Hydrogen molecular ions and the violent birth of the Solar System

Abstract

Many pieces of evidence indicate that the Solar System youth was marked by violent processes: among others, high fluxes of energetic particles (greater than or equal to 10 MeV) are unambiguously recorded in meteoritic material, where an overabundance of the short-lived ¹⁰Be products is measured. Several hypotheses have been proposed to explain from where these energetic particles originate, but there is no consensus yet, mostly because of the scarcity of complementary observational constraints. In general, the reconstruction of the past history of the Solar System is best obtained by simultaneously considering what we know of it and of similar systems nowadays in formation. However, when it comes to studying the presence of energetic particles in young forming stars, we encounter the classical problem of the impossibility of directly detecting them toward the emitting source (analogously to what happens to galactic cosmic rays). Yet, exploiting the fact that energetic particles, such as cosmic rays, create ${{\text{H}}_3}^{+}$ and that an enhanced abundance of ${{\text{H}}_3}^{+}$ causes dramatic changes on the overall gas chemical composition, we can indirectly estimate the flux of energetic particles. This contribution provides an overview of the search for solar-like protostars permeated by energetic particles and the discovery of a protocluster, OMC-2 FIR4, where the phenomenon is presently occurring.

This article is part of a discussion meeting issue ‘Advances in hydrogen molecular ions: H₃⁺, H₅⁺ and beyond’.

1. The violent birth of the Solar System

Understanding our past history seems to be an overwhelming human need. Humans have always tried to understand what happened on Earth and also before our planet was formed. Like archaeologists dig the ground to discover long-vanished civilizations, astronomers peruse space to find traces of the first steps of the Solar System formation. On the one hand, we have the present Solar System objects: planets and the small bodies like comets and asteroids. On the other hand, we have astronomical objects that are in the phase of eventually forming solar-like planetary systems. Past experience has repeatedly demonstrated that we can learn more ‘comparing the two hands than focusing on only one’.

An illustrative and powerful example is provided by the violent birth of the Solar System. Traces of it are contained in meteoritic material, specifically in the measured overabundance of ¹⁰Be when compared to that of the Galaxy (at the time of the Solar System formation) [1,2]. Indeed, ¹⁰Be is a short-lived radionuclide whose half-life is only 1.5 Myr so that the measured overabundance implies that it was incorporated in the meteoritic rocks exactly when they formed, namely at the very beginning of the Solar System formation process, 4.5 billion years ago. In other words, ¹⁰Be provides us with a snapshot of the very early Solar System. That in itself might not be such important information if it were not for the fact that ¹⁰Be is only formed by spallation reactions, namely the breaking of O and Si atoms when they are hit by protons and nuclei with more than 10 MeV of energy [3–5]. Therefore, the overabundance of ¹⁰Be in meteoritic material tells us that at the very beginning of the Solar System there were energetic (with greater than or equal to 10 MeV) particles and that their overall dose was about 10¹⁹–10²⁰ particles cm⁻² [6].

However, meteorites cannot tell us why and where the energetic particles were present. In order to know more, we need to search for solar-like planetary systems that are forming nowadays in the Galaxy and where a similar process is taking place. If our hunt is successful, we will have systems where we can observe and study the energetic particle phenomenon and, consequently, we will have the possibility to understand what accelerated particles at such huge velocities in a young Solar System that had ‘only’ gravitational energy to spend.

2. How to probe the presence of protostellar energetic particles

The most obvious way to find a solar-like progenitor irradiated/permeated by energetic particles with energies greater than or equal to 10 MeV would be to look for the particles themselves. But unfortunately, this is not possible because, even assuming that we had a super-sensitive telescope, the accelerated protons and nuclei would be deviated by the Galactic magnetic fields, removing therefore the spatial origin of the particles arriving on Earth. The problem is similar to the one of searching for the sources of acceleration of cosmic rays and, as in that case, we have to use the effects of the interaction of energetic particles on the surrounding matter: (i) the emission of greater than or equal to 280 MeV γ-rays, due to the decay of the pions π⁰ created when nuclei hit H atoms [7,8]; (ii) the ionization of the gas [9–11]; and (iii) the heating of the gas. The first method applies only to particles with energies greater than or equal to 280 MeV and, even assuming that they are present, is out of reach of present instruments; the last one is too generic to be a fingerprint of the presence of energetic particles. The second one, the enhanced ionization of the gas is, instead, a promising one. As a matter of fact, it was used to reveal the presence of freshly accelerated cosmic rays in molecular clouds close to supernovae remnants, where cosmic rays are believed to be accelerated [10–12].

In molecular gas, the ionization is usually measured by observing molecular ions [10–14]. Obviously, in the context of this discussion meeting issue, ${{\text{H}}_3}^{+}$ is one of them, as it is directly formed by cosmic rays (they actually ionize either H or H₂ which both react with other H₂ and form ${{\text{H}}_3}^{+}$). But ${{\text{H}}_3}^{+}$ lines in cold gas can only be observed in absorption (e.g. [10]), so that ${{\text{H}}_3}^{+}$ observations need an IR source behind the object to study. Consequently, ${{\text{H}}_3}^{+}$ observations are not adapted to obtain maps of ionization in protostars, whose envelopes are less than a few thousand AU in diameter. On the contrary, we propose the use of two molecular ions whose rotational lines can be observed in emission and which are almost direct products of the interaction of energetic particles with the gas: HCO⁺ and N₂H⁺ [15,16].

The chemistry of these two hydrogen molecular ions is very simple and is illustrated in figure 1. Briefly, in UV-shielded regions, the reaction of ${{\text{H}}_3}^{+}$ with CO and N₂ results in the formation of HCO⁺ and N₂H⁺, respectively, at very similar rates. The two ions are then destroyed by reactions with abundant neutrals or electrons, again at similar rates. Given the actual numbers, in ‘normal’ molecular clouds, the reaction with CO is the dominant destruction route, and it destroys N₂H⁺ forming more HCO⁺. As a consequence, in normal clouds the HCO⁺/N₂H⁺ abundance ratio is much larger than the CO/N₂ abundance ratio. On the contrary, in the presence of a high electron abundance (greater than or equal to 10⁻⁶ with respect to H₂), electron recombination dominates both HCO⁺ and N₂H⁺ destruction. Being now both the formation and destruction of HCO⁺ and N₂H⁺ caused by the same species and at the same rates, the HCO⁺/N₂H⁺ abundance ratio is equal to the CO/N₂ one. We emphasize that this situation can only occur in cosmic-ray or energetic-particle irradiated molecular gas,¹ so that the HCO⁺/N₂H⁺ abundance ratio can be a very direct probe of the cosmic-ray/energetic-particle irradiation.

Figure 1.

Scheme of the reactions involving HCO⁺ and N₂H⁺ in UV shielded molecular gas. In standard molecular gas, the reaction with CO dominates the destruction of N₂H⁺ which leads to additional formation of HCO⁺ and, consequently, an abundance ratio HCO⁺/N₂H⁺ ≫ CO/N₂. Conversely, in highly cosmic-ray/energetic-particle irradiated gas, electrons are the major destroyers of both HCO⁺ and N₂H⁺, so their abundance ratio tends to be equal to CO/N₂. (Online version in colour.)

One caveat has to be spelled out: the method only applies in regions where the CO and N₂ depletion is negligible. In cold regions, where the two species are frozen onto the grain mantles, or in UV-illuminated regions, where they are selectively photo-dissociated, the HCO⁺/N₂H⁺ abundance ratio is altered by the differential gaseous CO/N₂ abundance ratio and the method cannot be applied. A way to get rid of the first problem is to select HCO⁺ and N₂H⁺ high J lines, which, by definition, probe warm gas. For the second case, the measured abundance of HCO⁺ and N₂H⁺ can be used to exclude an enhanced electron abundance caused by UV photons, because the latter increase the HCO⁺ and N₂H⁺ destruction rate without increasing their formation, contrarily to the cosmic-ray/energetic-particle irradiation case: the result is a low abundance of both species in UV-illuminated regions.

3. The hunt for protostellar energetic particles

Armed with this method, we used the HIFI instrument on board the Herschel Space Observatory to observe protostellar sources in the high-J lines of HCO⁺ and N₂H⁺ [15,16]. We obtained observations of almost a dozen low-luminosity sources in almost a dozen HCO⁺, H¹³CO⁺ and N₂H⁺ lines.² The targeted sources (in table 1) were selected to represent candidate solar-like forming planetary systems. Therefore, we selected nearby sources with luminosities from 1 to 500 L_o,³ so as to cover a relatively wide range of mass and evolutionary stages. The observed lines had upper level J_up ranging from 6 to 13, which correspond to upper level energies ranging from 90 to 400 K and frequencies from about 500 to 1200 GHz. Specifically, in nine sources we observed HCO⁺ J_up = 6, 8 and 12, H¹³CO⁺ J_up = 6, 7 and 9, and N₂H⁺ J_up = 6, 8 and 11. In OMC-2 FIR4, which was a target of the Herschel Key Project CHESS (Chemical Surveys of Star formation regions [17]), we obtained an unbiased line survey that covered the frequency range 480–1240 GHz and, consequently, we observed HCO⁺, H¹³CO⁺ and N₂H⁺ lines with J_up from 6 to 13 [18]. More details on the observations are reported in [15–18].

Table 1.

List of sources targeted by the Herschel observations to search for the presence of energetic particle ionization in warm gas. In the first column the name of the source is reported, in the second its bolometric luminosity. The last three columns report the J_up of the HCO⁺, H¹³CO⁺ and N₂H⁺ observed, but not necessarily detected, lines.

source name	luminosity (Lo)	HCO+ line Jup	H13CO+ line Jup	N2H+ line Jup
VLA1623	1	6, 8, 11, 12	6, 7, 9	6, 8, 11
L1527	2	6, 8, 11, 12	6, 7, 9	6, 8, 11
L1157-mm	11	6, 8, 11, 12	6, 7, 9	6, 8, 11
NGC1333-IRAS2A	16	6, 8, 11	6, 7, 9	6, 8, 11
Serpens-FIRS1	33	6, 8, 11	6, 7, 9	6, 8, 11
L1641-S3-mm1	67	6, 8, 11, 12	6, 7, 9	6, 8, 11
CepE-mm	100	6, 8, 11, 12	6, 7, 9	6, 8, 11
IC1396N	150	6, 8, 11, 12	6, 7, 9	6, 8, 11
NGC7129-FIRS2	500	6, 8, 11	6, 7, 9	6, 8, 11
OMC-2 FIR4	500	6 to 13	6 to 13	6 to 13

The measured HCO⁺, H¹³CO⁺ and N₂H⁺ line fluxes were compared with grids of theoretical predictions obtained with a non-LTE radiative transfer code (specifically, the Large Velocity Gradient code by Ceccarelli et al. [19]) to derive the density and temperature of the gas as well as the column density of HCO⁺ and N₂H⁺, respectively. From the latter, we estimated the HCO⁺/N₂H⁺ abundance ratio for each source of table 1. Figure 2 shows the results of the analysis, namely the derived HCO⁺/N₂H⁺ abundance ratio, with the bars indicating the range of values within 1 sigma, as a function of the bolometric luminosity of each table 1 source. The values of the HCO⁺/N₂H⁺ abundance ratio have relatively large error, because of the relatively small number of observed lines, except in the case of OMC-2 where a total of 17 lines could be detected.

Figure 2.

Measured HCO⁺/N₂H⁺ abundance ratio as a function of bolometric luminosity towards a sample of solar-like protostars (adapted from [16]). The measurements are derived from Herschel observations of high lying (J_up ≥ 6) lines. The names of the targeted sources are marked in the plot. The bars represent the range of values within 1 sigma derived by the analysis for each source (see text). The dashed red lines show the predicted HCO⁺/N₂H⁺ abundance ratio for a gas at 40 K and with a density of 1 × 10⁶ cm⁻³, for a cosmic-ray/energetic-particle ionization rate ζ equal to 5 × 10³ and 10² times the standard one of 1 × 10⁻¹⁷ s⁻¹, as marked in the plot. Note that ζ depends approximately on the square root of the gas density (see text). The code ASTROCHEM² was used to derive the plotted predictions. (Online version in colour.)

In the same plot, we also show the theoretical predictions of the HCO⁺/N₂H⁺ abundance ratio, obtained with the astrochemical code ASTROCHEM (see figure 2 caption) and assuming that the gas has an average temperature of 40 K and density of 1 × 10⁶ cm⁻³. The HCO⁺/N₂H⁺ abundance ratio is predicted to be 10 and 4 for a value of the cosmic-ray/energetic-particle ionization rate ζ equal to 1 × 10⁻¹⁵ and 5 × 10⁻¹⁴ s⁻¹, respectively. Note that increasing/decreasing the density would increase/decrease the latter value of ζ by approximately the density square root because the flux of ionizing particles remains the same while the rate of recombination of the major positive charge carrier (${{\text{H}}_3}^{+}$) increases/decreases with the square root of the density. In other words, the same ζ leads to a predicted HCO⁺/N₂H⁺ abundance ratio approximately proportional to the square root of the gas density.

Based on the discussion of the previous section, we consider that only the sources with HCO⁺/N₂H⁺ ≤ 10 are candidates for the presence of energetic particles. Five sources satisfy this tight criterion:⁴ NGC1333 IRAS2A, L1641-S3-mm1, IC1396N, NGC7128N and OMC-2 FIR4. However, the derived HCO⁺/N₂H⁺ abundance ratio in the first four sources has a large uncertainty so that more observations are necessary to firmly establish the presence of energetic particles in them. On the contrary, the small error bar in OMC-2 FIR4 leads us to conclude that an enhanced flux of ionizing energetic particles is present. We will discuss this source in detail in the next section.

A first result of the study is that of the ten targeted sources only one clearly shows the presence of energetic particles, while four others might do: potentially, the phenomenon of energetic particle irradiation might be a relatively common phenomenon, with 10% as lower limit. A second conclusion is that there is no evidence for the HCO⁺/N₂H⁺ abundance ratio to depend on the source luminosity, in the range 1–500 L_o spanned by our study. However, both conclusions have to be taken with prudence, as the observed sample is obviously very small and not unbiased in terms of sources. Unfortunately, new observations of high J HCO⁺, H¹³CO⁺ and N₂H⁺ lines are not easy to obtain with the presently available facilities. From the ground, the best possibility is offered by ALMA: however, the atmospheric absorption prevents us from observing most of the high J HCO⁺, H¹³CO⁺ and N₂H⁺ lines, leaving insufficient to obtain strong enough constraints. In this context, it is important to identify alternative modes to probe the presence of energetic particles in protostars: we will discuss this point in the next sections.

4. OMC-2 FIR4: a protocluster permeated with energetic particles

OMC-2 FIR4 is a bright FIR source belonging to the Orion Molecular Complex, at a distance of (393 ± 25) pc [20]. The large-scale (greater than or equal to 10′′ equivalent to about 2000 AU) structure of OMC-2 FIR4 was reconstructed from single-dish continuum observations analysed via a self-consistent radiative transfer model [21]. It consists of a roughly spherical envelope of 30 M_o which extends about 1.2 × 10⁴ AU⁵ in radius and has a density that goes from 6 × 10⁵ cm⁻³ at the border of the condensation to 4 × 10⁶ cm⁻³ at 440 AU, where the dust temperature is predicted to reach 100 K. Millimetre interferometric observations have shown the presence of several protostars embedded in this dense envelope [22,23]. The most recent observations, obtained within the IRAM-NOEMA Large Program SOLIS (Seeds Of Life In Space [24]) count eight embedded sources [25], which makes OMC-2 FIR4 a protocluster with a density of approximately 10⁴ sources pc⁻³.

The CHESS Herschel HCO⁺, H¹³CO⁺ and N₂H⁺ lines (table 1) probe the densest and warmest part of the OMC-2 FIR4 envelope [15]. It is approximately constituted of two regions: the first one has an average density of approximately 1 × 10⁶ cm⁻³, temperature of approximately 40 K and radius of approximately 3700 AU, consistent with the values predicted by the continuum modelling above [21]; the second one is four times denser, three times warmer and twice smaller in size, and was not remarked by the continuum analysis. In both regions the derived HCO⁺/N₂H⁺ abundance ratio is very small, between 3 and 4. The comparison with the model predictions obtained with the astrochemical model ASTROCHEM⁶ implies a highly enhanced cosmic-ray/energetic-particle ionization rate ζ, larger than 10³ times the commonly adopted cosmic-ray ionization rate of 1 × 10⁻¹⁷ s⁻¹ (see also figure 2). Given the difference in density, the higher density zone has also a higher ζ (which goes approximately as the square root of the density, as said in the previous section). This leads one to think that the source of particle acceleration is embedded in the OMC-2 FIR 4 envelope and it is not external to it.

Unfortunately, the Herschel observations had a limited spatial resolution (the smallest beam size was 18′′ equivalent to ≈4000 AU) so that they could not provide any information on where the source of accelerated particles lies in the OMC-2 FIR4 envelope. For that, interferometric observations would be necessary. However, this is easier said than done, as the ground-based interferometers are limited by the presence of the atmospheric absorption bands that fall where most HCO⁺, H¹³CO⁺ and N₂H⁺ J_up ≥ 6 lines, unfortunately, lie (see previous section). Fortunately, recent IRAM-NOEMA interferometric observations obtained for different scopes within the Large Program SOLIS [24] allowed us to obtain a finer view of where the energetic particle-emitting sources lie in the OMC-2 envelope [26,27]. The observations covered a band where we detected lines from two cyanopolyynes, HC₃N and HC₅N, and one hydrocarbon, c-C₂H₃. Their analysis, surprisingly, revealed, in both cases, the presence of an enhanced cosmic-ray/energetic-particle ionization rate in the east region of the OMC-2 FIR4 envelope.

Figure 3 shows the 82 GHz continuum map of OMC-2 FIR4 and the distribution of the HC₃N and HC₅N line emission, obtained by the SOLIS observations [26]. While HC₃N lines are detected over the entire FIR4 envelope, probed by the 82 GHz continuum, HC₅N only emits in the eastern part. The derived HC₃N/HC₅N abundance ratio in this eastern region is very low, between 4 and 12, with respect to that expected for a molecular gas with the density and temperature reported above (40–120 K and 1–4 × 10⁶ cm⁻³) and normal characteristics, namely a gaseous C/O elemental abundance lower than one and a cosmic-ray ionization rate of 1 × 10⁻¹⁷ s⁻¹. The only way to reconcile the observations with the Nahoon⁷ astrochemical model predictions is assuming a cosmic-ray/energetic-particle ionization rate ζ more than 10³ times higher.

Figure 3.

SOLIS observations of the 82 GHz continuum emission map (brown colour map and black contours), overlapped on the region of emission from both HC₃N and HC₅N (cyano shadowed), and the region with only HC₃N (blue shadowed) [26]. c-C₂H₃ emits mostly in the HC₅N only region [27]. The triangles show the positions of the three FIR sources of the region: FIR3, FIR4 and FIR5. Details of the HC₃N, HC₅N and c-C₂H₃ observations are reported in [26,27]. (Online version in colour.)

The SOLIS map of the emission from three lines of c-C₂H₃ combined with the analysis of 24 lines of c-C₂H₃ obtained with single-dish observations within the IRAM 30 m Large Program ASAI (Astrochemical Surveys At IRAM [28]) led to similar conclusions [27]. Specifically, the temperature derived from a non-LTE analysis of the 24 lines (40–50 K) and the c-C₂H₃ column (6–8 × 10¹² cm⁻²) can only be reproduced by the Meudon PDR (Photo-Dissociation Region) code⁸ if the region is illuminated by a strong UV field (approx. 1700 times the interstellar one) and a high cosmic-ray/energetic-particle ionization rate ζ, again about 4 × 10⁻¹⁴ s⁻¹. The SOLIS map of the three c-C₂H₃ lines shows that the emission is dominated by the eastern envelope and approximately coincides with the emission from HC₅N (figure 3).

To summarize, three different datasets of five different species, whose two ions are directly created by reactions with ${{\text{H}}_3}^{+}$, analysed by three different codes (and chemical networks) all lead to the same conclusion: the eastern region of the OMC-2 FIR4 envelope is permeated by an intense flux of energetic particles, which enhances the ionization rate by more than one thousand times with respect to the standard rate due to cosmic rays. The source/sources emitting the energetic particles is/are likely embedded in the FIR4 envelope, as the larger the density the higher the derived ζ.

5. Concluding remarks

As mentioned in §1, the overabundance of ¹⁰Be is equivalent to a dose of about 10¹⁹–10²⁰ particles cm⁻² received by the rocks incorporated in the meteoritic material [6]. As discussed in §4, the energetic-particle ionization rate ζ measured in OMC-2 FIR4 is higher than about 4 × 10⁻¹⁴ s⁻¹ [15,26,27]. This ionization rate can be translated into a flux of ionizing particles, assuming an energy spectrum of the impinging particles, computing the energy losses due to their penetration throughout the gas and accounting for the cross section of the particle ionization at a given particle energy (see [15] for details). If we then consider that the energetic particles are emitted by a single source, the ζ measured in OMC-2 FIR4 would correspond to a flux of 1–3 × 10¹⁹ particles cm⁻² yr⁻¹ hitting a target at 1 AU distance from the emitting source (where the particle flux scales by the square of the distance from the emitting source). Therefore, a flaring of 1–10 years of the particle-emitting source would be enough to account for the dose experienced by the young Solar System, if the Sun was the source. One could use the, admittedly poor, statistics of §3 to have an order of magnitude of the flaring rate/duration. Assuming that a protostar spends about 10% of its early life in a flaring status, this would yield an integrated dose more than 100 times larger than that received by the meteoritic material. In other words, the ‘impacted’ rocks could have been more distant than 1 AU from the particle accelerating source(s). Thus, in order to unveil the mystery of the energetic particles in the early Solar System, two related questions need to be answered: where and how are the particles accelerated?

Various theories have been invoked for the Solar System [2–5,29,30]. One strong constraint put by the observations presented here is that the energetic particles travel for more than ≈4000 AU: therefore, they are not confined to the equatorial plane of a circumstellar disc, as some models assume [4,6,30]. Recently, new theories have been put forward to explain the case of OMC-2 FIR4 [31,32]. First, a promising possibility is that the energetic particles are accelerated in one or more shocks caused by the supersonically outflowing gas crashing into the quiescent surrounding one [31]. In favour of this possibility, the cosmic-ray/energetic-particle ionization rate ζ in the outflow shock L1157-B1 is ≈3 × 10⁻¹⁶ s⁻¹, namely ten times larger than standard one [33].

Another promising explanation is that the energetic particles are accelerated at the accretion shock, therefore at a smaller scale but from a more intense shock than the outflow ones [32]. Specifically, a protostar with a present mass equal to 10 M_o and that will eventually become a 20 M_o star would produce an amount of cosmic-ray/energetic-particle ionization rate as that measured by the observations reported here. The radio emission detected towards OMC-2 FIR4 in the eastern part [23] would be in agreement with this hypothesis, as remarked by [32]. Another possibility is that more than one source is accelerating the particles. Gaches & Offner [32] computed the average ionization rate ζ in the case of a cluster of protostars of lower mass and found that it would increase with increasing number of protostars in the cluster. In their study, the distance between protostars is a parameter self-consistently derived from a generic model of the cluster. We, thus, would argue that the observed high density of protostars in OMC-2 FIR4 might also provide an explanation to the measured ζ.

Finally, most of the studies about the possible earliest phases of the Solar System have been so far carried out towards low-luminosity protostars in relatively isolated and quiet environments, like in the case of IRAS16293-2422, the prototype of low-mass Class 0 sources [34–36]. However, many independent pieces of evidence show that the Solar System was born in a crowded environment, where the density was about 2–25 × 10⁴ sources pc⁻³ [37]. In this respect, therefore, OMC-2 FIR4 shares a first property with the Sun cradle protocluster: a high density, ≈10⁴ sources pc⁻³, of forming stars. A phase of intense irradiation from energetic particles is the second important property shared by OMC-2 FIR4 and the Solar System cradle. Our final comment is that, waiting for the discovery of other sources bathed in energetic particles, studies that compare cometary and meteoritic with present protostellar material should focus their attention towards the protostars of OMC-2 FIR4, which are very likely more similar to our earliest Solar System.

Data accessibility

This article has no additional data.

Authors' contributions

All authors contributed to the reported work, in the acquisition, reduction, analysis and interpretation of the data, as well as the writing of the manuscript.

Competing interests

We declare we have no competing interests.

Funding

This project has received funding from the European Research Council (ERC) under the European Union's Horizon 2020 research and innovation programme, for the project ‘The Dawn of Organic Chemistry’ (DOC), grant agreement no. 741002. This work is supported by the French National Research Agency in the framework of the Investissements d'Avenir programme (ANR-15-IDEX-02), through the funding of the ‘Origin of Life’ project of the Univ. Grenoble-Alpes.