Frontiers in Polymer Chemistry

804 CHIMIA 2013, 67, Nr. 11 Polymers, Colloids, and interfaCes

doi:10.2533/chimia.2013.804 Chimia 67 (2013) 804–810 © Schweizerische Chemische Gesellschaft Frontiers in Polymer Chemistry

A. Dieter Schlüter*

Abstract: The article shows how the initial concept of Staudinger on linear macromolecules was expanded topologically by increasing the cross-section diameter of polymer chains and by introducing sheet polymers with planar rather than the commonly known linear repeat units. The two concrete projects addressed are the synthesis of dendronized and of two-dimensional polymers. It is explained how these novel macromolecules were achieved and which obstacles had to be overcome but also where these frontiers in polymer chemistry might lead to new insights in polymer science in general and novel applications in particular. The article also provides insights into analytical issues because both target macromolecules are in an extraordinarily high molar mass range and contrast/sensitivity issues can turn rather serious in particular for the two-dimensional polymers. Keywords: Dendronized polymers · Two-dimensional polymers

Introduction access to extremely highly isotactic poly- After intense thinking and brain-storming propylene (it-PP), which is an important with my co-workers and colleagues two In 1912, thus one hundred and one engineering plastic particularly in the au- research targets were in fact identified and years ago, Hermann Staudinger com- tomotive industry. Note that today’s plas- considered worth pursuing: ‘thick’ linear menced his professorship at ETH Zürich tics make up 50 percent of the volume macromolecules and two-dimensional and soon thereafter published ground- of new cars but only 10 percent of the (sheet-like) polymers (2DP). This short re- breaking papers on his perception of mac- weight.[4] But also tuning the molar mass view article tries to introduce the reader to romolecules.[1] Despite serious disputes and the molar distribution of many vinyl the merits of these two targets and where with those of his colleagues who did not polymers under rather robust conditions is we see their potential importance. It begins believe that there are molecules larger than not so much an issue anymore as it was with a few introductory comments, while a crystallographic unit cell, his experimen- some 25 years back. Controlled radical the major part of the article will go more tally carefully substantiated view of the polymerization procedures have opened up intensely into thickness and dimensional- existence of long linear chain molecules a new avenue into this field and to some ity matters in connection with polymers. (rather than colloidal assemblies of small degree replaced the experimentally quite Linear polymer chains are available molecules) became increasingly more ac- demanding anionic polymerizations.[5] Of with various different main chain struc- cepted worldwide, and ultimately proved the many polymers with novel structures tures as well as various different lateral indeed to be correct.[2] Nowadays, mac- that were synthesized in the past few de- substituents. Examples are poly(hydroxyl romolecules, also referred to as polymers cades, polyfluorenes (PF) deserve some acrylate), 1, and polystyrene, 2, which are – or more mundanely as plastic, and the attention.[6] They were made available important components in superabsorbent products made with them can be found in through recently developed transition met- hydrogels and for packaging applications, almost every aspect of life ranging from al-mediated cross-coupling protocols, one respectively (Fig. 1). The cross-section di- nutrition to healthcare and from trans- of which is Suzuki polycondensation.[7] ameters (thickness) of their backbones are portation to clothing [World plastics pro- PFs are typically used in form of copoly- increased by the lateral substituents, which duction 2012: 265 million tonnes. http:// mers e.g. as active components in organic are hydroxyethylester for 1 and phenyl www.plasticseurope.org], and are of an light emitting diodes (OLED). Finally, ex- for 2. While these substituents naturally enormous economic importance. Because amples of novel processing schemes are, affect somewhat backbone properties, of the huge impact of linear (and cross- for instance, the development of spinning such as persistence length (a measure for linked) macromolecules to all of us, there fibers from liquid-crystalline (lyotropic) chain stiffness) and entanglement length has been an intense research effort both in solutions of the rigid macromolecule (describing the average chain length be- industry and academia aiming at ever bet- poly(para-phenylene therephthalamide) tween entanglements), in a first approxi- ter synthesis procedures, structurally new (aramids) from sulphuric acid[8] and so- mation their impact on main chain cross- polymers and novel processing schemes called gel-spinning of ultra-high molecu- section diameter is small, irrespective of leading to both new fields of application lar weight (>3.106 g/mol) polyethylene[9] the fact that such a diameter is difficult as well as replacement of other materials that have a stiffness and strength 5–10 to define anyway. This becomes evident (e.g. metals) in known applications. The times that of steel on an equal weight basis. when comparing these polymers with the improvements in classical and supported These fibers have found numerous applica- linear polymethylmethacrylate (PMMA) metallocene catalyst systems shall serve tions in the construction, protection, sport derivative 3,[11] in which the ‘tiny’ methyl as a first example here.[3] They provided and medical industries. group of the ester is replaced by a large, Having this impressive triumph of lin- regularly branched (dendritic) substituent ear macromolecules laid out in front of (Fig. 1). Such a huge substituent at ev- *Correspondence: Prof. Dr. A. D. Schlüter us, we wondered whether there was still ery other backbone C-atom increases the Laboratory of Polymer Chemistry room for progress from the fundamental cross-section diameter enormously and by Department of Materials side. Were there challenges that had not this, changes the polymer fundamentally. ETH Zurich, HCI J 541 Wolfgang Pauli Strasse 10, CH-8093 Zürich. been looked into and which promised to Polymer 3 is not a PMMA anymore with E-mail: [email protected] lead to useful developments once tackled? a property profile making it the famous Polymers, Colloids, and interfaCes CHIMIA 2013, 67, Nr. 11 805

commercial product PlexiglassR; rather it Fig. 1. (a) Chemical is the lateral substituents in 3 which con- structures of typical trol everything and downgrade the back- linear polymers with bone from a major player, particularly low cross-section in terms of entanglement formation, to a diameter, 1 and 2, and a dendronized simple scaffold just holding all this mass polymer (DP) 3 (PG5) together. The backbone cannot undergo the with a huge regularly slightest conformational change anymore branched substitu- without ‘permission’ from the voluminous ent at every other mass surrounding it. Clearly something backbone carbon new has been created. All of a sudden one atom resulting in a can ask, for example, whether small com- ‘thick’ polymer with pounds such as solvent molecules can be an unprecedentedly put into a polymer main chain (a stupid large cross-section question in regard to 1 and 2), whether the diameter. (b) Results of molecular dynam- cross-section of a main chain is responsive ics simulations of six to external stimuli or whether such mac- dendronized poly- romolecules are molecular objects that mers differing by the can be rolled across solid substrates. This dendron generation may suffice for a moment to frame what is g from g = 1–6 (PG1– meant by thick polymer chains and where PG6). Courtesy of the importance of this aspect may lie. Prof. C. Aléman and The second target of our group’s re- Dr. Oscar Bertran, TU search efforts aims at a dimensional exten- Barcelona, Spain. See sion of Staudinger’s so successful concept ref. [10]. of macromolecules. He viewed macromolecules as long, linear, covalently bonded arrays of repeat units (RU) and, thus, defined a polymer chain by the chemical structure of a RU, the number of RUs Fig. 2. (a) Illustration and the two end groups that terminate the of the difference chain at both ends. This concept has been between a linear expanded over the decades to branched and (top) and a two- cyclic polymers as well as various forms dimensional polymer of copolymers, which however all have (2DP) (bottom). topologically linear RUs (except for the While the former has linearly connecting branching points). Looking at macromol- RUs, the latter has ecules under dimensionality aspects, we laterally connecting wondered whether it would not be possi- RUs resulting in a ble to create representatives with topologi- periodic monolayer cally planar RUs. Such RUs would give sheet with internally polymers a second dimension and would tessellated (periodic) free macromolecules from being restricted molecular structure. to topologically one-dimensional (linear) (b) TEM micrograph geometry. The term macromolecule would of a 2DP. (c) Cartoon then also include monomolecular sheet- representations of a conventional ultrathin like objects, the entire structure of which polymer film and a is a tessellated array of RUs terminated by 2DP. While the me- an ‘infinite’ number of end groups posi- chanical strength of tioned at the sheet circumference. Fig. 2 the former rests upon illustrates the difference between a classi- the chemical and cal Staudinger polymer and a two-dimen- physical netpoints sional polymer (2DP) as we see it. (cross-links and Because of the many insightful inves- entanglements, re- tigations carried out on ultrathin polymer spectively), the latter films, it is important to avoid confusion forms a mechanically coherent entity be- with this particular class of materials. We cause of its regularly therefore emphasize again that the second tessellated internal target of our studies does not aim at yet structure consisting another ultrathin polymer film but rather at of planar RUs. solving the fundamental synthetic problem associated with the generation of monolayer sheets solely consisting of topologically planar RUs. Fig. 2 pinpoints the difference. While currently known ultrathin films cannot be described by RUs and often even have a completely disordered, randomly 806 CHIMIA 2013, 67, Nr. 11 Polymers, Colloids, and interfaCes

cross-linked internal structure,[12] the sheets aimed at here are in principle characterized by a periodic array of covalent cross-links over their entire expanse. Such a complex target raises, of course, the point why one should go through all the effort with devising synthetic strategies and ana- lyzing (intrinsically low contrast) structures when slightly thicker films with no internal order can easily be accessed? Or in other words: What are 2DP good for? The answer to this question is twofold. On the one hand, these unique polymers are expected to add to the fundamental under- standing of polymer science and, on the other, they will lead to whole bouquet of novel applications. The fundamental un- derstanding would be advanced because the entire polymer physics was developed to describe and predict properties of linear chains and – with regular systems at hand can now be reasonably expanded and ad- justed to polymers of higher dimensionality, the sheets. Regarding the applications, Fig. 3. (a) Recursive chemical equation explaining how to synthesize DPs with ever increasing g. 2DP might prove useful as nano-scale (b) Comparison of the molar mass of PG5 with 10600 RUs with other synthetic polymers as well membranes (where they offer monodis- as the biological macromolecule amylopectin. (c) AFM height image of co-prepared PG1-PG5 perse holes), for sensing, catalysis, molec- chains to visualize the increased chain stiffness (decreasing undulations) and increased chain ular landscaping in general and all other cross-section dimensions (thickness) with increasing g. (d) AFM height images of a single PG5 aspects that leverage their very particular chain on an array of Tobacco Mosaic Virus (TMV) and of a single chain embracing a single TMV on mica. molecular organization.

agent DG1.After supposedly each terminal creation of such a mass has any meaning Thick Polymers and Cylindrical amine functional group has reacted with because it leads into novel, unexplored Molecular Objects DG1, the g = 2 DP PG2 is obtained. This scientific terrain. As the AFM images in procedure can then in principle be repeated Fig. 3 suggest, DPs with increasing g turn The divergent growth concept, as first over and over again until the DP with de- more and more into molecular objects to reported by Vögtle,[13] is a powerful syn- sired g is reached. In the case mentioned which an interior and an envelope can be thetic measure to increase the size of any above, Dr. Baozhong Zhang, a senior sci- assigned. However, increasing of g cannot molecule that carries at least a few func- entist in our group, carried the sequence be continued forever: Mass increase per tional groups to start the growth from. This through to PG5 which – after careful de- divergent growth step is exponential while has impressively been shown by Tomalia fect quantification[17] – had a molar mass increase in cross-section span is linear.[18,19] for dendrimers, which are more or less of approximately 200 MDa.[11] In the last Thus, there must be a point where all spherical macromolecules, the diameter step, when going from PG4 to PG5, the branches are densely packed around the of which depends on the number of con- mass increase amounted to 100 MDa. This backbone and further growth can only be secutive divergent growth steps.[14] Their increase, which proceeds in a controlled achieved at the expense of reduced struc- divergent synthesis starts from a small fashion, is the largest mass that has ever ture perfection. For packing reasons not oligofunctional core molecule to which been generated in a single reaction step to all terminal amine groups that could react branches are added step-by-step, genera- give a structurally defined, covalent mol- with a dendronization agent have the space tion (g) by generation. A bench mark syn- ecule and is more than any polymerization available anymore to actually do so. It is thesis in terms of size achieved and num- reaction has so far afforded. Also in terms thus noted that DPs have two different g ber of growth steps is Majoral’s dendrimer of total molar mass, this reaction is remark- regimes. They need to be differentiated, with twelve branching points and thus g able and shows the enormous potential of in order to understand what DPs actually = 12.[15] This tool was available when our divergent growth. With its molar mass of are, what thick molecular object actually group in 1994 started to do first steps to- 200 MDa, PG5 not only by far exceeds all means, in terms of structure perfection but wards systematically thickening polymer synthesized macromolecules (Fig. 3b) but also in terms of property profile. For the chains.[16] Fig. 3a shows the outline of the also surpasses most biological macromol- first regime g < g applies and for the max applied protocol. It starts from a first gen- ecules. Amylopectin, the storage form of second g ≥ g . The straight (black) line max eration polymer (g = 1; PG1) whose molar glucose, ranges in the same mass category in Fig. 4c marks the maximum cross-sec- mass had precisely been determined by and only the DNAs, which can contain mil- tion span as it develops with g, while the light scattering and gave the equivalent of lions of nucleotides, are still much higher green line is the span resulting from pure 10600 RUs. PG1 carries two protected ter- in molar mass. The AFM height image in density considerations (assuming in a first minal amine groups, the (Boc-) protection Fig. 3c shows semi-quantitatively that the approximation a density ρ ~1.1 g/cm3).[19] groups, which needed to be completely re- DPs become increasingly more rigid and g is at their intersection and thus at ap- max moved under acidic conditions before the also thicker with increasing g. proximately g = 6.5 (for this particular resulting ammonium-terminated interme- While an average molar mass of 200 molecular structure). Below this point both diate (not shown) could be brought to re- MDa for a single molecule may sound lines enclose an area which means that DPs action with an excess of the dendronization impressive, one needs to ask whether the below g should be responsive to external max Polymers, Colloids, and interfaCes CHIMIA 2013, 67, Nr. 11 807

stimuli.[20] Their maximum cross-section is not yet reached in the dry state and their interior has capacity to accommodate solvent or guest molecules (such as drugs) to the point that the branch work is fully stretched out and reaches the black line. On the contrary, at and beyond g , DPs max should (ideally) not be responsive anymore with interesting consequences in terms of properties and potential applications. But now let’s first concentrate on the g < g max regime. Below g the branch work around the max backbone of DPs has capacity to take up solvent and other molecules.[21] This is a unique property of DPs and sets them apart from common linear polymers, the backbones of which have no internal volume. This useable volume of DPs qualifies them as molecular containers which can in principle be used to selectively transport cargo from one side of a membrane to another, reminiscent of DNA translocation through nanopores.[22] The uptake is driven by the osmotic pressure or specific interactions Fig. 4. (a) ChemDraw structures of parts of dendrons of generation g emanating from the back- between guest molecules and the DP host. bone in a stretched conformation and thus defining the corresponding DPs maximum cross- DPs can also squeeze out their guest mol- section span (multiplied by factor 2). (b) Molar masses of the RUs of PG1-PG5 and the diameter ecules under certain conditions and thus resulting from such masses based on density (from measurements with the shown model com- pound and molecular dynamics (MD) simulation). (c) The g dependencies resulting from (a) and (b) force the DP chains to precipitate from so- and their intersection at g = 6.5 which marks g . lution (loss of osmotic stabilization). This max squeezing out, which is an important process e.g. in regard to unloading of cargo, can to fascinating findings there is no concrete tions aiming at their properties become be achieved by temperature changes if the knowledge yet about the g > g regime. available. Importantly, a combined analyti- max dendron/dendron and dendron/guest mol- The reason for this is lack of experimen- cal/theoretical tool has already been devel- ecule interactions are carefully chosen.[20] tal availability. Only very recently in our oped to try to quantify defects also in this The property of temperature-dependent group the doctoral student Hao Yu reached delicate range.[24] Fig. 5b displays a dream loading and unloading and its direct impact and surpassed g DPs synthetically and of the author in this regard. As mentioned max on persistence lengths may lead to new ma- one needs to wait until the first publica- above, at and beyond g DPs should have max terials based on slightly cross-linked DPs. Such networks may have an extraordinary temperature dependence of their elastic moduli. Because for g < g the backbone max is not in its fully stretched conformation, the contour length is subject to change.[23] Changes can be induced by external stimuli such as pH value. If one uses DPs with un- protected terminal groups the pH value of the surrounding aqueous medium controls the level of charging. The lower the pH the more effective charges will be placed on the DP resulting in a stretching of the backbone by charge repulsion. If the pH is increased, however, the number of effective charges on the DP is reduced allowing its backbone to contract for entropic reasons. Such changes in contour lengths have been determined by AFM pulling experiments. It was found that contractions by 20% are within reach and that the forces associated with this contraction are in the range of 1 nN.[23] This in principle qualifies DPs for Fig. 5. (a) Typical responsiveness in the g < g regime: (top) loading capacity of the cylindri- applications in molecular machines, e.g. to max cal objects as defined by unused volume compared to fully stretched structure; (middle) strong lift a weight that is fixed to one end of a DP response of temperature on backbone stiffness for DPs in theta solvents; (bottom) charged DPs chain. Fig. 5a displays cartoons describing as molecular machines which, depending on salt concentration change can contract or elongate. these properties of g < g DPs. max (b,c) Two (future) examples for the g > g regime: Rolling of DPs to be used for anisotropic lubri- While the exploration of the g < g re- max max cation and creation of a homologous series of single molecule colloidal particles with quantized gime is well underway and has already led diameter difference of ~1 nm between next neighbor DPs. 808 CHIMIA 2013, 67, Nr. 11 Polymers, Colloids, and interfaCes

a densely packed dendritic branch work recent case of a differing definition shall be monomer structure is given, which com- surrounding their backbone, ideally not mentioned, however, to illustrate how im- bines a collection of important features. allowing for molecules to penetrate into portant it is to find a consensus. Colson and First of all it is shape-persistent, which their interior because of a densely packed Dichtel[27] suggested in a 2013 publication means its conformational space is re- structure. This dense packing results in to not only consider macromolecules as duced, a feature that can facilitate crystal- molecular objects, which are characterized 2DP that meet our definition but to also lization. Also it contains three anthracene by large molecules whose shape is inde- include “all covalently linked networks units and six acetylene units (marked in pendent of the environment the objects are of monomers with periodic bonding in red) that are held together by triphenylene exposed to. If this picture holds true, the two orthogonal directions” irrespective of bridges. Both anthracenes and acetylenes adsorptive forces DPs are exposed to when whether they have been isolated in “layered are photosensitive groups while the rest of placed on solid substrates should be small crystals, multilayer or monolayer films”. the monomer should be photostable at the because of the small contact area. Ideally This view was supported by comparing to wavelength used for polymerization later one can therefore expect the objects to be linear polymers for which the authors cor- on. Chart (a) also depicts that the mono- rollable on the substrate, similar to as it rectly stated that linear (one-dimensional) mers in their hexagonally shaped single was shown for carbon nanotubes.[25] This polymers remain as such whether they are crystals are arranged up and down in an would be an attractive feature because it found in the bulk, dissolved in solution, alternating sequence, whereby – and this should open the way to molecular lubrica- or isolated as individual macromolecules. is key! – each anthracene of one mono- tion and, if the DPs can be oriented, even What this comparison however lacks is the mer faces exactly one acetylene unit of to anisotropic lubrication (Fig. 5b). The g important aspect that for linear polymers it the neighboring monomer such that the > g regime however is expected to offer was proven countless times that, irrespec- acetylene is positioned right across the an- max even more.As mentioned, the first DPs with tive of whether they are in bulk, solution or thracene’s 9,10-positions. These positions g beyond g have been synthesized. This as individualized entities, they are in fact are prone to undergo cycloaddition reac- max is to be considered the first step towards clearly identifiable individual chains. This tions with 2π- or 4π-systems if the latter molecular colloidal particles. Because was achieved, e.g., by dissolving a (non- observe the so-called Schmidt distance,[31] there are countless procedures leading to cross-linked) bulk polymer into a solvent which in essence means that the orbitals of colloidal particles based on polymers,[26] it in which the then untied chains enjoy their both components need to have a minimum is important to consider the features DP- conformational freedom. The same has not overlap to bring the reaction about. Chart based particles would be offering. In the normally been achieved for synthetic lay- (b) illustrates a major breakthrough that all first place this is their cylindrical shape ered crystals though.[28] In other words, to those involved in the group’s 2DP project, combined with monodisperse diameters. consider any layered material a 2DP misses including Priv. Doz. Dr. Junji Sakamoto Everybody who has experience in the self- the point in the same way that graphite dif- and the doctoral student Patrick Kissel assembly of particles will appreciate it if fers from graphene, the prototypical 2DP celebrated. This breakthrough was not as- the particles used are strictly monodisperse provided by Nature. Graphite would not sociated with lots of glamour and pomp but and size segregations are avoided to begin normally be considered a 2DP. A discus- rather with a simple solubility difference. with. In addition to this, entire homolo- sion on this seemingly subtle discrepancy As soon as monomer single crystals on a gous series would be available (Fig. 5c), between definitions was chosen as the en- glass plate were reached by a spreading next neighbor members of which differ trée for this small chapter because it is in solvent front, they behaved as any single by approximately 1 nm in diameter only. fact not subtle at all but rather touches the crystal suddenly surrounded by solvent This allows for an extreme fine tuning in very heart of what a 2DP is and what not, would do: they dissolved. If, however, the terms of supramolecular aggregates com- and is therefore believed to be instructive single crystal prior to this experiment was pared to what is possible these days with to a reader not so familiar with this matter. photoirradiated at a wavelength to bring conventional colloids.[26] Finally, it should While there may be many synthetic about the cycloaddition reaction between be mentioned that the internal segment dis- paths to 2DP, we concentrated on ap- the opposing reactive units of neighbor- tribution of colloidal DP particles is much proaches where specially designed mono- ing monomers, the crystals remained in- better known than in any other cross-linked mers were pre-oriented in two dimensions tact! Thus, something fundamental must nanoparticles. While the precise spatial prior to being connected with one another have happened with the monomers that position of a certain branch point within in a regular fashion. Pre-orientation can be prevented them from going into solution. a given particle can of course not be fore- achieved, e.g. in liquid crystalline phases, Obviously we suspected that the desired seen, the probability distribution function at an air/water interface and in layered cycloaddition may have happened, which can well be established. This is an unheard single crystals. Our laboratory pursued and has the potential to convert the entire structure precision for colloidal particles is still pursuing the last two approaches. monomer layers within the crystal into and will be of value for future applications. Because structure characterization is par- polymer layers. At this point a bit of care is ticularly complicated with monolayers, appropriate. The monomer crystals were in the interfacial experiments have not yet the size range of 10–100 µm which may be Two-dimensional Polymers (2DP) reached the maturity[29] of the ones based considered relatively small. In terms of the on layered single crystals.[30] This is why number of monomers being present in each The term 2DP has often been used in this space-limited article concentrates ex- layer ranging from one end of the crystal to the literature but in a less strict sense than clusively on a recently published, first re- the other, however, they were huge. Such the one advanced by our laboratory, which alization of a 2DP based on photopolymer- a span easily involves 10’000–100’000 – as mentioned above - is based on two in- ization within a monomer crystal followed monomers. What now is the significance of gredients, the concept of RUs originating by exfoliation of the resulting crystal this number? Assume there is a positional from Staudinger and the monolayer nature. composed of 2DPs into individual sheets. change between next neighbor monomers This short highlight cannot possibly offer This selection does therefore not reflect a resulting from the chemical reaction be- a comprehensive treatment of the available superiority of one of the approaches over tween them. Even if this change is tiny it literature to this rather complex matter and the other. will add up to a sizeable quantity for a huge the reader is therefore referred to a 2009 Fig. 6 shows important intermediate number of monomers, which results in review article by Sakamoto et al.[12] A very steps towards the first 2DP. In chart (a) the formation of small cracks or phase bound- Polymers, Colloids, and interfaCes CHIMIA 2013, 67, Nr. 11 809

aries. This phase change will terminate growth! Thus, it cannot be automatically expected that the 2DP monolayer sheets obtained from such an approach have the same size as the crystals from which they are generated. Also it would not be clear from the very beginning whether crystal geometry (here hexagonal) will actually translate into sheet geometry. After this somewhat cautioning comment it is time to explain what happened when the po- lymerized crystals were subjected to exfoliation procedures, that is to procedures in which the crystals are exposed to inter- calating solvents.[32] Such solvents creep in between the layers, swell the crystals and eventually disintegrate them. Initially sheet packages are obtained which then further exfoliate all the way down to single sheets (presuming sheets are not intercon- nected through defects). Chart (c) of Fig. 6 shows the process schematically. In fact, monomer crystal could be exfoliated into Fig. 6. Converting a monomer single crystal into a 2DP. (a) Chemical structure of monomer. packages as well as single sheets. The for- Optical micrograph (OM) of its single crystals. Internal packing into layers. (b) OM images of mer were used for analysis by transmission monomer crystals which dissolve in N-methylpyrrolidone (NMP) contrasted with an irradiated electron microscopy (TEM) and electron crystal (false colors) which does not dissolve anymore under the same conditions. (c) Sketch of diffraction (ED) (chart d) and found to be the exfoliation process. (d) Wiener-filtered TEM image of an irradiated crystal, a model explaining in excellent agreement with the proposed the contrast and an electron diffractogram proving the order. structure of a 2DP. The final breakthrough was achieved by atomic force microscopy Andri Schütz, a postdoc in our group, is (Lacey carbon). Without this network all (AFM) analysis which gave the expected presently investigating how to perform ex- sheets would simply pass through the grid height of a single sheet for the latter (not foliations such that a maximum of single holes and the filtering effect would be zero. shown). As expected, Raman spectroscopy sheets can be obtained. This turned out to But even with this arrangement it cannot revealed the molecular structure, thus, re- be a rather challenging nano-engineering be assured that single sheets, which are moving last doubts. In this particular case, matter and it is still not clear whether a expected to be rather flexible, do not pass the sheets of 2DPs were not yet spanned fully satisfactory solution can eventually through the much, much smaller pores of over holes but must have a considerable be found. Fig. 7 shows an overview scan- the carbon network. Despite these compli- mechanical stability because they survived ning electron microscopy (SEM) image of cations the SEM image is remarkable in the stressful exfoliation as well as filtration the transient situation, currently attained that it shows many, many features in a size steps through Lacey carbon grids requiring by Andri. Irradiated crystals (irradiation range that reasonably corresponds to the them to crumple and fold up and expand length is important) were subjected to mi- crystals used for exfoliation. Also the fea- again. crowave heating (heating/cooling profile tures, which are believed to be sheet pack- Now that the first case of a 2DP is es- and applied temperature are important) ages, are more or less transparent which tablished, can one lean back and relax? for a certain time (time is important) and means that they comprise of a few up to Certainly not! The proof of principle was the obtained dispersions (difficult to de- ~50 sheets. Note that the features burn achieved, yes, but nothing more. Countless termine quantitatively how many crystals away in the electron beam the easier the issues still have to be addressed, which have disappeared) filtered through a TEM thinner they are. Single sheets do so within sketchily include: to create broader struc- Cu-grid, the holes of which were covered a matter of seconds. This can be used as a tural basis both for the single crystal as with a holey network of amorphous carbon qualitative criterion to differentiate single well as the interface approach; to facilitate Fig. 7. SEM overview monomer synthesis; to develop large-scale image of 2DP sheets monomer syntheses for selected cases; to and sheet packages optimize crystal size; to establish exact captured on a filter polymerization conversions; to establish composed of a TEM analytical tools for routine 2DP analysis grid spanned with a [HR-TEM, Raman-mapping, molecular fine but irregular net- scale AFM and STM (where appropriate), work of amorphous etc.]; to standardize exfoliation procedure carbon (Lacey). A and determine exfoliation yield; to obtain feature with crumpled stable single sheet dispersions for polymer top is circled. physical studies; to investigate sheet properties e.g. the in-plane elastic modulus; to compare synthetic 2DPs with natural 2DPs such as graphene. Coming to the end, a recent result from an exfoliation study involving the above irradiated crystals shall be mentioned. Dr. 810 CHIMIA 2013, 67, Nr. 11 Polymers, Colloids, and interfaCes

sheets from packages. Many of the pack- noted that two further crystal-based 2DPs [10] O. Bertran, B. Zhang, A. D. Schlüter, A. ages exhibit a structure on their top remi- are about to be published and also the in- Halperin, M. Kröger, C. Aleman, RSC Adv. 2013, 3, 126; O. Bertran, B. Zhang, A. D. niscent of gathering a tablecloth with one’s terfacial approach is coming along very Schlüter, A. Halperin, M. Kröger, C. Aleman, J. hand. This is considered an indication of well. Thus, it seems as if a gold mine was Phys. Chem. B 2013, 117, 6007. the top layers being caught while undergo- discovered providing access to sheets with [11] B. Zhang, R. Wepf, K. Fischer, M. Schmidt, S. Besse, P. Lindner, B. T. King, R. Sigel, P. ing further exfoliation. structurally different RUs under ambient Schurtenberger, Y. Talmon, Y. Ding, M. Kröger, conditions allowing for rational organic A. Halperin, A. D. Schlüter, Angew. Chem., Int. synthesis. Ed. 2011, 50, 737. [12] J. Sakamoto, J. Van Heijst, O. Lukin, A. D. Summary and Outlook Schlüter, Angew. Chem., Int. Ed. 2009, 48, Acknowledgements 1030. Polymer chemistry is an exciting and This work was only possible because of [13] E. Buhleier, W. Wehner, F. Vögtle, Synthesis creative field of research. It allows put- the engaged work and creative input by my 1978, 155. [14] D. A. Tomalia, A. M. Naylor, W. A. Goddard III, ting fantasy and imagination into reality coworkers and colleagues. The names not (yet) contained in the publications are Chiara Angew. Chem., Int. Ed. 1990, 29, 138. and eventually societal use. Once target [15] M.-L. Lartigue, B. Donnadieu, C. Galliot, Gstrein (DP), Wenyang Dai, Payam Payamyar, structures have been identified the whole A.-M. Caminade, J.-P. Majoral, J.-P. Fayet, Marco Servalli, and Dr. Zhikun Zheng (2DP). Macromolecules 1997, 30, 7335. Also, see: J. repertoire of organic and polymer synthe- Particulary important contributions came from Lim, M. Kostiainen, J. Maly, V. C. P. da Costa, sis combined with the analytical arsenal Prof. Afang Zhang and Dr. Baozhong Zhang in O. Annunziata, G. M. Pavan, E. E. Simanek, J. available these days is at one’s disposal the field of thick polymers and from Priv. Doz. Am. Chem. Soc. 2013, 135, 4660. [16] R. Freudenberger, W. Claussen, A. D. Schlüter, and everything has just to be combined in Dr. Junji Sakamoto and Dr. Patrick Kissel in H. Wallmeier, Polymer 1994, 35, 4497. the proper way to achieve them. The two the field of 2DP. Both activities were further [17] L. Shu, I. Gössl, J. P. Rabe, A. D. Schlüter, projects presented here were started 1992 prominently supported by Profs. Avraham Macromol. Chem. Phys. 2002, 203, 2540; R. Al- (DPs) and 2002 (2DP), which is men- Halperin, Grenoble, and Martin Kröger, ETHZ, Hellani, A. D. Schlüter, Helv. Chim. Acta 2006, 89, 2745; R. Al-Hellani, J. Barner, J. P. Rabe, tioned to make clear that without stamina as well as em. Prof. Gerhard Wegner, MPI-P, Mainz. A fruitful collaboration on 2DP with A. D. Schlüter, Chem. Eur. J. 2006, 12, 6542; Y. exploring new territory may not lead any- Guo, J. D. Van Beek, B. Zhang, M. Colussi, P. Prof. Benjamin T. King, UN Reno, is also where. The author considers the present Walde, A. Zhang. M. Kröger, A. Halperin, A. D. gratefully acknowledged. We rely on a net- Schlüter, J. Am. Chem. Soc. 2009, 131, 11841. trend for short financing schemes and work of specialists who help us with structure Also, see: B. Zhang, H. Yu, A. D. Schlüter, A. demand for immediate success (whatever and property analysis. This includes Drs. and Halperin, M. Kröger, Nat. Commun. 2013, in this is) detrimental to progress in science. Profs. Carlos Aleman, TU Barcelona, Michal press. DOI: 10.1038/ncomms2993. [18] P. G. De Gennes, H. Hervet, J. Phys. Lett. 1983, Breakthroughs require a mental, financial Borkovec, U Geneva, Paola Ceroni, U Bologna, 44, L351. and operational environment without re- Rolf Erni, Empa Dübendorf, Lay-Theng Lee, [19] B. Zhang, R. Wepf, M. Kröger, A. Halperin, A. strictions, deadlines and the like. Also it is CEA Saclay, Jean-Christophe Leroux, ETHZ, D. Schlüter, Macromolecules 2011, 44, 6785. not considered helpful to evaluate scien- Raffaele Mezzenga, ETHZ, Antonella Rossi, [20] B. Zhang, A. Kröger, C. Rosenauer, A. D. Schlüter, G. Wegner, Coll. Polym. Sci. 2013, tific performance by bibliometric indices ETHZ, Manfred Schmidt, U Mainz, Andreas Stemmer, IBM/ETHZ, Motomu Tanaka, U accepted. and other questionable measures such as [21] D. Kurzbach, D. R. Kattnig, B. Zhang, A. D. Heidelberg, Joost VanderVondele, ETHZ, the h-index because this may discourage Schlüter, D. Hinderberger, Chem. Sci. 2012, 3, Dimitris Vlassopoulos, FORTH Heraklion, 2550. people to go for risky, long-term goals. Renato Zenobi, ETHZ, and all their involved [22] A. Meller, Adv. Chem. Phys. 2012, 149, 251; A. Both projects have now reached the state coworkers without whom we would not be Oukhaled, L. Bacri, M. Pastoriza-Gallego, J.- in which they have gained their own mo- where we are! I would like to cordially thank M. Betton, J. Pelta, ACS Chem. Biol. 2012, 7, 1935. mentum. Experiencing this transition from all these supporters for their continued inter- [23] I. Popa, B. Zhang, P. Maroni, A. D. Schlüter, the period in which one always has to push est, support and encouragement. The author M. Borkovec, Angew. Chem., Int. Ed. 2010, 49, and push in order to keep a project going to thanks Profs. Ivan Gitsov, SUNY-ESF, and 4250. the period in which the same runs all by it- Paul Smith, ETHZ, for helpful comments to [24] M. Kröger, A. D. Schlüter, A. Halperin, Macromolecules 2013, submitted. self is a wonderful and rewarding moment the article. Financial support came from the Swiss National Science Foundation (SNF) and [25] M. R. Falvo, R. M. Taylor II, A. Helser, V. Chi, making one forget about all the trouble and F. P. Brooks Jr., S. Washburn, R. Superfine, through ETH grants which is most gratefully frustration one had to go through. Nature 1999, 397, 236. acknowledged. [26] J. C. Daniel, R. Audebert, ‘Small Volumes and In regard to DPs, synthesis has reached Large Surfaces: The World of Colloids’, in: the dimensions of smaller biological func- Received: July 4, 2013 ‘Soft Matter Physics’, Eds. M. Daoud, C. E. tional units and it may well be in future Williams, Springer, Berlin/Heidelberg, 1999. [1] H. Staudinger, Ber. Dtsch. Chem. Ges. 1920, [27] J. W. Colson, W. R. Dichtel, Nat. Chem. 2013, that such units can be mimicked not on- 53, 1073; H. Staudinger, J. Fritschi, Helv. Chim. advanced online. ly in their size but also in their function. Acta 1922, 5, 785. [28] There are a few claims of having achieved Obviously, macromolecules that should [2] H. F. Mark, Naturwiss. 1980, 67, 477; H. exfoliations down to the single layer sheet serve such purposes need to be far more in- Ringsdorf, Angew. Chem., Int. Ed. 2004, 43, without, however, providing a mass balance: 1064. J.-C. Tan, P. J. Saines, E. G. Bithell, A. K. telligently designed than the present, rather [3] L. Resconi, L. Cavallo, A. Fait, F. Piemontesi, Cheetham, ACS Nano 2012, 6, 615; A. Gallego, primitive cases. Additionally, it will be an Chem. Rev. 2000, 100, 1253; G. Wilke, Angew. C. Hermosa, O. Castillo, I. Berlanga, C. J. important future goal to use the divergent Chem., Int. Ed. 2003, 42, 5000; L. L. Böhm, Gómez-García, E. Mateo-Martí, J. I. Martínez, Angew. Chem., Int. Ed. 2003, 42, 5010; P. D. F. Flores, C. Gómez-Navarro, J. Gómez- growth method to generate molecular col- Hustad, Science 2009, 325, 704. Herrero, S. Delgado, F. Zamora, Adv. Mat. loidal particles. This will open up access to [4] http://plastics.americanchemistry.com/ 2013, 25, 2141. as yet unexplored size ranges which so far Automotive [29] T. Bauer, Z. Zheng, A. Renn, R. Enning, A. [5] For example, see: K. Matyjaszewski, J. Xia, Stemmer, J. Sakamoto, A. D. Schlüter, Angew. were inaccessible for single molecules and Chem. Rev. 2001, 101, 2921. Chem., Int. Ed. 2011, 50, 7879. also make available soft matter nano-sized [6] D. Neher, Macromol. Rapid Commun. 2001, 22, [30] P. Kissel, R. Erni, W. B. Schweizer, M. D. particles with an unheard level of internal 1365; U. Scherf, E. J. W. List, Adv. Mater. 2002, Rossell, B. T. King, T. Bauer, S. Götzinger, A. structure control. 14, 477; O. Inganas, F. Zhang, M. R. Andersson, D. Schlüter, J. Sakamoto, Nat. Chem. 2012, Acc. Chem. Res. 2009, 42, 1731. 4, 287; R. Bhola, P. Payamyar, D. J. Murray, Regarding the 2DP project a critical [7] J. Sakamoto, M. Rehahn, G. Wegner, A. D. B. Kumar, A. J. Teator, M. U. Schmidt, S. contemporary will point to its present nar- Schlüter, Macromol. Rapid Commun. 2009, 30, Hammer, A. Saha, J. Sakamoto, A. D. Schlüter, rowness. Having just one proven case is 653. B. T. King, J. Am. Chem. Soc. 2013, 135, 14134. [8] S. Kwolek, US patent RE30,352, July 29, 1980. [31] G. M. J. Schmidt, ‘Solid State Photochemistry’, in fact a severe weakness which has to be [9] P. Smith, P. J. Lemstra, J. Mater. Sci. 1980, 15, Verlag Chemie, Weinheim, Germany, 1976. worked on with all energy. It is therefore 505. [32] J. N. Coleman, Acc. Chem. Res. 2013, 46, 14.