HYDROXYACETONITRILE (HOCH₂CN) FORMATION IN ASTROPHYSICAL CONDITIONS. COMPETITION WITH THE AMINOMETHANOL, A GLYCINE PRECURSOR

Potassium cyanide, formaldehyde, dichloromethane, and potassium hydroxide were purchased from Sigma Aldrich, stearic acid from Fluka analytical, and ammonia (99.9% purity) from Air Liquide. Dichloromethane was distilled over phosphorus pentoxide. Reagents were mixed in different ratios in a Pyrex vacuum line using standard manometric techniques. The gaseous mixture was deposited at a rate of 6 × 10⁻¹ mol min⁻¹ on a gold-plated surface kept at 40 K with a model 21 CTI cold head. The warming up of the samples was performed at a heating rate of 5 K min⁻¹ using a resistive heater along with a Lakeshore model 331 temperature controller. The infrared spectra of the sample were recorded in reflection mode between 4000 and 600 cm⁻¹ using a Nicolet Magna 750 FTIR spectrometer with an MCT detector. Each spectrum was averaged over one hundred scans with 1 cm⁻¹ resolution. The mass spectra were monitored using an RGA quadrupole mass spectrometer (MKS Microvision-IP plus) as the products are desorbed during a controlled temperature ramp. The ionization source is a 70 eV impact electronic source and the mass spectra are recorded between 1 and 70 amu for a full scan, or recorded for specific mass using the SIM (selected ion monitoring) mode. The VUV irradiation source is an H₂ flux lamp irradiating in a range from 120 nm until a continuum in the visible through an MgF₂ window directly connected on our cryostat. The VUV flux is 2 × 10¹⁵ photons cm⁻² s¹, which is 10¹² times more important than the flux of cosmic rays induced by UV photons in the dense molecular clouds of the interstellar medium.

The hydrogen cyanide (HCN) gas used in these experiments was synthesized directly on the Pyrex line at 10⁻³ mbar using the following protocol (Gerakines et al. 2004). In a manifold, 378 mg of stearic acid (CH₃(CH₂)₁₆COOH, 1.32 mmol) are mixed with 86 mg of potassium cyanide (KCN, 1.32 mmol). The manifold is then directly connected to the Pyrex line and left under vacuum for 24 hr. HCN is synthesized by slowly heating the manifold including CH₃(CH₂)₁₆COOH/KCN and cooled with a water bath (room temperature); after that the pressure sought is reached in the Pyrex line. Usually, 5 mbar of pure HCN are synthesized per experiment, which allows five to six experiments per CH₃(CH₂)₁₆COOH/KCN tube. The hydroxyacetonitrile, also named glycolonitrile, was synthesized following Gaudry (1974). The hydroxyacetonitrile is deposited directly from its synthesis tube. Therefore, the amount deposited can only be estimated from the infrared spectra.

The amount of formaldehyde is obtained from the band at 1717 cm⁻¹ (which has a band strength of $\mathcal {A}$ = 9.6 × 10⁻¹⁸ cm molecule⁻¹). The amount of NH₃ is obtained from the band at 1106 cm⁻¹ ($\mathcal {A}$ = 1.3 × 10⁻¹⁷ cm molecule⁻¹). For the amount of HCN, we consider that all the HCN is converted in CN⁻, since when HCN and NH₃ in excess are mixed in the same ramp and deposited, only the [NH⁺₄ ⁻CN] salt is formed (Danger et al. 2011a). Therefore, the initial ice can be reduced in our experimental conditions to a CH₂O:NH₃:[NH⁺₄ ⁻CN] ice. Furthermore, since the CN⁻ band is difficult to integrate, its amount was obtained from the ammonium band at 1460 cm⁻¹ ($\mathcal {A}$ = 4.4 × 10⁻¹⁷ cm molecule⁻¹). However, as the ammonium band overlaps with the formaldehyde feature (1494 cm⁻¹; $\mathcal {A}$ = 3.9 × 10⁻¹⁸ cm molecule⁻¹), the entire band was integrated and subtracted from the formaldehyde amount obtained from the band at 1717 cm⁻¹, giving the amount of [NH⁺₄ ⁻CN] salt in the ice. The band strength of the hydroxyacetonitrile can be derived from the amount of hydroxyacetonitrile formed during the warming. The amount of formaldehyde consumed has been calculated between 60 K and 80 K from the decrease of the band located at 1717 cm⁻¹ ($\mathcal {A}$ = 9.6 × 10⁻¹⁸ cm molecule⁻¹). The amount of aminomethanol formed has been calculated from its band at 1008 cm⁻¹ ($\mathcal {A}$ = 3.5 × 10⁻¹⁷ cm molecule⁻¹). By subtracting those two values, the amount of hydroxyacetonitrile formed during the warming is obtained. Those values are reported in Table 1. It is important to note that this band strength can only be used for the hydroxyacetonitrile embedded in ices with similar composition to the one used here. Indeed, when the hydroxyacetonitrile is deposited as a pure ice, all the band strengths present large intensity variations depending on the deposit, and consequently the band strength calculated in this experiment cannot be used.

The abundances of aminomethanol and hydroxyacetonitrile have been determined in various environments, such as cold molecular clouds, comets, or young stellar objects, for different ratios of ammonia and cyanide, using a numerical resolution of a system of first-order ordinary differential equations to model the formation and destruction of each species, the water abundance being kept constant (Figure 6 and Equation (1)). These equations can be sorted into three different classes: thermal formation, thermal desorption, and photodestruction. The chemical reaction network and the ordinary differential equation system are shown in Figure 6 and Equation (1). For two-body thermal reactions (k₁ and k₂) and thermal desorptions (k₃, k₅, k₇, k₉), the rate coefficients are given in the form k = Aexp (− E/RT), where T is the ice temperature, A the pre-exponential factor, and E the energy (activation or desorption). Photodissociation rates (k₄, k₆, k₈, k₁₀) are written as k = σf, where σ represents the photodissociation cross-section for a given species and f the ultraviolet radiation field depending on the astrophysical environment (Woodall et al. 2007). Kinetic data are used as well as the photodissociation cross-section obtained in this present work (σ _photo (HOCH₂CN) = 5.8 × 10⁻²⁰ photons⁻¹ cm²) to estimate kinetic rates of the formation (k₁) and photodissociation reactions (k₁₀) of HOCH₂CN. The photodissociation rates of H₂CO (k₄), NH₃ (k₆), and HOCH₂NH₂ (k₈) are also taken into account, using the kinetic rates given in the literature. We also included the desorption rate for all the reactants and all the products. All thermal reaction, photodissociation, and desorption rates are given in Table 2:

$$\begin{eqnarray} &&\left\lbrace \begin{array}{@{}llll}\dsty\frac{d[{\rm H_2CO}]}{dt}=-k_1. [{\rm H_2CO}].[{\rm CN^-}]-k_2.[{\rm H_2CO}].[{\rm NH_3}]\\\ms\ms \quad\quad\quad\quad\quad\quad {-}\;k_3.[{\rm H_2CO}]-k_4.[{\rm H_2CO}]\\\ms\ms \dsty\frac{d[{\rm NH_3}]}{dt}=-k_2. [{\rm H_2CO}].[{\rm NH_3}]-k_5.[{\rm NH_3}]-k_6.[{\rm NH_3}]\\\ms\ms \dsty\frac{d[{\rm HOCH_2NH_2}]}{dt}=k_2. [{\rm H_2CO}].[{\rm NH_3}]-k_7.[{\rm HOCH_2NH_2}]\\\ms \quad\quad\quad\quad\quad\quad\quad\quad {-}\;k_8.[{\rm HOCH_2NH_2}]\\\ms\ms \dsty\frac{d[{\rm CN^-}]}{dt}=-k_1. [{\rm H_2CO}].[{\rm CN^-}]\\\ms\ms \dsty\frac{d[{\rm HOCH_2CN}]}{dt}=k_1. [{\rm H_2CO}].[{\rm CN^-}]-k_9.[{\rm HOCH_2CN}]\\\ms\ms \quad\quad\quad\quad\quad\quad\quad\quad {-}\;k_{10}.[{\rm HOCH_2CN}].\\ \end{array} \right.\nonumber\\ &&[-12pt]\end{eqnarray} \tag{ 1 }$$

Table 2. Parameters for Thermal Formation, Photodestruction, and Desorption Rates

	Species	Rate Constant	$\mathcal {A}$	E	Ref
			(s−1)	(kJ mol−1)
Thermal formation ratea	HOCH2NH2	k1	0.5 × 10−2	4.4	(Bossa et al. 2009)
	HOCH2CN	k2	2.8 × 10−1	3.8	This work
	NH3	k5	3 × 1012	25.5	(Sandford & Allamandola 1993)
Desorption ratea	H2CO	k3	3 × 1012	20	Unpublished
	HOCH2CN	k9	3 × 1012	58	c
	HOCH2NH2	k7	3 × 1012	58	(Bossa et al. 2009)
		Rate Constant	σ (photons−1 cm2)
	HOCH2CN	k10	5.8 × 10−20		This work
Photodestruction rateb	NH3	k6	3.2 × 10−20		(Cottin et al. 2003)
	H2CO	k4	10−19		(Woodall et al. 2007)
	HOCH2NH2d	k8	4.5 × 10−19		(Duvernay et al. 2010)

Notes. ^aThe rate coefficients are given in the form k = Aexp (− E/RT), where T is the ice temperature, A the pre-exponential factor, and E the energy (activation or desorption). ^bPhotodissociation rates are written as k = σf, where σ represents the photodissociation cross-section and f the interstellar ultraviolet radiation field. ^cSince HOCH₂CN and HOCH₂NH₂ have the same desorption temperature under our vacuum conditions, we use the same desorption rate. ^dWe used the experimental value determined for NH₂CH(CH₃)OH.

Download table as:  ASCIITypeset image

The aminomethanol (HOCH₂NH₂) is easily formed at 70 K from an ice mixture containing formaldehyde (CH₂O), ammonia (NH₃), and water (H₂O) (Bossa et al. 2009). This reaction is the first step of the Strecker synthesis toward the aminoacetonitrile formation. However, in order to form the aminoacetonitrile, hydrogen cyanide has to be present in the medium. In aqueous solution, it is known that the aminoacetonitrile (NH₂CH₂CN) formation is counterbalanced by the formation of the hydroxyacetonitrile (HOCH₂CN), which is formed by the direct reaction of hydrogen cyanide on formaldehyde (Taillades et al. 1998). Therefore, in order to estimate the possibility of forming the aminoacetonitrile in astrophysical-like conditions, it is important to investigate the possibility of the competition between the aminomethanol and the hydroxyacetonitrile formation (Figure 1(b)). For that purpose, we deposit at 40 K an ice mixture containing CH₂O:NH₃:[NH⁺₄ ⁻CN]. NH₃ and HCN are mixed in the same vacuum line and then co-deposited with CH₂O (see the experimental section for details). Figure 2 displays one of these ices containing CH₂O:NH₃:[NH⁺₄ ⁻CN] with a ratio 1.0:0.2:0.1. At 40 K, the infrared spectrum is the sum of the individual spectra of the starting materials, indicating the absence of reactivity at this temperature. This ice formed at 40 K is then warmed to 300 K with a temperature ramp of 5 K min⁻¹. Whereas bands of CH₂O (1717 cm⁻¹) and of [NH⁺₄ ⁻CN] (2087 cm⁻¹) decrease to their sublimation at 110 K and 160 K, respectively, some new bands appear at 884 cm⁻¹, 986 cm⁻¹, 1052 cm⁻¹, 1110 cm⁻¹, 1390 cm⁻¹, 1432 cm⁻¹, 1600 cm⁻¹, 1673 cm⁻¹, and 2247 cm⁻¹ (Figure 2, 170 K).

Zoom In Zoom Out Reset image size

At 205 K, only the new infrared bands relative to new products remain (Figure 2). Some of these infrared features are related to aminomethanol (986, 1110, 1432, and 1600 cm⁻¹) (Bossa et al. 2009) and hydroxyacetonitrile (884, 1052, 1673, and 2247 cm⁻¹) as displayed in Table 1, and as shown by the dashed and dotted lines in Figure 3(a) for HOCH₂NH₂ and HOCH₂CN, respectively. Infrared analysis at solid state cannot be directly reduced to the sum of the spectra of individual components. In order to strengthen the presence of aminomethanol and hydroxyacetonitrile in the solid, we perform a complementary analysis using mass spectrometry analysis. The mass spectrum (Figure 3(b)) is monitored at 214 K when both HOCH₂NH₂ and HOCH₂CN sublimate. On this spectrum, two different mass patterns are present: one between m/z = 54 and 57, and another one between m/z = 45 and 47. The comparison of the relative intensity of these patterns with the ones of pure HOCH₂CN and pure HOCH₂NH₂ (Figure 3(b)) confirms the previous infrared attributions, since m/z = 54, 55, 56, and 57 are related to the fragment pattern of HOCH₂CN, and m/z = 46, and 47 to that of HOCH₂NH₂ (Figure 3(b)).

Zoom In Zoom Out Reset image size

Those results demonstrate that from an H₂O:NH₃:[NH⁺₄ ⁻CN] ice mixture, both aminomethanol and hydroxyacetonitrile are formed (Figure 4). However the ratio between them is directly connected to the initial composition of the ice mixture, and more precisely to the initial NH₃/CN⁻ ratio. Indeed from an ice mixture containing H₂O:NH₃:[NH⁺₄ ⁻CN] in proportion 1.0:0.2:0.1, the branching ratio has been estimated to be 87% in favor of the hydroxyacetonitrile (see the experimental section). When the amount of ammonia is decreased in the initial ice, as for an ice mixture containing H₂O:NH₃:[NH⁺₄ ⁻CN] in proportion 1.0:0.1:0.2, the reaction evolves in favor of the hydroxyacetonitrile giving a branching ratio of 97% for the hydroxyacetonitrile. Therefore, the branching ratio between hydroxyacetonitrile and aminomethanol effectively depends on the relative amount of [NH⁺₄ ⁻CN] salt compared to ammonia in the initial ice (Figure 4).

Zoom In Zoom Out Reset image size

The activation energy for the hydroxyacetonitrile formation has been estimated from an ice mixture containing only traces of free ammonia. Consequently, it can be assumed that, during the ice warming, the formaldehyde will only be consumed by hydroxyacetonitrile formation. This ice has been warmed at 5 K min⁻¹, and the evolution of the formaldehyde amount has been followed using its band at 1717 cm⁻¹ ($\mathcal {A}$ = 9.6 × 10⁻¹⁸ cm molecule⁻¹). By tracing this evolution as a function of the temperature (Figure 5) and fitting using Equation (2), the activation energy for the hydroxyacetonitrile formation has been obtained. The corresponding value is 3.86 kJ mol⁻¹. This value is lower than that obtained for the aminomethanol, which is 4.5 kJ mol⁻¹ (Bossa et al. 2009), meaning that in our experimental conditions hydroxyacetonitrile formation is favored over aminomethanol:

$$\begin{equation} \left\lbrace \begin{array}{@{}llll}\dsty\frac{d[{\rm H_2CO}]_t}{dt}=-k_T. [{\rm H_2CO}]_t\\\ms\ms [{\rm H_2CO}]_t= [{\rm H_2CO}]_0. e^{-k.t}\\\ms T=T_0 + \beta t \;\;{\rm and}\;\; k_T=A_0 e^{-\frac{E_a}{R.T}} \\\ms [{\rm H_2CO}]_t= [{\rm H_2CO}]_0. e^{\frac{A_0.(T-T_0)}{\beta }.e^{-\frac{E_a}{R.T}}},\\\ms \end{array} \right. \end{equation} \tag{ 2 }$$

with T being temperature (K), E_a the activation energy (kJ mol⁻¹), A₀ the pre-exponential factor (s⁻¹), β the ramp rate (K s⁻¹), and the gas constant R = 8.31 J K⁻¹ mol⁻¹.

Zoom In Zoom Out Reset image size

In order to estimate the efficiency of the aminonitrile formation through the Strecker reaction, it is important to take into account the concurrent reactions that could occur (Taillades et al. 1998). This is particularly important for the first step of the Strecker reaction in the condensed phase, which concerns the formation of aminoalcohols from aldehydes/ketones and ammonia (Figure 1(b)). In the general scheme of the Strecker reaction, the starting material includes aldehydes/ketones, ammonia, and hydrogen cyanide. Consequently, the condensation of ammonia with aldehyde or ketone is in direct competition with the condensation of hydrogen cyanide with aldehyde. Therefore, aminoalcohol formation occurs with hydroxynitrile formation (Figure 4).

In this contribution, we used formaldehyde, ammonia, and hydrogen cyanide, three molecules detected in various astrophysical environments, and clearly demonstrated that aminomethanol formation with hydroxyacetonitrile formation. We have also tested these reactions in the presence of a small amount of water, and the same reaction products were obtained. Activation energies of formation of these two compounds are relatively similar (Table 2: 3.8 kJ mol⁻¹ for the hydroxyacetonitrile and 4.4 kJ mol⁻¹ for the aminomethanol; Bossa et al. 2009; Duvernay et al. 2010). This implies that the ratio between the aminomethanol and hydroxyacetonitrile formed will be mainly determined by the initial ratio of the amount of ammonia to cyanide. Consequently, since the formation of the aminoacetonitrile depends on the remaining cyanide in the solid phase (Figure 1(b), reaction of cyanide with imine), it is important to understand the influence of the initial ratio between ammonia and cyanide on hydroxyacetonitrile and aminomethanol formation. In order to estimate the evolution of the relative abundances of aminomethanol and hydroxyacetonitrile in more realistic interstellar ices, we simulate the evolution of the formation of these compounds as a function of time (Figures 6–8) using a restricted chemical system that gives information on the competition between the different processes (experimental section, Equation (1)).

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Figure 7(a) displays the evolution of thermal processes leading to the formation of hydroxyacetonitrile and aminomethanol from the initial ice, which contains NH₃, H₂CO, and CN⁻ in proportion 10%, 1%, and 1% with respect to H₂O, respectively, at a constant temperature of 30 K without UV photons. The initial amounts of NH₃ and H₂CO have been chosen from observational data (Dartois 2005) while we chose the upper limit of detection for CN⁻. At 30 K, thermal production of hydroxyacetonitrile and aminomethanol begins in less than one year, and reaches a maximum in 10 years. The branching ratio is 80% in favor of the hydroxyacetonitrile, which demonstrates the difference in selectivity between these two compounds, and 0.8% of hydroxyacetonitrile with respect to water is formed. Considering the initial amount of cyanide in the ice (1% with respect to water), the remaining cyanide in the ice is 0.2% with respect to water. Therefore, with this ice mixture, it remains cyanide in order to form the aminoacetonitrile. The total amount of (HOCH₂CN + HOCH₂NH₂) compounds formed represents 1% with respect to water. This implies that all the formaldehyde is consumed, since its initial proportion in the ice is 1%. Furthermore, in a time range coherent with astrophysical environments (up to 10⁷ years), the amounts of HOCH₂CN and HOCH₂NH₂ are constant. Since the thermal desorption is taken into account in the model, this implies that these two compounds remain in the solid phase at the surface of grains.

Figures 7(b) and (c) display the same thermal evolution in the presence of a photon flux of 10³ photons cm⁻² s⁻¹ (Prasad & Tarafdar 1983) and of 10⁸ photons cm⁻² s⁻¹ (Mathis et al. 1983), respectively. Figure 7(b) displays the simulation of the matter evolution inside molecular clouds, whereas Figure 7(c) simulates the evolution that could occur at the edge of molecular clouds, such as in photodissociation regions. In each case, the amount of hydroxyacetonitrile and aminomethanol decreases. Since no thermal desorption has been observed in Figure 7(a), this disappearance is due only to photodegradation processes. In molecular cloud conditions (Figure 7(b)), each compound reaches its maximum of thermal production at the grain surface, and the photodegradation only begins at 3 × 10⁶ years for HOCH₂CN and at 10⁶ years for HOCH₂NH₂. Therefore, in the presence of a low photon flux, the hydroxyacetonitrile and aminomethanol remain stable at the surface of grains. It should be noted that aminomethanol is photodegradated faster than hydroxyacetonitrile. When the photon flux is increased to 10⁸ photons cm⁻² s⁻¹ (Figure 7(c)), both compounds also reach their maximum thermal formation after 10 years. The aminomethanol is totally photodegradated after 2200 years, whereas the hydroxyacetonitrile totally disappears from the grain surface after 15,000 years. Consequently, hydroxyacetonitrile and aminomethanol could be detected at the grain surface inside molecular clouds. In photodissociation regions, the photon flux is more important, leading to a rapid photodegradation, which limits their lifetime at the grain surface. Furthermore, in order to define the effect of the initial composition of ices, a simulation is performed with an ice containing at 30 K NH₃, CH₂O, and CN⁻ in proportion 10%, 1%, and 0.1% with respect to water, respectively. As shown in Figure 8, when the amount of CN⁻ is decreased by a factor of 10, the branching ratio is no longer in favor of the hydroxyacetonitrile, but at 90% favors the aminomethanol. Using these conditions, the formaldehyde is also entirely consumed, since amounts of HOCH₂CN and HOCH₂NH₂ represent 1% of matter relative to water, which is the initial amount of formaldehyde. Furthermore, the amount of hydroxyacetonitrile is 0.1% with respect to water, which means that all the cyanide is consumed. In this environment, there is no cyanide to perform the aminoacetonitrile formation. Another reservoir of cyanide thus has to be taken into account. Consequently, as observed experimentally, the initial ratio in the ice between ammonia and cyanide is the major parameter that balances the ratio between hydroxyacetonitrile and aminomethanol formed at the grain surface. Furthermore, the possibility of forming the aminoacetonitrile in this ice also strongly depends on this initial ratio, since its formation is directly related to the cyanide remaining in the ice after this first reaction step (Danger et al. 2011a).

Figure 9 presents the different environments in which amino acids, hydroxyacids, and their precursors could form. Some hydroxyacids and amino acids detected inside meteorites (Pizzarello et al. 2010) could originate from their grain syntheses inside molecular clouds. Hydroxynitriles and aminoalcohols can also be formed inside comets and asteroids, since formaldehyde, ammonia, and hydrogen cyanide can be incorporated inside such objects when planetary systems are formed. Consequently, hydroxynitriles and aminoalcohols could have been formed in various astrophysical environments showing important variability in temperature, pressure, or hydration conditions. This variability could explain the differences observed in their subsequent products detected in meteorites (Throop 2011), which are hydroxyacids and amino acids (Pizzarello et al. 2010). It is obvious that the Strecker reaction is not the only pathway that leads to the formation and detection of amino acids inside meteorites. The Michael addition has been proposed for the formation of some amino acid derivatives. Further reactions can lead to their formation without the requirement of an aqueous phase, such as the Fischer–Tropsch-type catalysis (Glavin et al. 2010). Furthermore, experimental simulations of interstellar or cometary ices have shown that amino acids could be formed from various chemical pathways (Elsila et al. 2007). This is confirmed by the fact that, depending on the chemical treatment applied to the organic residue obtained at the end of these experiments (Bernstein et al. 2002; Caro et al. 2002), various amounts of amino acids are detected. Before hydrolysis, only small amounts are identified, while after the residue hydrolysis, the quantity of amino acids is much more important (Meinert et al. 2012). The same behavior is observed during meteorite analyses (Glavin et al. 2010). Therefore, some amino acids could be formed directly through reactions like the Strecker synthesis, whereas others could be obtained after the hydrolysis of macromolecules, such as organic residues obtained from experimental simulations of cometary ices, or from the soluble organic matter of meteorites.

Zoom In Zoom Out Reset image size