Oxidative polymerization of perfluoroolefinc monomers like tetrafluoroethylene (TFE) by molecular oxygen into P-PFPEs (A) described in Figure 1, is a well-known reaction [9].

The forming peroxidic polymer has the following general structure: TO-(CF₂CF₂O)_p-(CF₂O)_r-(O)_q-T’, comprising fluoroether repeating units (-CF₂CF₂O- and -CF₂O-), interspersed peroxy units (-CF₂CF₂OO- and -CF₂OO-) and perfluorinated alkyl groups, acyl fluoride or fluoroformate as chain end-groups (T, T’). Practically, this oxypolymerization can be activated either by high energy UV light or by employing elemental fluorine or perfluoroalkyl hypofluorites, especially CF₃OF, as chemical initiators [9]. The peroxidic polymer is the result of a complex free radical oxidation reaction dominated by the chemistry of perfluoro oxyradicals [10]. The composition and the reaction yields of forming PFPE polymers depend on irradiation condition, TFE concentration, fluid dynamics of the system and other classical physical parameters like temperature and pressure. Commonly, the average molecular weight of formed P-PFPEs (A) is more than 10⁴ amu [11]. Carbonyl fluoride and tetrafluoroethylene oxide are the main by-products.

α,ω-Diacyl PFPE can be obtained by thermal treatment or chemical reduction of peroxide bonds present in the above mentioned peroxidic precursor. Under the appropriate thermal treatment, homolysis of fragile peroxide bonds in P-PFPEs (A) produces PFPE alkoxy radicals; subsequently, the cleavage of the bond in ß-position to the oxygen-centered radicals leads to the formation of reactive carbon-centered PFPE radicals [12,13]. The fast cage-combination reaction of these radicals takes place and leads to a PFPE material in which acyl fluoride and fluoroformyl moieties are nearly 50% of the end groups (Figure 2). Generally, above 150 °C those fluoroformate groups can also be converted into acyl fluorides. After the thermal treatment of the peroxidic PFPE, the reduced PFPE products can be distinguished into two fractions: The major one includes PFPE with monoacid end-groups, while the minor one comprises completely neutral (perfluoroalkyl end-groups) and diacyl fluoride PFPEs.

P-PFPEs (A) can also be converted directly to diacyl fluoride PFPEs, F(O)CCF₂O(CF₂CF₂)_n(CF₂O)_mCF₂C(O)F, by chemically induced reduction of the peroxide unites with better yield compared with thermal treatment (Figure 3). The chemical reduction is performed with reducing agents, such as hydrogen iodide or with molecular hydrogen in the presence of a noble metal catalyst (e.g., Palladium) [14].

Finally, hydrofluoropolyethers DA-FPEsare prepared through the reaction of diacyl fluoride PFPEs (DAF-PFPEs) with electrophiles in the presence of a source of fluorine ions, generally, metal fluorides. This process actually comprises two steps (Figure 4): Firstly, in aprotic polar solvent, DAF-PFPEs react with fluoride ions released from metal fluorides to form metal perfluoroalkoxides M⁺⁻OCF₂RFCF₂O⁻M⁺ as intermediates; secondly, metal perfluoroalkoxide is alkylated by a suitable alkylating agent X-R_H leading to the introduction of alkoxy groups at the end of the perfluoropolyether chain.

Metal perfluoroalkoxides are known as compounds characterized by low nucleophilicity [15] and they decompose into acylfluorides and metal fluorides, generally around room temperature. In order to increase the stability and the nucleophilic strength of these species, polar aprotic solvents (e.g., diglyme, tetraglyme) are generally required. The stability of these species depends on the nature of metal fluoride [16], temperature, and the solvent [17]. Metal perfluoroalkoxides are also important industrial intermediates in the synthesis of hypofluorites like CF₃CF₂OF [18] and fluorinated vinylethers [19,20]. Completely anhydrous reaction conditions are required, because in presence of water the hydrolysis of acyl fluoride groups generates hydrofluoric acid, HF, which can be absorbed by the solid metal fluoride giving MHF₂ [16]. Because of the low nucleophilicity of perfluoroalkoxides, strong alkylating agents with high electrophilicity are needed to effectively react with metal perfluoroalkoxides and form DA-FPEs.

The boiling points (b.p.) of DM-FPEs are reported in Figure 10 as a function of the number of backbone chain atoms (N), i.e., carbon plus oxygen atoms. They increase regularly with N. In Figure 10, DM-FPEs boiling points are also compared with those of α,ω-dihydrofluoroethers (HFEs) [31] and perfluoropolyethers (PFPEs) [32].

The highest boiling points are those of DM-FPEs, the intermediates are HFEs and the lowest ones are PFPEs. The above-mentioned result reflects the typical low intermolecular interactions of perfluorinated compounds. Partially fluorinated ethers, like HFEs and DM-FPEs, display higher boiling temperatures and this is indicative of greater intermolecular polar interactions [11].

Considering three similar fluoromethylethers, CF₃OCF₃ (b.p. 214 K; μ = 0.49 D), CF₂HOCF₃ (b.p. 238 K, μ = 1.18 D) and CH₃OCF₃ (b.p. 250 K, μ = 2.56 D), the boiling points and the dipole moments have analogous tendencies [33]. The progressive substitution of fluorine atoms with hydrogen in the methoxy end caps leads to a stronger dipole and consequently, to an increase of the boiling point, as shown in Figure 10. The boiling point differences between PFPEs and HFEs and between PFPEs and DM­FPEs are not constant but decrease continuously with the chain length. Since the attractive forces are due to polar terminal groups, their relative intensity decreases as N increases and, as a result, the boiling points of HFEs, DM-FPEs and PFPE tend to an asymptote at higher N values.

A comparison of the specific volume (V_sp) at 298 K of DM-FPEs, HFEs and PFPEs series is shown in Figure 11. The DM-FPEs sequence shows the highest specific volume and the PFPEs the lowest at identical number of chain atoms, N. The V_sp of PFPEs is almost constant and independent of N. The specific volumes of DM-FPEs and HFEs decrease continuously to a plateau that approximately coincides with the constant V_sp of PFPEs. This behavior can also be explained on the basis of the differences between terminal groups. Moving from –CF₃ to –CF₂H and to –CH₃, the molecular weight of the groups diminishes; on the contrary, the variation of volume is less noticeable because the Van der Waals radii of fluorine and hydrogen are similar [34]: 0.147 and 0.12 nm, respectively. Consequently, the difference in specific volume between the three series is mainly due to the variation of molecular weight, explaining the tendencies of V_sp observed in Figure 11. Furthermore, as the weight fraction of the chain ends decreases with N, it is logical that the specific volume of HFEs and DM­FPEs tends to the V_sp values of PFPEs when the chain is sufficiently long.

The vapor pressure data, plotted as Clausius-Clapeyron equation in Figure 12, show a deviation from linearity at low temperatures, which is common for many oligomeric substances. Vaporization enthalpies (ΔHv) are also derived from the Clausius-Clapeyron plot in the range of temperatures in which the function is linear.

The ΔH_v data of DM-FPEs, reported as a function of number of chain atoms, N, are compared with those of HFEs and PFPEs in Figure 13. These data indicate the highest values for DM-FPEs, intermediate for HFEs and the lowest for PFPEs. As the effect of the polar terminal groups (–OCF₂H and –OCH₃) diminishes with the length of the chain, ΔH_v also decreases and tends to the values of PFPEs when N is above 15 [4].

The refractive indices of the series DM-FPEs and HFEs plotted as a function of the number of chain atoms, N, indicate two opposed behaviors (Figure 14). In particular the refractive index n increases for DM-FPEs and decreases for HFEs. In both cases, increasing the chain length, the refractive indices tend to a plateau value, close to the characteristic refractive index of linear PFPEs (1.290–1.295) [11]. The tendencies of DM-FPEs and HFEs refractive indices are mainly due to the terminal groups that contribute to n in an opposite way. As reported in the literature, the group influence of –OCH₃ is higher than that of –OCF₃, while that of –OCF₂H is lower [35].

In 2004, the first understanding of the atmospheric chemistry for DM-FPEs was achieved by studying the kinetics and mechanism of CH₃O(CF₂CF₂O)_nCH₃ (n = 1–3) oxidation which is initiated by Cl atoms and OH radicals in the atmosphere; the impact of their degradation products on the climate were also discussed [1]. Like most organic compounds, the primary tropospheric degradation of CH₃O(CF₂CF₂O)_nCH₃ is mainly dominated by the attack of OH radicals [38]. Understanding of reaction kinetics is essential for the evaluation of the atmospheric lifetime of these molecules. In addition to the reaction with hydroxyl radicals (OH), organic compounds can also be removed from the atmosphere via photolysis, wet deposition, and reaction with NO₃ radicals, Cl atoms, and O₃ [38,39,40]. For saturated compounds such as CH₃O(CF₂CF₂O)_nCH₃, reactions with NO₃ radicals and O₃ are typically too slow to be of importance [1]. The average concentration of Cl atoms in the troposphere is orders of magnitude less than that of OH radicals [41]. Reaction with Cl atoms will be a less significant atmospheric loss mechanism. Since ethers do not absorb at UV wavelengths > 200 nm, photolytic sink will be not be an important degradation pathway [42] either. Highly fluorinated molecules, such as CH₃O(CF₂CF₂O)_nCH₃, are hydrophobic and wet deposition is unlikely to be of importance. In conclusion, the atmospheric lifetime of CH₃O(CF₂CF₂O)_nCH₃ is determined by its reaction with OH.

Smog chambers equipped with FTIR spectrometers were used to study the Cl atom and OH radical initiated oxidation of CH₃O(CF₂CF₂O)_nCH₃ in 720 Torr of air at 296.3 K. Relative rate methods were used to measure k(Cl+CH₃O(CF₂CF₂O)_nCH₃)) = 3.7 (±0.7) × 10⁻¹³and k(OH+CH₃O(CF₂CF₂O)_nCH₃) = 2.9 (± 0.5) × 10⁻¹⁴ cm³ molecule⁻¹·s⁻¹. The relative rate method is widely used to measure the reactivity of OH radicals with organic compounds [43]. The hydrogen abstraction reaction from the terminal groups has been assumed, according to established literature data, as the rate-determining step for the decomposition of the entire molecule [44]. It is well known that –CF₂CF₂– units are unreactive toward Cl and OH radicals. Thus, the reactivity of CH₃O(CF₂CF₂O)_nCH₃ is confined to the C–H bonds on either end-groups of the molecule. Indeed, appropriate studies showed that the reactivity of the methyl group did not change noticeably by increasing the number n of –(CF₂CF₂O)– units from 1 to 3 [45]. Nevertheless, the average concentration of Cl atoms in the troposphere is orders of magnitude less than that of OH radicals [41]. Thus, knowing the k(OH+CH₃O(CF₂CF₂O)_nCH₃)_s rate constant, the DM-FPEs atmospheric lifetime has been evaluated to be two years [1].

Tuazon et al. claimed that in air the oxidation of HFEs with difluoromethoxy ending groups, OCF₂H, like HCF₂OCF₂OCF₂CF₂OCF₂H, HCF₂OCF₂CF₂OCF₂H, and HCF₂OCF₂OCF₂H, gives C(O)F₂ as the only carbon-containing product [46]. For DM-PFEs which end with methoxy-groups, fluoroformates CH₃O(CF₂CF₂O)_nC(O)H were recognized as the only oxidation decomposition products with a 100% yield; the formed fluoroformates can be further oxidized to di-formates H(O)CO(CF₂CF₂O)_nC(O)H [1]. Concerning the impacts of fluoroformates on the climate, since fluorinated esters are known to be easily hydrolyzed [46,47], it is usually assumed that they can be easily removed though wet deposition; therefore Andersen et al. [1] deduced that formate CH₃O(CF₂CF₂O)_nC(O)H is not expected to be persistent or pose any significant environmental hazard. However, Bravo et al. [48] assumed that the highly fluorinated nature of fluoroformates may decrease their solubility in water and could contribute a significant indirect GWP. Up to now, due to the lack of experimental data like gas-to-water equilibrium and solubility test of fluoroformates, it is still unclear how rapidly fluoroformates will be removed via uptake and hydrolysis in rain/cloud/seawater [49]. Further experimental examinations are needed for the determination of the atmospheric implication of fluoroformats.

In 1995, Pinnock et al. outlined a method to determine the instantaneous forcings (IF) from the IR absorption spectra such as those reported in Figure 14. As the values of GWP can be estimated on the basis of the IF, in Table 3 GWP relative to carbon dioxide of DM-FPEs at 100 year time horizons has been reported, and the comparison with other greenhouse gases, such as CF₃Cl (CFC-11), CF₂HOCF₃ (HFE125), CH₃OCF₃ (HFE143a) and other HFEs, is reported in Table 3.

It is evident from Table 3 that DM-FPEs have relatively high instantaneous forcing (IF) but much lower GWPs, if compared with CFC11 and HFE compounds. In addition, the higher the number of –(CF₂CF₂O)– units, the higher the absorption in the 800–1200 cm⁻¹ range, that is responsible for the high IF. Thus the reason of the mitigation of DM-FPEs GWP is mostly due to their short atmospheric lifetimes and the emission of DM-FPEs into the atmosphere should not contribute significantly to the climate change.

On the basis of the data of DM-FPEs which contain only –(CF₂CF₂O)– units and the studies on series of HFPE with –CF₂H end-groups [51], it seems reasonable to expect that the reactivity of C–H bonds is independent from the number of -CF₂O- units in the DM-FPE molecule. However, there are no available kinetic data for DM-FPEs with structure CH₃O(CF₂CF₂O)_n(CF₂O)_mCH₃ to confirm this assumption.

In addition, DM-FPE commercial products are mixtures of homologous copolymers with similar lengths but different unites, i.e., different n and m; further experimental studies are most likely necessary for a better understanding of the effects of the (CF₂CF₂O)_n/(CF₂O)_m ratio on atmospheric chemistry of DM-FPEs.