 October 2021 photopolymer in the wavelength range of ∼0.2–1.1 μm. It was uniquely defined by the chiral index, a pair of integers (n, m)that found that neither fullerite nor C60 photopolymer exhibits any describes the length and orientation of the nanotube’s circumference strong absorption feature around ∼2175 Å. Instead, fullerite shows vector within the graphene sheet. SWNTs are classified into three prominent absorption bands at ∼2520–2670 Å, similar to that seen types: (i) armchair (n, n) nanotubes, (ii) zigzag (n, 0) nanotubes, and in some hydrogen-deficient and carbon-rich supergiant R CrB stars (iii) chiral (n, m) nanotubes with n = m.1 Electrically, SWNTs can be (e.g. R CrB, RY Sgr, V348 Sgr) that peaks around ∼2400–2500 Å either metallic or semiconducting, depending on their geometry, i.e. (Hecht et al. 1984; Drilling et al. 1997). Unlike fullerite, C60 on the (m, n) integer-pair, or on the way of the rolling up. Armchair photopolymer exhibits three absorption peaks at ∼2710, 3890, and SWNTs are always metallic and exhibit no energy bandgap, as are 5100 Å that are not seen in the ISM. zigzag SWNTs with m = 3q,whereq is an integer.2 On a statistical As the basic structural element of fullerenes, graphene is in- basis, one-third of the nanotubes are metallic, and two-thirds are timately related to fullerenes as demonstrated by Chuvilin et al. semiconducting. (2010) experimentally and by Berne&Tielens(´ 2012) computa- Each SWNT species, as labelled by a unique pair of integers tionally that C60 could be formed from a graphene sheet. However, (n, m), has well-defined electronic and spectroscopic properties. whether graphene is present in the interstellar and circumstellar Because of their distinct physical and chemical properties, different space is less certain. Garc´ıa-Hernandez´ et al. (2011, 2012) detected (n, m) structural species may be considered separate chemical a set of unusual IR emission features in several planetary nebulae substances. However, despite the importance of (n, m)-specific (PNe), both in the Milky Way and in the Magellanic Clouds, and absorption spectra, reliable experimental determination of nanotube attributed them to planar C24, a piece of graphene sheet. Based absorption cross-section at the individual-tube level has been ham- on the absence of the 2755 Å absorption feature in the interstellar pered by the difficulty of sorting as-grown mixtures into structurally extinction curve characteristic of the π–π ∗ electronic transition pure fractions and by the challenge of determining absolute SWNT of graphene, Li, Li & Jiang (2019) argued that in the ISM as concentrations in suspensions that often also contain surfactants or much as ∼20 ppm of C/H could be tied up in graphene (also see polymer coatings (see Guo et al. 2004; Murakami & Maruyama Chen, Li & Zhang 2017). In addition, Sarre (2019) attributed the 2009; Weisman & Kono 2019). Whereas the interstellar extinction widespread extended red emission (ERE; see Witt & Vijh 2004) is the strongest in the UV, to the best of our knowledge, exist- – a broad, featureless emission band at ∼5400–9500 Å – to the ing measurements of the absorption cross-sections for individual photoluminescence of graphene oxide nanoparticles.