- 1Centro de Astrobiología (CAB), CSIC-INTA, Madrid, Spain
- 2Grupo Interdisciplinar de Sistemas Complejos (GISC), Madrid, Spain
- 3Instituto de Investigación Tecnológica (IIT), Universidad Pontificia Comillas, Madrid, Spain
Phosphorus-bearing molecules are fundamental for life on Earth, yet their astrochemical origins remain poorly understood. Their formation in the interstellar medium has been challenging to elucidate due to limited spectroscopic detections and the reliance on theoretical models that depend on numerous kinetic parameters whose values are very uncertain. Multi-parameter models often suffer from “sloppiness”, where many parameter combinations exhibit negligible influence on model outcomes, while a few dominate system behavior. In this study, we introduce the Fisher Information Spectral Reduction (FISR) algorithm, a novel and computationally efficient method to reduce the complexity of such sloppy models. Our approach exposes the strong parameter hierarchy governing these systems by identifying and eliminating parameters associated with insensitive directions in the parameter space. Applying this methodology to the phosphorus astrochemistry network, we reduce the number of reaction rate coefficients from 14 to 3, pinpointing the key reactions and kinetic parameters responsible for forming PO and PN, the main phosphorus-bearing molecules typically detected in interstellar space. The simplified model retains its predictive accuracy, offering deeper insights into the mechanisms driving phosphorus chemistry in the interstellar medium. This methodology is applicable to multi-parameter models of any kind and, specifically in astrochemistry, facilitates the development of simpler, more realistic and interpretable models to effectively guide targeted observational efforts.
1 Introduction
Phosphorus (P) is an element of significant astrobiological importance due to its abundance in biomass and its critical role in essential biochemical functions: the backbone of nucleic acids contains sugar-phosphates, phosphorylated molecules act as energy carriers, and cellular membranes contain phospholipid head groups (Walton et al., 2023). Interestingly, despite its ubiquity in life on Earth, phosphorus is far less abundant on cosmic scales than other essential elements for life, such as hydrogen (H), carbon (C), oxygen (O), and nitrogen (N) —a phenomenon often referred to as the phosphorus enigma. In fact, phosphorus ranks 18th in cosmic abundance, orders of magnitude lower than other biogenic elements (Maciá-Barber, 2020). With regard to the origin of terrestrial phosphorus, it is worth noting that the chemical complexity of early Earth was augmented by the influx of molecular compounds inherited from its parental molecular cloud during the process of planetary formation and the subsequent impact of asteroids and comets on its surface. Interplanetary dust, meteorites, and large impactors may have deposited reactive, reduced oxidation state phosphorus onto early Earth’s surface. This process, supplemented by terrestrial P-reduction pathways, enriched prebiotic environments with reactive P, including organic phosphonates with carbon–phosphorus bonds, as identified in the Murchison meteorite (Cooper et al., 1992). Furthermore, the recent detection of PO in comet 67P/Churyumov–Gerasimenko (Rivilla et al., 2020) and of phosphorus-rich grains in asteroid Ryugu samples (Pilorget et al., 2024), indicates that minor Solar System bodies can act as reservoirs for phosphorus-bearing compounds. Therefore, the detection of phosphorus sources in the interstellar medium (ISM) is key to trace the evolution of phosphorus during the Solar System formation and unveil how this element became available for the origin of life.
In recent decades, the interstellar chemistry of phosphorus has emerged as a promising area of research driven by high-sensitivity instrumentation such as the Atacama Large Millimeter/submillimeter Array (ALMA). To date, more than 300 molecules have been detected in interstellar and planetary environments, including complex organic molecules (COMs) with potential prebiotic significance. Despite this impressive detection capacity, only seven P-bearing molecules have been unambiguously identified in the ISM: PO (Tenenbaum et al., 2007; Rivilla et al., 2016),
Besides the discovery of new chemical species in the ISM through spectroscopic techniques, astrochemistry also relies on laboratory experimental work and computational models. Laboratory experiments replicate interstellar physical conditions to obtain rotational spectra for known molecular compositions. Computational models, on the other hand, integrate data from astronomical measurements and laboratory experiments, providing a framework to hypothesize the underlying physical-chemical processes responsible for the observed molecular abundances. In this context, astrochemical models such as UCLCHEM (Holdship et al., 2017) or Nautilus (Ruaud et al., 2016) have been developed to simulate the time evolution of abundances for numerous interstellar species, which interact through extensive networks of chemical reactions under various physical conditions and energetic physical processes (e.g., UV/cosmic rays irradiation, protostellar heating or stellar feedback in the form of high-velocity winds inducing shock waves) in both the gas phase and on dust grain surfaces. These models typically consist of applying the law of mass action to all reactions in the chemical network and solving numerically the associated systems of ordinary differential equations. However, the precision of mass-action kinetics depends heavily on the accuracy of the reaction rate coefficients, which remain poorly determined for the majority of reactions.
In the past decade, it has become clear that astrochemical models present severe limitations in reproducing the [PO]/[PN] ratios observed across the ISM (predicted ratios
Sloppy models are prime candidates for model reduction techniques. In general, these techniques aim to simplify large-scale dynamical systems through the elimination of unnecessary parameters while preserving essential characteristics (Antoulas, 2005). The primary techniques can be categorized into (i) eigenvector projection methods (using Singular Value Decomposition) and (ii) moment matching methods. Some eigenvector projection methods diagonalize the system and truncate those states that are difficult to control and observe (Moore, 1981), while others aim to minimize an error measure of the system (Glover, 1984) or separate fast and slow timescales (Gugercin and Antoulas, 2004). On the contrary, moment-based methods focus on reducing the complexity while approximating accurately some statistics (often central moments estimated from the data). Overall, both categories of methods strive to achieve a balance between computational efficiency and accuracy, making them suitable for large-scale models but relying too strongly on linearization or on global measurements. Alternatively, M.K. Transtrum and collaborators, following a different approach, developed a decade ago the Manifold Boundary Approximation Method (MBAM) (Transtrum et al., 2011; Transtrum and Qiu, 2014; Transtrum and Qiu, 2016; Quinn et al., 2023). This elegant algorithm is based on information geometrical arguments, balancing the data with the complexity of the model exploiting the low sensitivity to some combinations of parameters.
In this paper, we introduce a methodology for reducing and simplifying models while retaining their predictive accuracy inspired by the MBAM reduction technique and apply it specifically to the phosphorus astrochemistry network. The proposed method, the Fisher Information Spectral Reduction method (hereafter, FISR method), is conceptually simpler than the MBAM and thus can be implemented by an easier and more efficient algorithm. It performs iterative dimensional reduction steps, progressively decreasing the number of model parameters. The final outcome is a reduced model with significantly fewer parameters, in which unnecessary complexity is removed.
Applying the FISR algorithm to the 14-reaction model of phosphorus chemistry in molecular clouds presented in Fernández-Ruz et al. (2023), we demonstrate that the observed abundances of PO and PN can be explained by a much simpler model comprising just 3 reactions and 3 parameters. This simplified model not only identifies the key reactions governing PO and PN formation but also offers a deeper comprehension of the hierarchical structure of the parameter space. Therefore, the goal of this work is not to obtain a model that improves the predictions of existing astrochemical models, but rather to derive simpler and interpretable models that maintain comparable predictive accuracy while eliminating unnecessary complexities. Our findings highlight that the complexity of a model should correspond to the complexity of the available knowledge with which it is constructed. This principle—also known in the sciences as Occam’s razor—, is broadly applicable beyond astrochemistry to other disciplines reliant on modeling.
The organization of this paper is as follows. In Section 2 we present the mathematical foundation of the complexity reduction method introduced here, the FISR method, and outline the algorithmic steps for its general implementation. In Section 3 we apply the FISR method to the phosphorus chemistry in the interstellar medium, uncovering the underlying chemical network dynamics that governs the formation of PO and PN across short and long timescales. In Section 4, we discuss the implications of the main findings for future phosphorus astrochemical research and examine the strengths and limitations of the FISR method within the context of complexity reduction methods.
2 Methods: the FISR algorithm
The method presented in this work is inspired by the Manifold Boundary Approximation Method (MBAM), originally developed and published by Transtrum and Qiu (2014), and shares its basic principles and primary objective. The reasoning behind this method is that models often do not respond significantly to changes in certain parameters. By applying concepts from information geometry, this approach identifies specific trajectories across the statistical manifold that represent meaningful combinations of the parameters. These trajectories follow the so-called sloppy directions, which are directions in the parameter space that are not informative and along which large parameter changes produce only a minimal impact on the model output. The term boundary in the name of the method connects with the practical observation made by the authors that, except in rare pathological cases, the topology of an
Our proposed algorithm, the Fisher Information Spectral Reduction (FISR) method, successfully achieves the same goals as MBAM but differs significantly from it in its technical details and algorithmic procedures. We name it FISR method because it is based on the eigenspectrum of the Fisher Information Matrix (FIM), which quantifies the amount of information a dataset carries about the model parameters, but also provides the directions on the statistical manifold carrying less information (the sloppiest directions). Since high information content is associated with strong parameter influence, the eigenspectrum of the FIM encodes the sensitivity of model outputs to changes in parameter values. In this Section, we present the mathematical framework of the FISR method and outline its algorithmic implementation from a general perspective, making it applicable to any sloppy model.
2.1 Theoretical framework and definitions
Let us consider a model
In addition, we define the cost function as
The cost quantifies how far the outputs obtained with parameter set
Finally, the Fisher Information Matrix (FIM)
In matrix form, the FIM and the Jacobian are related through
In cases where the parameters are positive by definition and/or their scales are significantly different, it will be very helpful to calculate
In practice, sloppy models are those whose FIM eigenvalues span along many orders of magnitude, from eigenvalues close to zero, whose associated eigenvectors pinpoint the parameter directions that hardly affect the system—the sloppy directions—, to large eigenvalues, whose eigenvectors denote directions of high parameter sensitivity. We will exploit this feature of the FIM for parameter reduction in the next Subsection.
2.2 Parameter reduction algorithm
The FISR algorithm is implemented through successive dimensionality reduction steps, reducing the original model iteratively until no further simplifications are possible. Each reduction step takes an initial model as the input and produces a simplified version as the output. This iterative process is carried out by an algorithm structured in the following three stages:
Stage 1. Navigation through the parameter space
Considering an initial model
where
Stage 2. Limit evaluation and generation of a new model
As mentioned above, stage 1 ends when one or more
(i) A parameter approaches zero in stage 1. This may indicate that the microscopic process it represents has become negligible in the system with the new values of the parameters, for example, because it is much slower than the rest of the processes. In such cases, the new model excludes this process and the corresponding parameter.
(ii) A parameter approaches infinity in stage 1. This is typical when the parameters are reaction rate coefficients, and implies that a mechanism is almost instantaneous in comparison to others. This provides an adiabatic elimination of the fast timescale, and this mechanism can now be described by a single process with the proper change in the initial conditions.
(iii) Two or more parameters tend to infinity or one to infinity while the other to zero in stage 1. Here, a new parameter that combines them but remains finite is introduced in the model.
Note, however, that more complicated scenarios are possible, and more than two parameters may tend to zero or infinity in the same reduction step. In any case, the interpretation of the limit to generate the new model is a task that cannot be automated and must be done by hand because it requires knowledge about the processes that are modeled and the physical meaning of the parameters. Note that the new model is now evaluated in the parameter set
Stage 3. Fitting of the new model
Once the new model is defined, it is fitted by finding the parameter set
Consecutive dimensional reduction steps. The whole process of parameter reduction starts with the original model

Figure 1. Structure of the FISR algorithm to reduce model complexity. The algorithmic flow can be divided in three stages: (stage 1) navigation through the parameter space of the initial model, constructing a path
3 Results
In this Section, we apply the FISR method introduced in this work to a phosphorus astrochemistry model comprising 14 chemical reactions involving 7 P-bearing molecules proposed by Fernández-Ruz et al. (2023). The parameters subject to reduction are the kinetic parameters, specifically the rate coefficients
3.1 Original model for the chemical evolution of phosphorus in the ISM
We adopt the phosphorus chemistry model for the ISM from our recent publication (Fernández-Ruz et al., 2023) as the original model of the complexity reduction process. We opt for this model because it has been proven to accurately describe the [PO]/[PN] ratios of the whole phosphorus network predicted by UCLCHEM, while remaining sufficiently small to apply the FISR method. The system, whose associated chemical network is shown in Figure 2, describes the evolution of phosphorus chemistry in a star-forming region over

Figure 2. Bipartite chemical network representing the 14 chemical reactions involved in the phosphorus chemistry in the interstellar medium studied in this work. Orange circular nodes correspond to P-bearing species, whose abundance evolves over time as described by Equation 6. The rest of the species are represented as blue circular nodes. Square nodes (labeled R1–R14) denote chemical reactions, with directed links pointing from reactants to reaction nodes, and from reaction nodes to products.

Table 2. Reactions and their rate coefficients. (a) The rate coefficients of these reactions were adjusted with Bayesian inference in Fernández-Ruz et al. (2023). (b) The Arrhenius parameters of these rate coefficients come from theoretical quantum chemical calculations.

Table 3. Initial gas-phase abundances that apply in the original model, extracted from Fernández-Ruz et al. (2023). Some initial abundances are missing because they are not needed as they do not appear in the models’ equations. (a) In cases where the source provided two values or we considered two sources, we used the geometric mean. (b) Up to date, CP has not been detected in the ISM, but it has been detected in a circumstellar shell envelope by Guelin et al. (1990). Thus, in our model we consider that CP is present but we fix its initial value to
We refer to the system of ordinary differential equations (ODEs) with 17 variables, derived from applying the law of mass action to the astrochemical model plotted in Figure 2, as the total system. This system is mathematically intractable and can only be solved using numerical methods. However, the linear effective system is a linearized approximation of the full system that can be theoretically solved under certain approximations for the 7 P-bearing species. The linearization is possible when the non P-bearing species are treated as constants, an assumption justified by their significantly higher abundances, as shown in Table 3 (see Supplementary Section S1.1 for details). The linear effective system solution consists of explicit equations for the time-evolving abundances:
where
The sensitivity analysis presented in Fernández-Ruz et al. (2023) revealed that the abundances of PO and PN at times
3.2 Toward the simplest model compatible with the chemical evolution of phosphorus in the ISM
3.2.1 Dimensionality reduction steps
Here, we apply the FISR algorithm to the original model of the phosphorus chemistry in the ISM through a series of iterative dimensionality reduction steps. As explained in Section 2.2, each reduction step

Figure 3. Evolution with
After each reduction step, a new model is built, and this process is repeated along 9 dimensionality reduction steps. Figure 4 shows the time evolution of the abundances for the P-bearing species of every new model

Figure 4. Evolution with time of the abundances relative to H of the P-bearing molecules for models 1 to 9 (solid lines), obtained in stage 3 of the FISR algorithm, and for the original model (O.M., dashed lines; Fernández-Ruz et al., 2023) of the phosphorus astrochemistry in the ISM. Models 1–9 (panels A–I) are progressively simpler approximations of the original model. Although only abundances of PO and PN beyond
As we can see in Figures 3A–C, corresponding to reduction steps 1, 2 and 3, respectively, the reaction rate coefficients
Figure 3D shows that the rate coefficient of reaction 8 (N + CP
In reduction step 5, the rate coefficient of reaction 12 (H+
Similarly to the former two steps, in the reduction step 6 (Figure 3F), the rate coefficient
The reduction step 7 is described in Figure 3G, where reaction rate coefficient
Interestingly, the reduction step 8 shown in Figure 3H involves 3 parameters. The rate coefficients of reactions 4 (O + PH
The last iteration of the FISR algorithm, reduction step 9, is shown in Figure 3I. Reaction rate coefficient
3.2.2 Predictive power and parameter sloppiness of the reduced models
As shown in the previous Subsection, at least one parameter tends toward an extreme value during each successive navigation through the multi-dimensional parameter space. However, it is important to note that the remaining parameters also undergo minor changes. This occurs because the vector

Figure 5. Evolution of the parameters, cost, and Fisher Information Matrix spectrum throughout the 9 dimensional reduction steps of the FISR algorithm applied to the phosphorus astrochemical system in the ISM. (A) Values of the rate coefficients at the parameter navigation (P.N.) and fitting of the new model (Fit.) (stages 1 and 3 of the FISR algorithm, respectively) for each reduction step. (B) Cost of the corresponding models after the parameter navigation stage and the fitting stage of the FISR algorithm. A dashed horizontal line is displayed at cost
Figure 5B shows the cost—calculated using Equation 2— after each parameter navigation and parameter fitting step. Since the cost associated with model
Finally, Figure 5C displays the eigenspectrum of each model, i.e., the set of eigenvalues of the Fisher Information Matrix
3.3 Final model for the chemical evolution of phosphorus in the ISM
The FISR method applied in this study enabled the reduction of the original 14-reaction model to significantly simpler models that still accurately reproduce the chemical evolution of PO and PN in the ISM beyond
Model 8 was constructed through eight consecutive reduction steps. Interestingly, while its dynamics is derived from reactions 1, 2, and 6 with rate coefficients almost identical to those in the original model, the overall output is not solely explained by the activity of these reactions. It also incorporates the influence of reactions 4, 8, 10, 12, 13, and 14, all treated as infinitely fast, through modifications in the initial conditions of PH, P, PO, and PN in model 8. These adjustments account for the reactants transformed into products at much faster rates than the remaining relevant reactions. Essentially, these fast processes involved the cascade of

Figure 6. Schematic representation of the evolution of phosphorus through the chemical processes leading to the formation of PO and PN in the ISM across two distinct time stages. During the early stage (up to
With one additional reduction step, reaction 1 is eliminated. Model 9 represents the direct transformation of P into PO and subsequently into PN, but with an associated cost of 0.55 that is
4 Discussion
In this work, we introduce a novel dimensionality reduction technique called the Fisher Information Spectral Reduction (FISR) method. It measures the sensitivity of the output on the parameters in any multi-parameter model, and we used it to reduce the complexity of a system describing the chemical network of phosphorus in the interstellar medium. This P-chemistry model, previously studied in Fernández-Ruz et al. (2023), demonstrates a strong parameter hierarchy, with high dependence on a few parameter directions, while exhibiting a pronounced insensitivity to the rest—commonly referred to as stiff and sloppy directions, respectively. Using the FISR method, we have exploited this property showing that a kinetic model with 3 chemical reactions (N + PO
We began this article by outlining the various challenges that phosphorus astrochemistry and its connection to the origin of life on Earth still present. Let us now examine how the drastic simplification of the phosphorus astrochemical network to its fundamental core, as presented here, enables a deeper understanding of its chemistry. Despite the minimal model 8 having only 3 rate coefficients as inputs, it accounts for the effects of other reactions without explicitly including their rate coefficients in the equations. The reduction steps in which one or more rate coefficients tend to infinity are interpreted as the corresponding reactions being instantaneous, and this translates into appropriate changes in the initial conditions. As shown in Figure 4, in the original model (dashed lines), by
It is worth noting that, although the FISR algorithm was applied using rate coefficients calculated for T = 100 K, the resulting reduced models remain good approximations (i.e., with low associated cost) over a certain temperature range. This is due to the system’s inherent sloppiness, which confers robustness to variations in the rate coefficients and, consequently, to changes in temperature. In Supplementary Section S6, we analyze the generalization of models 1–9 —originally obtained at T = 100 K—to temperatures ranging from 40 K to 160 K, without reapplying the full algorithm at each temperature. This robustness is advantageous, as it removes the need to re-run the computationally expensive FISR algorithm for every temperature. We find that the dynamics of phosphorus chemistry in the 40–160 K range can still be interpreted as a two-stage process, being the approximation of the early stage as instantaneous particularly accurate at higher temperatures. At very low temperatures, applying the full FISR method becomes necessary to obtain precise predictions. Accordingly, Supplementary Section S8 shows the costs of models 1–9 at T = 40, 70, 100, 130, and 160 K.
Relative to other dimensionality reduction methods currently available in the literature, the FISR method proposed here is significantly simpler. Specifically, in comparison to the MBAM algorithm (Transtrum and Qiu, 2014), upon which it is based, the FISR algorithm employs a streamlined methodology that is easier to implement, conceptually simpler, and admits larger output vectors (see Supplementary Section S7). Nevertheless, it is important to note that neither the FISR method nor the MBAM is currently suitable for application to very large systems due to several limitations: (i) both are based on a streamlined algorithmic structure that cannot be parallelized, (ii) each point of the trajectory navigating the parameter space requires the diagonalization of the Fisher Information Matrix computed numerically from the solution of the ODEs—in the phosphorus network, we were able to solve the model analytically, but this will not generally be feasible—and (iii) manual analysis of each reduced model is required at the end of each reduction step. In summary, it is crucial to continue advancing this promising line of research by developing new algorithms capable of analyzing more complex systems. Such advancements will enable a comprehensive study of the chemistry of the interstellar medium, where we believe parameter sloppiness is the rule rather than the exception.
Data availability statement
The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author.
Author contributions
MF-R: Conceptualization, Formal Analysis, Investigation, Methodology, Software, Writing – original draft, Writing – review and editing, Data curation, Visualization. IJ-S: Conceptualization, Investigation, Methodology, Writing – original draft, Writing – review and editing. MC: Conceptualization, Investigation, Methodology, Writing – original draft, Writing – review and editing, Formal Analysis, Software. MR-B: Investigation, Writing – original draft, Writing – review and editing, Visualization. JA: Investigation, Writing – original draft, Writing – review and editing, Conceptualization, Formal Analysis, Funding acquisition, Methodology, Project administration, Software, Supervision.
Funding
The author(s) declare that financial support was received for the research and/or publication of this article. JA and MF-R received support from grant No. PID2021-122936NB-I00, MC from grant No. PID2022-140217NB-I00, MR-B from grant PID2022-140180OB-C22, and IJ-S from grant PID2022-136814NB-I00, all of them funded by the Spanish Ministry of Science and Innovation/State Agency of Research MCIN/AEI/10.13039/501100011033 and by “ERDF A way of making Europe''. JA acknowledges financial support from grant No. PIE2024ICT085 and IJ-S from grant No. PIE 202250E155, both of them funded by the Spanish National Research Council (CSIC). MR-B is supported by the Instituto Nacional de Técnica Aeroespacial ”Esteban Terradas'' (INTA). MF-R is supported by a predoctoral contract from INTA. Authors benefited from the interdisciplinary framework provided by CSIC through the “LifeHUB.CSIC” initiative (PIE 202120E047-Conexiones-Life). IJ-S also acknowledges funding from the ERC grant OPENS (project number 101125858) funded by the European Union. Views and opinions expressed are however those of the authors only and do not necessarily reflect those of the European Union or the European Research Council Executive Agency. Neither the European Union nor the granting authority can be held responsible for them.
Acknowledgments
The authors are indebted to S. Ares, J. A. Cuesta and M. Martínez-Jiménez for fruitful comments.
Conflict of interest
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
The reviewer FF declared a past co-authorship with the author(s) IJ-S to the handling editor.
Generative AI statement
The authors declare that Generative AI was used in the creation of this manuscript exclusively to review the English precision of some sentences in the manuscript.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fspas.2025.1570525/full#supplementary-material
Footnotes
1Note that with this criterion, the orientation of
2From now on, we will refer to model
3Recall that the non P-bearing species are assumed to be constant due to their higher abundances and the P-bearing species become the only dynamical variables. Therefore, for all the cases in this Subsection, we only consider changes in the initial conditions of the P-bearing species, as the changes in the non P-bearing species are negligible.
References
Agúndez, M., Cernicharo, J., and Guélin, M. (2007). Discovery of phosphaethyne (HCP) in space: phosphorus chemistry in circumstellar envelopes. Astrophysical J. Lett. 662, L91–L94. doi:10.1086/519561
Agúndez, M., Cernicharo, J., Pardo, J. R., Guélin, M., and Phillips, T. G. (2008). Tentative detection of phosphine in IRC +10216. Astronomy Astrophysics 485, L33–L36. doi:10.1051/0004-6361:200810193
Agúndez, M., and Wakelam, V. (2013). Chemistry of dark clouds: databases, networks, and models. ChRv 113, 8710–8737. doi:10.1021/cr4001176
Aota, T., and Aikawa, Y. (2012). Phosphorus chemistry in the shocked region L1157 B1. Astrophysical J. 761, 74. doi:10.1088/0004-637X/761/1/74
Bernal, J. J., Koelemay, L. A., and Ziurys, L. M. (2021). Detection of PO in orion-KL: phosphorus chemistry in the plateau outflow. Astrophysical J. 906, 55. doi:10.3847/1538-4357/abc87b
Brown, K. S., and Sethna, J. P. (2003). Statistical mechanical approaches to models with many poorly known parameters. Phys. Rev. E 68, 021904. doi:10.1103/physreve.68.021904
Chantzos, J., Rivilla, V. M., Vasyunin, A., Redaelli, E., Bizzocchi, L., Fontani, F., et al. (2020). The first steps of interstellar phosphorus chemistry. Astronomy Astrophysics 633, A54. doi:10.1051/0004-6361/201936531
Charnley, S. B., and Millar, T. J. (1994). The chemistry of phosphorus in hot molecular cores. MNRAS 270, 570–574. doi:10.1093/mnras/270.3.570
Cooper, G. W., Onwo, W. M., and Cronin, J. R. (1992). Alkyl phosphonic acids and sulfonic acids in the murchison meteorite. Geochimica Cosmochimica Acta 56, 4109–4115. doi:10.1016/0016-7037(92)90023-C
de la Concepción, J. G., Cavallotti, C., Barone, V., Puzzarini, C., and Jiménez-Serra, I. (2024). Relevance of the P+O2 reaction for PO formation in astrochemical environments: electronic structure calculations and kinetic simulations. Astrophysical J. 963, 142. doi:10.3847/1538-4357/ad1ffa
Douglas, K. M., Gobrecht, D., and Plane, J. M. C. (2022). Experimental study of the removal of excited state phosphorus atoms by H2O and H2: implications for the formation of PO in stellar winds. Mon. Notices R. Astronomical Soc. 515, 99–109. doi:10.1093/mnras/stac1684
Faro, J., Von Haeften, B., Gardner, R., and Faro, E. (2019). A sensitivity analysis comparison of three models for the dynamics of germinal centers. Front. Immunol. 10, 2038. doi:10.3389/fimmu.2019.02038
Fernández-Ruz, M., Jiménez-Serra, I., and Aguirre, J. (2023). A theoretical approach to the complex chemical evolution of phosphorus in the interstellar medium. Astrophysical J. 956, 47. doi:10.3847/1538-4357/acf290
Fontani, F. (2024). Observations of phosphorus-bearing molecules in the interstellar medium. Front. Astronomy Space Sci. 11, 1451127. doi:10.3389/fspas.2024.1451127
Fontani, F., Mininni, C., Beltrán, M. T., Rivilla, V. M., Colzi, L., Jiménez-Serra, I., et al. (2024). The GUAPOS project: G31.41+0.31 Unbiased ALMA sPectral Observational Survey. IV. Phosphorus-bearing molecules and their relation to shock tracers. Astronomy Astrophysics 682, A74. doi:10.1051/0004-6361/202348219
Fontani, F., Rivilla, V. M., Caselli, P., Vasyunin, A., and Palau, A. (2016). Phosphorus-bearing molecules in massive dense cores. Astrophysical J. Lett. 822, L30. doi:10.3847/2041-8205/822/2/L30
Furuya, K., and Shimonishi, T. (2024). Deep search for phosphine in a prestellar core. Astrophysical J. Lett. 968, L19. doi:10.3847/2041-8213/ad50cc
García de la Concepción, J., Puzzarini, C., Barone, V., Jimenez-Serra, I., and Roncero, O. (2021). Formation of phosphorus monoxide (PO) in the interstellar medium: insights from quantum-chemical and kinetic calculations. Astrophysical J. 922, 169. doi:10.3847/1538-4357/ac1e94
Glover, K. (1984). All optimal Hankel-norm approximations of linear multivariable systems and their L-error bounds. Int. J. control 39, 1115–1193. doi:10.1080/00207178408933239
Goldsmith, P. F., Liseau, R., Bell, T. A., Black, J. H., Chen, J.-H., Hollenbach, D., et al. (2011). Herschel measurements of molecular oxygen in Orion. Astrophysical J. 737, 96. doi:10.1088/0004-637x/737/2/96
Gomes, A. C. R., Souza, A. C., Jasper, A. W., and Galvão, B. R. L. (2023). The P(4S) + NH(3Σ–) and N(4S) + PH(3Σ–)reactions as sources of interstellar phosphorus nitride. Astronomical Soc. Aust. 40, e011. doi:10.1017/pasa.2023.13
Guélin, M., Cernicharo, J., Paubert, G., and Turner, B. E. (1990). Free CP in IRC +10216. Astronomy Astrophysics 230, L9–L11.
Gugercin, S., and Antoulas, A. C. (2004). A survey of model reduction by balanced truncation and some new results. Int. J. Control 77, 748–766. doi:10.1080/00207170410001713448
Halfen, D. T., Clouthier, D. J., and Ziurys, L. M. (2008). Detection of the CCP radical (X2 Π r) in IRC +10216: a new interstellar phosphorus-containing species. Astrophysical J. 677, L101–L104. doi:10.1086/588024
Holdship, J., Viti, S., Jiménez-Serra, I., Makrymallis, A., and Priestley, F. (2017). UCLCHEM: a gas-grain chemical code for clouds, cores, and C-shocks. Astronomical J. 154, 38. doi:10.3847/1538-3881/aa773f
Jiménez-Serra, I., Viti, S., Quénard, D., and Holdship, J. (2018). The chemistry of phosphorus-bearing molecules under energetic phenomena. Astrophysical J. 862, 128. doi:10.3847/1538-4357/aacdf2
Lagarias, J., Reeds, J., Wright, M., and Wright, P. (1998). Convergence properties of the nelder--mead simplex method in low dimensions. SIAM J. Optim. 9, 112–147. doi:10.1137/S1052623496303470
Larsson, B., Liseau, R., Pagani, L., Bergman, P., Bernath, P., Biver, N., et al. (2007). Molecular oxygen in the ρ Ophiuchi cloud. A&A 466, 999–1003. doi:10.1051/0004-6361:20065500
Lefloch, B., Codella, C., Montargès, M., Vastel, C., Podio, L., Viti, S., et al. (2024). Spatial distributions of PN and PO in the shock region L1157-B1. Astronomy Astrophysics 687, A75. doi:10.1051/0004-6361/202245338
Lefloch, B., Vastel, C., Viti, S., Jimenez-Serra, I., Codella, C., Podio, L., et al. (2016). Phosphorus-bearing molecules in solar-type star-forming regions: first PO detection. MNRAS 462, 3937–3944. doi:10.1093/mnras/stw1918
Maciá-Barber, E. (2020). The chemical evolution of phosphorus: an interdisciplinary approach to astrobiology. Oakville, Ontario, Canada: Apple Academic Press.
Moore, B. C. (1981). Principal component analysis in linear systems: controllability, observability, and model reduction. IEEE Trans. Automatic Control 26, 17–32. doi:10.1109/tac.1981.1102568
Pilorget, C., Baklouti, D., Bibring, J.-P., Brunetto, R., Ito, M., Franchi, I., et al. (2024). Phosphorus-rich grains in Ryugu samples with major biochemical potential. Nat. Astron. 8, 1529–1535. doi:10.1038/s41550-024-02366-w
Quinn, K. N., Abbott, M. C., Transtrum, M. K., Machta, B. B., and Sethna, J. P. (2023). Information geometry for multiparameter models: new perspectives on the origin of simplicity. Rep. Prog. Phys. 86, 035901. doi:10.1088/1361-6633/aca6f8
Rivilla, V. M., Drozdovskaya, M. N., Altwegg, K., Caselli, P., Beltrán, M. T., Fontani, F., et al. (2020). ALMA and ROSINA detections of phosphorus-bearing molecules: the interstellar thread between star-forming regions and comets. MNRAS 492, 1180–1198. doi:10.1093/mnras/stz3336
Rivilla, V. M., Fontani, F., Beltrán, M. T., Vasyunin, A., Caselli, P., Martín-Pintado, J., et al. (2016). The first detections of the key prebiotic molecule PO in star-forming regions. Astrophysical J. 826, 161. doi:10.3847/0004-637X/826/2/161
Rivilla, V. M., García De La Concepción, J., Jiménez-Serra, I., Martín-Pintado, J., Colzi, L., Tercero, B., et al. (2022). Ionize hard: interstellar PO+ detection. Front. Astronomy Space Sci. 9, 829288. doi:10.3389/fspas.2022.829288
Rivilla, V. M., Jiménez-Serra, I., Zeng, S., Martín, S., Martín-Pintado, J., Armijos-Abendaño, J., et al. (2018). Phosphorus-bearing molecules in the galactic center. MNRAS Lett. 475, L30–L34. doi:10.1093/mnrasl/slx208
Ruaud, M., Wakelam, V., and Hersant, F. (2016). Gas and grain chemical composition in cold cores as predicted by the Nautilus three-phase model. Mon. Notices R. Astronomical Soc. 459, 3756–3767. doi:10.1093/mnras/stw887
Rugel, M. R., Beuther, H., Bihr, S., Wang, Y., Ott, J., Brunthaler, A., et al. (2018). OH absorption in the first quadrant of the Milky Way as seen by THOR. Astronomy Astrophysics 618, A159. doi:10.1051/0004-6361/201731872
Scibelli, S., Megías, A., Jiménez-Serra, I., Shirley, Y., Bergner, J., Ferrer Asensio, J., et al. (2025). First detections of PN, PO, and PO+ toward a shocked low-mass starless core. Astrophysical J. Lett. 985, L25. doi:10.3847/2041-8213/add344
Sil, M., Srivastav, S., Bhat, B., Mondal, S. K., Gorai, P., Ghosh, R., et al. (2021). Chemical complexity of phosphorous-bearing species in various regions of the interstellar medium. Astronomical J. 162, 119. doi:10.3847/1538-3881/ac09f9
Tenenbaum, E. D., Woolf, N. J., and Ziurys, L. M. (2007). Identification of phosphorus monoxide (X2Πr) in VY Canis majoris: detection of the first P-O bond in space. Astrophysical J. 666, L29–L32. doi:10.1086/521361
Tenenbaum, E. D., and Ziurys, L. M. (2008). A search for phosphine in circumstellar envelopes: PH3 in IRC +10216 and CRL 2688? Astrophysical J. 680, L121–L124. doi:10.1086/589973
Transtrum, M. K., Machta, B. B., and Sethna, J. P. (2011). Geometry of nonlinear least squares with applications to sloppy models and optimization. Phys. Rev. E 83, 036701. doi:10.1103/PhysRevE.83.036701
Transtrum, M. K., and Qiu, P. (2014). Model reduction by manifold boundaries. Phys. Rev. Lett. 113, 098701. doi:10.1103/PhysRevLett.113.098701
Transtrum, M. K., and Qiu, P. (2016). Bridging mechanistic and phenomenological models of complex biological systems. PLOS Comput. Biol. 12, e1004915. doi:10.1371/journal.pcbi.1004915
Turner, B. E., and Bally, J. (1987). Detection of interstellar PN: the first identified phosphorus compound in the interstellar medium. Astrophysical J. Lett. 321, L75. doi:10.1086/185009
Wakelam, V., Herbst, E., Loison, J.-C., Smith, I. W. M., Chandrasekaran, V., Pavone, B., et al. (2012). A kinetic database for astrochemistry (KIDA). Astrophysical J. Suppl. Ser. 199, 21. doi:10.1088/0067-0049/199/1/21
Walton, C., Ewens, S., Coates, J., Blake, R., Planavsky, N., Reinhard, C., et al. (2023). Phosphorus availability on the early earth and the impacts of life. Nat. Geosci. 16, 399–409. doi:10.1038/s41561-023-01167-6
Keywords: phosphorus astrochemistry, interstellar medium, astrobiology, dynamical systems, chemical reaction networks, kinetic parameters, complexity reduction, parameter sensitivity
Citation: Fernández-Ruz M, Jiménez-Serra I, Castro M, Ruiz-Bermejo M and Aguirre J (2025) Strong parameter hierarchy in the interstellar phosphorus chemical network. Front. Astron. Space Sci. 12:1570525. doi: 10.3389/fspas.2025.1570525
Received: 03 February 2025; Accepted: 10 July 2025;
Published: 30 July 2025.
Edited by:
David Benoit, University of Hull, United KingdomReviewed by:
Kotomi Taniguchi, National Astronomical Observatory of Japan (NAOJ), JapanFrancesco Fontani, Osservatorio Astrofisico di Arcetri (INAF), Italy
Copyright © 2025 Fernández-Ruz, Jiménez-Serra, Castro, Ruiz-Bermejo and Aguirre. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Jacobo Aguirre, amFndWlycmVAY2FiLmludGEtY3NpYy5lcw==