Tuning the Molecular Packing of Self‐Assembled Amphiphilic PtII Complexes by Varying the Hydrophilic Side‐Chain Length

Abstract Understanding the relationship between molecular design and packing modes constitutes one of the major challenges in self‐assembly and is essential for the preparation of functional materials. Herein, we have achieved high precision control over the supramolecular packing of amphiphilic PtII complexes by systematic variation of the hydrophilic side‐chain length. A novel approach of general applicability based on complementary X‐ray diffraction and solid‐state NMR spectroscopy has allowed us to establish a clear correlation between molecular features and supramolecular ordering. Systematically increasing the side‐chain length gradually increases the steric demand and reduces the extent of aromatic interactions, thereby inducing a gradual shift in the molecular packing from parallel to a long‐slipped organization. Notably, our findings highlight the necessity of advanced solid‐state NMR techniques to gain structural information for supramolecular systems where single‐crystal growth is not possible. Our work further demonstrates a new molecular design strategy to modulate aromatic interaction strengths and packing arrangements that could be useful for the engineering of functional materials based on PtII and aromatic molecules.


Introduction
Self-assembled structures of p-conjugated systemsh ave long been an area of active research in materials science,where molecular assemblies in solution,i nt he solid state and in the liquid crystalline state can be created. [1] It is widely known that the functionalp roperties of self-assembled structures are highly dependento nt he molecular structure of the monomers, and that minor changes in the molecular design can result in dramatic differences in the features and functionalities of the resulting materials. [2] Therefore, tuning the balance of attractive noncovalent interactions (such as hydrogen bonding and aromatic interactions) and steric effects by molecular design is necessary to generate self-assembled materials with distinctp roperties. [2] Outside purely organic compounds, the introduction of am etal ion like Pt II to p-conjugated molecules can further extendt he range of possible intermolecular interactions and enablea dditional applicationsi no ptoelectronics and biomedicine. [3] Hence, control of the resulting structures through molecular designi sd esirable.
As previously shown, tuning the length of peripherala lkyl chains-which are commonly used to achieve solubility of the compounds-can dramatically affect the self-assembly behavior of Pt II complexes. For instance,w er eportedt he self-assembly of two oligophenyleneethynylene (OPE)-based bispyridyldichlorido Pt II complexes featuring terminal dodecyloxyv ersus methoxyc hains. [4] We found that shortening the side chains causesad rastic change from as lipped arrangement of the molecules, both in solution andthe solidstate, to almost parallel p-stacksi nt he solid state with enhanced aromatic interactions and shortenedP t ÀPt distances of % 4.4 .I na nother elegant example, Yamand co-workers showed that the hydrophobic tail length of amphiphilic Pt II complexes can alter the morphologies and emission behavior of self-assemblies in aqueous medium. [5] Additionally,for different luminescentPt II mesogens, it has been demonstrated that the variation of alkyl chains can be used to modulate the mesophases with regards to their stability and photophysical behavior. [6] Interestingly,m odification of the alkyl tail length for some of these compounds has led to the emergence of multistimuli-responsive polymorphism in the solid state, [6a,b] thus demonstrating the key impact of small structuralchanges (e.g.,C 14 vs. C 16 chain [6a] )onthe final properties of the systems.
Apart from nonpolar alkyl chains, polar oligo-or poly(ethylene glycol)chains (OEG, PEG) are commonly used as hydrophilic counterparts for achieving solubility in aqueous media. [7] However, thee ffecto fE Gc hain length variationo nt he self-assembly of Pt II complexeshas been scarcely investigated. [8] To the best of ourk nowledge,t he only exampleo fs elf-assembling Pt II complexes featuringP EG n chains of different length was reported by the group of Manners. [9] In this work, tetrazolebased tridentate Pt II complexes with ancillary ligands based on PEG n were observed to form 1D fibers (n = 16) or 2D platelets (n = 7) in polar solvents, depending on the PEG chain length.
In this work, we demonstratep recise control over the supramolecular packingo fa mphiphilic Pt II complexes by systematic variation of the hydrophilic side-chain length. This understanding could contribute to establishing ac orrelation between molecular design and packing modes,w hich remainso ne of the greatest challenges in the field of self-assembly. [10] To this end, we herein investigated the solid-states tructures of as eries of rod-like bispyridyldichlorido Pt II complexes 1-4 featuring either three OEG chains on both termini, namely triethylene glycol (TEG, 1), diethylene glycol( DEG, 2), ethylene glycol (EG, 3), or methoxy groups (4,S cheme 1). Complexes 1-4 are smaller structurala nalogues of ap reviously reported Pt II complex 5 that was found to exhibit au nique molecular organization based on its crystal structure:t he presence of voluminous TEG chains hinders ap arallel molecular arrangement and ultimately enablest he formation of ac haracteristic handshake motif, where the TEG groups of two adjacent Pt II units are in close contact through multiple CH···O interactions (Scheme 1a nd Figure 1A). [11] To simplify our system, we shortened the aromatic backbone of the ligand to two rings (pyridine-ethynylpheny-lene) in order to reduce the amount of potential interactions, highlighting the influence of the glycol chains with identical psurface. We envisioned that ag raduald ecrease of the EG chain length whilek eeping the aromatic surfaceo ft he molecules unchanged could progressively alleviatet he stericd emand of the chains and,c onsequently,r einforce the extent of p-overlap between the stackedm olecules. This, in turn, is expected to lead to ag radualt ransformation from am ore slipped to more parallelp acking upon shortening the EG side chains. Thus, in this report, we focus on the self-assembly structures in the solid state, whereas investigationso nt he processes in solution are still ongoing.
Usually,t he solid-state structure and molecular organization are determined by single-crystal X-ray diffraction, andw ew ere able to obtain suitable single crystalsf or complexes 1 to 3. Crystallization of complex 4,h owever,r esulted in polycrystalline powders unsuitable for X-ray analysis. Therefore, ac omplementary method was requiredf or obtaining as ystematic comparisonofa ll compounds.T ot his end, we employed solid-state 1 Ha nd 13 Cm agic-angle spinning (MAS) NMR in combination with 2D, "through-space" NMR correlation techniques based on the direct magnetic dipole-dipole (or dipolar) interaction between nuclear spins to elucidate the aggregation of the methoxy-functionalized complex 4. [12] Our approachc onsisted in identifying through-space 1 H, 13 Cc orrelations characteristicf or certain structuralm otifs by comparing the 1 Ha nd 13 CM AS NMR results obtained from complexes 1 to 3 with the crystallographic data. On this basis, we are able to readily identify or exclude previously observed packing structuresf or this specific series of OPE-based Pt II complexes,f illing in the missing structural information from crystallographic data.

Results and Discussion
Synthesis Pt II complexes 1-4 were synthesized in moderate to good yields by complexation reaction of the precursor salt [PtCl 2 (PhCN) 2 ]w ith the respective pyridine-based ligands in toluene at 100 8C( Scheme 1a nd the Supporting Information). Complexes 1-3 resulted in crystalline solids upon concentration from Et 2 Oo rp entane-containingC H 2 Cl 2 as well as EtOAc solutions, whereas all other used solvents led to highlya morphous structuresw iths oft texture. Compound 4,h owever, could only be obtained as ap owdery solid, which is poorly soluble in most organic solvents due to the short side chains. All complexes couldb ec haracterized by 1 H/ 13 CNMR in solution, solid-state NMR, and gave good results for elemental analyses.

Crystal structure analysis
By slow vapor diffusion of Et 2 Oi nto solutions of 1-3 in EtOAc, pale yellow single crystals suitable for X-ray diffraction analysis could be grown. 1 and 2 both were solved and refinedi nt he triclinic space group P-1, whereas 3 could be solved in the monoclinic space group P2 1 /n (for details of crystal structure determinations see Figures S1-S6 andT ables S1-S3).I nt he following,t he arrangements of the stacked molecules in the solid state will be characterized in terms of al ongitudinal shift along the molecular axis as well as by al ateral offset as depicted in Scheme 1. Figure 1s hows the obtained molecular structures and packingo fc omplexes 1 to 3 in their crystal structures. For comparison, an excerpto ft he crystal structure of previously reported analogue 5 is also shown. Relevant close intermolecular contacts are highlighted. In the crystal structure of 1 ( Figure 1B), the packing can be viewed as an arrangement of longitudinallys hiftedm olecular units mainly stabilizedb y CÀH arom ···O (2.31 ,2 .58 )i nteractions and ac lose Pt···H contact (2.91 )b etween the central Pt atom of one molecule and the protono ft he inner glycolc hain of an eighboringm olecule. Such Pt···H interactions have been observed in earlier reports and can be explained by donation of electron density from the d z 2 orbital of the Pt to the s*o rbitalo ft he CÀH. [13] Interestingly,n oa romatic interactions are observed in the crystal structure of 1 (for at op view of the p-system illustrating the orientation of the aromatic rings, see Figure S2). Instead, the structure formation is dominated by the TEG chains, which form al arge, bulky structures urrounding the neighboringa romatic backbonei naso-called handshake motif (Figure1C), similarly to that observedf or complex 5 ( Figure 1A). [11] The lack of close p-p contacts for 1 compared to the higher homologue 5 can be explained by the considerably larger aromatic surfacefor the latter (six vs. four aromatic rings), which enables am ore efficient p-overlapo ft he OPE backbone in the crystal. The relatively short p-systemo f1 must orient in at wisted manner with at orsion angle of the two aromatic rings (phenyl and pyridine) of 53.88 to form the present structure. Stabilization of the handshake motif is achieved by intermolecular CÀH arom ···O interactions exhibitingd istances in the range of previously observed weak hydrogenb onding for such systems (2.41-2.56 ). [11] The closest PtÀPt distance that can be found in the packing of 1 is 9.854 ,w hereas the PtÀPt distance illustratingt he longitudinals hift of two molecules is 18.120 (Figure 1B).
In the single crystal structure of 2,ad istinct change in the arrangement compared to 1 is apparent ( Figure 1D). Although the molecules still arrange in as lipped manner,t he change from TEG to DEG side chainsf acilitates am olecular stacking through p-p interactions, leading to an assembly with ad ecreasedl ongitudinal moleculars hift (inner stack PtÀPt distance 8.126 vs. 18.120 in 1)a nd as light lateral offset (see also Figure S4). The parallel displaced p-stacking of two pyridine units has ad istance of % 3.2 .I nc ontrast to the structure of 1,w here the Pt II Cl 2 fragment is surrounded by TEG chains, the Pt II Cl 2 moiety in 2 is located on top of the CCb ond of the next molecule in thes tack.T he structure is furthers tabilizedb y CÀH glycol ···O (2.44 )a swell as CÀH glycol ···C arom (2.90 )interactions. Thus,i nstead of a handshake motifa si n1,t he DEGs idec hains in 2 interdigitatew ith neighboring sidec hains ( Figure S4). Furthermore, an intermolecularw eak CÀH glycol ···Cl (2.84 )i nteraction is observed within adjacent stacks ( Figure S4).
In contrast to compounds 1-3,s ingle crystals of 4 could not be obtained. However,c onsidering the trend followed by 1-3, we would expect am ore parallel arrangement of the molecules in 4 (shorter PtÀPt distances) and enhanced aromatic interactions compared to 1-3 due to the lack of any side chains. Due to no availables ingle crystal X-ray diffraction dataf or complex 4,acomplementary technique (solid-state NMR) was required to elucidate the molecular arrangement in the solid state.

Insights into the molecular packingfrom solid-state NMR
Our approach consisted in identifying intermolecular contacts characteristic for the packing motifs observed in 1-3 and 5.B y comparing these contacts with those potentially observed (or alternatively not observed) for 4,apacking arrangement for this complex can be derived. Alls amples usedf or solid-state NMR measurements were powdered samples andw ereo btained under identical conditions as those used for the preparation of the single crystals discussed in the previous section (EtOAc/Et 2 Os olutions).W et herefore assumed identical (or at the very least comparable) arrangements of the molecules in the samples for solid-state NMR studies and in the crystalline state. Additionally,w eh ave conducted powder XRD analysiso f 1-4 to furtherp robe this assumption.Adetailed summary and analysiso ft he XRD resultsc an be found in the Supporting Information. Compound 1 shows the lowest crystallinity of all samples obtained by the precipitation protocol ( Figure S18). While the positions of the reflexes are in relativelyg ood agreement with the predicted XRD pattern (based on the corresponding single-crystal data), the overall intensity is greatly diminishedi np articular for the range corresponding to the longrange order.T he sterically demanding TEG chains, whicha dopt ah ighly ordered orientation in the single crystal, most likely preventt his long-range order in the powder sample;n evertheless, as imilar short-range order could be confirmed. On the other hand, the higher crystallinity of 2 is reflected in the sharp reflexes, which match the predicted XRD pattern well ( Figure S19). Yet, the observation of some minor discrepancies, in particularf or the immediate environment of the Pt II moiety, indicates that the highl evel of interdigitation of the DEG chains might not be perfectly transferred to the powdered sample used for solid-state NMR studies. Nevertheless,b ased on the reasonablyg ood agreement betweent he experimental and predicted XRD patterns, we infert hat the orientation within the 1D stack is largely preserved. For the stericallyl east demanding compound for which single-crystal analysisw as possible (3), the predicted XRD pattern reproduces the experimental results without discrepancies( FigureS20). This result is logical considering the absence of significant steric effects when the chains are reducedt oE Gu nits. On this basis, we infer that the molecular orientationi nt he bulk as well as in the single crystal matches. For compound 4,s harp reflexes could be appreciated as well, which suggests that the sample has ah ighly crystalline nature,w hich matches the observed trend for compounds 1-3 ( Figure S21). Figure 2A summarizes the 1 HM AS NMR spectra of 1-4. Overall,b road 1 Hl ines with few discernable detailsa re observedd ue to the strong 1 H, 1 Hd ipolar couplings present in the solid state. Nevertheless, apart from the broad peak around d iso ( 1 H) = 3.5 ppm assigned to the protons of the EG moieties, two signals in the aromatic region are observed: while the signal at higher ppm values (d iso ( 1 H) = 9.1 to 8.1 ppm) can be assigned to the aromatic protons adjacent to the pyridine nitrogen (H a ,p osition Ha in Figure 2), the 1 H signal at lower values (d iso ( 1 H) = 7.8-6.8 ppm) corresponds to the aromatic protons Hb (H b -pyridyl) and Hc.T he exact position of the two peaks varies between complexes 1 to 4,g oing from d iso ( 1 H a ) = 9.1 ppm and d iso ( 1 H b,c ) = 7.8 ppm in complex 1, to d iso ( 1 H a ) = 8.1ppm and d iso ( 1 H b,c ) = 6.8 ppm in complexes 2 and 3.A st he moieties are structurally identical, the difference can be solely attributed to packing effects, in this case to aromatic ring current effects caused by the neighboring aromatic backbonei n2 and 3,o rt he lack thereof in 1.T he shift to lower ppm values is often indicative of p-p interactions as observed for example in p-conjugated polymers and otherc onjugated systems. [14] Replacing the EG moieties with methoxy groups as in complex 4 leads to am arkedly different 1 HM AS NMR spectrum, cf. Figure2A. The signal at d iso ( 1 H) = 3.5 ppm is now significantly reduced in intensity and width, and strong shifts of the 1 Hs ignals in the aromatic region are observed. The upfields hifto f proton c suggests an even stronger participation of the outer ring in p-p interactionsc ompared to complexes 2 and 3.I n contrast, the 1 Hs ignal a is situated at ah igher ppm value than that of 2 and 3,w hichm ight lead to the wrong conclusion that the pyridyl ring of 4 undergoes weaker aromatic interactions than that of 2 and 3.H owever, this effect is typical for bispyridyldichlorido Pt II andP d II complexes exhibiting a pseudo-parallel packing,a nd can be attributed to CÀH a ···Cl interactions between the a-pyridyl protons of one molecule and the electron-withdrawing Cl ligands of an adjacent molecule. [4] To furtherp robe the molecular packingo fc omplex 4,2 D 1 H, 1 Hd ouble-quantum single-quantum correlation (DQ-SQ) MAS NMR experiments were performed ( Figure 2B,CandS 9). This kind of experiment allows to identify 1 H, 1 Hs pin pairs in close spatial proximity to each other (up to % 3.5 in rigid organic solids and longerd istances if the systems are flexible or proton diluted). [15] The 2D spectra of complexes 1 to 3 ( Figures  S8 and S9) are characterizedb yabroad auto-correlation signal situated at d SQ = 3.5 ppm and d DQ = 7.0 ppm (attributedt ot he EG units) with some smaller auto-correlation peaks visible at lower ppm values most likely originating from the residual trapped solvent. However,n oa utocorrelation signals in the aromatic region (around d SQ = 7ppm) are observed. In analogy to the 1 HM AS NMR spectrum,r eplacing the EG units with methoxy groups leads to the appearance of three aromatic 1 Hs ignals in the DQ-SQ correlation spectra (Figure 2B,C ). While the cross-correlation signals of protons Ha and Hc are readily discernible at d SQ = 8.6 ppm and d SQ = 5.9 ppm (d DQ = 14.6 ppm, dark blue, marked ac and ca), the broad Gaussian peak associated with protons Hb at around d SQ = 7ppm is mostly obscured. In addition to the cross-correlation between the methoxy protons anda romatic proton Hc at d DQ = 9.4 ppm-whichi se xpected duet ot he close spatial proximity of the methoxy groups on the same ring-a crosscorrelation between the methoxy protons and proton Ha is observeda td DQ = 12.2 ppm. Interestingly,n oc ross-correlation between the aromatic proton Hb and the methoxy protons is observable, thought his could be relatedt ot he broad nature and comparably low amplitude of the peak at d SQ = 7.0 ppm. These results suggest that the methoxy groups are placed in relativelyc lose proximity to the complex center, though whether this is from above/below the aromatic backbone or from other neighboring complex is unclear at this point. The most significant differenceb etween the spectra of 4 and the other three samples is the appearance of severala uto-correlation peaks in the aromatic chemical shift range (marked orange aa,b lue-green bb,a nd yellow-green cc). While these are barely discernible after one rotor period of DQ excitation ( Figure 2B), their amplitude increases when going to longer (four rotor periods) DQ excitation times ( Figure 2C). This indicates a 1 H-1 Hd istance of around 3.5-4.0 ,p ointingt op-p stacking of the monomers with ap arallel arrangement of the molecular units ( Figure 2D).
To summarize the resultsf rom 1 HM AS and 1 H, 1 HD Q-SQ NMR correlation data, 1 HM AS NMR revealed changes in chemical shift of the aromatic protons associated with differences in aromaticr ing current effects, with 1 showing weaker and 2 and 3 showing stronger ring current effects. Protons Ha and Hb in 4 show am oderate effect of % 0.5ppm, which would indicate ap ackingm otif where the protons are further away from the neighboringa romatic ring center pointing towards the edges of the said ring, thoughp roximity of Ha to the electron-withdrawing Cl (H a ···Cl) could lead to the shift to higher ppm. These observations fit well with ap acking model similar to the pseudo-parallel molecular arrangement observed by Allampallye tal. for al onger, methoxy-functionalized OPE-based Pt II complex. [4] To confirm the resultsf rom 1 HM AS and 1 H, 1 HD Q-SQ NMR correlation spectroscopy, 13 C{ 1 H} cross-polarization (CP/MAS) and 2D 13 C{ 1 H} heteronuclear correlation (HETCOR) NMR experiments were performed (detailsa re given in Figures S12-S17). Figure 3s ummarizes the 13 C{ 1 H} CP/MASN MR spectra of complexes 1-4 along with the peak assignment. Note that the assignment uses the numbering given in Figure 3, which is different from the numberingo ft he carbonsi nt he crystal structures. The assignment is based on ac ombination of variable contact time CP measurements forc omplex 1 as well as the corresponding 2D 13 C{ 1 H} HETCORN MR data ( Figure S14). To betterv isualize the changes in isotropic 13 Cc hemical shift for the different complexes, the 13 Cs ignals associated with the OPE backbone are plottedi nt he lower part of Figure 3. The most striking differenceb etween the spectrumo fc omplex 4 and those of complexes 1-3 is the splitting of the broad 13 C signal at d iso ( 13 C) = 129 À122 ppm, assigned to the b-carbono f the pyridine moiety (C b ;p osition C2). This carboni sd irectly bonded to proton Hb,w hich was previously attributed to a broad Gaussian-shaped peak in the 1 HM AS NMR spectra (vide supra). Thus, the doublet character of this 13 Cs ignal indicates a more complex 1 Hl ine shape, with the Gaussian distribution being merely an approximation, which further explainst he complex cross-correlation peaks observed in the 1 H, 1 HD Q-SQ NMR spectrum of Figure 2B,C .A st he substitution of the EG groups occurs on the outer phenylr ing, these changes observedf or C2 and Hb must be related to intermolecular packing effects. Another interesting feature of Figure 3i st he lowfield shift of the alkyne carbons C4 and C5 for complexes 2 and 3 compared to complex 1.T his low-field shift is not observed for complex 4.B ecause the packingm ode switches from the handshake motif with al ong-shifted arrangementi n complex 1 to al ess shifted, staircase pattern in complexes 2 and 3 with the Pt II Cl 2 center being close to the alkyne triple bond,w ec an attribute the low-field shiftt oachange in packing.
Several of the 13 Cs ignals across all four complexes in Figure3 are split into doublets, most notably the 13 Cs ignals assigned to carbon C7 at d iso ( 13 C) =~109 ppm, as well as those for C3, C4,a nd C8 in the case of complex 1.T he splitting of carbons C7 and C8 can be explained by two inequivalent carbon sites in the asymmetric unit of the crystal structure. However, according to the single-crystalX RD data, carbons C3 and C4 only occur once in the asymmetric unit of complex 1. Therefore, it is likely that the sample used for MAS NMR has a slightly different packing and additional disorder compared to the single crystal used for XRD (cf. Figures S18-S20). Figure 4 summarizes the 2D 13 C{ 1 H} HETCOR spectra for 1, 2,a nd 4 acquired with 0.5 ms (green contour lines) and 3.0 ms CP contact time (black contourl ines)t od ifferentiate between directly bonded 1 H- 13 Cand intermolecular 1 H-13 Ccontacts, respectively. Ac omparison of these 2D HETCOR NMR datasets( green and black)a llows us to identify the packing structure for each com-plex as illustrated by the structuralf ragments and corresponding assignments of 1 H, 13 Cc orrelations below these spectra (see captiono fF igure 4f or details of the assignment). The spectrum of complex 1 shows two 13 Cs ignals for carbon C8 due to the existence of two inequivalent positions in the crystallographic asymmetric unit. As for complex 5 (data shown in Figures S12 and 13), this also results in different intermolecular 1 H-13 Cd istances fort he two positions, manifested in the 2D HETCOR NMR spectra as correlations HbC8 and HbC8' labeled as b8 and b8',r espectively,i nF igure 4A (complex 1). The large longitudinal shift between two molecules in 1 is also evident from the correlation HbC9,w hich links proton Hb (located at the complex center) to the outermost OPE carbon C9,ac orrelation that is only possible via intermolecular contacts. Af urther indicator for the magnitude of the longitudinals hift is the absence of potential intermolecular correlations HbC6 and HaC5,w hich would be expected for an intermediate shift (these are present in the case of complexes 2 and 3,s ee Figure 4B).
Since the solid-state 2D HETCOR NMR spectrao f2 and 3 are very similar,o nly those for 2 will be discussed in the following ( Figure 4B). As mentioned before,t he crystal structures of 2 and 3 lack the characteristic TEG handshake motif and feature am oderately shifteda romatic p-p stacking arrangement. This packing can be confirmed from the 2D HETCOR datasetsi n Figure 4B for 2 (see Figure S16 for 3), which shows numerous intermolecular correlations, most notably HbC6, HaC5, HaC4, HbC9, HcC2,a nd HcC1.A lthough HbC9 and HbC8 were also observed for complex 1,c orrelations HcC1,a nd HcC2 place the pyridine moiety close to the second phenyl ring of the OPE backbone. Interestingly,t he correlationsb etween C7 (bonded to proton Hc)a nd protons Ha or Hb (attached to carbons C1 and C2)a re not observed. In addition, correlations HaC5 and HaC4-which were not observed for complex 1are indicators for the Pt II Cl 2 complex center being in the proximity to the alkyne triple bond of the neighboringm olecule.
The 2D HETCOR spectra of 4 together with the proposed molecular arrangement are shown in Figure4C. In contrast to the 2D HETCOR spectraf or 1 and 2,n one of the intermolecular correlationss uch as HbC9, HaC9, HaC6, HaC5,a nd HaC4 are observed. This can only be achieved for as tructural arrangement with ap arallel or nearly parallel p-p stacked arrangement of the OPE backbones, similart ot he one described by Allampally et al. for the larger OPE-based complexw ith methoxy functionalization. [4] Such as tructural arrangementi sa lso supported by the 1 H, 1 HD Q-SQ NMR correlation data ( Figure 2). Thus, through ar igorousc omparison and careful assignment of observed and unobserved 1 H, 13 Cc orrelationsf or complexes 1-3 and 5 with knownc rystal structures, it can be concluded that the aromatic interactions for the OPE backbone of complex 4 occur in an early parallel fashion as illustrated by the molecular fragments in Figure4C.

Conclusions
In summary,w eh ave reported the synthesis and molecular packing of four new amphiphilic Pt II -based complexes 1-4 with short aromatic OPE backbones and hydrophilic oligo(ethylene glycol)t ermini of different length.Asystematic length variation was achieved by gradually removing one (2), two (3)o rt hree (4)O CH 2 -CH 2 units from the TEG-substituted derivative 1.D ata from single-crystal X-ray diffraction showed  . 2D 13 C{ 1 H} HETCOR spectra and crystal structure fragments for (a) 1 and (b) 2 depicting the relative orientation of two successively stacked molecules. (c)Proposed molecular packing and 2D 13 C{ 1 H} HETCORs pectra for 4.Side chains have been omitted for clarity. All 2D HETCOR spectrai nclude two different CP contactt imes of 3.0 ms (black)a nd 0.5 ms (green)contour lines to differentiateb etweeni ntermolecular 1 H-13 Ccontacts and directly bonded 1 H-13 C correlations, respectively. The assignment uses numbersfor 1 Ha nd letters for 13 Cp ositions as indicated in each molecular fragment. Red/green arrows and assignments to specific 1 H-13 Ccontactsi ndicateobserved correlations; dark blue arrows and circles indicate correlationst hat are not observeda sd iscussed in the maintext. Asterisks indicatespinnings ide bands. that complex 1 retainedt he characteristic handshake motif previously described for complex 5,w hich features the same TEG side chains, but has al arger aromatics urface than 1.T he strong similarity between the packingm otifs of 1 and 5 indicate that the TEG groups with their associatedh ydrophilic interactions are the primary forces dictating the solid-state structure. However,i nc ontrast to 5,c omplex 1 shows no significant aromatic interactions duet ot he reduced aromatic surface. Shortening the hydrophilic side chains in complexes 2 and 3 resultedi ns tructures with longitudinally shifted molecules involvingm ore pronounced p-p interactions, thought he bulky Pt II Cl 2 moiety in the complex centerp revents efficient p-p stacking between neighboringc omplexes.T wo-dimensional solid-state 1 HM AS and 13 CM AS NMR experiments for complexes 1, 2, 3,a nd 5 allowed us to identify 1 H, 13 Ct hroughspace correlationsc haracteristic for each packing motif. These, in turn, led us to proposeapacking motif for the only complex (4)f or which single crystalsc ould not be grown, emphasizing the potential of solid-state NMR spectroscopya sakey tool for structural determination in the solid state. Complex 4 was found to exhibit am ore parallela rrangement of the molecules stabilized by p-p interactions, mostl ikelyw ith as hort longitudinal and lateralo ffset to accommodate the stericallyd emanding Pt II Cl 2 moiety at the complex center. On this basis, we conclude that the steric demand of the central Pt II Cl 2 moiety limits the effectiveness of p-p interactions;t his is supported by the fact that even short mono(ethylene glycol) terminia ppear to prevent an efficient p-overlap of the ligands aromatic surface. Am ore pronounced p-overlapi so nly possible after eliminating hydrophilic interactions by replacing the ethylene glycol termini with methoxy groups. Therefore, if p-p interactions are frustrated by steric demands, hydrophilic interactions will become the dominant factor in the self-assembly process, leading to structurala ggregation. Thus, our investigations illustrate that even minor changes in the glycol chain length can be used to modulate and control aromatic interaction strengths and packing arrangements, which in turn might be useful for engineering electronic devices based on Pt II and aromatic molecules.

Experimental Section
See the Supporting Information for details on single-crystal diffraction and solid-state NMR experiments. Deposition numbers 2004413 (1), 2004414 (2), and 2004415 (3)c ontain the supplementary crystallographic data for this paper.T hese data are provided free of charge by the joint Cambridge Crystallographic Data Centre and Fachinformationszentrum Karlsruhe Access Structures service.