Skip to main content
NASA Author Manuscripts logoLink to NASA Author Manuscripts
. Author manuscript; available in PMC: 2019 May 31.
Published in final edited form as: Chem Commun (Camb). 2018 May 31;54(45):5716–5719. doi: 10.1039/c8cc01391h

First identification of unstable phosphino formic acid (H2PCOOH)

Cheng Zhu a,b,, Robert Frigge a,b, Andrew M Turner a,b, Ralf I Kaiser a,b,, Bing-Jian Sun c,, Si-Ying Chen c, Agnes H H Chang c,
PMCID: PMC6340708  NIHMSID: NIHMS1005172  PMID: 29774326

Abstract

The hitherto elusive phosphino formic acid molecule (H2PCOOH) was detected for the first time in the gas phase. Theoretical calculations revealed an unexpected kinetic stability of H2PCOOH compared to the isovalent carbamic acid (H2NCOOH) although the replacement of a single nitrogen atom by phosphorus decreases the bond order from a partial double (–C=N–) to a single (–C–P–) bond. This work provides a fundamental framework to explore the synthesis and stability of derivatives of carbonic acid (H2CO3), in which one or both hydroxyl groups (OH) are replaced by hydride moieties involving third row atoms.


For the last century, Langmuir’s concept of isovalency in which ‘two molecular entities with the same number of valence electrons have similar chemistries’1 has been instrumental in developing modern concepts of chemical bonding2 and in understanding basic principles of molecular structure and reactivity of isovalent systems.2,3 Specific attention has been devoted to contrasting the nitrogen chemistry with the analogous chemistry of phosphorus.4 Located in main group 15, nitrogen and phosphorus have both five valence electrons and hence are isovalent. Nevertheless, their chemical bonding can be quite distinct as evident from the structures of the diatomic nitrogen molecule (N2) compared to the tetra phosphorus molecule (P4) carrying nitrogen-nitrogen triple and phosphorus-phosphorus single bonds, respectively (Scheme 1). Likewise, fully saturated hydrides of nitrogen are limited to hydrazine (N2H4) and triazene (N3H5),5 whereas phosphanes with up to eight phosphorus atoms (octaphosphane; P8H10) have been prepared recently.6 The divergent chemical stability of nitrogen versus phosphorus compounds is illustrated in particular when comparing carbamic acid (H2NCOOH) with the isovalent phosphorus analogue phosphino formic acid (H2PCOOH) (Scheme 1). Carbamic acid was successfully synthesized in low temperature matrices containing ammonia (NH3) and carbon dioxide (CO2) by annealing the ices from 10 K to 80 K. The overall pathway involved an initial formation of an ammonia-carbon dioxide van der Waals complex followed by the reaction of the Lewis base (NH3) with the Lewis acid (CO2) and a successive proton transfer from the nitrogen to the oxygen atom.7

Scheme 1.

Scheme 1

Structures of tetraphosphorus (P4, Td), dinitrogen (N2, D∞h), phosphino formic acid (H2PCOOH, C1), carbamic acid (H2NCOOH, C1), and carbonic acid (HOCOOH, C2v).

Despite extensive interest in the isovalent phosphino formic acid (H2PCOOH) from the preparative inorganic and also physical chemical viewpoint as isovalent systems of the nitrogen analogous carbamic acid (H2NCOOH) molecule, free phosphino formic acid (H2PCOOH) has not been isolated yet although an inherent barrier of 292 kJ mol−1 should inhibit the decomposition to carbon dioxide (CO2) and phosphine (PH3).8,9 Diemert et al. stabilized phosphino formic acid (H2PCOOH) within the coordination sphere of transition metal complexes): (CO)5M(H2PCOOH) and [(CO)5M(H2PCOOH)]2 (M = Cr, W).10 Derivatives, in which the hydrogen atoms of the phosphino moiety were replaced by organic groups such as phenyl (C6H5), could be stabilized in chromium and tungsten based transition metal complexes as well.11 The stability of free derivatives of phosphino formic acid-carbonic ester phosphides (RHPCOOR′) with R and R′ being alkyl or aromatic groups12 and sodium diphenylphosphinoformate ((C6H5)2PCOONa)13 - was attributed to the lack of acetic hydrogen atoms at the carboxyl moiety, which likely hydrolyze the phosphino formic acid and their derivatives in solution.

Here, we report on the very first experimental detection of the hitherto elusive phosphino formic acid (H2PCOOH) molecule in the gas phase. Our experiments are merged with ab initio electronic structure calculations to verify the stability of gas phase phosphino formic acid (H2PCOOH). The experiments were conducted in an ultrahigh vacuum (UHV) surface science machine operated at a base pressure of a few 10−11 Torr.6,1418 Binary ice mixtures of phosphine (PH3) and carbon dioxide (CO2), phosphine (PH3) and [13C]-carbon dioxide (13CO2), as well as phosphine (PH3) and [18O]-carbon dioxide (C18O2) were prepared in separate experiments with thicknesses of 830 ± 50 nm at 5.5 ± 0.2 K. The non-equilibrium chemistry and bond-cleavage processes were stimulated by exposing these ices to energetic electrons for one hour at an average dose of 2.8 ± 0.6 eV per molecule. After the irradiation, the ices were heated at a rate of 1 K min−1 to 320 K (temperature programmed desorption; TPD). During the TPD phase, the subliming neutral molecules were first ionized via tunable vacuum ultraviolet (VUV) photoionization (PI), mass resolved in a reflectron time-of-flight mass spectrometer (PI-ReTOF-MS), and eventually identified by their arrival times through a multichannel plate via their mass-to-charge ratios. By systematically tuning the photoionization energies from 10.86 eV to 9.60 eV, we selectively ionize specific structural isomers based on their ionization energies (IE) as obtained from the electronic structure calculations and determine which isomer(s) is(are) produced thus ultimately identifying phosphino formic acid (H2PCOOH).

The mass spectra of the subliming molecules released from the irradiated phosphine (PH3)-carbon dioxide (CO2) ices are revealed in Fig. 1 as a function of temperature recorded at photon energies of 10.86 and 9.60 eV. In addition to the precursor phos-phine (PH3) (m/z = 34; IE = 9.87 eV), phosphanes like diphosphine (P2H4) (m/z = 66; IE = 8.8 eV), and hydroxyphosphine (H2POH) (m/z = 50; IE = 9.22–9.23 eV), ion signal was also observed at m/z = 78.19,20 Considering the molecular weight of phosphine (34 amu) and carbon dioxide (44 amu), the molecular weight of 78 amu is consistent with the formation of a molecule of the formula PH3CO2. The corresponding TPD profile at m/z = 78 exhibits an onset of close to 180 K and a maximum of the ion counts at 192 K (Fig. 2(a)). To confirm that the species observed via m/z = 78 are linked to a molecule with the formula PH3CO2, two additional isotopic experiments were carried out by replacing carbon dioxide with [13C] and [18O] carbon dioxide in separate experiments. Here, the TPD profiles recorded at m/z = 79 and 82 (Fig. 2(b) and (c)) - shifted by 1 amu and 4 amu in the [13C] and [18O] carbon dioxide experiments, respectively. These findings reveal that the product molecules have one carbon atom and two oxygen atoms, a finding which is consistent with the molecular formula PH3CO2. It is worth mentioning that we also conducted a control experiment, that is, an experiment under identical conditions with phosphine (PH3) and carbon dioxide (CO2), but without exposing the ice to energetic electrons. In this control study, no signal was observed at m/z = 78 (Fig. 2(d)), which demonstrates that the PH3CO2 product is the result of the radiolysis of the ices, but not an artifact from the photoionization laser and/or gas phase ion-molecule reactions of the precursors.

Fig. 1.

Fig. 1

Mass-to-charge dependent temperature programmed desorption (TPD) profiles obtained after photoionizing the desorbing molecules of irradiated (a) phosphine (PH3) and carbon dioxide (CO2) ice with 10.86 eV photons, (b) phosphine (PH3) and [13C]-carbon dioxide (13CO2) ice with 10.86 eV photons, (c) phosphine (PH3) and [18O]-carbon dioxide (C18O2) ice with 10.86 eV photons, and (d) phosphine (PH3) and carbon dioxide (CO2) ice with 9.60 eV photons.

Fig. 2.

Fig. 2

Temperature programmed desorption (TPD) profiles recorded via PI-ReTOF-MS for (a) m/z = 78 (PH3-CO2 system, PI = 10.86 eV), (b) m/z = 79 (PH3-13CO2 system, PI = 10.86 eV), (c) m/z = 82 (PH3-C18O2 system, PI = 10.86 eV), (d) m/z = 78 (blank PH3-CO2, PI = 10.86 eV), and (e) m/z = 78 (PH3-CO2 system, PI = 9.60 eV).

Having established the formation of a new molecule with the molecular formula PH3CO2 (78 amu) along with its 13C and 18O substituted counterparts, we are shifting our attention now to the identification of the PH3CO2 product isomer(s). This objective is achieved by selectively photoionizing specific structural isomers based on their ionization energies. Our electronic structure calculations expose the existence of ten closed shell local minima, whose adiabatic ionization energies and relative energies obtained at the CCSD(T)/CBS level with B3LYP/cc-pVTZ zero-point vibrational energy corrections are presented in Fig. 3. These species can be arranged in three groups: the thermodynamically most stable phosphino formic acid isomers (H2PCOOH; 1a-1b) [IE = 9.74–9.75 eV; ΔE = 0.00–0.20 eV], its enol tautomers (HPC(OH)2; 2a-2d) [IE = 8.23–8.42 eV; ΔE = 0.85–1.10 eV], and the formic phosphinous anhydride species (HCOOPH2; 3a-3d) [IE = 9.34–9.50 eV; ΔE = 0.00–0.16 eV]; the ionization energies of these three groups are well separated by about > 0.24 eV (group I and III) and 1.1 eV (group II and III). At a photon energy of 10.86 eV, all isomers can be ionized (Fig. 2(a)). A photoionization energy of 9.60 eV allows only the ionization of isomers belonging to group II and III; considering the ionization energy of 9.74–9.75 eV, phosphino formic acid (H2PCOOH; 1a-1b) cannot be ionized by 9.60 eV photons. The TPD profile of the 9.60 eV experiments is depicted in Fig. 2(e) and does not show any ion counts at m/z = 78. Therefore, we can conclude that isomers belonging to group II and III are not ionized at 9.60 eV, and that signal at m/z = 78 in the 10.86 eV experiment is linked solely to the previously elusive phosphino formic acid (H2PCOOH), but not to HPC(OH)2 or HCOOPH2.

Fig. 3.

Fig. 3

Molecular structures of H2PCOOH isomers along with relative energies in eV (black), adiabatic ionization energies in eV (blue), and point groups (red).

It is interesting to rationalize the exclusive synthesis of the phosphino formic acid isomer (H2PCOOH). Upon interaction of energetic electrons, the phosphine molecule (PH3) can undergo a unimolecular decomposition to form predominantly the phosphino (PH2) radical plus suprathermal atomic hydrogen (H) (reaction (1)).6,21 The suprathermal hydrogen atom can add to the carbon or oxygen atom of the carbon dioxide (CO2) molecule.2224 These pathways lead to the formyloxyl (HCOO) and cis/trans-hydroxy-carbonyl (HOCO) isomers, respectively, with the latter being more stable by 0.51/0.59 eV (reaction (2)). Recent experiments in carbon dioxide-methane23 and carbon dioxide-water24 ices confirm a preferential hydrogen atom addition to the oxygen atom of carbon dioxide, and solely the cis/trans-hydroxycarbonyl (HOCO) radical was detected spectroscopically. In the low temperature carbon dioxide-phosphine ices, the cis/trans-hydroxycarbonyl (HOCO) radical could recombine without barrier with the phosphino (PH2) radical forming the experimentally detected phosphino formic acid isomer (H2PCOOH) (reaction (3)) if both radicals hold a favorable recombination geometry. Note that any synthesis of the non-detected formic phosphinous anhydride species (HCOOPH2) requires the formation of the formyloxyl (HCOO) radical intermediate, which then recombines with the phosphino (PH2) radical. However, the aforementioned discussion excluded the formation of the formyloxyl (HCOO) radical, which in turn eliminates the synthesis of the formic phosphinous anhydride species (HCOOPH2). Finally, the failed detection of the enol tautomer (HPC(OH)2) deserves some attention. Previous studies on the acetaldehyde (CH3CHO)-vinyl alcohol (H2CCH(OH)) keto-enol isomer pair in electron irradiated carbon monoxide-methane ices suggests that the enol tautomer is formed via atomic hydrogen shift from the carbon to the oxygen atom in acetaldehyde upon radiolysis of the latter.15 The non-detection of the enol tautomer (HPC(OH)2) in the present experiments proposes that the keto-enol tautomerization is absent possibly due yields of the phosphino formic acid isomer (H2PCOOH) being too low to form detectable fractions of the enol tautomer of (HPC(OH)2.

PH3 PH2+ H ΔRG= 3.37 eV (1)
+ CO2 HCOO ΔRG= 0.37 eV (2a)
+ CO2cis/transHOCO ΔRG= 0.15/0.22 eV (2b)
cis/transHOCO+ PH2 H2PCOOH ΔRG= 2.70/2.63 eV (3)

Having established the synthesis of the hitherto elusive phosphino formic acid molecule (H2PCOOH) along with its 13C and 18O substituted counterparts, we are turning our attention now to the stability of this compound. Our electronic-structure calculations predict that phosphino formic acid (1a) can undergo unimolecular decomposition forming phosphine (PH3) and carbon dioxide (CO2) via a weakly bound van der Waals complex in an overall exoergic reaction (ΔRG = −0.53 eV; ΔRG = −51 kJ mol−1) after isomerizing to 1b via a barrier of 0.45 eV (43 kJ mol−1) and passing a transition state located 1.94 eV (187 kJ mol−1) above 1b (Fig. 4(a)). These data demonstrate that although phosphino formic acid (H2PCOOH) is less stable than the phosphine (PH3) and carbon dioxide (CO2) reactants, phosphino formic acid (H2PCOOH) is kinetically stable due the significant barrier separating the reactants and products. The features of the underlying potential energy surface toward decomposition (Fig. 4(a)) are alike to those of the surfaces of carbamic acid (H2NCOOH)25 and carbonic acid (HOCOOH), which have been isolated previously.2630 The calculated phosphorus-carbon bond distance of 1.87–1.89 Å (1a-1b) (Table S2, ESI) correlates with the existence of a phosphorus-carbon single bond (1.84 Å), but not with a double bond (1.71 Å), which is shorter by about 0.03–0.05 Å. On the other hand, the nitrogen-carbon bond length in the isovalent carbamic acid (H2NCOOH) was determined to be 1.36–1.38 Å (Table S3, ESI), i.e. between a typical nitrogen-carbon single (1.47 Å) and double bond (1.29 Å) due to resonance and hence partial double bond character of the-C=N- moiety. Although the replacement of the nitrogen atom by phosphorus increases the bond length to carbon from a partial double (–C=N–) to a single (–C–P–) bond, the barrier to a gas phase decomposition increases from 149 kJ mol−1 to 187 kJ mol−1 from carbamic acid (H2NCOOH) to phosphino formic acid (H2PCOOH) (Fig. 4). This unexpected enhancement of stability can be rationalized in terms of distinct partial charges distributions. Here, the nitrogen atom in carbamic acid (H2NCOOH) holds a charge of −0.264 e and can easily capture the adjacent acidic proton on oxygen to pass the transition state (Fig. 4b), while the positive charges on the phosphorous atom (0.003 e) in phosphino formic acid (H2PCOOH) may hinder the approach of the proton and lead to a higher barrier for the decarboxylation of phosphino formic acid (H2PCOOH). Consequently, the replacement of single nitrogen atom by lower electronegative phosphorus enhances the stability of the molecule carrying the carbon-phosphorus bond (phosphino formic acid) compared to the carbon-nitrogen moiety (carbamic acid). Finally, considering the average velocity of 240 ms−1 of the phosphine formic acid (H2PCOOH) subliming at 190 K and the distance between the ice and the photoionization laser of 2 mm, the lifetime of the subliming phosphino formic acid (H2PCOOH) has to be at least 6 μs to survive the time between the sublimation and the photoionization event.

Fig. 4.

Fig. 4

Molecular decomposition pathway of phosphino formic acid (H2PCOOH) (a) and of the isovalent formic acid (H2NCOOH) (b). Energies were obtained at the CCSD(T)/CBS level with B3LYP/cc-pVTZ zero-point vibrational energy corrections. Partial charges (e) of the species related to the decomposition of phosphino formic acid (H2PCOOH) and carbamic acid (H2NCOOH) are also indicated.

In conclusion, exploiting single photon vacuum ultraviolet photoionization reflectron time of flight mass spectrometry (PI-ReTOF-MS), the present study provides compelling evidence on the exclusive synthesis of the phosphino formic acid (H2PCOOH) molecule - the hitherto elusive isovalent species of carbamic acid (H2NCOOH). Utilizing isotopically labelled reactants, this approach allowed the determination of the molecular formula of the newly formed molecule based on the isotopic shifts and the fragment-free detection of the neutral parent molecule upon photoionization. The non-equilibrium nature of the reaction pathway via a phosphorus-hydrogen bond rupture in phosphine followed by addition of the suprathermal hydrogen atom to the oxygen atom of carbon dioxide and recombination of the cis/trans hydroxy carbonyl (HOCO) with phosphino (PH2) leads exclusively to a directed synthesis of the free phosphino formic acid (H2PCOOH) molecule.

Supplementary Material

P2SI

Acknowledgments

This work was supported by NASA under Grant NNX16AD27G to The University of Hawaii. The Hawaii group would also like to thank the W. M. Keck Foundation to fund the surface science machine. AHHC, BJS, and SYC thank the National Center for High-performance Computer in Taiwan for providing the computer resources for the calculations.

Footnotes

Electronic supplementary information (ESI) available: Experimental and theoretical details, Fig. S1 and Tables S1–S5. See DOI: 10.1039/c8cc01391h

Conflicts of interest

There are no conflicts to declare.

References

  • 1.Langmuir I, J. Am. Chem. Soc, 1919, 41, 868–934. [Google Scholar]
  • 2.Bellus D, Science of Synthesis: Houben-Weyl Methods of Molecular Transformations Guidebook, Georg Thieme Verlag, Stuttgart and New York, 2000. [Google Scholar]
  • 3.Parker DSN, Mebel AM and Kaiser RI, Chem. Soc. Rev, 2014, 43, 2701–2713. [DOI] [PubMed] [Google Scholar]
  • 4.Rossi L, Science of Synthesis: Houben-Weyl Methods of Molecular Transformations, Georg Thieme Verlag, Stuttgart and New York, 2005. [Google Scholar]
  • 5.Förstel M, Maksyutenko P, Jones BM, Sun BJ, Chen SH, Chang A and Kaiser RI, ChemPhysChem, 2015, 16, 3139–3142. [DOI] [PubMed] [Google Scholar]
  • 6.Turner AM, Abplanalp MJ, Chen SY, Chen YT, Chang AH and Kaiser RI, Phys. Chem. Chem. Phys, 2015, 17, 27281–27291. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Bossa JB, Theule P, Duvernay F, Borget F and Chiavassa T, Astron. Astrophys, 2008, 492, 719–724. [Google Scholar]
  • 8.Diemert K, Hahn T, Kuchen W and Tommes P, Phosphorus, Sulfur Silicon Relat. Elem, 1993, 83, 65–76. [Google Scholar]
  • 9.Fu Y, Yu TQ, Wang YM, Liu L and Guo QX, Chin. J. Chem, 24, 299–306. [Google Scholar]
  • 10.Diemert K, Hahn T, Kuchen W, Mootz D, Poll W and Tommes P, Z. Naturforsch., B: Chem. Sci, 1995, 50, 209–212. [Google Scholar]
  • 11.Diemert K, Hahn T and Kuchen W, J. Organomet. Chem, 1994, 476, 173–181. [Google Scholar]
  • 12.Thamm R and Fluck E, Z. Naturforsch., B: Chem. Sci, 1981, 36, 910–916. [Google Scholar]
  • 13.Diemert K, Hahn T and Kuchen W, Phosphorus, Sulfur Silicon Relat. Elem, 1991, 60, 287–294. [Google Scholar]
  • 14.Jones BM and Kaiser RI, J. Phys. Chem. Lett, 2013, 4,1965–1971. [DOI] [PubMed] [Google Scholar]
  • 15.Abplanalp MJ, Gozem S, Krylov AI, Shingledecker CN, Herbst E and Kaiser RI, Proc. Natl. Acad. Sci. U. S. A, 2016, 113, 7727–7732. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Abplanalp MJ and Kaiser RI, Astrophys. J, 2017, 836, 195. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Bergantini A, Maksyutenko P and Kaiser RI, Astrophys. J, 2017, 841, 96. [Google Scholar]
  • 18.Gobi S, Bergantini A, Turner AM and Kaiser RI, J. Phys. Chem. A, 2017, 121, 3879–3890. [DOI] [PubMed] [Google Scholar]
  • 19.Elbel S, Tom Dieck H, Becker G and Ensslin W, Inorg. Chem, 1976, 15, 1235–1238. [Google Scholar]
  • 20.Berkowitz J, Curtiss LA, Gibson ST, Greene JP, Hillhouse GL and Pople JA, J. Chem. Phys, 1986, 84, 375–384. [Google Scholar]
  • 21.Turner AM, Abplanalp MJ and Kaiser RI, Astrophys. J, 2016, 819, 97. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Song X, Li J, Hou H and Wang B, J. Chem. Phys, 2006,125, 094301. [DOI] [PubMed] [Google Scholar]
  • 23.Bennett CJ and Kaiser RI, Astrophys. J, 2007, 660, 1289–1295. [Google Scholar]
  • 24.Zheng W and Kaiser RI, Chem. Phys. Lett, 2007, 450, 55–60. [Google Scholar]
  • 25.Ramachandran B, Halpern AM and Glendening ED, J. Phys. Chem. A, 1998, 102, 3934–3941. [Google Scholar]
  • 26.Loerting T, Tautermann C, Kroemer RT, Kohl I, Hallbrucker A, Mayer E and Liedl KR, Angew. Chem., Int. Ed, 2000, 39, 891–894. [DOI] [PubMed] [Google Scholar]
  • 27.Ludwig R and Kornath A, Angew. Chem., Int. Ed, 2000, 39, 1421–1423. [DOI] [PubMed] [Google Scholar]
  • 28.Kohl I, Winkel K, Bauer M, Liedl KR, Loerting T and Mayer E, Angew. Chem., Int. Ed, 2009, 48, 2690–2694. [DOI] [PubMed] [Google Scholar]
  • 29.Bernard J, Seidl M, Kohl I, Liedl KR, Mayer E, Galvez O, Grothe H and Loerting T, Angew. Chem., Int. Ed, 2011, 50, 1939–1943. [DOI] [PubMed] [Google Scholar]
  • 30.Bucher G and Sander W, Science, 2014, 346, 544–545. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

P2SI

RESOURCES