Synthesis of Nitrogen Oxyfluorides at Low Temperature Monitored by Barometry and Infrared Spectroscopy

Abstract

Nitryl fluoride (FNO₂) and nitrosyl fluoride (FNO) are two nitrogen oxyfluorides that are rarely described in the literature. Their respective syntheses, by combination of nitrogen oxide (NO₂ or NO) and molecular fluorine (F₂), are known to be particularly energetic, which can lead to a flame even at low temperatures. This reactivity, however, has not prevented their use to synthesize exotic inorganic compounds. The simultaneous presence of nitrogen oxides, fluorine, and the nitrogen oxyfluorides formed, all gaseous and very reactive, strongly limits the number of analytical techniques available to precisely monitor the reaction. This article proposes to revisit the synthesis of these two compounds, by very precise monitoring of the pressure inside the reactor coupled with infrared spectroscopy, in a setup specifically built and dedicated to their study.

1. Introduction

The very first mention of nitryl fluoride (FNO₂) dates back to the work of Moissan and Lebeau in 1905, obtained from nitrogen monoxide (NO) and fluorine (F₂) at room temperature according to a very strongly exothermic reaction. , At the same time, Ruff and Stäuber described the same compound, prepared from a mixture of nitric acid (HNO₃) and hydrogen fluoride (HF), but also another compound nitrosyl fluoride (FNO), prepared from silver fluoride and nitrosyl chloride. Their preparation posed many problems, due to the high reactivity of the compounds obtained with glass (SiO₂) and paraffin used in their devices. Moissan’s work was reproduced in 1932 by Ruff et al. and showed that the main product of the reaction between nitric oxide (NO) and elemental fluorine (F₂) was nitrosyl fluoride FNO, in contradiction with Moissan’s observations. They speculated that the F₂ used by Moissan was perhaps not free of O₂, and would have led to the formation of FNO₂ to the detriment of FNO. They also mention a reactivity of NO₂ with F₂ to form FNO₂, which had also not been observed by Moissan. These observations were later confirmed by Faloon and Kenna in 1951. Few other synthesis routes are described in the literature to synthesize FNO₂ and FNO. Ogg and Ray proposed using nitrogen pentoxide (N₂O₅) with sodium fluoride (NaF) to prepare FNO₂ at 35 °C. Davis and Rausch passed nitrogen dioxide over cobalt­(III) fluoride (CoF₃) at 300 °C, yielding FNO₂ (purity: 89.5%). Ratcliffe and Shreeve succeeded in synthesizing FNO with nitrogen dioxide (NO₂) mixed with alkaline fluoride (CsF, KF). The reaction is slow and incomplete, although it can be speeded up by heating the mixture at 90 °C. Styger et al. suggest using a pulsed electric discharge on mixtures containing SF₆ and NO _x (NO or NO₂) to obtain FNO or FNO₂. As a result, reactions between NO and NO₂ with F₂ were favored for preparing these compounds. Although having no known application in industry, FNO and FNO₂ were occasionally mentioned for carrying out the synthesis of exotic compounds such as nitronium or nitrosylium salts of halides, , transition metals, − actinides, − platinum group metals, , xenon, used as reagent in organic chemistry, − or proposed as an etching agent for silicon. − Beyond the use of these compounds as fluorinating agents in our laboratory for multiple applications, the aim of this work is to re-examine the synthesis of these compounds from NO _x and F₂. An experimental setup was specifically built to allow the synthesis and study of these compounds in a highly controlled environment.

2. Materials and Methods

Caution! Fluorine (F₂), nitrogen dioxide (NO₂), and nitric oxide (NO) to a lesser extent are toxic, corrosive and oxidizers (Immediately dangerous to life or health concentrations: 25, 13, and 100 ppm, respectively), yielding hydrofluoric acid (HF) and nitric acid (HNO₃) upon contact with water or moisture. Reactions between fluorine and nitrogen oxides are extremely exothermic and yield nitrosyl fluoride (FNO) and nitryl fluoride (FNO₂), whose have a reactivity and toxicity close to that of fluorine. The use of fluorine and nitrogen oxides should only be allowed if all safety precautions are scrupulously observed.

2.1. Experimental Setup

Due to the toxic and corrosive nature of the gases used, a dedicated setup was built that meets all the criteria in terms of current safety standards. Extensive care has been taken to guarantee the lowest possible leak rate (<0.1 mbar/h; admissible standard: 4 mbar/h). A schematic representation of the assembly is detailed in Figure . The tubes are made of 316L steel (Swagelok) in 1/4 in. diameter (internal diameter: 0.18 in.), as well as bellows valves (Swagelok) and the ferrules for tube fitting (Swagelok). 316L steel, already used on other fluorination setups present in our laboratory, resists fluorine by forming a passivation layer as long as the temperature does not exceed 100 °C. Higher temperatures would have required Nickel or Nickel-based alloys (Monel) to withstand hot fluorine up to 600 °C. All elements have undergone careful degreasing before installation. Part of the assembly was installed in an UN110 oven (Memmert). The gas cylinders being stored outside the assembly, two NO _x and F₂ tanks, made from 1/4 in. spiral tubes, were installed inside the oven to have a reserve of gas at the right temperature. The temperature set point for the oven was 80 °C, however, measurement by thermocouples at different points indicate a lower temperature, between 72 and 75 °C. This choice of a set point of 80 °C was made with the aim of minimizing on the one hand the significant temperature differences that could exist in the room and on the other hand to limit the dimerization of nitrogen dioxide (described in Section and beyond), which would require to inject a quantity of fluorine significantly greater than that of nitrogen oxide (well above atmospheric pressure), potentially posing a risk when using the setup. The section of the assembly located outside the oven was traced with heating wire, controlled by a 4-way controller box (Omicron Technologies), which was also set at 80 °C. The temperature measured at different points on the network outside the oven was between 75 and 83 °C. The reactors were machined from a PCTFE rod. This polymer has the advantage of having a very low water vapor transmission rate, of being impermeable to gases, and of being usable up to the temperature of liquid nitrogen. An infrared spectrometer (Nicolet Summit X) equipped with a temperature-controlled Mercury gas cell (cell path: 10 cm) completes the setup. The infrared windows are made of CaF₂, resistant to F₂, allowing the recording of spectra over a wavenumber range from 1000 to 4500 cm^–1 with a 0.5 cm^–1 resolution and performing 16 scans (recording time: 1 min). A cold trap, made from a machined stainless-steel rod, separates the setup and a dry scroll vacuum pump (nDXS series, model 10iC, Edwards) resistant to corrosive atmospheres was installed. Digital pressure sensors (Keller Series 33X) were installed at key points (injection loop and PCTFE reactor) to monitor and record pressure evolution. A passive trap filled with soda lime completes the setup. All the setup has undergone several passivation steps with F₂ gas before use.

1.

Schematic representation of the setup.

2.2. The NO₂–N₂O₄ Equilibrium

Nitrogen dioxide (NO₂) is in permanent equilibrium with its dimer, dinitrogen tetroxide (N₂O₄), according to eq :

$$2\mathrm{N}{\mathrm{O}}_2\rightleftarrows {\mathrm{N}}_2{\mathrm{O}}_4$$ 1

The dimerization constant K ₂ and Gibbs free energy of the reaction Δ_r G° are linked according to the following relationship (eq ). Δ_r G° is temperature dependent. A first approximation consists of expressing it as a function of the standard molar entropies S° ₂₉₈ and the standard enthalpies of formation Δ_f H° ₂₉₈ of the two species from data tabulated in the literature (eq ). Figure S1 on the Supporting Information shows the percentage difference between the dimerization constant K ₂ expressed according to eqs and and that expressed from the formula proposed by Chao et al., over the range 262.15– 373.15K. At 73.5 °C, the relative difference is only 0.51%, which allows the application of the simplified eqs and :

$$K_2={\mathrm{e}}^{-{\mathrm{\Delta }}_{\mathrm{r}}G°/\mathit{RT}}$$ 2

$${\mathrm{\Delta }}_{\mathrm{r}}{G}^{°}={\mathrm{\Delta }}_{\mathrm{f}}{H}_{298}^{°}({\mathrm{N}}_2{\mathrm{O}}_4)-2{\mathrm{\Delta }}_{\mathrm{f}}{H}_{298}^{°}(\mathrm{N}{\mathrm{O}}_2)-T(S_{298}^{°}({\mathrm{N}}_2{\mathrm{O}}_4)-2S_{298}^{°}(\mathrm{N}{\mathrm{O}}_2))$$ 3

The dimerization constant K ₂ and the thermodynamic activity a of the species are related according to eq :

$$K_2=\frac{a_{{\mathrm{N}}_2{\mathrm{O}}_4}}{{a_{{\mathrm{NO}}_2}}^2}$$ 4

Thermodynamic activity a is directly linked to their pressure, denoted P, which can be expressed according to the ideal gas law. Figure S2 shows the deviation between the pressure of gases NO₂ and N₂O₄ considered ideal compared to that calculated from the van der Waals equation for amount of substance (1 mmol) and volumes (50 mL) of the order of magnitude of the experiments carried out. It is sufficiently small (0.37% for N₂O₄ and 0.18% for NO₂ at 80 °C) to consider, in the context of the study, that NO₂ and N₂O₄ are ideal gases. It is the same for F₂ and NO (0.08 and 0.10%, respectively). The expression for K ₂ is then simplified (eq ):

$$K_2=\frac{P_{{\mathrm{N}}_2{\mathrm{O}}_4}/P^{°}}{{P_{{\mathrm{NO}}_2}}^2/P^{°2}}={\mathrm{e}}^{-{\mathrm{\Delta }}_{\mathrm{r}}G°/\mathit{RT}}$$ 5

The measured pressure P _mes is equal to the sum of the partial pressures of N₂O₄ and NO₂, by application of Dalton’s law (eq ). The use of the two previous equations leads to the resolution of the following quadratic equation (eq ), whose unique chemically possible solution (the second mathematical solution is a strictly negative number) is expressed according to eq .

$$P_{\mathrm{meas}}=P_{{\mathrm{NO}}_2}+P_{{\mathrm{N}}_2{\mathrm{O}}_4}$$ 6

$${\mathrm{e}}^{57240-175.9T/\mathit{RT}}{P_{{\mathrm{NO}}_2}}^2+P_{{\mathrm{NO}}_2}-P_{\mathrm{meas}}=0$$ 7

$$P_{{\mathrm{NO}}_2}=\frac{-1+\sqrt{1+4P_{\mathrm{meas}}{\mathrm{e}}^{57240-175.9T/\mathit{RT}}}}{2{\mathrm{e}}^{57240-175.9T/\mathit{RT}}}$$ 8

It is then possible to plot the NO₂ fraction $x=\frac{P_{{\mathrm{NO}}_2}}{P_{\mathrm{meas}}}$ as a function of the measured pressure P _mes for different temperatures (Figure S3). At a given temperature, the fraction x is strongly dependent on the pressure: the lower it is, the more the equilibrium favors the NO₂ species.

2.3. Gases Used

Nitrous oxide (N₂O, Linde) is quality 4.8, i.e., purity greater than 99.998%. Infrared spectra do not show any impurities detectable (Figure S4a). Nitrogen dioxide (Linde) is quality 2.0, i.e., purity greater than 99%. Its infrared spectrum (Figure S4b) shows in addition to the characteristic bands of the molecule, an additional band centered on 1734 cm^–1 characteristic of N₂O₄ dimer, according to the balance established by the previous eq . Nitric oxide (NO) and nitric acid (HNO₃) are also detected in trace amounts (less than 0.5%). Nitric oxide (NO) has a quality of 2.5, i.e., a purity greater than 99.5%. However, infrared analysis revealed the significant presence of NO₂ and N₂O (Figure S4c). The literature reports that NO disproportionate at high pressures from 30 °C according to the following reaction (eq ):

$$3\mathrm{NO}\to \mathrm{N}{\mathrm{O}}_2+{\mathrm{N}}_2\mathrm{O}$$ 9

Tsukahara et al. have reviewed the reaction’s kinetics at different pressures and at 25 °C, a temperature easily reached during storage. The conversion rate reaches 0.045 (or 4.5%) after one year for an initial NO pressure of 20 bar. In our case, the operating pressure of the NO cylinder is 33 bar, stored in an air-conditioned room (21 °C). By performing calibration curves on characteristic peaks of N₂O (ν­(NN)), NO₂ (ν_as(NO)) and NO, with consideration of the existing balance between NO₂/N₂O₄, one can estimate NO purity at 95%, the remaining 2.5% being made up of NO₂ and N₂O in equal parts. Purification, which will be explained in Section 2.4, made it possible to improve the purity of NO, by reducing the quantity of NO₂ by 80% and that of N₂O, more volatile, by almost 50% (composition: 98% NO, 1.5% N₂O, 0.5% NO₂). The purity of the elemental fluorine F₂ used is guaranteed at 98% according to its manufacturer (Solvay). Analysis by infrared spectroscopy at 80 °C shows HF, CF₄ and SiF₄ as impurities (Figure S4d).

2.4. Experimental Procedure

The condensation of gases takes place according to the following procedure. The cold trap is in operation, the PCTFE reactor is cooled to the temperature of liquid nitrogen with a Dewar, and the vacuum is reestablished. Nitrogen is first admitted, and the vacuum is created again. Several cycles are carried out to eliminate any impurities that may have been adsorbed in the pipes (water vapor, but also HF or NO _x ). NO _x is then admitted into the injection loop (V = 54.3 mL) and the pressure is recorded. It is then condensed into the reactor (V = 44.7 mL). The reactor is closed, and the upper part of the setup is flushed with nitrogen, before being put back under vacuum. F₂ is then admitted in excess into the injection loop. Due to the high volatility of F₂ even at the temperature of liquid nitrogen (P _sat = 364.2 mbar at −196 °C), the initial pressure of F₂ must be at least equal to that of the NO _x injected to prevent any gas backflow. It is then slowly injected into the reactor. The pressure difference measured before and after the injection then makes it possible to precisely determine the quantity of F₂ injected. The reactor is naturally returned to room temperature by removing the Dewar. The gas mixture obtained can then be analyzed by infrared spectroscopy. An example of pressure monitoring before the reactivity study is detailed in Figure S5 and allows to visualize all the steps upstream. For nitric oxide, a prepurification step was carried out by injecting a large excess of NO by performing several injection cycles, before recovering some of the vapor phase (less than 1/4 of the total NO injected) into the injection loop as the NO vaporizes after removing the Dewar. Excess NO is removed by flushing the reactor with nitrogen. This procedure is very effective for NO₂ removal (over 80% decrease) due to the major difference in the boiling points of the constituents: −152 °C for NO and 21 °C for NO₂/N₂O₄. N₂O removal is less effective (50% decrease) because of its lower boiling point (−88 °C).

3. Results and Discussion

3.1. NO₂/N₂O₄ + F₂

The NO₂/N₂O₄ equilibrium is important to determine the minimum quantity of fluorine to inject into the reactor. N₂O₄ being the dimer of NO₂, it can either directly participate in the reaction with fluorine, or dissociate to form NO₂, whose reactivity with F₂ is proven. In the hypothetical case of a mixture consisting solely of NO₂, 0.5 equiv of F₂ are necessary to allow a complete reaction. Considering the NO₂/N₂O₄ pressures used during the study (between 100 and 800 mbar) and the injection temperature (at least 70 °C), the minimum fraction x of NO₂ is at least around 0.8 (Figure S3), which means that at least 80% of the mixture is made up of NO₂, the remainder being N₂O₄. In this least favorable scenario, 0.6 equiv of F₂ are necessary to convert all of the nitrogen oxide. It is this value which will be retained subsequently.

Figure a shows the barometric profile of a NO₂/N₂O₄ + 0.6 F₂ reaction (P _NOx = 605.2 mbar; P _F2 = 361.0 mbar; V _inj = 54.3 mL; 73 ± 1 °C) after Dewar removal. It should be compared to the theoretical profile of the NO _x /F₂ mixture for which no reaction would take place. The two curves practically overlap during the first 5 min. They correspond to the natural reheating of F₂ after the withdrawal of the Dewar. In the interval of 5 to 16 min, the measured pressure is significantly higher than that predicted if no reaction would take place. Considering the melting and boiling temperatures of FNO₂ (−166 and −72 °C, respectively according to), this overpressure can only be explained by the presence of this species. The reactivity between N₂O₄ (the only species existing at these temperatures) and F₂ at low temperatures is therefore proven. After 16 min, the measured pressure becomes lower than that predicted, explained by the fact that the stoichiometry balance of the reaction between nitrogen oxide is zero (N₂O₄ + F₂ → 2 FNO₂) or strictly negative (NO₂ + 1/2 F₂ → FNO₂). At thermal equilibrium, the pressure reaches 728.4 mbar (V _reactor = 44.7 mL; 17 °C), which is higher than expected (719.0 mbar). A theoretical material balance can then be established on the system, assuming that the impurities present in the gases do not react with F₂ (Table ). The fraction x in NO₂ (and by extension the P _NO2 and P _N2O4 pressures) contained in the nitrogen oxide mixture can be calculated by solving the previous eq from the P _NOx pressure measured experimentally and at 73 °C. Due to a zero gas balance in the case of the reaction between N₂O₄ and F₂, the parameter β is necessarily equal to n_N2O4. Despite the excess F₂ used, the reaction is not complete (FNO₂ purity: 91.8%, unreacted NO₂: 2.8%). This is confirmed by the infrared analysis of the gas mixture after the reaction, which shows the presence of NO₂. N₂O₄ is completely absent from the spectrum, most likely due to the low pressure of NO₂ (Figure b). Its absence also demonstrates its strong reactivity toward F₂. Comparable results are obtained if we consider that the minor species NO and HNO₃ (less than 0.5% of the quantity of NO₂/N₂O₄ injected each) have reacted with F₂, yielding FNO and FNO₃ respectively (FNO₂ purity: 91.3%, unreacted NO₂: 3.2%).

2.

(a) Barometric profile of the reaction NO₂/N₂O₄ + 0.6 F₂ after Dewar removal. (b) Infrared spectrum of the gas mixture after warming to room temperature.

1. Material Balance (in mmol) of the Reaction NO₂/N₂O₄ + 0.6 F₂ .

	NO2	N2O4	F2	FNO2	impurities (0.01 NO2 and 0.02 F2)
injected at 73 °C	n NO2  = 0.980	n N2O4  = 0.148	n F2  = 0.667	0	0.025
At t + 45 min	n NO2  – α	n N2O4  – β	n F2  – 1/2 α – β	α + 2 β	0.025
	0.038	0	0.047	1.239	0.025

To validate the hypothesis of a NO₂/N₂O₄ reactivity at low temperature, several cycles of condensation and return to ambient temperature are carried out from an initial mixture consisting of 789.3 mbar of NO₂/N₂O₄ and 459.3 mbar of F₂. The reactor was immersed in a water bath thermostatically controlled at room temperature (18 °C) after removal of the Dewar to accelerate the return at RT. At each new equilibrium, the measured pressure is systematically lower than the previous one (Figure a). The infrared analysis carried out on the gas mixture in the fifth cycle does not show the presence of NO₂, proof that all the nitrogen oxides (NO₂ and N₂O₄) have been consumed (Figure b). It is then possible to go back to the theoretical NO _x and F₂ injection temperature, based on a total reaction between NO₂/N₂O₄ and F₂ by performing a mass balance of the reaction after the fifth cycle. Because the equilibrium between NO₂ and N₂O₄ is temperature-dependent, an iteration must be performed to find the theoretical temperature. The result of this iteration gives a calculated temperature of 74.9 °C, close to the one measured experimentally (74 ± 0.5 °C). The injection temperature is comparable if the reactive impurities NO and HNO₃ are considered (74.6 °C).

3.

(a) Barometric profile of a NO₂/N₂O₄ + 0.6 F₂ reaction by performing 5 cycles of cooling-warming to room temperature. (b) Infrared spectrum of the gas mixture after the 5th cycle.

From the previous spectrum, an attribution of the peaks observed over the range 1000–4500 cm^–1 in relation to data from the literature can be proposed (Table ). The modes ν₁ to ν₆ correspond respectively to ν_s(NO), ν­(NF), δ_sc(ONO), ν_as(NO), δ_wag(ONO) and δ_r(ONO). The frequencies of the fundamental vibrations are in agreement with the literature, − regarding the resolution of the spectrometer (0.5 cm^–1). However, wavenumbers of combination bands are significantly lower than reported.

2. Assignment and Frequencies (in cm^–1) of Infrared Absorption Bands of FNO₂ .

assignment	literature −	this work	intensity (%)
2 ν5	1140	1140.8	4.3
ν2 + ν3	1282	not observed
ν1	1312/1310.75	1311.8	63.6
ν2 + ν5	1396	1393.2	4.1
2 ν6	1483	1484.4	0.9
ν4	1793/1793.47	1793.0	100
ν1 + ν5	1875	1874.5	4.1
ν1 + ν6	2045	2043.0	0.1
ν1 + 2ν3	2237	not observed
3 ν2	2450	2450.2	<0.1
ν2 + 2 ν4/2 ν1	2615	2610.7	0.7
ν1 + 2 ν6	2780	2776.8	0.4
ν1 + ν4	3090	3084.4	6.7
ν5 + 2 ν1	3180	3172.0	0.3
ν4 + 2 ν6	3270	3257.1	0.1
2 ν4	3575	3561.6	0.3
3 ν1	3910	not observed
ν4 + 2 ν1	4380	4362.9	0.2

Perrine and Johnston studied the reaction kinetics between NO₂ and F₂ for temperatures between 27.2 and 70.2 °C. According to their work, only the NO₂ species was reactive toward F₂, leading to the formation of FNO₂. Their experiment was reproduced at room temperature (19.0 ± 0.5 °C), with a mixture of approximately 47% NO₂ and 53% N₂O₄ (198.8 mbar; 44.7 mL) and 1 equiv of F₂ (1048.5 mbar; 8.5 mL). The valve separating the F₂ from the nitrogen oxide is then opened quickly. The typical brown coloring of NO₂ disappears instantly, and is associated with a very faint detonation noise coming from inside the reactor. The pressure curve presents an atypical profile, with a slow and regular decrease of the pressure over time, associated with an equilibrium still not reached after 1 h (Figure a). The measured pressure is surprisingly much higher than predicted by considering a total reaction of NO₂/N₂O₄ with F₂ to form FNO₂ (theoretical: 289 mbar). Infrared analysis of the gas mixture shows the coexistence of the FNO₂ species along with the nitrosyl fluoride FNO in a significant amount, generally the product of the reaction between NO and F₂. , NO₂ is thermodynamically unstable past 507 °C, according to thermodynamic data reported in. However, it has been reported in the literature that nitrogen dioxide can decompose into NO and O₂ even at lower temperatures. , The formation of FNO₂ from NO₂ + 1/2 F₂ or 1/2 N₂O₄ + 1/2 F₂ being an exothermic process, such temperature can easily be reached locally. When quickly adding F₂ to the NO₂/N₂O₄ mixture, some of the NO₂ or N₂O₄ should therefore decompose into NO and O₂. The NO formed then reacts with F₂ to form FNO via an exothermic reaction, which then reacts back with O₂ to yield FNO₂. The infrared analysis of the gas mixture confirms this hypothesis, where the relative intensity of the FNO peak decreases very slightly compared to that of FNO₂ (Figure b). The global reactions envisaged in the case where only the NO₂ species is considered are

$$\mathrm{N}{\mathrm{O}}_2+\frac{1}{2}{\mathrm{F}}_2\to \mathrm{FN}{\mathrm{O}}_2({\mathrm{\Delta }}_{\mathrm{r}}{H}_{298.15}^{°}=-141.9\mathrm{kJ}/\mathrm{mol})$$ 10

$$\mathrm{N}{\mathrm{O}}_2\stackrel{\mathrm{\Delta }}{\to }\mathrm{NO}+\frac{1}{2}{\mathrm{O}}_2$$ 11

$$\mathrm{NO}+\frac{1}{2}{\mathrm{F}}_2\to \mathrm{FNO}({\mathrm{\Delta }}_{\mathrm{r}}{H}_{298.15}^{°}=-156.0\mathrm{kJ}/\mathrm{mol})$$ 12

$$\mathrm{FNO}+\frac{1}{2}{\mathrm{O}}_2\to \mathrm{FN}{\mathrm{O}}_2({\mathrm{\Delta }}_{\mathrm{r}}{H}_{298.15}^{°}=-43.1\mathrm{kJ}/\mathrm{mol})$$ 13

4.

(a) Barometric profile of the reaction NO₂/N₂O₄ + 1 F₂ performed at room temperature. (b) Decrease in the relative intensity of the FNO absorption band compared to that of FNO₂ (I = 100%) at t + 3 min, t + 10 min, t + 30 min, t + 45 min following the valve opening. The color code is the same as that of the previous Figure a.

FNO₂ formation and NO₂ decomposition into NO (or FNO if F₂ is present) and O₂ are two distinct phenomena. However, it is possible to estimate the proportion of the two species by establishing the following material balance at constant temperature (19.0 ± 0.5 °C during this experiment, Table ). To simplify the calculations, it is assumed that the N₂O₄ species has converted into 2 equiv of NO₂, and NO₂ has fully reacted (or decomposed). A system of linear equations can therefore be written and solved for each measured pressure P _mes (Figure S6).

3. Material Balance (in mmol) of the Reaction NO₂/N₂O₄ + 1 F₂ Performed at Room Temperature (19 °C).

	NO2	F2	FNO	FNO2	O2	impurities (0.01 NO2 + 0.02 F2)
initial	n NO2  + 2 n N2O4  = 0.551	n F2  = 0.360	0	0	0	0.011
FNO2 formation	0.551 – x	0.360 – 1/2 x	0	x	0	0.011
NO2 decomposition	0.551 – x – y	0.360 – 1/2 x – 1/2 y	y	x	1/2 y	0.011
equilibrium	0.551 – x – y	0.360 – 1/2 x – 1/2 y	y	x	1/2 y	0.011
$$\{\begin{array}{l}0.551-x-y=0\\ 0.922-\frac{1}{2}x=\frac{P_{\mathrm{mes}}{V}_{\mathrm{eq}}}{\mathit{RT}}\end{array}$$

3.2. NO + F₂

The literature reports that NO + 1/2 F₂ reaction was associated with the formation of a flame upon contact of the two gases at −164 °C. The previous experiment also showed a vigorous reaction between NO₂ and F₂ at room temperature, leading to the formation of FNO. During the various tests carried out on the NO–F₂ system, detonations were sometimes heard, a few tens of seconds after the withdrawal of the Dewar, or even during F₂ injection at −196 °C when introduced in large excess compared to NO (more than 1 equiv). This phenomenon is independent of the amount of F₂ or NO (purified or not) and seemingly random. However, it can be predicted by reading the pressure during or just after the injection of F₂: if it is high, this means that it has not yet reacted with NO as expected. Such reactivity is explained by the very low activation energy of the reaction NO + F₂ → FNO + F, less than 6.2 kJ/mol or even zero according to Tajima et al. Detonations act on the barometric profiles by a sudden increase in pressure followed by a pressure drop, characteristic of a shockwave associated with the freezing of the FNO (Figure a).

5.

Barometric profiles of two NO + ≈ 0.53 F₂ reactions displaying (a) or not (b) a shockwave after Dewar removal.

Figure a displays the barometric profile of a stoichiometric purified NO + 1/2 F₂ mixture (P _NO = 507.6 mbar; P _F2 = 253.8 mbar; V = 54.3 mL; 73 ± 1 °C), along with an infrared spectrum collected at the end of the experiment (Figure b). The species FNO is well observed, as well as some unreacted NO₂. A slight excess of F₂ (0.55 equiv) managed to convert all of the NO _x into FNO and FNO₂. This experiment was associated with detonation upon warming. A third species was also formed: trifluoroamine oxide F₃NO. The literature mentions this compound, , which can be obtained in low yield by reaction between FNO and F₂ at very low temperatures. Estimation based on the establishment of a material balance (Table ) indicates that the mixture NO + 0.55 F₂ consists of 92.2% FNO, 4.2% F₃NO, 1.5% N₂O, 1% of impurities from F₂ gas, and 0.5% of F₂ and FNO₂.

6.

(a) Barometric profile of the reaction NO + 0.5 F₂ (blue) and NO + 0.55 F₂ (green) after Dewar removal. (b) Infrared spectra of the respective gas mixtures after warming to room temperature.

4. Material Balance (in mmol) of the Reaction NO + 0.55 F₂ .

	NO	NO2	N2O	F2	FNO	FNO2	F3NO	impurities from F2
initial	0.8752	0.0045	0.0134	0.4829	0	0	0	0.0099
t + 60 min	0	0	0.0134	0.0430 – α	0.8752 – α	0.0045	α	0.0099
n eq = 0.9361 – α = 0.9076
	0	0	0.0134	0.0047	0.8369	0.0045	0.0383	0.0099

From the previous spectrum, an attribution of the peaks observed over the range 1000–4500 cm^–1 can be proposed (Table ). The modes ν₁ to ν₃ correspond respectively to ν­(NO), ν­(NF) and δ_sc(FNO). The observed bands are in excellent agreement with the literature, − regarding the resolution of the spectrometer. The characteristic band of the N–O bond in F₃NO, centered on 1690.3 cm^–1, is also consistent with the work of Nectoux et al. (1690.70 cm^–1). Its frequency, intermediate between those of the asymmetric vibrations of NO₂ (1617.0 cm^–1) and FNO₂ (1793.0 cm^–1) is surprisingly high compared to nitrogen oxides presenting a simple N–O bond (1285.1 cm^–1 for N₂O and 863.1 cm^–1 for N₂O₅). It is explained by the presence of three N–F bonds with a strong inductive effect which results in a [F₂NO]⁺[F]⁻ type resonance structure, inducing a bond order of 2.

5. Assignment and Frequencies (in cm^–1) of Infrared Absorption Bands of FNO.

assignment	literature −	this work	intensity (%)
2 ν3	1037 , /1035.584	1035.6	0.5
ν2 + ν3	1290/1283.3/1282.953	1283.0	6.0
2 ν2	1522.2	1521.9	0.5
ν2 + 2 ν3	1796.766	not observed
ν1	1844.03/1843.5/1844.099	1843.3	100
ν1 + ν3	2365 , /2365.903	2365.8	4.3
ν1 + ν2	2612/2610.8/2610.473	2610.1	2.9
ν1 + ν2 + ν3	3130.462	not observed
2 ν1	3652 ,	3652.9	4.9
2 ν1 + ν3	4174/4177	4177.7	0.1
2 ν1 + ν2	4441/4420	4420.5	0.1

To test the hypothesis of reactivity of FNO with F₂ at low temperature as reported by Passmore et al., an experiment involving 1 equiv of frozen purified NO (310.2 mbar) and 1.5 equiv of F₂ (467.5 mbar) was conducted. Such a quantity of F₂ accumulated in the reactor caused it to react unexpectedly with NO during its injection, recorded by the pressure sensor (Figure a). The reactor was first warmed up to room temperature and an infrared spectrum was recorded. Then, it was cooled again for 3 h to the temperature of liquid nitrogen with fillings of the Dewar every hour or so to compensate for natural evaporation. A second spectrum was recorded after the reactor warmed to room temperature (Figure b). Despite the excess of F₂, no notable reactivity between FNO and F₂ was observed during the experiment. Equilibrium pressures at room temperature are almost identical before (591.3 mbar) and after (585.3 mbar) the prolonged cooling at −196 °C and the relative intensities of the F₃NO signal in infrared spectroscopy compared to that of the FNO has almost not changed (0.052 versus 0.051). The formation of the F₃NO species in low yield from FNO and F₂ at −196 °C, as observed by Passmore et al., would therefore rather be linked to the sometimes violent combination between NO and F₂ (detonation). The energy released by the reaction could then be sufficient to allow the fluorination of the FNO species into F₃NO. To give an idea of the temperatures and pressures potentially reached during the reaction, it is possible to determine approximately the temperature rise at the end of the reaction, by returning the system to an isochoric process (no volume variation) at room temperature (298 K), and for 1 mol of reagents (i.e., 0.4 mol of NO and 0.6 mol of F₂)­

$$T_{\mathrm{f}}-T_{298\mathrm{K}}=-\frac{{\mathrm{\Delta }}_{\mathrm{r}}{U}_{298}^{°}\times 0.4{\mathit{n}}_0}{0.4{\mathit{n}}_0(C_{v,\mathrm{F}2}^{°}+C_{v,\mathrm{FNO}}^{°})}$$ 14

7.

(a) Barometric profile of the reaction NO + 1.5 F₂ performed at room temperature. t₀ corresponds to the Dewar removal. F₂ injection started at t-260s, shockwave was recorded at t – 84s, and F₂ injection ended at t – 56s. (b) Relative intensity of the F₃NO absorption band compared to that of FNO (I = 100%) at t + 50 min (green) and t + 306 min (magenta).

The internal energy of the reaction Δ_r U ₂₉₈ can be estimated from the enthalpy of the reaction Δ_r H ₂₉₈ (−156.0 kJ/mol at 298 K) assuming an ideal system

$${\mathrm{\Delta }}_{\mathrm{r}}{U}_{298}^{°}={\mathrm{\Delta }}_{\mathrm{r}}{H}_{298}^{°}-\mathit{RT}{\mathrm{\Delta }}_{\mathrm{r}}\nu$$ 15

The reaction stoichiometry Δ_rν is −1/2. The molar heat capacity at constant volume of FNO is unknown, but it must be between the one of an ideal polyatomic gas (C _v = 3R = 24.9 J/mol/K) and the molar heat capacity at constant pressure C _p,298 K (41.8 J/mol/K). That of F₂, a diatomic gas considered ideal, is 5/2 R (20.8 J/mol/K). The calculation gives a temperature range between 2500 and 3400 K. The temperature rise actually reached must however be significantly lower, taking into account the fact that the process is not perfectly isolated, the reaction FNO + F₂ → F₃NO is no longer spontaneous beyond 1443 K according to the thermodynamic data in the literature, and the fact that part of the energy of the reaction is converted into light by the formation of a flame. Nevertheless, a temperature rise of several hundred K° can briefly allow a reaction between FNO and F₂ by crossing the activation barrier, leading to the formation of F₃NO in small quantities. The fact that F₃NO was only detected when a shockwave was recorded by the pressure sensor also supports this hypothesis (Figures , and S7).

4. Conclusions

Behind the apparent simplicity of the reactions studied, numerous parameters must be considered to prepare the compounds with high purity. Precise measurement of temperature, combined with in situ pressure monitoring provides extensive information on the reactivity of species in the gas phase. It must be considered as an analytical tool in its own right, especially if no other technique can be used. The reactivity of N₂O₄ with F₂ at very low temperature could thus be demonstrated for the first time and led to the formation of nitryl fluoride FNO₂. At room temperature and with excess F₂, nitrosyl fluoride FNO is coformed with FNO₂. Due to the disproportionation of nitric oxide, purification had to be carried out to reduce nitrogen dioxide and nitrous oxide. The very strong reaction between NO with F₂ is also confirmed, yielding FNO and F₃NO in small quantities, its presence reflecting the intensity of the reaction. Mastery of the synthesis of FNO₂ and FNO provides interesting perspectives for their use as a fluorinating agent. A liquid nitrogen/ethanol mixture (−116 °C) directly placed at the outlet of the NO bottle could eliminate (or greatly reduce) these two species by freezing them (mp −11 °C and −91 °C, respectively), thus reducing a step during the synthesis of FNO.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.5c05613.

• Figure S1: Dimerization constant K₂ calculated from ref and estimated from Gibbs free energy of the reaction at 298 K and relative error made using the estimated constant; Figure S2: Relative errors made on pressure measurement using ideal gas law rather than the van der Waals model for the studied gases (F₂, NO, NO₂, N₂O₄); Figure S3: Mole fraction x of the NO₂ species as a function of measured pressure of a NO₂/N₂O₄ mixture at different temperatures (20, 70, 75, 80 °C); Figure S4: Infrared spectra of (a) N₂O, (b) NO₂, (c) NO and (d) F₂ recorded at 80 °C; Figure S5: Typical barometric profiles of a NO _x + F₂ experiment before Dewar removal; Figure S6: FNO conversion into FNO₂ during the NO₂ + 0.5 F₂ performed at room temperature as a function of time; Figure S7: Barometric profiles and corresponding infrared spectra of two NO + ≈ 0.5 F₂ reactions presenting (a, c) or not (b, d) a shockwave following the Dewar removal (PDF)

The authors declare no competing financial interest.