NMR Investigation of Aniline Oligomers Produced in the Early Stages

The products obtained within early stages of the oxidative polymerization of aniline in solutions of various weak organic acids or in water, and anili...
6 downloads 0 Views 564KB Size
Download PDF
6666

J. Phys. Chem. B 2009, 113, 6666–6673

NMR Investigation of Aniline Oligomers Produced in the Early Stages of Oxidative Polymerization of Aniline Jaroslav Krˇizˇ,* Larisa Starovoytova, Miroslava Trchova´, Elena N. Konyushenko, and Jaroslav Stejskal Institute of Macromolecular Chemistry, Academy of Sciences of the Czech Republic, HeyroVsky Sq. 2, 162 06 Prague 6, Czech Republic ReceiVed: January 27, 2009; ReVised Manuscript ReceiVed: March 19, 2009

The products obtained within early stages of the oxidative polymerization of aniline in solutions of various weak organic acids or in water, and aniline oligomers produced by the oxidation of aniline and aniline-15N in acetic acid (0.4 M) with a limited amount of oxidant were analyzed using 1H, 13C, and 15N 1D and 2D NMR spectroscopy and 1H PFG NMR. Such products are virtually identical in all cases, according to 1H NMR. They are always a mixture of products, among which one of them is prominent. Both native and neutralized forms of the products were examined. As shown by a combination of 1H DQF COSY, 1H NOESY, 1 H-13C and 1H-15N HSQC, and 1H-13C and 1H-15N HMBC spectra, both forms of this product contain an oligoaniline moiety ended mostly by phenylamino groups. In a significant amount, the chains containseither as an inner or terminal groupsan unexpected six-member ring with an oxygen-containing substituted quinoneimine structure. The most probable structure of the major product is given. The difference between the native and neutralized forms of the product was examined. It is shown that the oligomeric chains, in particular quinoneimine units of the former one, are protonated. Both forms of the product exhibit a slight paramagnetism, and contain about 2 × 10-9 mol g-1 of unpaired electron spins. 1. Introduction Conducting polymers, such as polyaniline (PANI), polypyrrole, and polythiophene, are attracting interest due to their “intelligent” behavior, the ability to respond to various external stimuli by the change in electrical, optical, or chemical properties. Among them, PANI is probably the most studied because of the variety of morphologies that are displayed,1 onedimensional structures such as nanowires2-4 and nanotubes5-11 being examples from the realm of nanostructures. Aniline oligomers are generally believed to be responsible for the selfassembly that guides the growth of polymeric structures.5,12 The chemical nature of such oligomers is still open to discussion. Although the polymerization of aniline has been studied for more than two decades13-15 and the most important features of its mechanism have been addressed,16 many details of the process have not been fully understood. The oligomers are produced at the early stages of aniline oxidation. It has been proposed that, based on the historical reminiscence of the first industrially produced synthetic dye, mauveine,17,18 such oligomers contain phenazine constitutional units.5 The support of this hypothesis comes mainly from FTIR19 and recently from Raman spectroscopies.20 The presence of such structures has also been supported by the quantum-chemical calculations.21,22 On the other hand, alternative structures have been suggested, such as azanes.23 Other methods are therefore needed to supply more detailed information. Also, the remarkable ability of PANI to form nanotubes or nanofibers if prepared in certain media,24-26 though widely studied and speculated on,27-32 seems to be dependent on the presence of such oligomers.1,12 One of the possible approaches, which does not solve these questions as such but could serve as a base for further understanding, is the structure analysis of oxidation products. Among other methods, NMR was widely used for this task.33-39

Unfortunately, most polymerization products were either not soluble so that rather unsophisticated methods of solid-state NMR were used or were not proper PANI but grossly similar products thought to be more or less probable models of the polymerization products. In this work, we decided to put under close scrutiny the products of early polymerization stages. The oxidation of aniline has indeed been followed40 by NMR, but its applicability to the characterization of PANI is limited, due to the paramagnetism of this polymer. It is, however, well suited for the characterization of oligomers, which are nonconducting and their paramagnetism is reduced. As is well-known, the oxidative polymerization can be stopped by fast neutralization and dilution of the polymerization mixture. The resulting oligomers form a precipitate, which can be purified and analyzed using molecular spectroscopy methods. As the products of very early oxidation stages are mostly soluble in dimethylsulfoxide (DMSO), one of the methods of choice is NMR spectroscopy. We present here the results of such analysis. 2. Experimental Section Sample Preparation. Aniline (0.2 M) was oxidized with ammonium peroxydisulfate (0.25 M) in its aqueous solution containing acetic (0.4 M), succinic (0.4 M), or sulfuric acid (0.1 M) or in the absence of acid. Solutions of the monomer and the oxidant were mixed at room temperature to start the polymerization. After the predetermined reaction time (2 or 4 min), the mixture was poured into an excess of 1 M ammonium hydroxide to convert any protonated intermediate forms into the corresponding bases and to stop the oxidation by diluting the reactants. The precipitate of the aniline oligomers was immediately filtered off, rinsed several times with water to remove residual reactants, and dried under a vacuum. The

10.1021/jp9007834 CCC: $40.75  2009 American Chemical Society Published on Web 04/15/2009

NMR Investigation of Aniline Oligomers product was then dissolved in DMSO-d6, hexamethyldisiloxane was added as an internal standard, and the solution was put into an NMR tube. The main sample of aniline oligomer investigated in the present study was prepared by the oxidation of 0.2 M aniline or aniline-15N (Aldrich) in an aqueous solution of 0.4 M acetic acid with a limited amount of potassium peroxydisulfate (0.1 M). Under such conditions, aniline oligomer is produced and the oxidant is consumed before the acidity reaches the level needed for the polymerization of aniline.1,16 The product was treated as above and is further referred to as sample A. A part of the native samples was deprotonated in 1 M ammonium hydroxide to potential bases. NMR Measurements. 1H, 13C, and 15N NMR spectra of about 2 wt % sample solutions in dimethyl sulfoxide (DMSO-d6) were measured at 300.13, 75.46, and 30.40 MHz, respectively, with an upgraded Bruker Avance DPX 300 spectrometer. Additionally, for better sensitivity and resolution, 13C NMR spectra were measured at 150.86 MHz with a new Bruker Avance III 600 spectrometer. For 1H NMR spectra, 64 scans of 32 k-points were collected with a 12.5 s repetition time. For 13C and 15N NMR single-pulse and distortion-excluding polarization transfer (DEPT135) spectra, at least 6000 scans of 64 k-points were collected with the broadband probe, using the same repetition time. Before Fourier transform, 1 Hz exponential weighting was applied. A 5% w/v solution of nitromethane-15N in DMSO was used as the external standard for 15N spectra. 1H correlation spectroscopy (COSY) (normal and long-range), double quantum filtered (DQF) COSY and nuclear Overhauser effect spectroscopy (NOESY), as well as 1H-13C and 1H-15N heteronuclear single-quantum coherence (HSQC) and heteronuclear multibond coherence (HMBC) NMR spectra were measured using a z-gradient inverse-detection probe. Two k-points in F2 and 256 increments in F1 were measured, respectively, and 32 (in 1H 2D spectra) and at least 160 (in 1H-13C or 1H-15N 2D spectra) scans were collected. A phase-shifted sine-bell weighting was applied before Fourier transform to 1028 (F2) and 512 (F1) points. Pulsed-field-gradient stimulated-echo (PGSTE) experiments were done using a water-cooled gradient probe and Bruker gradient unit; the gradient pulse length δ was 1 ms, the diffusion delay was held constant (20 ms), and the field gradient was incremented in 16 steps in the range 0-500 G cm-1. 3. Results and Discussion In this study, oligomers obtained by stopping the oxidative polymerization of aniline in its very early stage were analyzed. In spite of the apparent good solubility of the oligomers in DMSO, we had to restrict their concentration to a maximum of about 2 wt % and remove the invisible heterogeneous phase by thorough centrifugation in order to obtain relatively wellresolved NMR spectra. Attempts to increase the maximum concentration and/or improve the resolution of the spectra by increasing the temperature of measurement led to marked changes of the spectra, indicating changes of structure. Therefore, all measurements were done at 296 K at which temperature the spectra were stable at least during the time of the experiment. However, some minor changes of the spectra were observed after prolonged storage of the samples in DMSO solution or, to a lower degree, even in solid form. The fact that the spectra rapidly deteriorate under increasing concentration and that the solution always contains microscopic heterogeneous objects which can be removed by centrifugation (at least 6000 rpm) indicates the tendency of the oligomers to aggregate and thus to avoid detection due to extreme signal

J. Phys. Chem. B, Vol. 113, No. 19, 2009 6667 broadening. Hence, it should be born in mind that not all features of structure of the given products might be apparent in our spectra owing to the well-known limitations of high-resolution NMR. 1 H NMR. Products obtained at the early stages of aniline polymerization by the oxidation of aniline with ammonium peroxydisulfate in the solutions of various acids or in water were analyzed by NMR in DMSO-d6 solution at 296 K. Relevant parts of 1H NMR spectra of the products are collected in Figure 1. The spectra do not contain the very broad signals of NH protons, which appear in the range 8.5-7.6 ppm (as shown in Figure 2) and do not give much information. As clearly seen, the spectra are virtually the same, differing only in the width of the signals and in some cases the broadening being probably produced by some higher-molecular-weight components. The major part of the NMR detectable products thus is identical or at least closely analogous, in accordance with the results of the earlier FTIR spectroscopic characterization.16 This was explained by the easy oxidation of neutral aniline molecules to oligomers,16 which preferentially takes place even in acidic media where the concentration of aniline molecules is low. The similar oxidation of anilinium cations is much more difficult and proceeds only later during the subsequent polymerization. In the following section, the analysis deals only with the oxidation product A obtained in 0.4 M acetic acid by using the limited amount of oxidant. Under such conditions, the reaction stops at the oligomeric stage before the polymerization of aniline starts. 1H NMR spectra of this product are shown in Figure 2 where the spectrum of a fully neutralized base form is on the bottom, whereas that of the native (or protonated) form is at the top. The signal assignment in the spectra corresponds to numbers for carbons and attached protons (if they exist) or to letters for nitrogen atoms and attached protons, as shown in Scheme 1 depicting the proposed prevailing structure of A. This proposed structure and the corresponding signal assignment are based on the arguments gradually explained in this study. 1 H NMR spectra in Figure 2 consist of three main regions: broad signals in the range 8.5-11.2 ppm (probably slowly exchanging NH protons of the oligomer chain), strongly overlapping aromatic proton signals in the range 6.6-8.0 ppm, and rather surprising signals in the range 5.4-6.5 ppm, from which we identify the most prominent two as 1 and 4. Prior to further discussion, it has to be noted that the well-developed aromatic signals clearly sit on a much broader signal, which probably corresponds to longer and less mobile conjugated oligoaniline chains. Some features of the structure thus are unfortunately lost. The attempts to improve the resolution by heating the sample to 340 K were only partly successful; at the same time, heating led to a marked change of the spectra (i.e., of structure) and was thus abandoned in further analysis. In a pulsed-field-gradient experiment, the signals including 1 and 4 exhibit virtually single-exponential decay corresponding to the diffusion coefficients in the range (2.37-2.40) × 10-10 m2 s-1, showing thus that (1) we see a relatively narrow oligomeric fraction in the spectra and (2) the protons corresponding to the observable signals belong to the same molecule (within experimental errors). The second conclusion is supported by the fact that 1 and 4 have NOESY cross-peaks with some of the signals in the aromatic region, as shown in Figure 3. The magnitude of the NOE effect is rather small, as can be expected in the case of an oligomer. The connectivity of the atoms corresponding to 1 and 4 with the rest of the molecule can be further proved by other spectra as shown below.

6668

J. Phys. Chem. B, Vol. 113, No. 19, 2009

Krˇizˇ et al.

Figure 1. 1H NMR spectra of the products of the oxidation of aniline in (a) 0.4 M acetic acid (reaction time 2 min), (b) the same (reaction time 4 min), (c) 0.1 M sulfuric acid, (d) 0.4 M succinic acid, and (e) the absence of acids (c-e: reaction time 4 min) (DMSO-d6, 296 K).

Remaining in the realm of proton spectra, let us consider the double-quantum filtered (DQF) COSY spectrum in Figure 4. The region 5.4-6.6 ppm containing signals 1 and 4 and their weaker variants is omitted, as these signals have no cross-peaks with the aromatic signals (showing thus that the groups containing protons 1 and 4 are not directly bonded to any of the aromatic groups, although they are near to them according to the NOESY spectrum shown in Figure 3). In spite of a great deal of signal overlap, one can see the connectivity between at least some of the proton signals. Thus, we have cross-peaks between 8a or 8b (ortho-protons in two kinds of a phenyl group) with 9 and 10 as well as between 12 and 13 or 16 and 17. There is a multitude of signals 12 which seemingly has no counterpart in signals 13 in the 1D proton spectrum, but the cross-peak pattern reveals that there is a multitude of mutually overlapped signals 13 in the region 7.38-7.48 ppm. Such variety has to be expected if the analogous aromatic units are repeated along the oligomer chain, as suggested in Scheme 1. 13 C NMR and 1H-13C NMR Correlation Spectra. Further information about the structure can be obtained from 13C NMR spectra and 1H-13C correlation. The 1D 13C NMR spectra of both the native and fully neutralized base product A are shown in Figure 5. The proposed signal assignment corresponding to

Scheme 1 is based on the subsequent (and previous) discussion of the 1H-1H, 1H-13C, and partly also 1H-15N correlation spectra. For sensitivity reasons, the spectra in Figure 5 were measured with a newly installed 600 MHz NMR spectrometer, about 2 months after all other spectra were recorded. As it can be seen by comparison with the F1-projections in the correlation spectra in Figures 6 and 7 (where earlier 75.46 MHz 13C NMR spectra were used), most of the signals (in particular 1, 4, and 2) are multiplied into groups of signals with slightly different chemical shifts. This cannot be merely the effect of a better resolution at the higher magnetic field intensity. As the repeated 1H NMR spectra at 300 MHz and 13C DEPT spectra at 75.46 MHz show an analogous (if less resolved) multiplying of signals, we have to consider it as an indication of some change in the samples. However, the cumulative intensity of the groups of signals remains approximately the same as that of the corresponding original single ones, suggesting thus that the chemical structure of the sample had not changed. At the same time, the repeated PFG NMR diffusivity measurement made at the same temperature no longer showed a uniform diffusion behavior of the sample but suggested the existence of at least four diffusion coefficients, three of them larger than the original one. Therefore, we suggest that the signal multiplication is a sign of some kind

NMR Investigation of Aniline Oligomers

J. Phys. Chem. B, Vol. 113, No. 19, 2009 6669

Figure 2. 300.13 MHz 1H NMR spectra of the native product A (top) and neutralized to corresponding base (bottom) collected in the limited oxidation of aniline in 0.4 M acetic acid (DMSO, 297 K, HMDSO ) 0.05 ppm).

SCHEME 1

of physical process slightly altering the magnetic shielding in some of the molecules (such as slow conformation change, aggregation or ordering) rather than of a chemical change. Unfortunately, there appears to be no way to examine the reversibility of the process, as heating of the sample leads to a structural change clearly reflected by NMR spectra. Part of the signal assignment in Figure 5 is based on the 1 H-13C direct correlation (HSQC) spectrum shown in Figure 6. The 13C signals correlated here to 1H signals assigned above on grounds of DQF COSY and NOESY spectra to either polyaniline units (12, 13, 16, 17) or to phenyl groups (8, 9, 10) have expected chemical shifts and are thus not surprising. The 13 C signals 1 and 4, like the corresponding 1H signals of the attached protons, have somewhat unusual chemical shifts virtually excluding the possibility that the corresponding carbon atoms are members of an aromatic or purely aliphatic moiety. 1 H-13C DEPT135 spectrum shows that both 1 and 4 correspond to C-H groups. Further information about these as well as other 13 C signals is obtained from the long-range 1H-13C HMBC correlation spectrum shown in part in Figure 7. Before starting discussion about this HMBC spectrum, let us point to the group of signals (originally mostly one signal) 2 in Figure 5. However misleading the chemical shifts of NMR signals might sometimes be, 2 must be a carbonyl group, as indicated in Scheme 1.

Figure 3. Relevant part of the 300.13 1H NOESY spectrum of the product A (DMSO, 296 K).

Inspecting now the HMBC spectrum in Figure 7, one can see that proton 1 correlates (in a decreasing value of the corresponding coupling constants, i.e., increasing distance) with carbons 2, 6, 3, and 5. Proton 4 correlates (in a decreasing value of coupling constants, again) with carbons 5, 3, 6, and 2. Such a coupling pattern can emerge only if the carbons 1, 2, 3, 4, 5, and 6 belong to a six-member ring. According to DEPT135 and HSQC spectra, carbons 2, 3, 5, and 6 have no proton attached to them. The proton signal 4 also correlates to other, originally minor signals at 147 and 154 ppm, the latter being assigned as 6a in Figure 5. The corresponding cross-peaks are perceptibly weaker,

6670

J. Phys. Chem. B, Vol. 113, No. 19, 2009

Krˇizˇ et al.

SCHEME 2

however. We assume that there is another weaker signal at approximately the same frequency as 4 belonging probably to an analogous proton in a similar structure attached to a slightly different molecular environment (e.g., R in Scheme 1 could be a proton, as shown below). Also there are other minor proton signals in the vicinity of either 4 or 1 correlating with 6, 6a, or the unassigned signals at 147 ppm, suggesting that there are more slightly different variants of the structure. It is quite common in NMR of oligomers that spectra corresponding to different chain lengths differ. In addition to it, the quinoid structure suggested in Scheme 1 could be at the end of the chain (R being a phenyl group) or somewhere inside the oligomeric chain. It is beyond our means to analyze all of these signals in detail. In the following, we concentrate on the main ones. As already stated, carbon 2 must be that of a carbonyl group. On the other hand, the other quaternary carbon 5 at 162 ppm cannot be a carbonyl; according to its shift, it cannot be NHsubstituted and thus has to belong to an imine group. In fact, the 15N NMR spectrum shown below contains the corresponding imine nitrogen c and the HMBC spectrum shows the expected coupling of proton 4 to it. The shifts of carbons 3 and 6 suggest that they are attached to a NH group, which is corroborated by the 1H-15N correlation, as shown below. The rest of the structure probably corresponds to attached phenyl groups and a not exceedingly populated oligoaniline chain, as proposed in Scheme 1 (for this, see the discussion below).

Figure 4. Relevant part of the 300.13 1H DQF-COSY spectrum of the product A (DMSO, 296 K).

15

N NMR and 1H-15N Correlation. Additional valuable information about the structure is offered by the 15N NMR spectra shown in Figure 8. According to the 1H-15N DEPT135 spectrum, a is a surprisingly intensive signal of NH2, b and e1 belong to NH, and c and c1 are signals of nitrogen atoms without a proton. The broad signal e cannot be identified as NH by DEPT135 thanks to a signal of the corresponding proton that is too broad, but it can be shown to be such using 15N APT spectrum. As one can see, the spectrum of the neutralized base product is dominated by the very broad signals of the oligoaniline chain d and e, the integral intensity of which is several times larger than that of the rest of the signals. The width of these signals could be caused by lower local mobility as well as local paramagnetism (see below). It is interesting that the signals of the oxygen-containing quinoneimine end group are clearly not affected by this broadening. All of the structural reasoning presented so far is further strengthened by the 1H-15N long-range correlation (HMBC) NMR spectrum in Figure 9. The spectrum was measured with the coupling-evolution delay assuming the 1H-15N coupling constant to be 6 Hz; shortening or prolonging of this delay had no qualitative effect on the spectrum. In Figure 9, we clearly see that the proton 4 is within coupling reach with the imine nitrogen c, in contrast to the proton 1. The latter is understandably coupled to nitrogen b, whereas the proton 4 apparently has several possibilities, starting with nitrogen a (in the form of a NH2 end group) and ending with e1 or even e if the quinoneimine moiety is not at the end of the chain. Summary of the Spectroscopic Arguments for the Proposed Structure. Although all of the spectroscopic information substantiating the proposed structure is contained in the above text, let us summarize the logic leading to its surprising part, namely, the carbonyl-containing quinoimine unit. The protons and attached carbons (correlated by HSQC) 1 and 4 belong to the same structure unit (shown by NOESY, 1H-13C HMBC) which is attached to the rest of the molecule (proved by NOESY). The chemical shifts of both proton and carbon signals 1 and 4 suggest a membership in a quinoid structure. 1H-13C HMBC shows that proton 1 is in a two-bond distance to quaternary carbons 2 and 6 and a three-bond distance to quaternary carbons 3 and 5. It also shows that proton 4 is in a two-bond distance to quaternary carbons 5 and 3 and threebond distance to quaternary 2 and 6 (the quaternary nature of 2, 3, 5, and 6 was proved by 1H-13C DEPT and APT spectra). From this, it is clear that carbons 1, 2, 3, 4, 5, and 6 form a six-member ring. Now, the chemical shift of carbon 2 shows without any doubt that it is a carbonyl. The chemical shift of 5 suggests that it is an imine carbon. This is proved by 1H-15N HMBC, where proton 4 correlates with the nitrogen c, the imine nature of which was proved by APT and 1H-15N DEPT spectra. In an analogous way, by 1H-15N HMBC, carbons 3 and 6 are

NMR Investigation of Aniline Oligomers

J. Phys. Chem. B, Vol. 113, No. 19, 2009 6671

Figure 5. 150.86 MHz 13C NMR spectra of the native (top) and neutralized (bottom) product A (DMSO, 297 K).

1

13

Figure 6. H- C HSQC (direct correlation) spectrum of product A (DMSO, 297 K).

proved to be bonded to nitrogen atoms a and b, respectively (a is correlated with proton 4, which is next to carbon 3, b with proton 1, which is next to carbon 6); a and b are shown to be NH (or NH2 in the case of a). 1H-15N HMBC correlation of c with the protons 8 further shows that the imine nitrogen c is further bonded to a phenyl group, as suggested in Scheme 1.

Figure 7. 1H-13C HMBC (direct correlation) spectrum of product A (DMSO, 297 K).

Thus, the quinoimine unit shown in Scheme 1 can be considered to be fully proved by NMR. Reliability of the Present Structure Analysis. As pointed out at the very beginning, there are signs that a part of the sample or possibly part of the observed structure could avoid NMR detection due to extreme broadening of the corresponding signals: the aromatic part of 1H NMR spectra clearly is superimposed on a broad, unresolved signal; some visible signals

6672

J. Phys. Chem. B, Vol. 113, No. 19, 2009

Krˇizˇ et al.

Figure 8. 30.40 MHz 15N NMR spectra of the native product A (top) and the product neutralized to corresponding base (bottom) (DMSO, 297 K, 5% nitromethane/DMSO ) 0.0 ppm).

Figure 9. 1H-15N long-range correlation (HMBC) NMR spectrum of the 15N-enriched neutralized product A (DMSO, 297 K, 5% nitromethane/ DMSO ) 0.0 ppm).

in 13C or 15N show perceptible broadening; when increasing the sample concentration, the integral signal intensity does not increase quite additively. Extreme broadening can be caused either by restricted mobility of the molecular segments (due to conjugation or self-aggregation) or by traces of paramagnetism. Indeed, ESR spectra found (2 ( 0.4) × 10-9 mol g-1 of unpaired electron spins in both forms of the sample, i.e., in the native and the fully neutralized one, in spite of the fact that the samples

do not conduct electric current.16 Although the level of paramagnetism is not high, one has to admit that some parts of the sample and/or structure could avoid detection due to extreme broadening caused by fast relaxation. All of these arguments point to the possibility that some structural features could be emphasized in our spectra, whereas others could be relatively suppressed. However, this possibility does not devalue the structural findings presented here.

NMR Investigation of Aniline Oligomers The rather surprising oxygen-containing quinoneimine structure suggested here on the base of the spectra could be formed by an attack of oxidized aniline on an ortho-position of an already bound aniline unit (either at the end of the chain or even somewhere inside it), as schematically shown in Scheme 2. The ortho-attack is known to be quite probable.16 The reason why its product is prone to oxidation in the suggested way could be that, due to conjugation, the radical preceding the final quinoneimine structure is better stabilized than the possible variant structures. This has to be cleared by theoretical tools, however. Somewhat surprisingly, we were not able to detect phenazine or phenylphenazine units found by vibrational spectra.19,20 The lack of the corresponding signals in all NMR spectra could be caused by local immobilization of these structural units and consequent extreme broadening of the corresponding signals. Finally, it is interesting to compare spectra of the native and neutralized product in Figures 2, 5, and 8. There is no doubt that the native product is protonated. Considering the rather low level of paramagnetism mentioned above, it is simply an effect of acid-base interaction. The chief site where the protons are bound are the imine nitrogen atoms d: the corresponding broad signal in 15N spectrum exhibits a huge upfield shift of about 58 ppm, and as one can expect, the signals of protons and carbons 16 and 17 are shifted, too. However, the NH sites in the oligoaniline chain are clearly affected as well: there is a slight shift of the broad signal e, and part of it is transformed into a narrower signal. The oxygen-containing quinoneimine moiety appears to be unaffected by protonation. 4. Conclusions Using various 1H, 13C, and 15N 1D and 2D spectra, we have shown that the products of early stages of oxidative aniline polymerization have quite analogous structures irrespective of the kind and concentration of the added acid, including its absence. In addition to the expected oligomeric chains with intermittent diamine and quinoneimine moieties, which are only partly detected by NMR due to their signal broadening, a markedly populated and hitherto not described oxygen-containing quinoneimine group was detected. The formation of such a group, which is partly terminal and partly embedded in the oligomeric chain, can be explained by an attack by an aniline molecule on a terminal or inner phenylenediamine group in an ortho-position and subsequent oxidation. It is not clear what role, if any, such a unit can have on the stereochemistry of the polymer growth, but obviously, it destroys the chain conjugation needed for the conductivity of products. It was found that the native polymerization products before neutralization are strongly protonated. The protonation affects mostly the regular oligoaniline chain, in particular its nitrogens in quinonediimine units. Acknowledgment. The authors thank the Czech Grant Agency (203/08/0686) and Grant Agency of the Academy of Sciences of the Czech Republic (IAA400500905) for financial support and J. Pilaø from the Institute of Macromolecular Chemistry in Prague for the EPR measurements.

J. Phys. Chem. B, Vol. 113, No. 19, 2009 6673 References and Notes (1) Sapurina, I.; Stejskal, J. Polym. Int. 2008, 57, 1295. (2) Xia, H.; Cheng, D.; Lam, P.; Chan, H. S. O. Nanotechnology 2006, 17, 3957. (3) Zhang, L.; Zhang, L.; Wan, M.; Wei, Y. Synth. Met. 2006, 156, 454. (4) Zhou, S.; Wu, T.; Kan, J. Eur. Polym. J. 2007, 43, 395. (5) Stejskal, J.; Sapurina, I.; Trchova´, M.; Konyushenko, E. N.; Holler, P. Polymer 2006, 47, 8253. (6) Konyushenko, E. N.; Stejskal, J.; Sˇedeˇnkova´, I.; Trchova´, M.; Sapurina, I.; Cieslar, M.; Prokesˇ, J. Polym. Int. 2006, 55, 31. (7) Zhang, L. J.; Zujovic, Z. A.; Peng, H.; Bowmaker, G. A.; Kilmartin, P. A.; Travas-Sejdic, J. Macromolecules 2008, 41, 8877. (8) Chiou, N.-R.; Lee, L. J.; Epstein, A. J. Chem. Mater. 2007, 19, 3589. (9) Hsieh, B.-Z.; Chuang, H.-Y.; Chao, L.; Li, Y.-J.; Huang, V.-J.; Tseng, P.-H.; Hsieh, T.-H.; Ho, K.-S. Polymer 2008, 49, 4218. (10) Sun, Q.; Deng, Y. Mater. Lett. 2008, 62, 1831. (11) Janosevic´, A.; C´iric´-Marjanovic´, G.; Marjanovic´, B.; Holler, P.; Trchova´, M.; Stejskal, J. Nanotechnology 2008, 19, 135606. (12) Tran, H. D.; Wang, Y.; D’Arcy, J. M.; Kaner, R. B. ACS Nano 2008, 2, 1841. (13) MacDiarmid, A. G.; Epstein, A. J. Faraday Discuss. Chem. Soc. 1989, 88, 317. (14) Stejskal, J.; Kratochvı´l, P.; Jenkins, A. D. Polymer 1996, 37, 367. (15) Stejskal, J.; Gilbert, R. G. Pure Appl. Chem. 2002, 74, 857. (16) Stejskal, J.; Sapurina, I.; Trchova´, M.; Konyushenko, E. N. Macromolecules 2008, 41, 3550. (17) Meth-Cohn, O.; Smith, M. J. Chem. Soc., Perkin Trans. 1994, 1, 5. (18) Meth-Cohn, O.; Travis, A. S. Chem. Br. 1995, 31, 547. (19) Trchova´, M.; Sˇedeˇnkova´, I.; Konyushenko, E. N.; Stejskal, J.; Holler, P.; C´iric´-Marjanovic´, G. J. Phys. Chem. B 2006, 110, 9461. ´ iric´-Marjanovic´, G.; Trchova´, M.; Stejskal, J. J. Raman Spectrosc. (20) C 2008, 39, 1375. 2008. (21) C´iric´-Marjanovic´, G.; Konyushenko, E. N.; Trchova´, M.; Stejskal, J. Synth. Met. 2008, 158, 200. (22) C´iric´-Marjanovic´, G.; Trchova´, M.; Stejskal, J. Int. J. Quantum Chem. 2008, 108, 318. (23) Venancio, E. C.; Wang, P. C.; MacDiarmid, A. G. Synth. Met. 2006, 156, 357. (24) Wei, Z.; Wan, M.; Lin, T.; Dai, L. AdV. Mater. 2003, 15, 136. (25) Dong, H.; Nyame, V.; MacDiarmid, A. G.; Jones, W. E. J. Polym. Sci., Part B: Polym. Phys. 2004, 42, 3934. (26) Dong, H.; Prasad, S.; Nayme, V.; Jones, W. E. Chem. Mater. 2004, 16, 371. (27) Qiu, H.; Wan, M.; Matthews, B.; Dai, L. Macromolecules 2001, 34, 675. (28) Zhang, Z.; Wei, Z.; Wan, M. Macromolecules 2002, 35, 5937. (29) Wei, Z.; Wan, M. AdV. Mater. 2002, 14, 1314. (30) Long, Y.; Zhang, L.; Ma, Y.; Chen, Z.; Wang, N.; Zhang, Z.; Wan, M. Macromol. Rapid Commun. 2003, 24, 938. (31) Zhang, L.; Long, Y.; Chen, Z.; Wan, M. AdV. Funct. Mater 2004, 14, 693. (32) Xia, H.; Chan, H. S. O.; Xiao, C.; Cheng, D. Nanotechnology 2004, 15, 1807. (33) Kulszewitz-Bajer, I.; Rozalska, I.; Kurylek, M. New J. Chem. 2004, 28, 669. (34) Kaplan, S.; Conwell, E. M.; Richter, A. F.; MacDiarmid, A. G. J. Am. Chem. Soc. 1988, 110, 7647. (35) Mathew, R.; Mattes, B. R.; Espe, M. P. Synth. Met. 2002, 131, 141. (36) Cotarelo, M. A.; Huerta, F.; Mallavia, R.; Morallon, E.; Vazquez, J. L. Synth. Met. 2006, 156, 51. (37) Zujovic, Z. D.; Gizdavic-Nikolaidis, M. R.; Kilmartin, P. A.; Idriss, H.; Senanayke, S. D.; Bowmaker, G. A. Polymer 2006, 47, 1166. (38) Zujovic, Z. D.; Zhang, L.; Bowmaker, G. A.; Kilmartin, P. A.; Travas-Sejdic, J. Macromolecules 2008, 41, 3125. (39) Hopkins, A. R.; Lipeles, R. A.; Hwang, S.-J. Synth. Met. 2008, 158, 594. (40) Gill, M. T.; Chapman, S. E.; DeArmitt, C. L.; Baines, F. L.; Dadswell, C. M.; Stamper, J. G.; Lawless, G. A.; Billingham, N. C.; Armes, S. P. Synth. Met. 1998, 93, 227.

JP9007834