Conformation and physics of polymer chains: a single-molecule perspective

Abstract

Single-molecule optical detection and spectroscopy has emerged in recent years as an excellent tool for structural and dynamic characterization of materials and biological systems on nanometer scales. Measuring light emitted by individual molecules reveals static and dynamic heterogeneities that are otherwise hidden in the ensemble average. The removal of ensemble averaging is especially important in the study of polymers, where the complexity of the system is a result of the large size of individual molecules, of the distribution of their molecular weights, and the unique conformational state of each polymer molecule. This review aims to give an insight into the latest results of the application of single-molecule spectroscopy to the study of individual polymer chains. In particular, we focus on the relationship between the conformation (or structure) of the chain on the one hand and its physical properties on the other, primarily with respect to a very important class of organic semiconductors, the conjugated polymers. We review how the conformation of an individual polymer chain is reflected in its photophysical properties, such as exciton migration and localization, and fluorescence spectra. The review also covers the relationship between the conformation of synthetic non-conjugated polymers and their rheological properties.

Main

Nanotechnology has been one of the driving forces of science and technology over the past 20 years. Scaling down the size of an object to nanometer dimensions can result in dramatic changes to its physical properties. With recent progress in the nanoscience and nanotechnology of polymeric materials, new methods are now required for the characterization of polymer structures and properties. Despite the great success of electron microscopy and scanning probe microscope techniques, there remains a need for a non-invasive method that could provide information not only on the structure but also on the physical phenomena occurring on the nanoscale level. One method that has recently emerged as an excellent tool for nanoscale characterization is single-molecule optical detection and spectroscopy. Single-molecule spectroscopy relies on the detection and measurement of the fluorescence signal and spectra from single isolated molecules or polymer chains (Figure 1). These can either be immobilized in a solid matrix or be diffusing in a solution or melt. In all cases, the concentration of the emitting molecules must be sufficiently low to ensure that the spatial separation between individual molecules is larger than the optical resolution of the setup, typically on the order of micrometers. As the fluorescence signal from a molecule is strongly influenced by the nanoscale environment, measuring many molecules one by one provides the distribution of physical properties of the respective nano-environments. Moreover, monitoring a molecule over a period of time reveals dynamic changes in its environment. These static and dynamic heterogeneities of nanoscale properties are averaged and hidden in the usual ensemble experiments. The strength of fluorescence detection is in its sensitivity and low background level — emissions of just a few hundred photons per second can be readily detected. The technique also offers remarkable dynamic range, from sub-nanosecond timescales to seconds and longer, and is non-invasive, making it possible to detect emissions from molecules buried deep in a sample.

Figure 1

Illustration of the principle of single-molecule fluorescence detection and spectroscopy.

Single-molecule spectroscopy has evolved since the beginning of the 1990s from low-temperature, high-resolution optical spectroscopic methods, such as fluorescence line narrowing and spectral hole burning^1,2. The early low-temperature single-molecule methods were soon successfully extended to room-temperature applications and have been employed with great success in biophysical and biochemical studies. Apart from biological applications, single-molecule spectroscopy has provided exceptional insights into the properties of soft and complex materials, and holds great promise for further development in this field. For recent reviews, see Refs^3–8. There will, no doubt, be further qualitative progress related to recently developed super-resolution microscopic techniques^9–10, which overcome the diffraction limit in optical imaging and bring the resolution closer to the molecular scale.

The aim of this article is to review the latest results from the application of single-molecule spectroscopy to the study of individual polymer chains. In particular, we focus on the relationship between the conformation of the chain and its physical properties. Typically, polymers are highly disordered and complex systems, in which each polymer chain finds itself in a unique conformational state. Exploring the conformation and physics of individual chains therefore provides the most direct correlation between polymer structure and properties. Knowledge of the physical properties of a single chain helps us to understand the properties of polymer nanostructures and leads to an advanced understanding of the macroscopic (bulk) state of the polymers. This review will focus on single polymer chains and will not deal with single small molecules as probes of polymer dynamics, which although an important area of research, has been well covered^6,7. The review is further confined to the use of far-field fluorescence microscopy, excluding non-optical scanning probe methods as well as near-field optical microscopy, and to spectroscopic studies performed at room temperature. Although cryogenic single-molecule spectroscopy provides unprecedented insights into the energy landscapes of conjugated polymers, for practical applications of polymers as functional materials, room-temperature spectral characteristics and excited-state dynamics are most relevant. Low-temperature work has also been reviewed extensively elsewhere⁸.

This review is divided into two parts. The first part deals with a very important class of organic semiconductors, the conjugated polymers. We look at how the conformation of individual polymer chains is reflected in their photophysical properties, such as exciton migration and localization, and fluorescence spectra. In the second part, we concentrate on the conformation or structure of synthetic non-conjugated polymers and the relationship with their rheological properties.

Conformation and (photo)physics of conjugated polymers

Conjugated polymers represent a class of novel organic materials with optical and electrical properties resembling those of inorganic semiconductors yet with the mechanical properties and processability of a plastic. Conjugated polymers absorb light in the visible part of the spectrum and often show bright emission. The study of their photophysics is complicated by the disordered nature of the material, which produces a wide distribution of microscopic optical properties and interactions. Such properties are determined by conjugated segments, sections of the polymer chain over which the π-electrons are delocalized. Conjugated segments are responsible for primary interactions of the polymer chain with light and act as more-or-less independent chromophores. Topological defects resulting from twisting of the chains, as well as chemical defects, shorten the conjugation segment length. Close proximity of segments located on different chains (or on different parts of the same chain) results in photophysical interactions, including energy transfer, ground- or excited-state aggregate formation, and charge transfer. These photophysical processes strongly modify the polymer optical properties. The microscopic structure resulting from the chain conformation is thus the determining factor for both the physical properties and practical applicability of conjugated polymers. The relationship between chain conformation (or film morphology) and the photophysics of conjugated polymers on the bulk level has been the subject of intense research, and has recently been reviewed¹¹. Some earlier single-molecule detection and spectroscopy studies of conjugated polymers have been introduced in Ref.¹².

Optical measurement of conformation and orientation of a single conjugated chain

Direct measurement of the relationship between structure and optical properties in individual chains is one of the most important problems to be dealt with in polymer nanoscience and nanotechnology. The problem is caused by the disparity between optical resolution, which is on the order of hundreds of nanometers, and the typical size of a polymer chain, which even for high-molecular weight polymers is on the order of a few tens of nanometers. It is thus not possible to directly measure the shape of a chain using an optical microscope and simultaneously measure its optical properties. In the case of conjugated polymers, the conformation of a chain can be related to the anisotropy in its absorption. Such anisotropy can be approximated by an absorption ellipsoid, as shown schematically in Figure 2. The ellipsoid is given by the sum of the absorption strengths (transition dipole moments) of individual conjugated segments of the chain, assuming that each segment is absorbing with the same probability and is oriented along the chain. In the simplest case, the ellipsoid is rotationally symmetric along one axis, and its shape is determined by the ratio of the two axes.

Figure 2

Typical conformations of conjugated polymers in (a) poor and (c) good solvents. Arrows in (a) shows the transition dipole moments μ_i of individual conjugated segments. (b,d) Corresponding rotationally symmetric absorption ellipsoid representations.

The first attempt to measure the conformation of single conjugated chains involved measuring the projection of the three-dimensional (3D) absorption ellipsoid into two dimensions of the sample plane by polarization-modulated excitation in wide-field fluorescence microscopy (Figure 3(a))¹³. However, since the individual chains are also oriented randomly in space, the two-dimensional (2D) projection does not represent the conformation uniquely. To obtain the conformation, it is necessary to measure a statistical ensemble of individual chains, plot the distribution of the absorption ellipsoid projection (as, e.g., a modulation depth parameter) and then use Monte Carlo simulations to test which of the known polymer conformations would best fit the results. Using this method, it was found that the prototypical conjugated polymer poly(2-methoxy-5-(2′-ethylhexyl)oxy–1,4-phenylenevinylene) (MEH–PPV) adopts a defect cylinder conformation when embedded in an inert polymer matrix (such as polycarbonate)¹³. In a strict sense, this method does mot measure the conformation of each chain individually but relies instead on data from a statistical ensemble of single chains, providing a typical conformation for this sub-ensemble. Despite this shortcoming, the method has been adopted by many groups, as reviewed later in this article.

Figure 3

Experimental techniques used in single-molecule spectroscopy of polymers. (a,b) Principle of measurement of absolute shape and orientation of three-dimensional absorption ellipsoid by a combination of (a) wide-field epi-illumination microscopy and (b) total internal reflection microscopy. Polarization-modulated fluorescence intensity traces from individual molecules (insets in microscopic images) are fitted to theoretical equations to obtain the ratio r of the ellipsoid axes and the polar angles of its orientation. (c) Principle of determination of the polar angles ϕ and θ of a single emitting dipole by defocused fluorescence imaging. Modified after Ref. 17, reproduced with permission (© 2009 Wiley-VCH).

Recently, a method to measure the absolute shape of a randomly oriented 3D rotational absorption ellipsoid was proposed¹⁴. In addition to the projection onto the sample plane, this method uses polarization-modulated total internal reflection illumination to measure the projection of the ellipsoid in a plane perpendicular to the sample (Figure 3(b)). This method has been used to measure the absorption ellipsoids of MEH–PPV in various polymer matrices¹⁵. The conformation of individual chains must still be reconstructed by molecular dynamics simulations, but the conformation is now available for each polymer chain individually, thus making it possible to study the conformation and properties on a truly single-molecule level. The method is not restricted to rotational ellipsoids; for polymer chains (or nanoparticles in general) with known orientation in space, the method provides the ratio of all three axes of a general ellipsoid¹⁶, information akin to 3D linear dichroism.

In many situations, information on the 3D orientation of the polymer chain is important. In cases where the orientation can be associated with the orientation of the emitting dipole, one can use techniques developed for measuring the 3D orientation of single small molecules, such as defocused imaging in wide-field epi-fluorescence microscopy (for a review, see Ref. 17). If the sample is defocused by approximately 1 μm from the microscope objective lens, an emitting dipole creates a characteristic pattern in the image that is unique for every spatial orientation (Figure 3(c)). Fitting of the pattern provides the polar angles of the dipole orientation. This method has been used successfully to confirm the perpendicular orientation with respect to the substrate of ordered MEH–PPV nanoparticles¹⁸. Alternatively, in samples with limited degrees of freedom, such as rod-like MEH–PPV chains in a nematic liquid crystal, the orientation can be monitored simply by measuring the degree of fluorescence anisotropy¹⁹.

For the general problem of 3D orientation of single chains, it is necessary to look at the orientation of the main axis of the absorption ellipsoid. In principle, any method that can measure the absorption transition dipole of a single molecule should be applicable for this purpose. This has been demonstrated using a combination of polarization-modulated epi-fluorescence and total internal reflection illumination with a wide-field microscope²⁰. It has thus been found that irrespective of conformation, MEH–PPV chains tend to orient parallel with the sample substrate during the spin-coating sample preparation process¹⁵.

Energy transfer and exciton localization

The processes of light absorption and excited state relaxation are crucial in any application of conjugated polymers and as such have been studied extensively (see, e.g., Ref. 21). The absorption of light occurs on individual conjugated segments and leads to the creation of a Frenkel-type singlet exciton, that is, an electronic excited state coherently shared by several molecules (monomers). The exciton is initially localized on the segment and can decay radiatively or nonradiatively via triplet states and other routes. The excitation energy can also be transferred by dipole–dipole interactions (incoherent hopping) to other conjugated segments, either along the polymer chain (intrachain transfer) or between segments located on adjacent chains (interchain transfer). Interchain energy transfer is more efficient than intrachain transfer²² due to the shorter inter-segment distances. Intrachain transfer is assumed to occur only on short length scales corresponding to a few conjugated segments²². On the other hand, intrachain energy transfer substantially exceeding the several-segment range has been reported for conjugated polymers used as chemical sensors, with exciton diffusion lengths of more than 90 nm²³. In the extreme case of rigid straight polymers, the wavefunction coherence length can reach micrometers in range²⁴. The energy transfer is also not necessarily simple dipole–dipole hopping; coherent energy migration, which can be interpreted as transfer of the conjugated segment, has been reported recently²⁵. Chemical structure, rigidity, conformation and the molecular environment appear to play decisive roles in the fate of these excitations. The ability of single-molecule spectroscopy to look at individual chains and their conformations provides much-needed insight into the problem.

The first detection of fluorescence from a single conjugated polymer chain²⁶, of a copolymer of poly(p-phenylene vinylene) (PPV) and poly(p-pyridylene vinylene) (PPyV), led to a surprising observation: the emission from the whole chain displayed single-step intermittency (blinking) and single-step photobleaching. These are features that have been observed previously for single small dye molecules but which were not expected for a conjugated polymer chain. This finding suggests that the whole chain can behave as a single chromophore—only one conjugated segment emits light and its temporary or permanent quenching causes quenching of the entire chain. Localization of the exciton on a single segment is caused by efficient energy transfer within the polymer chain, but this effect has not been observed in bulk solution experiments. The PPV–PPyV copolymer has relatively low molecular weight (ca. 20,000), but the same phenomena have since been observed for high molecular weight (ca. 900,000) MEH–PPV as well²⁷.

Since these initial observations, considerable effort has been devoted by various groups to clarifying the fluorescence blinking effect and its relationship with the conformation of the polymer chain. The conformation in dilute solution depends on the quality of the solvent. In good solvents, interchain interaction is weaker than the chain–solvent molecule interaction, and the chain takes a random coil conformation. In poor solvents, the attractive interaction between parts of the chain prevails and the polymer assumes a molten globule conformation in the case of flexible polymers, or a rod (or toroid) conformation in the case of rigid polymers. The conformation in solution is retained even after depositing individual polymer chains onto a solid substrate. This has been elegantly demonstrated for single chains of MEH–PPV cast from toluene (a poor solvent) and chloroform (a good solvent)²⁸, as shown in Figure 4(a). In MEH–PPV/toluene samples, all chains show step-wise intensity changes, whereas MEH–PPV/chloroform chains decay continuously as the conjugated segments bleach one by one and there is little interaction between the segments in an extended chain. When the conjugated polymer chain is doped into a solid matrix of an inert polymer, it can develop an extended or collapsed conformation depending on the solvent and the matrix polymer. The collapsed conformation is best described by a defect cylinder¹³. For example, low-molecular-weight polystyrene (PS) is a matrix that supports the extended conformation of poly(2-methoxy-5-(2′,6′-dimethyloctyloxy)–p-phenylenevinylene (OC₁C₁₀–PPV)²⁹. In this matrix, the fluorescence intensity changes are continuous, in contrast to the case of a more polar poly(vinyl butyral) matrix (Figure 4(b)). However, matrix polarity is not the only parameter; fluorescence blinking of single OC₁C₁₀–PPV chains reappears in high-molecular-weight PS. Other work on MEH–PPV³⁰ and fluorene-based copolymers³¹ indicates that high-molecular-weight PS is a poorer-solvent matrix than polynorbornene (Zeonex), even though the polarities of the two polymers are similar. More work will be needed to clarify the problem of conjugated polymer conformation and related blinking in inert polymer matrixes. In the enclosed spaces of hollow silica nanospheres, single polymer chains of poly-2,6- naphthylidenevinylene interestingly show both continuous bleaching and discrete blinking, pointing to the coexistence of different conformations³².

Figure 4

Comparison of fluorescence time traces of conjugated polymers in different conformational states. (a) MEH–PPV cast from poor solvent of toluene (top) and good solvent of chloroform (bottom). Modified after Ref. 28, reproduced with permission (© 2000 NAS). (b) OC₁C₁₀–PPV in thin-film matrix of poly(vinyl butyral) (PVB) (top) and of low-molecular-weight polystyrene (bottom). Modified after Ref. 29, reproduced with permission (© 2003 ACS). (c) Polythiophene without (top) and with (bottom) grafted side-chains of polystyrene.

In all the above studies, the link between the conformation and the degree of exciton localization is indirect. It was assumed that the observation of fluorescence blinking must indicate that the chain is collapsed. Recently, a method for 3D determination of absorption ellipsoids was used to examine the relationship between the directly measured conformation and fluorescence blinking¹⁵. The most important finding was that MEH–PPV in low-molecular-weight PS can adopt two different conformations: an extended defect coil and a compact defect coil. At the same time, two types of fluorescence decay were observed, discrete blinking and continuous bleaching, and a strong correlation was observed between the conformations and the fluorescence behavior, as summarized in Figure 5. The extended defect coils corresponded to those chains that showed continuous bleaching, thus providing for the first time a direct link between the polymer conformation and photophysics. Molecular dynamics simulations predicted that in the compact defect coils, there are domains with higher local density that serve as local exciton traps. The existence of such domains was later confirmed by direct super-resolution localization imaging of emission sites from individual MEH–PPV chains³³. Super-resolution imaging makes it possible to localize a single emitting species in a microscopic image with accuracy of a few nanometers (Figure 6). Different MEH–PPV chains can accommodate one to several distinct emitting sites. Moreover, it was found that these sites are highly dynamic—individual sites on a single chain can be switched on and off, resulting in jumps of the emitting sites along the chain. The jump sizes showed a large distribution from a few nanometers to more than 15 nm. Another new technique that appeared recently, 2D polarization imaging³⁴, has also been used to study differences in energy transfer in MEH–PPV in poly(methyl methacrylate) (PMMA) cast from different solvents.

Figure 5

Photophysics and conformation of MEH–PPV in a thin-film matrix of low-molecular-weight PS. (a,b) Fluorescence time traces of two different molecules. (c,d) Histograms of the shape of the absorption ellipsoids corresponding to blinking (c) and continuously bleaching (d) single MEH-PPV chains. (e,f) Chain conformations reconstructed from the two types of ellipsoids using molecular dynamics simulations.

Figure 6

Mapping of the emission sites within a single polymer chain. (a) Principle of super-resolution localization microscopy. The emission intensity profile from one molecule is fitted using a 2D Gaussian function (left) to determine the centroid position (right). (b) Fluorescence intensity trajectory of a single MEH–PPV molecule embedded in a thin film of Zeonex. Each intensity level is represented by a different color. (c) Left: Centroid positions obtained from the trajectory in (b). Large circles show positions of emitting sites. The radii of circles correspond to standard deviations of the centroid positions. Each color represents the position obtained for the intensity level of the same color as in (b). Right: Schematic conformation of the molecule. Emitting sites within the molecule are indicated by light-blue dots.

Even for the same conjugated polymer chain in the same environment, the energy transfer and exciton localization depend on the molecular weight of the polymer. This effect has been studied extensively by several groups. Long chains of dioctyloxy–PPV (DOO-PPV), a poorly soluble PPV compound known for its sensitivity to the environment, show discrete fluorescence blinking superimposed on a gradually changing fluorescence background³⁵. Absorption polarization measurements indicate that the blinking originates from regions on the chain that are more isotropic than the regions showing the gradual weakening of fluorescence, pointing to conformational disorder on a single chain. Large MEH–PPV chains dissolved in different solvents have been studied using photon arrival correlation measurements³⁶. In these experiments, short laser pulses are used to excite fluorescence. The short pulse-length ensures that only one photon is absorbed in one chromophore upon each excitation. The inter-photon arrival time distribution measures the probability that more than one photon emitted by the polymer chain will arrive at the detector at the same time. Single MEH–PPV chains showing discrete blinking display a decrease in arrival probability. This method has since been used on an MEH–PPV series with increasing molecular weight (from 55,000 to 2,600,000)³⁷. The probability that only one photon is emitted upon each excitation was found to increase with decreasing molecular weight and also to be greater for polymers adopting a collapsed conformation. Even some single chains of the highest molecular weight polymer showed emission from a single conjugated segment despite an average total number of chromophores in such chains of 670.

Apart from the influence of solvent or solid environment, the conformation of a chain can also be controlled by its chemical composition. To prevent the effects of coiling or collapsing in semi-flexible polymers such as MEH–PPV, attempts have been made to use rigid-main-chain polymers. One such example is methyl-substituted ladder-type poly(paraphenylene), single chains of which in a PS matrix can exhibit both emission from a single conjugated segment due to efficient directional energy transfer, or more typically, emission from multiple segments due to bidirectional energy transfer along the rigid chain³⁸.

The collapsed conformation can also be prevented by grafting bulky non-conjugated side chains on the conjugated main chain. It has been shown that the presence of long PS side chains on a polythiophene chain completely suppresses the fluorescence blinking that is otherwise observed in the unsubstituted single polythiophene chains (Figure 4(c))³⁹. Molecular dynamics simulations have confirmed that the compact conformation of the unsubstituted polythiophene causes any two conjugated segments on the chain to find themselves within the Förster radius for efficient energy transfer, resulting in exciton localization and fluorescence blinking⁴⁰. The PS side-chains effectively force the polythiophene main-chain into an extended coil conformation, thus restricting the energy transfer to a few neighboring segments along the chain. A similar effect of enhancing the effective rigidity of the main chain can be achieved by decorating the chain with macrocycles, as demonstrated for the conjugated polymer of poly(p-phenylene-ethynylene-butadiynylene)⁴¹.

Detailed insight into the energy transfer process can be gained from experiments on custom-designed systems, such as conjugated polymers end-capped with organic dyes as energy acceptors. A study of perylene-endcapped poly(indenofluorene) revealed that while on average the energy transfer along the backbone is inefficient, individual molecules may exhibit extremely efficient energy transfer along the conjugated polymer chain and to the perylene dyes even without spectral overlap between the energy donor and acceptor⁴². This efficient energy transfer in some of the molecules is likely to be due to a specific conformation of the main chain resulting from the presence of kinks on the otherwise rigid-rod chain.

Very recently, it has been reported⁴³ that in long MEH–PPV chains, part of the chain may stay in a permanently quenched state. For some chains, this quenched state was suggested to account for more than 90% of the chain length, and the quenched parts would remain invisible to single-molecule spectroscopy in fluorescence mode. Other experiments (unpublished data from our group) do not support this observation, indicating that the phenomenon might be dependent instead on sample preparation or other factors. More work will be needed to clarify this point.

Fluorescence spectra and lifetimes

There has been continuing interest in the fluorescence spectral properties and lifetimes of conjugated polymers, especially those of the PPV family. The relationship between chain conformation and the fluorescence spectrum has been studied extensively both in bulk solutions and in spin-cast films. It has been found, for example, that the hydrodynamic radius of MEH–PPV in a poor solvent (tetrahydrofuran), is approximately half that in a good solvent (chlorobenzene), indicating an extended conformation in the latter, and at the same time that the fluorescence spectrum in tetrahydrofuran is shifted toward blue⁴⁴. In contrast, a red shift of the fluorescence spectra of MEH–PPV in toluene solutions has been observed upon addition of increasing amounts of hexane (poor solvent). The poor solvent also causes a decrease in fluorescence quantum efficiency, yet no significant changes in the fluorescence lifetime⁴⁵.

The first measurements of fluorescence spectra on single chains of MEH–PPV indicated an important deviation from the bulk studies: different individual MEH–PPV chains in an inert polymer matrix have different spectra that can be roughly classified into two groups: red and blue⁴⁶. The histogram of peak wavelengths shows two distributions with maxima at around 555 and 580 nm, as seen in Figure 7. Two-dimensional polarization measurements have not revealed differences in the chain conformations that are sufficient to produce the two spectral types. The origin of the red type was initially suggested to be aggregate emission. It was also found that the relative amount of red-type emitters increased with the molecular weight of MEH–PPV⁴⁷. A direct comparison of fluorescence spectra for MEH–PPV cast from different solvents²⁸ has since shown that single chains in MEH–PPV/toluene samples (poor solvent, collapsed conformation) typically produce fluorescence peaks at around 570 nm (red type), whereas MEH–PPV/chloroform chains (good solvent, extended conformation) are of the blue type with peaks around 550 nm (Figure 7(a,b)). Moreover, the red-type spectra undergo significant changes during photobleaching, shifting up to 50 nm to shorter wavelengths. The red-type spectra have been attributed to an extension of the conjugation length in well-ordered regions of the closely packed conformations through localized suppression of torsional motion of the chain, which would otherwise interrupt the conjugation. The longer conjugated segments are characterized by efficient energy transfer and undergo photobleaching at first, causing a shift of emission to the next-longest segments and a gradual blue-shift of the single-chain spectra. Interestingly, MEH–PPV in PMMA has been found to produce only one type of spectrum⁴⁸, with a distribution maximum at approximately 565 nm. Similarly, MEH–PPV in a Zeonex matrix produces only single-peak spectral distributions, whereas the same conjugated polymer exhibits the usual two spectral types in a matrix of low-molecular-weight PS⁴⁹. The two types have been assigned to two slightly different defect coil conformations in the PS. The red-type chains have been attributed to a more extended conformation, with the red shift due to the existence of the longer conjugated segments that are supported in the extended chain.

Figure 7

Spectral characterization of single chains of the PPV family of conjugated polymers. (a,b) Fluorescence spectra and evolution in time for MEH–PPV cast from (a) toluene (poor solvent) and (b) chloroform (good solvent). Modified after Ref. 28, reproduced with permission (© 2000 NAS). (c) Relative spectral shift S_r during photobleaching of long-chain MEH–PPV and short-chain DOO–PPV. Negative values of S_r correspond to blue shift. Modified after Ref. 36, reproduced with permission (© 2006 APS). (d) Bimodal histogram of fluorescence peak wavelengths of single chains of MEH–PPV in a thin-film polycarbonate matrix. Modified after Ref. 46, reproduced with permission (© 2000 AAAS). (e) Multimodal histogram of fluorescence peak wavelengths of single oriented nanoparticles of MEH–PPV. Modified after Ref. 50, reproduced with permission (© 2003 ACS).

The decisive role of structural ordering of the chain in determining the spectral properties of MEH–PPV has been demonstrated in an example of single-MEH–PPV nanoparticles formed by ink-jet printing from dilute solution⁵⁰. The particles were oriented on a glass substrate such that the emission dipoles were perpendicular to the substrate surface. The histogram of the florescence peaks has four maxima, at 599, 608, 613 and 619 nm — all at wavelengths higher than that for single MEH–PPV chains dispersed in a matrix (Figure 7(e)). The four maxima were assigned to the emissions of highly ordered conjugated segments with increasing length, corresponding to 8, 9, 10 and 11 mers.

The interpretation of the red-type spectra as being due to longer conjugated segments in planar sections of the chain is supported by recent theoretical work that also ruled out aggregation as a cause of the red shift⁵¹. Another important recent development is the finding that even short isolated PPV oligomers can show large differences in fluorescence spectral shapes and peak positions — bent oligomers of the same length have spectra that are red-shifted from those of the straight versions by as much as 20 nm (at low temperatures)⁵².

Apart from static spectral properties, analysis of the spectral shift during prolonged irradiation of a single chain can provide further information on the energy landscape of conjugated polymers. Analysis of the relative spectral shift of the fluorescence of short-chain rigid-rod DOO–PPV and long-chain flexible MEH–PPV (Figure 7(c)) has shown that in MEH–PPV, there is a large distribution of conjugated segment lengths, which causes a continuous blue-shift during the irradiation of most chains, whereas very little dispersion in the conjugation length is observed for DOO–PPV³⁶. On the other hand, solvent polarity (toluene vs. chloroform) has surprisingly little effect on the relative spectral shift in MEH–PPV. The blue spectral shift during continuous irradiation is a relatively common phenomenon that is also observed for other conjugated polymers, such as OC₁C₁₀–PPV²⁹ and polythiophene³⁹.

Further insight into the energy structure and dynamics of the PPV family of polymer can be gained from studying block copolymers of energy donors and acceptors. The fluorescence spectra of single chains of DMOS–co-MEH–PPV has shown that the presence of the 2-dimethyloctylsilyl (DMOS) block causes ordering of the MEH blocks, leading to a significantly larger fraction of red-light-emitting MEH sites in the block copolymer compared to the MEH–PPV homopolymer⁵³.

An interesting perspective can be obtained from measuring the emission spectra as a function of concentration. A systematic study of changes of the MEH–PPV spectra upon dilution from bulk film to isolated chains in a matrix demonstrated a blue shift and spectral broadening of the spectra upon isolation of the chains⁵⁴.

Bimodal distributions of fluorescence are not restricted to the PPV family. Similar distributions have been found for polythiophenes⁵⁵ as well as for the polyfluorene copolymer poly(9,9-di-n-octylfluorene-alt-benzothiadiazole)⁵⁶, which have been explained similarly as being due to the existence of two types of chain conformation, where the red-shifted emission originates from regions of higher order and/or closer chain packing. Further detailed study of polyfluorene oligomers with increasing length has shown that the three known spectrally distinct phases, i.e. the glassy (α), γ and β phases, are represented differently in oligomers of different length and that a minimum of nine monomer repeat units is necessary for the β phase to start forming⁵⁷.

Compared to single-chain fluorescence spectra, there have been relatively few studies of conjugated polymer fluorescence lifetimes. It was found early on that the fluorescence lifetime of MEH–PPV changes as a result of blinking: high fluorescence intensity states ('on' states) are characterized by single-exponential decay (with the decay time of ca. 310 ps), whereas in low-intensity states (dim or 'off' states), the lifetime shortens and the decay becomes multi-exponential with a sub-50 ps component and a 230 ps component⁴⁶. Shortening of the lifetime is a result of transient quenching (blinking) of the fluorescence by a nearby quencher, most likely a hole polaron. Later detailed studies at higher temporal resolution confirmed the appearance of a short-lifetime component in the dim states. Together with polarization-modulation experiments, these results have been used to propose a model of a single MEH–PPV chain in which the chain consists of several energy-transfer domains and the energy migration is efficient within a domain and inefficient between domains⁵⁸. The entire domain can be quenched if it is within the effective radius of a quencher. Lifetime measurements on short-chain MEH–PPV supporting only 1–3 conjugated segments have revealed, on the other hand, that the lifetime of the chain increase after permanent bleaching of one segment, accompanied by a blue spectral shift⁵⁹. This result can be rationalized by assuming that the permanent quenching (most likely photooxidation) of a segment leads to a shift of its absorption spectrum and, as a result, the segment no longer functions as an energy acceptor. This in turn leads to an increase in the lifetime of the energy donor (unbleached segment).

Structure/conformation and physics of non-conjugated polymers

Here we review reports results on the conformation and related physics obtained mainly for non-conjugated synthetic polymers. Studies on single chains of natural polymers, such as DNA, have made great contributions to the field of polymer physics, and have been reviewed elsewhere (e.g. Refs 60,61). In particular, the conformation and relaxation of large polymer chains^62,63, as well as interaction with other chains⁶⁴, have been investigated in great detail by single-molecule techniques. Further, as cyclic and star-branched DNA are relatively easily obtainable, topology effects on polymer dynamics have also been investigated^65,66. The study of large DNA is facilitated by its size—a fully labeled DNA helix is easily imaged by fluorescence microscopy and changes in position and shape can be easily and precisely followed. Synthetic polymers, on the other hand, typically have dimensions smaller that the resolution of optical microscopy, rendering single-chain studies more challenging. In addition, it is often more difficult to prepare well-defined and -purified labeled synthetic chains of uniform size (molecular weight distribution). As a result, compared to natural and biopolymers, the field of single-molecule polymer physics of synthetic molecules is still in its infancy.

Detection and monitoring of non-conjugated polymer chains

Unlike conjugated polymers, non-conjugated polymers do not absorb and emit light in the visible part of the spectrum. For fluorescence detection, the chains must be labeled by covalently bonding a fluorescent dye. An often-used dye is perylene diimide (PDI; Figure 8(a)) which is preferred for its brightness and photostability. The dye can be attached to the chain end or incorporated at an arbitrary position inside the chain^67,68. In immobilized chains, the orientation of the dye (measured, for example, by defocused imaging¹⁷) can serve as a measure of the chain conformation and any change in orientation as a measure of local dynamics. In solutions or melts, the diffusion of labeled chains can be followed by direct fluorescence imaging in the case of slow diffusion or by single-molecule fluorescence correlation spectroscopy in the case of fast diffusion⁶⁹. For other applications, the chain can be labeled with multiple dyes, such as by capping both ends^70,71 or labeling with three dyes along the chain¹⁷. Nanometer-scale localization of the dyes at both ends by super-resolution microscopy then gives the end-to-end distance directly for a single polymer chain⁷⁰, information that is unavailable by any other non-invasive analytical method. If the two dyes are spectrally distinct, the end-to-end distance of the chain can be determined by the fluorescence resonance energy transfer between them⁷². Alternatively, the dyes can be selectively probed using different excitation wavelengths, making it possible to measure their orientation independently.

Figure 8

Diffusion of labeled linear (top) and cyclic (bottom) polymer chains in semi-dilute solution of unlabeled linear chains of the same polymer. (a) Linear and cyclic poly(tetrahydrofuran) (PTHF) molecules containing a perylene diimide (PDI) moiety. (b) Configuration of PDI–linear PTHF and PDI–cyclic PTHF surrounded by linear PTHF. (c) Two-dimensional trajectories of self-diffusing molecules of PDI–linear PTHF and PDI–cyclic PTHF. (d) Histograms of diffusion coefficients for PDI-linear PTHF and PDI-cyclic PTHF. Lines show calculated theoretical statistical distributions corresponding to the diffusion of single molecules in a homogeneous environment using the diffusion coefficients from the respective histograms.

Conformation and dynamics of immobilized single chains

Studying the conformation and dynamics of individual labeled chains in the solid state has great potential for revealing local polymer segmental dynamics. Unlike individual dyes doped into a polymer, which probe mainly the polymer free volume and its changes, the fluorescence from dyes covalently attached to the polymer chains more accurately reflects the dynamics of the polymer itself.

There have been a few attempts to examine differences in the dynamics at the end and in the center of a single chain. Perylene diimide has been attached in different positions to polybutadiene or polyisoprene and studied in unlabeled matrices of polybutadiene, polyisoprene and triblock poly(styrene-b-isoprene-b-styrene) rubber. However, the results so far have been somewhat disappointing: PDI in the center or at the ends of a chain show no differences in orientation or dynamic intensity behavior, at least on the time scale of the experiments (0.2–2.0 s)⁷³. Similarly, PDI in the center or at the end of PMMA or poly(butyl methacrylate) has been found to show only weak orientational dynamics⁷⁴. However, it was possible in the same study to obtain the distribution of the orientations of PDI in the center of a PMMA chain, which reflects the orientational distribution of the center PMMA segment. It was found that the center segments are distributed randomly, that is, the conformations of individual chains correspond to a bulk conformation and are not perturbed by the dimensions of the thin film. A 100 nm-thick film of PMMA with a molecular weight of 10⁵ can therefore still be regarded as 'bulk'.

Capping both ends of a polymer chain with fluorescent dyes can provide information on other parameters related to chain conformation. This strategy has been used to measure the end-to-end distances of polyfluorene⁷¹. After capping the polyfluorene with perylene imide and embedding the polymer in a Zeonex matrix, individual perylene imide molecules could be localized in space with nanometer accuracy during photobleaching. The distance between the dyes could then be measured for a statistical ensemble of polyfluorene chains with an average chain length of 31 nm to obtain the distribution of end-to-end distances of polyfluorene, in this case yielding a mean value of 22.7 nm. Fitting the distribution to various theoretical models of chain conformation led to the conclusion that polyfluorene takes essentially a linear conformation with a few kinks caused by rotation between monomer blocks.

A similar approach using two spectrally different dye molecules, PDI and terrylene diimide, has also been used to allow independent measurements of the spatial orientation of a p-phenylene oligomer⁷⁵. The relative orientation of the two end-cap dyes also made it possible to estimate the average bending angle, in this case 22°, of the short oligomer — a value that is independent of the kind of inert polymer matrix (PMMA or Zeonex). Extrapolation of this value to p-phenylene polymers affords a persistent length of 19 nm, in a good agreement with measurements by other methods.

Conformation/structure and diffusion in solution

The conformation and shape of a polymer chain are directly related to its dynamics or rheological properties. This fact can be used in studies of single polymer chains in solutions in two ways: polymers of known shape can be used to examine diffusion properties, or diffusion characteristics can be measured to obtain information on the chain conformation.

The first approach has been demonstrated using single poly(tetrahydrofuran) (PTHF) chains of different topologies, linear and cyclic, labeled with PDI dyes. The diffusion of the polymer chains was studied in a semi-dilute solution of unlabeled linear PTHF (Figure 8)⁷⁶. It was found that whereas the linear PTHF–PDI chains underwent simple homogeneous diffusion characterized by a single diffusion coefficient, the cyclic PTHF–PDI chains exhibited a more complex, multimode diffusion process described by two diffusion coefficients. The observed diffusion heterogeneity was interpreted as being due to partial threading of some of the cyclic chains with the linear matrix, which results in slower diffusion on the one hand, and to a more compact conformation of the unthreaded cyclic chains and faster diffusion than the linear counterparts on the other.

The second approach has been used to study the conformational changes of single polyelectrolyte chains as a function of the pH of the solution. Poly(2-vinylpyridine), a weak electrolyte, was labeled with the dye Alexa 488 and its diffusion was followed in aqueous solution by fluorescence correlation spectroscopy⁷⁷. An abrupt change of the diffusion coefficient and an associated sudden decrease in hydrodynamic radius from 22.6 to 6.8 nm with increasing pH was observed, interpreted as a first-order transition of the molecular conformation from an extended coil at low pH to a collapsed globule at high pH. Labeling the poly(2-vinylpyridine) chains with a pH-sensitive dye, such as Oregon Green 488, also made it possible to estimate the local pH at the position of the chain. Surprisingly, it was found that the local pH near the dye, as inferred from its brightness, was consistently three orders of magnitude higher than the bulk pH of the solution⁷⁸.

Summary and perspectives

The characterization of complex systems such as polymers by macroscopic (ensemble) methods can be likened to looking at a complex picture through an opaque screen. Using single-molecule spectroscopy, we can open holes in the screen and see the details behind. However, the holes we have so far managed to open are too few, too far apart and too random. We still do not know very well what happens to a conjugated polymer chain in a solid environment (or in any environment other than a good solvent). We also know little about the conjugated segment itself — how it is defined or how much it is fixed in space. And we keep getting contradicting data on the spectral properties of conjugated polymers. Applications of single-molecule detection to study non-conjugated chains have just begun. This review has shown a few examples of the great potential we believe that single-molecule spectroscopy holds for the study of polymer physics, and we hope it serves to motivate other researchers to use single-molecule spectroscopy in many so-far-unforeseen applications.