PROBING THE CARBON-PHOSPHORUS BOND COUPLING IN LOW-TEMPERATURE PHOSPHINE (PH₃)-METHANE (CH₄) INTERSTELLAR ICE ANALOGUES

Abstract

Phosphine, which has now been confirmed around the carbon-rich star IRC+10216, provides the first example of a phosphorus-containing single bond in interstellar or circumstellar media. While four compounds containing both phosphorus and carbon have been discovered, none contain a carbon-phosphorus single bond. Here, we show that this moiety is plausible from the reaction of phosphine with methane in electron-irradiated interstellar ice analogues. Fractional sublimation allows for detection of individual products at distinct temperatures using reflectron time-of-flight mass spectrometry (ReTOF) coupled with vacuum ultraviolet photoionization. This method produced phosphanes and methylphosphanes as large as P₈H₁₀ and CH₃P₈H₉, which demonstrates that a phosphorus-carbon bond can readily form and that methylphosphanes sublime at 12–17 K higher temperatures than the non-organic phosphanes. Also, irradiated ices of phosphine with deuterated-methane untangle the reaction pathways through which these methylphosphanes were formed and identified radical recombination to be preferred over carbene/phosphinidene insertion reactions. In addition, these ReTOF results confirm that CH₃PH₂ and CH₆P₂ can form via insertion of carbene and phosphinidene and that the methylenediphosphine (PH₂CH₂PH₂) isomer forms in the ices, although methylphosphine (CH₃P₂H₃) is likely the more abundant isomer and that phosphanes and organophosphanes preferentially fragment via the loss of a phosphino group when photoionized. While the formation of methylphosphine is overall endoergic, the intermediates produced by interactions with energetic electrons proceed toward methylphosphine favorably and barrierlessly and provide plausible mechanisms toward hitherto unidentified interstellar compounds.

1. INTRODUCTION

The recent discovery of phosphine (PH₃) in the circumstellar envelope of the carbon-rich star IRC+10216 (CW Leonis) (Agundez et al. 2008, 2014a; Tenenbaum & Ziurys 2008) at abundances of 10⁻⁸ compared to molecular hydrogen (H₂) has revitalized the interest in the interstellar phosphorus chemistry. Besides phosphine, only six phosphorus-bearing molecules have been discovered in interstellar and/or circumstellar environments (Figure 1). These are phosphorus nitride (PN) (Turner & Bally 1987; Ziurys 1987; Guélin et al. 2000; Milam et al. 2008), carbon phosphide (CP) (Guélin et al. 1990; Milam et al. 2008), phosphaethyne (HCP) (Agundez et al. 2007; Milam et al. 2008), phosphorus monoxide (PO) (Tenenbaum et al. 2007), dicarbon phosphide (CCP) (Halfen et al. 2008), and cyanophosphaethyne (NCCP) (Agundez et al. 2014b). Each of these, except for phosphorus monoxide (PO), has been detected along with phosphine (PH₃) in IRC+10216 with abundances compared to molecular hydrogen of 3 × 10⁻¹⁰ for phosphorus nitride (PN) (Milam et al. 2008), 5 × 10⁻⁹ for carbon phosphide (CP) (Milam et al. 2008), 3 × 10⁻⁸ for phosphaethyne (HCP) (Milam et al. 2008), 10⁻⁹ for dicarbon phosphide (CCP) (Halfen et al. 2008), and an upper abundance of 3 × 10⁻⁸ for the tentative detection of cyanophosphaethyne (NCCP) (Agundez et al. 2014b). Furthermore, phosphaethyne (HCP) and phosphine (PH₃) account for 5% and 2%, respectively, of the total phosphorus budget around IRC+10216 (Agundez et al. 2014a). Given the carbon-rich nature of this circumstellar envelope, it is not surprising that four of the six phosphorus-bearing compounds around IRC+10216 also contain carbon. Using the phosphaethyne to hydrogen cyanide (HCN) ratio (HCP/HCN = 0.001, compared to the solar ratio of 0.003) (Milam 2007) to estimate the phosphorus to nitrogen ratio, the relatively high abundance of nitrogen also rationalizes the presence of phosphorus nitride (PN) and cyanophosphaethyne (NCCP). Phosphorus monoxide (PO), on the other hand, was discovered in the oxygen-rich circumstellar envelop of the supergiant star VY Canis Majoris (Tenenbaum et al. 2007). Unlike phosphine, these molecules are notable in that phosphorus is bonded only to elements of the second period of the periodic table of the elements and that each compound contains a strong double or triple bond with phosphorus holding bond energies between 510 and 620 kJ mol⁻¹ (Johnson III 2015). In contrast, phosphine only contains phosphorus-hydrogen single bonds with a bond energy of only 343 kJ mol⁻¹ (Cottrell 1954).

Figure 1.

Phosphorus-bearing molecules detected in the interstellar medium.

It is not surprising that, with the exception of cyanopho-sphaethyne (NCCP), analogues to each of these compounds have been discovered in the interstellar medium in which phosphorus is substituted by its isovalent element: nitrogen. Given the discovery of phosphine, it follows that larger phosphorus-containing compounds analogous to those formed from isoelectronic ammonia should exist in interstellar environments. For example, methylamine (CH₃NH₂) was first discovered in the hot cores Sagittarius B2 and Orion A with abundances of 1 × 10⁻⁹ and 3 × 10⁻⁹, respectively, compared to molecular hydrogen (Fourikis et al. 1974; Kaifu et al. 1974). Recent laboratory experiments exposing ices of ammonia (NH₃) and C1 to C6 hydrocarbons (Kim & Kaiser 2011) to energetic electrons, which mimicked the interaction of secondary electrons generated in the track of galactic cosmic rays while penetrating ice-coated interstellar grains, demonstrated that methylamine (CH₃NH₂) can be formed via the barrierless recombination of methyl radicals (CH₃) with amino radicals (NH₂) at 10 K. If ammonia is replaced by phosphine, the phosphorus analogue, methylpho-sphine (CH₃PH₂) is expected to form in interstellar analogue ices and predicted to exist toward Sagittarius B2 and Orion A along with methylamine (Lafont et al. 1982). However, as of today, interstellar methylphosphine (CH₃PH₂) has remained elusive (Halfen et al. 2014). In the present work, we investigate to what extent carbon-phosphorus bond coupling can lead to the formation of methylphosphine (CH₃PH₂) and potentially higher-order organophosphorus compounds in ices of phos-phine (PH₃) and methane (CH₄) upon interaction with energetic electrons generated in the track of galactic cosmic ray particles penetrating ice-coated particles in cold molecular clouds.

2. EXPERIMENTAL

The experiments were conducted in a stainless steel chamber operating under ultra-high vacuum pressures of 5 × 10⁻¹¹ Torr by exploiting oil-free turbomolecular pumps and dry scroll backing pumps (Bennett et al. 2013; Jones & Kaiser 2013; Jones et al. 2014a, 2014b; Kaiser et al. 2014, 2015; Maity et al. 2014a, 2014b, 2015; Maksyutenko et al. 2015). The ices were prepared on a reflective silver substrate mounted to a rotatable cold finger manufactured using oxygen-free high-conductivity copper capable of achieving temperatures as low as 5.5 ± 0.2 K by a closed-cycle helium refrigerator (Sumitomo Heavy Industries, RDK-415E). Methane (Advanced Specialty Gases, 99.999%) and phosphine (Sigma-Aldrich, 99.9995%) were premixed in a gas mixing chamber at 110 Torr each and then introduced into the main recipient with the help of a glass capillary at a pressure of 5 × 10⁻⁸ Torr for 8 minutes. A Nicolet 6700 Fourier Transform Infrared Spectrometer (FTIR) probed the deposited ices on the silver substrate on line and in situ from 6000 cm⁻¹ to 500 cm⁻¹ with 4 cm⁻¹ resolution, and the integrated infrared peak areas $\left(\underset{{\tilde{v}}_1}{\overset{{\tilde{v}}_2}{\int }}A(\tilde{v})d\tilde{v}\right)$ of the v₂ band of phosphine (987 cm⁻¹) and v₁ + v₄ combination band of methane (4195 cm⁻¹) were used with their integrated absorption coefficients (A_exp) of 5.1 × 10⁻¹⁹ cm molecule⁻¹ and 3.5 × 10⁻¹⁹ cm molecule⁻¹, respectively (Brunetto et al. 2008; Turner et al. 2015), in the following equation to determine the column density (N) of each reactant in the ice:

$$N=\frac{\text{ln}(10)\underset{{\tilde{v}}_1}{\overset{{\tilde{v}}_2}{\int }}A(\tilde{v})d\tilde{v}}{2A_{\text{exp}}}\text{cos}\left({\text{sin}}^{-1}\frac{\text{sin}\alpha }{n_{\text{ice}}}\right).$$ (1)

The angle at which light passes through ice (β) is related to the angle of the incoming beam (α) by Snell’s law: n₁sin α = n_icesin β. This was incorporated into Equation (1) along with a factor of 2 to account for the incoming and outgoing beams and by assigning the refractive index of vacuum to be n₁ = 1. To determine the refractive index of the ice mixture and the ice thickness, laser interferometry (Hudgins et al. 1993; Westley et al. 1998; Fulvio et al. 2009; Turner et al. 2015) was utilized during deposition by reflecting a helium-neon laser (λ = 632.8 nm) off the silver substrate and ice surfaces. The relative intensity between the maxima and minima of the interference fringes was used (Goodman 1978; Westley et al. 1998; Babar & Weaver 2015) to determine a refractive index of n_ice = 1.44 ± 0.04, which ranges between the refractive indices of pure phosphine $\left(n_{{\text{PH}}_3}=1.51\right)$ (Turner et al. 2015) and pure methane $\left(n_{{\text{CH}}_4}=1.33\right)$ (Brunetto et al. 2008). The column densities of 3.8 ± 0.3 × 10¹⁸ molecules cm⁻² for phosphine and 1.2 ± 0.1 × 10¹⁸ molecules cm⁻² for methane indicate that the deposited ice mixture had a 3.2 ± 0.6:1.0 phosphine to methane ratio. The ice thickness (d) was measured using the equation:

$$d=\frac{m\lambda }{2\sqrt{n_{\text{ice}}^2-{\text{sin}}^2\theta }}.$$ (2)

Using the laser’s angle of incidence (θ = 4°± 1°) and by counting interference fringes (m = 4.45 ± 0.05 fringes), 980 ± 40 nm of ice was deposited. One hour after the deposition, which was sufficient time to re-establish the base pressure, the ice was irradiated with 5 keV electrons at a flux of 2 × 10¹⁰ electrons s⁻¹ cm⁻² at a 70° angle of incidence over an area of 1.0 ± 0.1 cm². Monte Carlo (CASINO) (Hovington et al. 1997) calculations, which exploited the weighted averaged density of 0.78 g cm⁻³ from the phosphine (0.90 g cm⁻³) (Francia & Nixon 1973; Turner et al. 2015) and methane (0.47 g cm⁻³) (Satorre et al. 2008) densities, were performed and determined the average absorbed dose with a 540 nm average penetration depth to be 3.5 ± 0.4 eV molecule⁻¹ for phosphine and 3.1 ± 0.3 eV molecule⁻¹ for methane. Note that the penetration depth is smaller than the thickness of the ices indicating that the electrons only interact with the ices but not the silver substrate. A temperature programmed desorption (TPD) protocol heated the irradiated ice to 300 K with a heating rate of 1 K minute⁻¹ and allowed the reactants and newly formed molecules to sublime. During the irradiation and heating, the FTIR monitored the ice on line and in situ. Also, two mass spectroscopic techniques analyzed the subliming species. A traditional quadrupole mass spectrometer (QMS) operating in residual gas analyzer (RGA) mode with 100 eV electrons at 1mA emission current offered detection of molecules via electron impact ionization. A more sensitive reflectron time-of-flight (ReTOF) mass spectrometer (Jordan TOF Products, Inc.) utilizing single photon photoionization (118.2 nm, 10.49 eV) (Jones & Kaiser 2013) was also used. The pulsed (30 Hz) coherent vacuum ultraviolet (VUV) light was generated via four-wave mixing with xenon (99.999%) as the nonlinear medium. The third harmonic (354.6 nm) of a pulsed neodymium-doped yttrium aluminum garnet laser (Nd:YAG, Spectra Physics, PRO-250, 30 Hz) underwent a frequency tripling process (ω_vuv = 3ω₁) to obtain the 118.2 nm light with about 10¹⁴ photons per pulse (Maity et al. 2014b). This light was spatially separated from the fundamental using a lithium fluoride (LiF) planoconvex lens (VonDrasek et al. 1988) (ISP Optics, LF-PX-38–150) exploiting distinct refractive indices of LiF for different wavelengths of 1.40 and 1.59, respectively (Li 1976). The VUV light was directed 1 mm above the ice surface, and the photoionized molecules were mass analyzed with a ReTOF mass spectrometer. Here, the arrival time of the ions to a multichannel plate is based on the mass-to-charge ratio, and the signal was amplified with a fast preamplifier (Ortec 9305) and recorded with a bin width of 4 ns triggered at 30 Hz (Quantum Composers, 9518). Previous studies (Turner et al. 2015) have shown that the ionization energy of phosphanes range from 9.8 eV for PH₃ to 7.5 eV for P₈H₁₀, and thus the chosen photon energy (10.49 eV) is capable of ionizing each of the phosphanes to be observed. Also, methylphosphine (CH₃PH₂) ionizes at 9.1 eV (Staley & Beauchamp 1974; Hodges et al. 1980), and following trends of both phosphanes and alkanes that ionization energies decline with increasing molecular size, more complex alkylphosphanes are also expected to have ionization energies below 10.49 eV. To obtain mechanistic information, additional experiments were performed replacing methane with deuterated methane, CD₄ (CDN Isotopes, 99.9% D atom) under otherwise identical experimental conditions.

3. RESULTS

3.1. Infrared Spectroscopy

During the irradiation, infrared spectra were recorded in two-minute intervals (Figure 2 and Table 1). Most notably, the strong phosphorus-hydrogen stretching modes of phosphine (PH₃) (v₁ and v₃) centered around 2310 cm⁻¹ decreased and broadened due to the appearance of the v₅ mode of diphosphine (P₂H₄) (2262 and 2287 cm⁻¹). Also, the v₁₁ mode of P₂H₄ emerged, albeit more subtly, at 1063 cm⁻¹ (Durig et al. 1996). Nearby, a tenuous peak at 1054 cm⁻¹ was assigned to the deformation band of the methyl group in methylphosphine (CH₃PH₂) (Kim et al. 2007), while a more distinguishable peak appears at 2973 cm⁻¹ caused by carbon-hydrogen stretching. Thus, only four new peaks emerged during irradiation and only two products, diphosphine and methylphosphine, could be assigned. A drawback of using infrared spectroscopy for methane-doped phosphine ices is that the most intense vibrations, the phosphorus-hydrogen stretching mode, occur in the same region of the spectrum for all products, i.e., typically from 2350 to 2250 cm⁻¹. In addition, carbonhydrogen stretching modes cannot be used to identify individual methylated phosphanes because the group frequencies overlap among each other in the range of 3000–2950 cm⁻¹. After the sublimation of methane and phosphine (80 K), only the phosphorus-hydrogen stretches centered around 2295 cm⁻¹ had significant intensity. Even after diphosphine sublimed (135 K), this peak slowly decreased in intensity and disappeared into the baseline as higher order products sublimed.

Figure 2.

(Top) Infrared spectrum of pristine methane–phosphine ice at 5.5 K. (Bottom) Spectra of phosphine (PH₃) and methane (CH₄) ice before irradiation (blue dotted), after irradiation (black), and after methane and phosphine sublimed (90 K, red). New peaks seen from the irradiation are labeled.

Table 1.

Infrared Absorption Assignments for Phosphine (PH₃) and Methane (CH₄) Ice at 5.5 K and the Irradiation Products

Assignment	Compound	Position (cm−1)	References
v2	PH3	983, 987	(1)
v4	PH3	1099, 1100, 1110sh	(1)
v4	CH4	1296, 1302	(2)
v2	CH4	1526	(2)
v2 + v4	PH3	2071, 2091	(1)
2v4	PH3	2193, 2209	(1)
v1	PH3	2305	(1)
v3	PH3	2313, 2326	(1)
v1/v3 + vL	PH3	2349, 2440, 2461	(1)
2v4	CH4	2589	(2)
v2 + v4	CH4	2809, 2816	(2)
3v2	PH3	2899	(1)
v3	CH4	3002, 3006, 3008	(2)
v3 + vL	CH4	3029, 3074	(2)
v1 + v2	PH3	3293	(1)
v1 + v4	PH3	3402	(1)
v3 + v4	PH3	3420	(1)
3v4	CH4	3841	(2)
v1 + v4	CH4	4193,4198	(2)
v3 + v4	CH4	4291, 4296, 4306	(2)
2v1	PH3	4519	(1)
v1 + v3	PH3	4547	(1)
New Peaks From Irradiation
δ(CH3)	CH3PH2	1054	(3)
v11	P2H4	1063	(4)
v5	P2H4	2262, 2287	(4)
v(C-H)	CH3PH2	2973	(3)

Note. v_L defines the lattice mode.

References.

⁽¹⁾Turner et al. (2015),

⁽²⁾Bennett et al. (2006),

⁽³⁾Kim et al. (2007),

⁽⁴⁾Durig et al. (1996).

3.2. Reflectron Time-of-flight Mass Spectrometry (ReTOF)

3.2.1. Phosphanes

The ReTOF data using a 10.49 eV photoionization energy provided the most useful and ample results for determining the products of irradiated ices of phosphine (PH₃) and methane (CH₄) (Figure 3, Tables 2 and 3), especially when compared to quadrupole mass spectrometry results (Appendix). This highly sensitive technique allowed molecular identification exploiting unique mass-to-charge ratios and well-defined sublimation temperatures from heating the ices to 300 K at a rate of 1 K minute⁻¹. Figure 4 depicts the ion count intensity for methane-doped phosphine ice as a function of temperature during warm-up of the irradiated ices to 300 K at all mass-to-charge ratios observed in the ReTOF. A few observations consistent with the irradiation of pure phosphine ice (Turner et al. 2015) can be highlighted. First, a series of saturated phosphanes including diphosphine (P₂H₄), triphosphane (P₃H₅), tetraphosphane (P₄H₆), and pentaphosphane (P₅H₇) were observed at progressively increasing sublimation temperatures via their parent ions that peaked at 130 K, 162 K, 190 K, and 208 K, respectively. Furthermore, the molecular ion counts for P₂H₄ and P₃H₅ were by far the highest of any product. Diphosphine showed no evidence of fragmentation, while triphosphane fragmented mostly into PH₄⁺ at a third of the parent ion intensity and also to minor amounts of P₂H₃⁺ and P₂H₂⁺ through PH₂ and PH₃ loss. Beginning with P₄H₆, fragmentation became dominant and the parent ion was about five times less intense than the combined major PH₂ and PH₃ loss fragments: P₃H₄⁺ and P₃H₃⁺. Pentaphosphane (P₅H₇) showed that the extent of fragmentation increased with molecular size as the parent ion intensity was only 10% that of the PH₂ loss fragment, P₄H₅⁺. A minor but notable protonated two-phosphorus fragment, P₂H₅⁺, also occurred for P₅H₇. Starting with hexaphosphane (P₆H₈), fragmentation was quantitative so that the molecular ion could no longer be observed in the mass spectra, making P₅H₇⁺ the largest observed parent ion. However, given the sequential order of sublimation temperatures as phosphanes increase in size along with the predictable pattern of fragmentation—predominantly from PH₂ loss—fragments can be used to infer the presence of their parent compounds (Figure 4). Specifically, the intensity and sublimation temperature of fragment ions P₅H₆⁺, P₆H₇⁺, and P₇H₈⁺ was exploited as a proxy for determination of hexaphosphane (P₆H₈), heptaphosphane (P₇H₉), and octapho-sphane (P₈H₁₀). In summary, phosphine in the irradiated phosphine-methane ice reacted to form saturated phosphanes as complex as P₈H₁₀.

Figure 3.

ReTOF mass spectrometry data as a function of sublimation temperature as irradiated phosphine (PH₃) and methane (CH₄) ice was heated from 5.5 K to 300 K at 1 K minute⁻¹.

Table 2.

Observed Ions in the ReTOF Mass Spectrometer for the Phosphine (PH₃) and Methane (CH₄) Irradiation

Mass	Formula	Comments	Molecular Formula of Parent Compound
35	PH4+	fragment	P3H5, P4H6, CH3P4H5
48	CH3PH2+	parent	CH3PH2
49	13CH3PH2+	isotope	CH3PH2
	CH3PH3+	fragment	CH3P3H4, CH3P4H5
66	P2H4+	parent	P2H4
67	P2H5+	protonated parent	P2H4
		fragment	P4H6, P5H7
78	CH3P2H+	fragment	CH3P3H4
79	CH3P2H2+	fragment	CH3P3H4
80	CH3P2H3+	parent	CH3P2H3
81	13CH3P2H3+	isotope	CH3P2H3
	CH3P2H4+	fragment	CH3P5H6
96	P3H3+	fragment	P4H6
97	P3H4+	fragment	P4H6
98	P3H5+	parent	P3H5
99	P3H6+	protonated parent	P3H5
		fragment	P6H8
110	CH3P3H2+	fragment	CH3P4H5
111	CH3P3H3+	fragment	CH3P4H5
112	CH3P3H4+	parent	CH3P3H4
126	P4H2+	fragment	P6H8
127	P4H3+	fragment	P6H8
128	P4H4+	fragment	P6H8
129	P4H5+	fragment	P5H7
130	P4H6+	parent	P4H6
143	CH3P4H4+	fragment	CH3P5H6
144	CH3P4H5+	parent	CH3P4H5
159	P5H4+	fragment	P6H8, P7H9
160	P5H5+	fragment	P6H8, P7H9
161	P5H6+	fragment	P6H8
162	P5H7+	parent	P5H7
175	CH3P5H5+	fragment	CH3P6H7
191	P6H5+	fragment	P7H9, P8H10
193	P6H7+	fragment	P7H9
207	CH3P6H6+	fragment	CH3P7H8
225	P7H8+	fragment	P8H10
239	CH3P7H7+	fragment	CH3P8H9

Table 3.

Fragmentation Patterns and Onset Sublimation Temperatures of Product Molecules for Phosphine (PH₃) with Methane (CH₄) Irradiation

Compound	Sublimation Temperature	Species Detected by ReTOF-MS	Ratio of Species Detected (Scaled to 100)
CH3PH2	87 K	CH3PH2+	100
P2H4	98 K	P2H5+, P2H4+	0.75:100
CH3P2H3	118 K	CH3P2H3+	100
P3H5	131 K	P3H6+, P3H5+, PH4+	0.3:100:34
CH3P3H4	148 K	CH3P3H4+, CH3P2H2+, CH3P2H+, CH3PH3+	100:53:31:4
P4H6	160 K	P4H6+, P3H4+, P3H3+, P2H5+, PH4+	35:85:100:2:12
CH3P4H5	175 K	CH3P4H5+, CH3P3H3+, CH3P3H2+, CH3PH3+	29:100:16:7
P5H7	185 K	P5H7+, P4H5+, P4H4+, P2H5+	10:100:40:10
CH3P5H6	197 K	CH3P4H4+, CH3P2H4+	100:10
P6H8	206 K	P5H6+, P5H5+, P5H5+, P4H3+, P4H2+, P3H6+	100:9:7:23:15:6
CH3P6H7	218 K	CH3P5H5+	100
P7H9	231 K	P6H7+, P6H5+, P5H5+, P5H4+	100:17:10:31
CH3P7H8	243 K	CH3P6H6+	100
P8H10	252 K	P7H8+, P6H5+	100:85
CH3P8H9	264 K	CH3P7H7+	100

Note. Because molecular ions for products larger than P₅H₇ and CH₃P₄H₅ were not observed, fragments were utilized to determine sublimation temperatures. The ratios of observed species assigned to each product are listed to illustrate the fragmentation patterns

Figure 4.

Time-of-flight mass spectra for the products of phosphine (PH₃) and methane (CH₄) irradiation as a function of temperature. Colored bands indicat sublimation events at similar temperatures. The intensity is listed on the left of each spectrum, while the mass-to-charge and ionic formula is on the right.

3.2.2. Organophosphorus Molecules

Further analysis revealed that each of the phosphanes observed in the ReTOF also correlated with an associated methylphosphane of the generic molecular formula CH₃P_xH_x+1, with x = 1–8. The most abundant products were molecules with the formula CH₃PH₂ and CH₃P₂H₃, which occurred in similar quantities, although CH₃PH₂ had two distinct sublimation events, which was unique among products, at 102 and 130 K. The first event was exploited to determine the onset sublimation temperature for methylphosphine (CH₃PH₂), while the second peak occurred coincidentally with diphosphine (P₂H₄) sublimation. Here, a significant portion of CH₃PH₂ remained trapped in the P₂H₄ matrix, which was the most abundant product, and was released when P2H4 sublimed. Neither CH₃PH₂ nor CH₃P₂H₃ showed evidence of fragmentation. However, considering the next member, nearly half of the sublimed CH₃P₃H₄ fragmented into CH₃P₂H₂⁺ and CH₃P₂H⁺, which result from PH₂ and PH₃ loss. The heaviest molecular ion observed for methylphosphanes was CH₃P₄H₅, and the combined PH₂ and PH₃ loss fragments from CH₃P₄H₅ had a 4:1 ratio compared to the parent ion. All methylphosphanes including CH₃P₄H₅ and larger sublimed at similar temperatures as their non-methylated phosphane analogues, which complicates fragmentation analysis because P₃H₃⁺ and P₃H₄⁺ could be fragments from both CH₃P₄H₅ and P₄H₆. However, given the observed ratios of P₂H₄⁺ to CH₃P₂H₃⁺ (6:1) and P₃H₅⁺ to CH₃P₃H₄⁺ (20:1), it is likely that P₄H₆ formed in larger amounts than CH₃P₄H₅ and thus the P₃H₄⁺ and P₃H₃⁺ fragments originated predominately from P₄H₆, which itself had a parent ion ratio of 8:1 with CH₃P₄H₅⁺. The parent ion for CH₃P₅H₆ was not observed, which is not surprising given the previously stated ratios between phosphanes and their methylated equivalents and that the intensity of P₅H₇⁺ was diminished due to increased fragmentation. However, CH₃P₅H₆ can still be detected since it sublimed at a similar peak temperature of 212 K as P₅H₇ and the major PH₂ loss fragment, CH₃P₄H₄⁺, was easily observed. This method also worked for the larger methylphosphanes CH₃P₆H₇, CH₃P₇H₈, and CH₃P₈H₉, which were identified using their fragments from PH₂ loss that appeared concurrently with the fragments from P₆H₈, P₇H₉, and P₈H₁₀, respectively. Thus, the irradiated methane-phosphine ices produced a series of phosphanes from P₂H₄ to P₈H₁₀ and methylated phosphanes from CH₃PH₂ to CH₃P₈H₉. Notably, neither alkylphosphanes more complex than methylphosphanes nor pure hydrocarbons were detected, which was likely a result of the three-to-one phosphine-to-methane ratio in the ice mixture.

While the larger phosphanes showed similar peak sublimation temperatures with their methylated analogues, the temperature at the onset of sublimation was distinct. Figure 5 and Table 3 show how onset sublimation temperatures increase with atomic weight and how, from PH₃ to P₈H₁₀ and from CH₃PH₂ to CH₃P₈H₉, the amount of temperature increase declined with each successive member in the series. For PH₃ to P₃H₅, the onset sublimation temperatures were 17–20 K lower than their corresponding methylated form: 69 K versus 87 K for PH₃/CH₃PH₂, 98 K versus 118 for P₂H₄/CH₃P₂H₃, and 131 K versus 148 K for P₃H₅/CH₃P₃H₄. The P₄H₆/CH₃P₄H₅ pair had a slightly lower difference of 15 K at 160 K versus 175 K, while P₅H₈ (185 K), P₆H₈ (206 K), P₇H₉ (231 K), and P₈H₁₀ (252 K) each began subliming 12 K below their methylated forms. Since higher order products were not observed via their parent ions, the fragments discussed previously were utilized to determine the onset sublimation temperature of the parent. The regression curves in Figure 5 used only the temperatures for directly observed ions, i.e., PH₃ through P₅H₇ and CH₃PH₂ through CH₃P₄H₅, and the curves were fit forward to higher order compounds. The regressions curves show good agreement with the assignments of parent ions from their fragments, which support the use of fragments when the intensity of parent ions is below the detection limit.

Figure 5.

Onset sublimation temperatures for the phosphanes (blue) and alkylphosphanes (red) observed in the ReTOF mass spectrometer.

These ReTOF findings are consistent with the results when CH₄ was substituted with CD₄ in the phosphine ices (Figure 6 and Table 4) with the highest observed molecular ions at m/z = 147 (CD₃P₄H₅⁺) and 162 (P₅H₇⁺). Using the fragments at m/z = 210 (CD₃P₆H₆⁺) and 225 (P₈H₁₀⁺), we inferred the largest products formed in these ices were CD₃P₇H₈ and P₈H₁₀. However, additional mass-to-charge ratios appear associated with the various isotopologues of the methylphosphanes. For example with the simplest product, methylphosphine (CH₃PH₂), three isotopologues appeared in a 2:10:1 ratio: m/z = 50 (CHD₂PH₂), 51 (CD₃PH₃), and 52 (CD₃PHD). Section 4.2 discusses these results in detail.

Figure 6.

Reflectron time-of-flight mass spectra for phosphine (PH₃) and deuterated methane (CD₄) irradiation. Colored bands indicate sublimation events at similar temperatures. The intensity is listed on the left of each spectrum, while the mass-to-charge and ionic formula is on the right.

Table 4.

Observed Ions in the ReTOF Mass Spectrometer for the Phosphine (PH₃) and Deuterated Methane (CD₄) Irradiation

Mass	Formula	Comments	Formula of Parent Compound
35	PH4+	fragment	P3H5, P4H6, CD3P4H5
50	CHD2PH2+	parent	CHD2PH2
51	CD3PH2+	parent	CD3PH2
52	CD3PHD+	parent	CD3PHD
64	P2H2+	fragment	P3H5
65	P2H3+	fragment	P3H5
66	P2H4+	parent	P2H4
67	P2H5+	protonated parent	P2H4
		fragment	P5H7
81	CH2DP2H3+	parent	CH2DP2H3
	CHD2P2H2+	fragment	CHD2P3H4
82	CHD2P2H3+	parent	CHD2P2H3
	CD3P2H2+	fragment	CD3P3H4
83	CD3P2H3+	parent	CD3P2H3
84	CD3P2H2D+	parent	CD3P2H2D
96	P3H3+	fragment	P4H6
97	P3H4+	fragment	P4H6
98	P3H5+	parent	P3H5
99	P3H6+	protonated parent	P3H5
111	CH3P3H3+	fragment	CH3P4H5
112	CH3P3H4+	parent	CH3P3H4
	CH2DP3H3+	fragment	CH2DP4H5
113	CH2DP3H4+	parent	CH2DP3H4
	CHD2P3H3+	fragment	CHD2P4H5
114	CHD2P3H4+	parent	CHD2P3H4
	CD3P3H3+	fragment	CD3P4H5
115	CD3P3H4+	parent	CD3P3H4
	CD3P3H2D+	fragment	CD3P4H4D
126	P4H2+	fragment	P6H8
127	P4H3+	fragment	P6H8
128	P4H4+	fragment	P5H7, P6H8
129	P4H5+	fragment	P5H7, P6H8
130	P4H6+	parent	P4H6
		fragment	P5H7
131	P4H7+	protonated parent	P4H6
		fragment	P5H7
145	CHD2P4H4+	fragment	CHD2P5H6
146	CD3P4H4+	fragment	CD3P5H6
147	CD3P4H5+	parent	CD3P4H5
159	P5H4+	fragment	P6H8, P7H9
160	P5H5+	fragment	P6H8
161	P5H6+	fragment	P6H8
162	P5H7+	parent	P5H7
		fragment	P6H8
177	CHD2P5H5+	fragment	CHD2P6H7+
178	CD3P5H5+	fragment	CD3P6H7+
193	P6H7+	fragment	P6H8
210	CD3P6H6+	fragment	CD3P7H8
225	P7H8+	fragment	P8H10+

4. ANALYSIS

4.1. Quantitative Analysis—Mass Balance

Although the infrared results provide limited information about the identity of the products, the significant changes in the area of the reactant peaks gleans information about the amount of reactants destroyed and the rate of these reactions. Figure 7 compiles the column densities of phosphine and methane during irradiation with an energy flux of 10¹⁴ eV cm⁻² s⁻¹ utilizing the infrared peaks at 987 cm⁻¹ and 4195 cm⁻¹ with integrated absorption coefficients of 5.1 × 10⁻¹⁹ cm molecule⁻¹ and 3.5 × 10⁻¹⁹ cm molecule⁻¹ (Brunetto et al. 2008; Turner et al. 2015). These column densities were fitted with the following first order rate equations:

$${\left[{\text{PH}}_3\right]}_t={\left[{\text{PH}}_3\right]}_{t=0}{e}^{-{\text{k}}_{1}t}$$ (3)

$${\left[{\text{CH}}_4\right]}_{\text{t}}={\left[{\text{CH}}_4\right]}_{t=0}{e}^{-{\text{k}}_{2}t}.$$ (4)

Figure 7.

Column density of phosphine (top) and methane (bottom) as a function of irradiation time at 10¹⁴ eV cm⁻² s⁻¹.

The rate at which phosphine and methane react are described by the rate constants k₁ = 4.7 ± 0.2 × 10⁻⁵ s⁻¹ and k₂ = 3.8 ± 1.7 × 10⁻⁵ s⁻¹. In pure phosphine ices, the rate constant for the destruction of phosphine leading to the formation of diphosphine, which accounted for 89 ± 4% of the products, was found to be k = 4.8 ± 0.1 × 10⁻⁵ s⁻¹, which is in agreement with the rate of destruction of phosphine in phosphine-methane ices.

In total, 6.4 ± 1.6 × 10¹⁷ molecules cm⁻² of phosphine were destroyed, which is a 17 ± 4% loss. A lower percentage of the initial methane reacted, 13 ± 5%, which is equivalent to 1.6 ± 0.6 × 10¹⁷ molecules cm⁻². This resulted in a loss rate of 0.11 ± 0.03 molecules eV⁻¹ for phosphine and 0.09 ± 0.03 molecules eV⁻¹ for methane. Since the v₁₁ band of diphosphine at 1063 cm⁻¹ grew too subtly and was partially overlapped by the v₄ band of phosphine, the temporal profile of diphosphine could not be monitored. However, the before- and after-irradiation spectra were compared to calculate this peak area and estimate the total diphosphine production. Using an integrated absorption coefficient of 7.0 × 10⁻¹⁹ cm molecule⁻¹ (Turner et al. 2015), ± 0.4 × 10¹⁷ molecules cm⁻² of diphosphine were produced. Thus, diphosphine accounted for 50 ± 13% of the phosphorus from phosphine destruction. Given that the irradiated ice had a three-to-one phosphine-to-methane ratio and that diphosphine contained 89% of the reacted phosphorus in pure phosphine ices, proportionally less diphosphine was formed in phosphine-methane ices, which indicates that either phosphine or diphosphine readily reacts with methane.

4.2. Reaction Pathways

4.2.1. Methylphosphine

The ReTOF results from irradiated phosphine and deuterated methane ice provide crucial information regarding the mechanism of formation for methylphosphanes by analyzing the intensities of various isotopologues. Figure 8 shows the possible formation routes that would lead to each of the three observed isotopologues of methylphosphine (CH₃PH₂). To obtain m/z = 50 (CHD₂PH₂), CD₄ has to decompose via the loss of molecular hydrogen or two deuterium atoms to form carbene (CD₂), which has been observed in previous irradiated ice studies (Holtom et al. 2005; Bennett & Kaiser 2007), and then insert into a phosphorus-hydrogen bond of phosphine (reaction (5)). If the carbene is formed in its first excited singlet state (a¹A₁), the insertion is barrierless (Gordon et al. 1987). For m/z = 51 (CD₃PH₂), methane and phosphine each lost a hydrogen or deuterium atom, and the resulting methyl (CD₃) (Kaiser et al. 1997) and phosphino (PH₂) radicals recombined barrierlessly (reaction (6)). Finally, the formation of m/z = 52 (CD₃PHD) mirrors that for CHD₂PH₂ but in this case phosphine lost two hydrogen atoms or molecular hydrogen to create the phosphinidene (PH) radical and then inserted into a carbon-deuterium bond of methane (reaction (7)). Phosphini-dene, like imidogen (NH) (Fueno et al. 1983), is expected to insert barrierlessly in its first excited singlet state (a¹ Δ):

$${\text{CD}}_4\to {\text{CD}}_2+2\text{D}/{\text{D}}_2$$ (5a)

$${\text{CD}}_2+{\text{PH}}_3\to {\text{CHD}}_2{\text{PH}}_2$$ (5b)

$${\text{CD}}_4\to {\text{CD}}_3+\text{D}$$ (6a)

$${\text{PH}}_3\to {\text{PH}}_2+\text{H}$$ (6b)

$${\text{CD}}_3+{\text{PH}}_2\to {\text{CD}}_3{\text{PH}}_2$$ (6c)

$${\text{PH}}_3\to \text{PH}+2\text{H}/{\text{H}}_2$$ (7a)

$$\text{PH}+{\text{CD}}_4\to {\text{CD}}_3\text{PHD}.$$ (7b)

Figure 8.

Retrosynthesis pathways from methylphosphine (CH₃PH₂) identifying the possible formulae from deuterated-methane substituted reactions. The masses shown were observed in the ReTOF.

Therefore, our results provide compelling evidence that methane decomposes not only to the methyl radical, but also to carbene. Likewise, phosphine was found to fragment to the phosphino radical and also to phosphinidene. The ratio of ion intensities for m/z = 50:51:52 is 2:10:1, indicating that radical recombination was the preferred formation pathway with CD₃PH₂ as the most abundant isotopologue. This could either be a result of the methyl and phosphino radicals reacting quickly or that more of these radicals were produced than carbene and phosphinidene.

4.2.2. Methyldiphosphine and Methylenediphosphine

The molecular formula CH₆P₂ can have two structural isomers: the carbon-terminated methyldiphosphine (CH₃P₂H₃) and the carbon-bridging methylenediphosphine (PH₂CH₂PH₂). Figures 9 and 10 show the pathways by which each of these isomers could be formed using the intermediates from the irradiation of methane, phosphine, and methylphosphine. From Section 4.2.1, it was shown that the isotopologues of methylphosphine were CHD₂PH₂, CD₃PH₂, and CD₃PHD. Noting that non-deuterated CH₆P₂ has m/z = 80, the observed peak at m/z = 81 contained only one deuterium atom. Because the results in Section 4.2.1 indicate that the deuterated methylidyne radical (CD) was not formed, carbene (CD₂) must be involved and methylphosphine (CHD₂PH₂) would first be formed by reaction (5). Reaction (8) shows that the loss of a deuterium atom from CHD₂PH₂ formed the core one-deuterium radical compound that recombined with a phosphino radical (PH₂) to form the methylenediphosphine isotopologue PH₂CHDPH₂. Given the starting materials and intermediates available, this is the only reaction pathway that can lead to m/z = 81 and thus the presence of methylenediphosphine is explicitly confirmed:

$${\text{CHD}}_2{\text{PH}}_2\to {\text{CHDPH}}_2+\text{D}$$ (8a)

$${\text{CHDPH}}_2+{\text{PH}}_2\to {\text{PH}}_2{\text{CHDPH}}_2.$$ (8b)

Figure 9.

Retrosynthesis pathways from methyldiphosphine (CH₃P₂H₃) identifying the possible formulae from deuterated-methane substituted reactions. The masses shown were observed in the ReTOF.

Figure 10.

Retrosynthesis pathways from methylenediphosphine (PH₂CH₂PH₂) identifying the possible structures from deuterated-methane substituted reactions. The masses shown were observed in the ReTOF.

The signal for m/z = 82 can be identified as three isotopomers (PH₂CD₂PH₂, PHDCHDPH₂, and CHD₂P₂H₃) with two deuterium atoms that can be formed through several pathways. Each of the three can be formed using CHD₂PH₂. Hydrogen loss from either the carbon or phosphorus atom in CHD₂PH₂ followed by recombination with the phosphino radical is shown in reaction (9) and reaction (10), respectively. Reaction (9) can also be completed by deuterium loss from CD₃PH₂:

$${\text{CHD}}_2{\text{PH}}_2\to {\text{CD}}_2{\text{PH}}_2+\text{H}$$ (9a)

$${\text{CD}}_3{\text{PH}}_2\to {\text{CD}}_2{\text{PH}}_2+\text{D}$$ (9b)

$${\text{CD}}_2{\text{PH}}_2+{\text{PH}}_2\to {\text{PH}}_2{\text{CD}}_2{\text{PH}}_2$$ (9c)

$${\text{CHD}}_2{\text{PH}}_2\to {\text{CHD}}_2\text{PH}+\text{H}$$ (10a)

$${\text{CHD}}_2\text{PH}+{\text{PH}}_2\to {\text{CHD}}_2{\text{P}}_2{\text{H}}_3.$$ (10b)

Also, insertion pathways involving CHD₂PH₂ are available in which the phosphinidene radical inserts into the phosphorus-hydrogen or carbon-phosphorus (reaction (11)) bond, the carbon-hydrogen bond (reaction (12)), or the carbon-deuterium bond (reaction (13)) of CHD₂PH₂:

$${\text{CHD}}_2{\text{PH}}_2+\text{PH}\to {\text{CHD}}_2{\text{P}}_2{\text{H}}_3$$ (11)

$${\text{CHD}}_2{\text{PH}}_2+\text{PH}\to {\text{PH}}_2{\text{CD}}_2{\text{PH}}_2$$ (12)

$${\text{CHD}}_2{\text{PH}}_2+\text{PH}\to {\text{PHDCHDPH}}_2.$$ (13)

A final reaction that results in m/z = 82 has carbene (CD₂) inserting into a phosphorus-hydrogen (reaction (14)) or phosphorus-phosphorus bond (reaction (15)) of diphosphine:

$${\text{P}}_2{\text{H}}_4+{\text{CD}}_2\to {\text{CHD}}_2{\text{P}}_2{\text{H}}_3$$ (14)

$${\text{P}}_2{\text{H}}_4+{\text{CD}}_2\to {\text{PH}}_2{\text{CD}}_2{\text{PH}}_2.$$ (15)

In summary, m/z = 82 can be assigned to two isotopomers of d₂-methylenediphosphine (PH₂CD₂PH₂ and PHDCHDPH₂) and one isotopologue of methyldiphosphine (CHD₂P₂H₃). We suggest that PH₂CD₂PH₂ through reaction (9b) is the most abundant contributor to m/z = 82 because CD₃PH₂ is the most abundant isotopologue of methylphosphine and phosphino radicals are readily available in irradiated phosphine-dominant ices.

Two isomers can contribute to m/z = 83: one for methyle-nediphosphine (PHDCD₂PH₂) and another for methyldipho-sphine (CD₃P₂H₃). Irradiation of CD₃PH₂ and CD₃PHD followed by recombination with the phosphino radical can result in either of these isomers depending on which hydrogen or deuterium atom is lost. Reaction ((16a) and (16b)) shows hydrogen and deuterium loss from the phosphorus atom on CD₃PH₂and CD₃PHD, respectively, and in reaction (17) the deuterium atom can be lost from carbon on CD₃PHD:

$${\text{CD}}_3{\text{PH}}_2\to {\text{CD}}_3\text{PH}+\text{H}$$ (16a)

$${\text{CD}}_3\text{PHD}\to {\text{CD}}_3\text{PH}+\text{D}$$ (16b)

$${\text{CD}}_3\text{PH}+{\text{PH}}_2\to {\text{CD}}_3{\text{P}}_2{\text{H}}_3$$ (16c)

$${\text{CD}}_3\text{PHD}\to {\text{CD}}_2\text{PHD}+\text{D}$$ (17a)

$${\text{CD}}_2\text{PHD}+{\text{PH}}_2\to {\text{PHDCD}}_2{\text{PH}}_2.$$ (17b)

The two isomers can also be formed via phosphinidene (PH) insertion into either a carbon-deuterium (reaction (18)) or a phosphorus-carbon/hydrogen (reaction (19)) bond of CD₃PH₂:

$${\text{CD}}_3{\text{PH}}_2+\text{PH}\to {\text{PHDCD}}_2{\text{PH}}_2$$ (18)

$${\text{CD}}_3{\text{PH}}_2+\text{PH}\to {\text{CD}}_3{\text{P}}_2{\text{H}}_3.$$ (19)

The final reaction mechanism (reaction (20)) involves the loss of a hydrogen atom from diphosphine (P₂H₄) and subsequent recombination with the methyl radical (CD₃):

$${\text{P}}_2{\text{H}}_4\to {\text{P}}_2{\text{H}}_3+\text{H}$$ (20a)

$${\text{P}}_2{\text{H}}_3+{\text{CD}}_3\to {\text{CD}}_3{\text{P}}_2{\text{H}}_3.$$ (20b)

Thus, the methylphosphines CD₃PH₂ and CD₂PHD are capable of producing both methylenediphosphine and methyldipho-sphine at m/z = 83, while diphosphine can only lead to methyldiphosphine. The dominant isotopomer is likely CD₃P₂H₃ as due to reaction (16a) and reaction (20). Not only are both radical recombination reactions, which have been shown to be the most favorable mechanism, but also reaction (16a) begins with the most abundant isotopologue of methylphosphine (CD₃PH₂), which then reacts with the radical from the most abundant reactant—the phosphino radical from phosphine. Similarly, reaction (20) shows the most abundant overall product, diphosphine, combining with a radical from the methane reactant.

Three isotopomers could be assigned to m/z = 84, and each originated from CD₃PHD: PHDCD₂PHD, CD₃PDPH₂, and CD₃PHPHD. The insertion of the phosphinidene radical (PH) into a carbon-phosphorus bond or phosphorus-deuterium bond of CD₃PHD is represented by reaction (21), while insertion into a carbon-deuterium or phosphorus-hydrogen bond is shown by reaction (22) and reaction (23), respectively:

$${\text{CD}}_3\text{PHD}+\text{PH}\to {\text{CD}}_3\text{PHPHD}$$ (21)

$${\text{CD}}_3\text{PHD}+\text{PH}\to {\text{PHDCD}}_2\text{PHD}$$ (22)

$${\text{CD}}_3\text{PHD}+\text{PH}\to {\text{CD}}_3{\text{PDPH}}_2.$$ (23)

Thus, phosphinidene insertion can explain each of the possible isotopomers. However, one radical recombination pathway is also possible by hydrogen loss from CD₃PHD followed by recombination with a phosphino radical (reaction (24)):

$${\text{CD}}_3\text{PHD}\to {\text{CD}}_3\text{PD}+\text{H}$$ (24a)

$${\text{CD}}_3\text{PD}+{\text{PH}}_2\to {\text{CD}}_3{\text{PDPH}}_2$$ (24b)

Reactions (21)–(24) provide little information about which product is most likely. While radical recombination has been previously identified as most probable, the low probability of hydrogen being removed from CD₃PHD makes this reaction unlikely. A comparison of the PH-insertion pathways shows that reaction (22) has three bonds available for insertion, reaction (21) has two bonds, and reaction (23) can only occur by insertion into only one bond. Thus, without further information about the ease at which the phosphinidene radical can insert into various bond types, PHDCD₂PHD may be the most abundant product at m/z = 84.

4.2.3. Summary

Some important conclusions from the irradiation of phosphine-deuterated methane ices are as follows.

First, the methylphosphine isotopologues have ratios for m/z = 50:51:52 of 2:10:1, indicating that radical recombination (CD₃PH₂) dominates and that carbene insertion (CHD₂PH₂) is twice as likely as phosphinidene insertion (CD₃PHD).

Second, the signals assigned to deuterated CH₆P₂ have ratios for m/z = 80:81:82:83 of 2:7:25:3. The most abundant product is thus CD₃P₂H₃ (at m/z = 83), which is formed by radical recombination of either CD₃PH with PH₂ (reaction (16)) or CD₃ with P₂H₃ (reaction (20)).

Third, the signal at m/z = 81 confirms that methylenedipho-sphine (PH₂CHDPH₂) must be present and that it formed via radical recombination of CHDPH₂ and PH₂ starting from CHD₂PH₂ and PH₃ (reaction (8)). This provides further evidence of the formation of CHDPH₂ (m/z = 50) via carbene insertion into a phosphorus-hydrogen bond of phosphine (reaction (5)).

Fourth, because the carbon-deuterium bond was cleaved in CHD₂PH₂, the decomposition of CHD₂PH₂ should also result in hydrogen loss to give CD₂PH₂, which can then recombine with the phosphino radical to form PH₂CD₂PH₂ (m/z = 82). Furthermore, carbon-deuterium bond cleavage should similarly occur in CD₃PH₂ and CD₃PHD, which when recombined with the phosphino radical would also form PH₂CD₂PH₂ (m/z = 82) in greater abundance than CHD₂PH₂ and also form PHDCD₂PH₂ (m/z = 83). Thus, compelling evidence exists for the formation three isotopologues of methylenediphosphine.

Fifth, all pathways forming m/z = 84 require CD₃PHD, which not only confirms this isotopologue of methylphosphine but also provides evidence that the phosphinidene radical inserts into a carbon-deuterium bond of methane.

4.3. Energetics

Using experimental reaction energies (Chase 1998), we now consider the energy necessary to form the observed products. The preferred pathway (reaction (6)) toward formation of methylphosphine (CH₃PH₂) requires the recombination of the phosphino (PH₂) and methyl (CH₃) radicals. To remove one hydrogen from their parent compounds, 339 kJ mol⁻¹ (3.51 eV) and 439 kJ mol⁻¹ (4.55 eV) are needed, respectively (reactions (25a) and (26a)). This energy necessary for bond cleavage is supplied by the energetic electrons. The barrierless methyl and phosphino radical recombination (reaction (27)) releases 291 kJ mol⁻¹ (3.01 eV), and thus the reaction energy for bond cleavage followed by methyl and phosphino radical recombination (reaction (28)) is +486 kJ mol⁻¹ (5.04 eV). The overall reaction including molecular hydrogen formation (reaction (29)) is endoergic by 50 kJ mol⁻¹ (0.52 eV). Compared to removing the first hydrogen, only slightly more energy is necessary to remove a second hydrogen from phosphine or methane and form phosphinidene and carbene (reactions (25b) and (26b)). However, an additional 90 kJ mol⁻¹ (0.94 eV) and 38 kJ mol⁻¹ (0.39 eV), respectively, is necessary to promote these radicals to their first excited singlet state (reactions (25c) and (26c)), which allows them to insert barrierlessly. With minimal thermal energy available at 5.5 K, the reaction must involve non-equilibrium chemistry to proceed. Non-irradiated blank experiments showed that no reactions occurred, confirming that thermal chemistry cannot create these products at such low temperatures. Each of the three mechanisms shown to produce methylphosphine—radical recombination, phosphinidene insertion, and carbene insertion—is capable of barrierless methylphosphine formation. The endoergic nature of the reaction indicates that non-equilibrium chemistry initiated by high energy particles such as galactic cosmic rays are necessary in interstellar conditions, but once this initial energy is supplied the reaction proceeds favorably and barrierlessly. Furthermore, these reaction intermediates can spontaneously produce the products seen in this study if they are present in the first monolayer and within close proximity without the need for ionizing radiation nor diffusion through the ice:

$$\begin{array}{cc}{\text{PH}}_3\to {\text{PH}}_2+\text{H}& +339\text{kJ}{\text{mol}}^{-1}\end{array}$$ (25a)

$$\begin{array}{cc}{\text{PH}}_3\to \text{PH}\left(X^3{\mathrm{\Sigma }}^{-}\right)+2\text{H}& +684\text{kJ}{\text{mol}}^{-1}\end{array}$$ (25b)

$$\begin{array}{cc}{\text{PH}}_3\to \text{PH}\left(a^1\text{Δ}\right)+{\text{H}}_2& +339\text{kJ}{\text{mol}}^{-1}\end{array}$$ (25c)

$$\begin{array}{cc}{\text{CH}}_4\to {\text{CH}}_3+\text{H}& +439\text{kJ}{\text{mol}}^{-1}\end{array}$$ (26a)

$$\begin{array}{cc}{\text{CH}}_4\to {\text{CH}}_2\left(X^3{\text{B}}_1\right)+2\text{H}& +897\text{kJ}{\text{mol}}^{-1}\end{array}$$ (26b)

$$\begin{array}{cc}{\text{CH}}_4\to {\text{CH}}_2\left(a^1{\text{A}}_1\right)+{\text{H}}_2& +499\text{kJ}{\text{mol}}^{-1}\end{array}$$ (26c)

$$\begin{array}{cc}{\text{CH}}_3+{\text{PH}}_2\to {\text{CH}}_3{\text{PH}}_2& -291\text{kJ}{\text{mol}}^{-1}\end{array}$$ (27)

$$\begin{array}{cc}{\text{CH}}_4+{\text{PH}}_3\to {\text{CH}}_3{\text{PH}}_2+2\text{H}& +486\text{kJ}{\text{mol}}^{-1}\end{array}$$ (28)

$$\begin{array}{cc}{\text{CH}}_4+{\text{PH}}_3\to {\text{CH}}_3{\text{PH}}_2+{\text{H}}_2& +50\text{kJ}{\text{mol}}^{-1}.\end{array}$$ (29)

5. CONCLUSION

Ices of phosphine with methane and deuterated methane that were irradiated with energetic electrons at 5.5 K produced a homologous series of phosphanes from P₂H₄ to P₈H₁₀ and methylphosphanes from CH₃PH₂ to CH₃P₈H₉. All observed products were phosphorus-containing but the products that also contain carbon have exactly one carbon atom. Because the group frequencies of the products from phosphine and methane irradiation overlap significantly with the parent peaks, FTIR had limited use in this type of study and diphosphine was the only product that could be quantified. Furthermore, quadrupole mass spectrometry with electron impact ionization observed only diphosphine, methylphosphine, methyldiphosphine, and triphosphane. On the other hand, ReTOF mass spectrometry could observe molecular ions as large as P₅H₇⁺ and CH₃P₄H₅⁺, while P₈H₁₀ and CH₃P₈H₉ could be identified from their fragments from PH₂ loss, which is the most common fragmentation pathway. Using deuterated methane results as evidence, a signal at m/z = 81 confirms the formation of methylenediphosphine (PH₂CHDPH₂) and, in addition to m/z = 51 (CHD₂PH₂), the formation of methylphosphine via carbene insertion. The signals at m/z = 52 (CD₃PHD) and m/z = 84 (CD₄H₂P₂) also confirm the formation of methyl-phosphine via phosphinidene insertion. However, radical recombination, which contributed to the intense signals at m/z = 51 (CD₃PH₂) and m/z = 83 (CD₃P₂H₃), was the most likely formation pathway. The overall reaction forming methylphosphine (CH₃PH₂) is endoergic by +50 kJ mol⁻¹ (0.52 eV), which makes this compound unlikely in cold environments that rely solely on thermal chemistry. However, our results show that energetic particles like galactic cosmic rays can induce non-equilibrium chemistry that not only forms methylphosphine but a suite of higher order phosphanes and methylphosphanes, and thus methylphosphine can be expected in cold interstellar environments with sufficient quantities of phosphine and methane. Methylphosphine contains a carbon-phosphorus single bond, which has yet to be observed in the interstellar medium but has been discovered in the C1 to C4 alkylphosphonic acids contained in the Murchison meteorite (Cooper & Cronin 1992), which verifies that the carbon-phosphorus single bond can be produced in extraterrestrial environments, although their ultimate origins remain elusive. These results also have potential implications to the chemistry of planetary atmospheres, as phosphine (Ridgway & Smith 1976; Larson et al. 1980) and methane (Ehrenfreund & Charnley 2000) have been discovered on Jupiter and Saturn. Future work can look into the abundance of individual isomers, such as methylenediphosphine (PH₂CH₂PH₂) and methyldi-phosphine (CH₃P₂H₃), by performing selective VUV photoionization experiments utilizing four-wave difference and sum mixing (Hilbig & Wallenstein 1982; VonDrasek et al. 1988) to further investigate the most likely reaction pathways. Also, more complex mixtures, such as the addition of water or carbon monoxide, can be explored to synthesize potential interstellar compounds in more astrophysically relevant ices.

APPENDIX

The use of an RGA (QMS) is common with experiments that detect products that sublime into the gas phase, and we utilize one in tandem with the ReTOF to compare the sensitivity of these techniques. It should be stressed that molecular sublimation and sputtering from the ice is negligible during the irradiation phase, and previous findings demonstrate that low currents of high energy electrons are inefficient at causing sputtering in low temperature ices at typically 5–10 K (Bahr et al. 2001; Baragiola et al. 2003). During the TPD phase, diphosphine (P₂H₄) contributed the most interesting results in the RGA mass spectra for phosphine and methane (CH₄) irradiated ice (Figure 11 and Table 5). Unlike the ReTOF, P₂H₄ fragmented even down to P₂⁺ via dissociative electron impact ionization. The other products seen, P₃H₅, CH₃PH₂, and CH₃P₂H₃, occurred at low intensities. The results for phosphine with deuterated methane (CD₄) (Figure 12 and Table 6) were identical for P₂H₄ and P₃H₅, and two isotopologues of CH₃PH₂ appeared: a strong m/z = 51 (CD₃PH₂) signal and an m/z = 50 (CHD₂PH₂) signal that barely appeared above background levels. Similar to the ReTOF results, CD₃PH₂ was the most abundant form of methylphosphine, although the RGA intensities are too small for a quantitative comparison. Only tenuous amounts of CD₃P₂H₃ were seen at m/z = 83. Thus, the ReTOF was a far more sensitive mass spectrometry method, as it was capable of detecting 15 products for the phosphine and methane ice mixture compared to only four products using the RGA, and three isotopologues of methyl-phosphine were seen and quantitatively compared using the ReTOF, while the RGA detected only two isotopologues.

Figure 11.

Quadrupole mass spectra of the products from phosphine (PH₃) and methane (CH₄) irradiation. The intensity (in thousands) is listed on the left while the mass-to-charge and ionic formula are on the right.

Table 5.

Observed Ions in the Quadrupole Mass Spectrometer for the Phosphine (PH₃) and Methane (CH₄) Irradiation

Mass	Formula	Comments	Formula of Parent Compound
35	PH4+	fragment	P3H5
48	CH3PH2+	parent	CH3PH2
62	P2+	fragment	P2H4
63	P2H+	fragment	P2H4
64	P2H2+	fragment	P2H4
65	P2H3+	fragment	P2H4, P3H5
66	P2H4+	parent	P2H4
80	CH3P2H3+	parent	CH3P2H3
93	P3+	fragment	P3H5
98	P3H5+	parent	P3H5

Figure 12.

Quadrupole mass spectra for the products of phosphine (PH₃) and deuterated methane (CD₄). The intensity (in thousands) is listed on the left while the mass-to-charge and ionic formula are on the right.

Table 6.

Observed Ions in the Quadrupole Mass Spectrometer for the Phosphine (PH₃) and Deuterated Methane (CD₄) Irradiation

Mass	Formula	Comments	Molecular Formula of Parent Compound
35	PH4+	fragment	P3H5
50	CHD2PH2+	parent	CHD2PH2
51	CD3PH2+	parent	CD3PH2
62	P2+	fragment	P2H4
63	P2H+	fragment	P2H4
64	P2H2+	fragment	P2H4
65	P2H3+	fragment	P2H4, P3H5
66	P2H4+	parent	P2H4
98	P3H5+	parent	P3H5