**3. Results and Discussion**

### *3.1. Structural Comparison with Previously Described Pyrophosphoramides*

Most intermolecular motifs that build up the supramolecular structure of a compound are due to the moieties within the molecule in question. Therefore, compounds with similar moieties and chemical structures may be expected to have similar intermolecular motifs and in turn, similar crystal structures [17,18]. In this regard, the various pyrophosphoramides discussed in this study can be divided into five main groups, dependent on their chemical structure (Table 2).


**Table 2.** Categorisation of pyrophosphoramides discussed in this publication including crystal structures and space groups published from data obtained by single-crystal X-ray diffraction for the solid products, along with the liquid O((Me2N)2PO)2.

> On comparison of the crystal structures of the compounds classified in each group, the most important molecular difference that seemed to affect the supramolecular structure was the presence of the N–H moiety. Its presence caused a significant difference between the structure of the mono-N-substituted pyrophosphoramides and the di-N-substituted pyrophosphoramides, which lack this moiety. In the crystal structures of mono-N-substituted pyrophosphoramides P=O···H–N hydrogen bonds were the most prominent and common intermolecular bonds in the structure. These formed numerous intermolecular and intramolecular synthons, which were the primary cause for the formation of the various supramolecular motifs observed, and which dictated both structure and symmetry [17–19]. The di-N-substituted pyrophosphoramides were shown to form crystalline structures wherein the pyrophosphoramide moieties did not interact with each other directly. Thus, no strong intermolecular interactions were found. The organic moieties therefore had a significant impact on the packing of molecules in the crystal structures obtained in contrast to the mono-N-substituted pyrophosphoramides (Table 3).


**Table 3.** Hydrogen bond distances in the crystal structure of **1.**

1 3/2 − X, +Y, 12+ Z; 2 +X, −1 + Y, +Z; 3 +X, 1 + Y, +Z.

### 3.1.1. Structural Comparison of Mono-N-Substituted Pyrophosphoramides.

Given that the compound characterised in this current study, **1**, was a mono-Nsubstituted pyrophosphoramide its novel structure is best discussed in relation to other mono-N-substituted pyrophosphoramides. Despite its chemical similarities to O((*t*BuNH)2PO)2, it not only crystallised in a different spacegroup, namely *Pca2*1, but it also showed a different supramolecular motif. First, the already published structures will be discussed and then compared with **1**.

The structures of Mono-N-substituted pyrophosphoramides O((RNH)2PO)2 were found to crystallise in two space groups: either *P*21/*c* or *Pccn*. Common supramolecular arrangements form part of the crystal structures of O((*t*BuNH)2PO)2 and the two O((2- MePhNH)2PO)2 polymorphs [10–12] (Figure 3). The first one is a ring synthon with a R22(8) graph set binding molecules through intermolecular hydrogen bonding, while in the other [17,19] a partially eclipsed conformations is reached (Figure 3). The second synthon

constricts the molecules from taking on other conformations and therefore minimises the number of possible supramolecular structures available. These two synthons together create the same supramolecular structure for all of the above crystal structures, namely a chain like packing as given in Figure 4. The respective structures for the three mono-Nsubstituted pyrophosphoramides are shown in Figure 5. This structure motif is repeated through translational symmetry to form infinite chains. The two molecules that form the actual supramolecular units are related to each other by the glide plane denoted in the *P*21/*c* spacegroup.

**Figure 3.** H-bonding synthons in *P2*1*/c* structures: (**a**) intermolecular ring synthon; (**b)** the two variants of the intramolecular synthon where A is the non-intermolecular bonding amine (all generated using ChemSketch [24]).

**Figure 4.** Basic unit for the supramolecular H-bonding motif that is common to all three mono-Nsubstituted pyrophosphoramides crystallising in *P*21/*c* (generated using ChemSketch [24]).

Formation of this hydrogen bonding motif seems to be independent of the nature of the organic substituent on the amide nitrogen, as it occurs for both the alkyl tert-butyl and aryl 2-methylphenyl analogues. The packing of these chains is, however, influenced by the organic substituents. The tert-butyl groups in O((*t*BuNH)2PO)2 pack in a staggered formation resulting in the closest possible packing of the chains [12]. No additional intermolecular interactions are detected within the usual limits.

The packing effects of the 2-methylphenyl groups are more complex than those of the tert-butyl groups. They give rise to two polymorphs of this compound. The two molecular components of these supramolecular units are related to each other by the glide plane imposed by the *P2*1*/c* spacegroup. In the two O((2-MePhNH)2PO)2 polymorphs however, the direction of the motif is different. The motif in the polymorph reported by Pourayoubi, M. et al. is built along the *c* axis while the same motif in the polymorph described by Cameron, S.T. et al. is perpendicular to the *c* axis. Given that the hydrogen bond interaction is identical in both polymorphs, the difference lies in the way the supramolecular chains pack. The polymorph discussed by Pourayoubi et al. [11] showed closer packing between the chains. In both polymorphs the 2-methylphenyl groups are oriented antiparallel to each

other (Figures 6 and 7). The stacking distances between the intramolecular 2-methylphenyl moieties are 3.539 Å and 3.933 Å [10,11]. In one polymorph [11] the position of the chains seems mainly influenced by the supramolecular structure given in the diagram shown in Figure 6. A square like motif is formed wherein each side is composed of the Ph(π)···H– C(meta) interactions [17]. Two contact distances of 3.159 Å and 3.188 Å are present, which are shorter than any contact distance reported in the polymorph described by Cameron et al. [10]. The latter forms similar inter- and intramolecular interactions with one much larger contact distances of 4.717 Å and a very oblique interaction between neighbouring molecules, not bound by hydrogen bonding, with a distance of 3.626 Å [10]. The former interaction is longer than any interactions described for the other polymorph and indicates a less efficient packing. Thus, the different modes of packing of the organic moieties are most probably the cause of the formation of the two polymorphs.

**Figure 5.** The hydrogen bonding motif common to all *P*21/*c* as noted in the published structures: (**a**) O((*t*BuNH)2PO)2; (**b**) O((2-MePhNH)2PO)2 published by Pourayoubi et al. [11]; (**c**) O((2-MePhNH)2PO)2 published by Cameron et al. [10] (all generated in Mercury [25]).

**Figure 6.** (**a**) O((2-MePhNH)2PO)2 polymorph published by Pourayoubi et al. showing both intermolecular and intramolecular C–H···Ph interactions (generated in Mercury [25]); (**b**) a diagram of the intermolecular C–H···Ph interactions for clarity (generated using ChemSketch [24]).

**Figure 7.** O((2-MePhNH)2PO)2 polymorph published by Cameron et al. showing both intermolecular and intramolecular C–H···Ph interactions (generated in Mercury [25]).

Experimental procedures on how to obtain only one of two polymorphs are not reported in the literature. The O((2-MePhNH)2PO)2 structure published by Cameron et al. was obtained by the reaction of phosphenyl chloride with ortho-methylaniline followed by recrystallisation in methanol. The synthesis of the polymorph obtained by Pourayoubi et al. was not described in the literature to the best of our knowledge. Interestingly, the crystallographic data of the two structures was collected at different temperatures (150 K for [11] and 295 K for [10,11]). It is, therefore, possible that a polymorphic transition occurs at lower temperatures. This is also in agreemen<sup>t</sup> with the difference in the corresponding unit cell volumes.

The mono-N-substituted pyrophosphoramides, which do not crystallise in the *P2*1*/c* spacegroup, show different inter- and intramolecular hydrogen bonding motifs. Given that the compounds crystallising in the same spacegroup (*P*21/*c*) contain both alkyl and aryl moieties and show a different packing of these moieties, a similar observation is expected for the 4-methylphenyl and the iso-propyl analogues. However, a different supramolecular structure was formed by the 4-methylphenyl derivative O((4-MePhNH)2PO)2, crystallising in spacegroup *Pccn* [13]. There is no intramolecular synthon present (Figure 3b), resulting in a different pattern of supramolecular interactions. The phosphoryl oxygen and the amide nitrogen, which typically form the intramolecular P=O···H–N motif in the *P*21/*c* structures, do not show any hydrogen bonding and therefore, the molecules are not limited to an eclipsed conformation (*vide supra*). They form a staggered conformation that enables the formation of the basic supramolecular building block for this structure. Two molecules are connected via two intermolecular P=O···H–N hydrogen bonds, forming a ring synthon with graph set R22(12) (Figure 8) [13,19]. Further N–H···Ph and P=O···H– C(Ph(C2)) interactions are observed. The N–H···Ph interactions are formed through the remaining amide moieties which do not interact in the R22(12) ring described prior and this additional set of interactions seems to stabilise the motif by increasing the packing efficiency and stopping this moiety from forming the previously mentioned intramolecular P=O···H–N bonding. The P=O···H–C(Ph(C2)) interactions also seem to add stability by further aiding the P=O oxygen atoms to obtain the orientation necessary to form this synthon [17,21].

The two molecules are related by a glide plane on the *a*–*c* plane along the *c* axis (Figure 9) which results in the formation of infinite chains along cell axis *c*. Neighbouring chains positioned anti-parallel to each other along the *b* axis and related to each other by a second glide plane along the diagonal of the *ab* plane. Chains neighbouring each other along the *a* axis are parallel and again related by translation. The main interaction responsible for arrangemen<sup>t</sup> in antiparallel chains is the Ph(π)···H–C(para-methyl) interaction between the closely situated 4-methyphenyl groups bound to different molecules in different chains [21], as noted in Figure 9 [13]. This guarantees the closest possible packing of the various phenyl groups in the molecule. Thus, the 4-methylphenyl groups seem to play a structure forming role in both the formation of a different hydrogen bonding motif and a different packing of the molecules in this structure compared to other pyrophosphoramides.

**Figure 8.** Hydrogen bonding motif for O((4-MePhNH)2PO)2 (**a**) as viewed along the *a*-axis (**left**) and offset to show the staggered conformation of the pyrophosphoramide (**right**) (generated in Mercury [25]); (**b**) The ring synthon responsible for the motif given in a simplified diagram (generated using ChemSketch [24]).

**Figure 9.** Intermolecular and intramolecular non-hydrogen bonding interactions in the structure of O((4-MePhNH)2PO)2 (generated in Mercury [25]).

The 2-methylphenyl and 4-methylphenyl derivatives both crystallise in different supramolecular structures and space groups. However, the 2-methylphenyl pyrophosphoramide crystallises in the same space group and shows the same supramolecular bonding as that for the tert-butyl analogue. O((4-MePhNH)2PO)2. The latter was collected at 90 K, whereas the O((2-MePhNH)2PO)2 polymorphs were collected at 150 K and 295 K [10,11,13]. Thus, it is unclear whether the occurrence of different structural motifs is mainly caused by the different nature of the side groups or simply because the single crystal data was collected for each compound at different temperatures

No intramolecular P=O···H–N hydrogen bonding is present in the crystal structure of **1**. This leads to the formation of a staggered arrangemen<sup>t</sup> of the pyrophosphoramide group similar to that observed for O((4-MePhNH)2PO)2. Thus again, the lack of intramolecular bonding between amide and phosphoroxide seems to hinder an eclipsed conformation (*vide supra*, Figure 3b) and related supramolecular motifs. The most important supramolecular motif in the structure of **1** is therefore the P=O···H–N interaction. This is in agreemen<sup>t</sup> with what was observed also in all other mono-N-substituted pyrophosphoramides, where intermolecular hydrogen bonding was a major factor in the formation of the relevant supramolecular motifs [10–13]. The actual synthon is shown in Figure 10. It is formed by two molecules of **1** related to each other by translation symmetry via two different P=O···H–N hydrogen bonding synthons. To the best of our knowledge this is the only mono-N-substituted pyrophosphoramide structure to exhibit multiple intermolecular hydrogen bonding synthons. The first synthon is a ring with a R<sup>1</sup> 2(8) graph set connecting one P=O group with two neighbouring N–H groups, which are bound to the two different phosphorus centres of the second molecule. The second synthon consists of a P=O···H–N interaction between the other P=O group of the first molecule and a neighbouring N–H group. The latter is not part of the previously discussed synthon; it shares, however, a phosphorus centre with one of the previously discussed amide groups. Neighbouring chains along the *a* axis pack anti-parallel to each other via a *2*1 screw axis (Figure 11). The remaining amide does not seem to participate in any type of intermolecular or intramolecular bonding, even though, theoretically, infinite chains along the *b* axis could be formed in a similar manner as observed in other mono-N-substituted pyrophosphoramides (*vide supra*). Therefore, the main mode of intermolecular interactions is the Van der Waals forces that caused the hydrophobic iso-propyl groups to stack.

**Figure 10.** Hydrogen bonding synthons present in the structure of O((*i*PrNH)2PO)2 (generated in Mercury [25] and ChemSketch [24]).

**Figure 11.** Antiparallel chains of O((*i*PrNH)2PO)2 noted along the a-axis (generated in Mercury [25]).

The cause for the formation of this different supramolecular motif is difficult to determine as the different molecular structure, synthesis, crystallisation techniques, and solvents used can all affect crystallization process and lead to this crystal structure. The temperature at which single crystal data was obtained is unlikely to be the cause of this as it falls into the same range as for the other compounds crystallising in spacegroup *P*21/*c* [10–12]. The synthesis and crystallisation approach used to obtain single crystals of **1** differs from published procedures to obtain crystalline material for the related compounds. Given the chemical similarities between the iso-propyl and tert-butyl moiety the difference in supramolecular motifs between the two is unexpected.

### 3.1.2. Structural Comparison of di-N-Substituted Pyrophosphoramides

Because of the lack of a N-H donor in di-N-substituted pyrophosphoramides the main supramolecular interactions present in the crystal structure derive from the organic substituents. A very good and well-known example of a di-N-substituted pyrophosphoramide O((R2N)2PO)2 is Schradan, O((Me2N)2PO)2 [3–9,22]. Schradan is a liquid at room temperature. No crystal structures collected of crystals below the melting point temperature are reported in the literature. Structural information from experimental data is only available for O((R1R2N)2PO)2 and the *N*,*N*--substituted diamine derivatives O((R1(NR2)2)2PO)2, i.e., O((BzMeN)2PO)2, O((C2H4(2,5-*i*PrPhN)2)2PO)2 and O((1,2-Cy(NaphN)2)2PO)2 [13–15]. No significant classical intermolecular interactions between the pyrophosphoramide or organic moieties are present in the corresponding published crystal structures. The pyrophosphoramide backbone forms a similar staggered conformation in all three compounds. In O((BzMeN)2PO)2, the main intermolecular interactions effecting the supramolecular structure are weak C–H···Ph interactions [13]. These form in two different synthons, namely N–Me···Ph(C2) and CH2(benzyl)···Ph. The molecules pack in layers linked along the axis *c* (Figure 12) [17,21].

**Figure 12.** Packing of O((BzMeN)2PO)2: (**a**) C–H···Ph interactions which act as the main supramolecular building block (generated in Mercury [25]); (**b**) simplified diagram of the supramolecular motif (generated using ChemSketch [24]).

The crystal structure of O((C2H4(2,5-*i*PrPhN)2)2PO)2 shows only very weak intermolecular interactions [14]. The pyrophosphoramide moieties seem to be isolated from each other by the bulky hydrophobic organic moieties (Figure 13).

In the crystal structure of O((1,2-Cy(NaphN)2)2PO)2 intermolecular motifs (C–H···Ph interactions) are present similar to those observed in the structure of O((BzMeN)2PO)2. Intermolecular interactions are noted between the benzyl methylene, the naphthyl groups and the cyclohexyl –CH2– groups. Three distinct C–H···Ph interactions are present namely CH2(benzyl methylene)···naphthyl(C4), C–H(Naphthyl(C8))···naphthyl(C10), and CH2(cyclohexyl)···napthyl(C3) [15]. These interactions are unidirectional, with the proton donors binding to the closest naphthalene along the *c* axis, which is typically the molecule diagonal to the proton donor molecule in the layers running along axis *a* (Figure 14). The

pyrophosphoramide group is connected to hydrate water by hydrogen bonding forming a R22(8) ring [15]. This ring motif separates the pyrophosphoramide molecule structurally from any further interaction with typical hydrogen bonding electron acceptors.

**Figure 13.** Structure of O((C2H4(2,5-*i*PrPhN)2)2PO)2 as viewed along the a-axis (generated in Mercury [25]).

**Figure 14.** Structure of O((1,2-Cy(NaphN)2)2PO)2 viewed along the c-axis showing hydrogen bonding (black contacts) and C–H···Ph interactions (light blue contacts) (generated in Mercury [25]).

In all three cases the supramolecular structure is dominated by the organic substituents. There is complete lack of π–π stacking interactions. A possible reason for this might be that the pyrophosphoramide backbone always takes on a more staggered conformation forcing the organic substituents into unfavorable positions to form π–π stacking. The staggered conformation itself is most likely caused by the lack of the amine N–H bonds, which usually force the molecule in a more eclipsed conformation through inter/intramolecular bonding as observed in some of the mono-N-substituted pyrophosphoramides.
