Ion-Irradiation-Induced Carbon Nanostructures in Optoelectronic Polymer Materials

The recent results obtained on the ion-irradiation-induced carbon nanostructures in optoelectronic polymer materials exemplified by boron-ion-implanted polymethylme‐ thacrylate (B:PMMA) with an energy of 40 keV, ion doses from 6.25 × 10 14 to 5.0 × 10 16 ions/ cm 2 , and current density <2 μA/cm 2 are reviewed. The positron annihilation spectrosco‐ py (slow positron beam spectroscopy based on Doppler broadening of positron annihilation gamma rays as a function of incident positron energy and positron annihilation lifetime at a positron energy of 2.15 keV, and temperature-dependent positron annihilation lifetime spectroscopy), optical UV-visible spectroscopy, Raman spectrosco‐ py, electrical (current–voltage) measurements, and nanoindentation test are chosen as the main experimental tools for the investigation of low-energy ion-induced processes in B:PMMA. The formation of carbon nanostructures is confirmed for the samples irradiated with higher ion fluences (>10 16 ions/cm 2 ) and the experimental results of the comprehen‐ sive study are found to be in a good agreement with SRIM (stopping and range of ions in matter) simulation results.


Introduction
Ion implantation is a powerful experimental approach for structural modification of materials. In case of organic media, the interest to ion-irradiated polymers is due to the ion implantation being one of the effective technological methods to turn dielectric polymers into semiconductors [1] as well as to improve surface-sensitive mechanical properties of polymers for hard-materials applications [2]. Substantial improvements in hardness, wear resistance, oxidation resistance, resistance to chemicals, and electrical conductivity are the main characteristics attained by polymeric materials after their low-to-medium-energy ion implantation [3][4][5][6][7].
The formation of free radicals at lower ion doses (<10 16 ions/cm 2 ) and carbonization at higher ion doses (>10 16 ions/cm 2 ) in the most polymeric materials are of general concept [8][9][10]. In particular, the ion irradiation of polymers leads to the scission and cross-linking of polymer chains, formation of volatile low-molecular fragments, and carbonization of the implanted layer. The carbonization process depends strongly on the implantation dose or/ and ion current density. As ion dose increases, the several stages can be distinguished for carbonization to be occurred. The formation of pre-carbon structures, nucleation, and growth of the carbonenriched clusters, aggregation of the clusters, formation of network of conjugated bonds, and transition to amorphous carbon or graphite-like material are the main stages predicted in literature [8][9][10]. Despite numerical studies of carbonization processes in polymers, the problem remains to understand better how the carbonaceous phase or carbon nanostructures formed under high-dose ion implantation could be dependent on the type of polymer matrix. In this respect, a lot of efforts have been made by researchers studying the B + -ion implantation into various polymer matrix such as polycarbonate, Kapton, polyethylene, polyamide, polyimide, poly(ethylene terephthalate), cellulose, polypropylene, polystyrene, polyethersulfone, and others ( [2,4,5,[11][12][13][14][15][16][17][18][19][20][21][22][23][24][25] and references therein). At the same time, the B + -ion implantation into widely used in practice polymer polymethylmethacrylate (PMMA) was not studied so far.
The interest to B + -ion implantation of polymers is due to formation of buried carbonaceous layer leading to increasing conductivity [20,23]. Also, B + -ion implantation into polymeric matrix results in significant increase of surface-sensitive mechanical properties. So, Lee et al. [4] reported the ion-induced improvement in hardness of Kapton irradiated by three different ion species, He, B, and Si at 200 keV to a dose of 3.5 × 10 19 ions/m 2 , where boron produced the largest improvement in hardness among the three, not following the increasing trend in atomic number.
The selection of PMMA matrix to be used for B + -ion implantation is due to an importance of this polymer for construction of many optical components (waveguides, lenses, prisms, etc.), lithography, biomedical applications, etc. ( [26][27][28][29][30] and references therein). PMMA was also a subject for implantation with Ag + -ions [9,31] to fabricate composite structures with silver nanoparticles for plasmonic applications as well as for implantation with C + , N + and Ar + -ions [29,30] that may find an extensive application in fabrication of various optoelectronic devices including organic light-emitting diodes, backlight components in liquid crystal display systems, diffractive elements, solar cells, waveguides, microcomponents for integrated optical circuits, etc. The well-known key performance and important characteristics of PMMA such as a long-term stability in outdoor environments, excellent surface hardness, light weight, outstanding transmittance and optical clarity, optical design flexibility and control, etc. [32], and high stability upon positron irradiation at room temperature [33] are also taken into account at the selection of basic polymeric matrix for low-energy ion implantation.
In this chapter, a review of recent results obtained using proper experimental and simulation techniques on the boron-ion-implanted polymethylmethacrylate (B:PMMA) is presented. As a result, the formation of ion-irradiation-induced carbon nanostructures in optoelectronic polymer materials exemplified by B:PMMA is evidently confirmed.

Sample preparation and SRIM simulation
The B + -ion implantation with the energy of 40 keV, doses from 6.25 × 10 14 to 5.0 × 10 16 ions/cm 2 and current density < 2 μA/cm 2 into the optically transparent PMMA plates (1.2 mm thickness) was performed under a pressure of 10 −5 Torr at room temperature by an "ILU-3" ion accelerator at the Kazan Physical-Technical Institute (KPTI, Russia) similar to as it was earlier done for Xe + and Ag + ions [34].
For comparative analysis of depth profiles of implanted ions and introduced vacancies in respect to ion mass, SRIM (stopping and range of ions in matter) simulations were carried out for 40 keV He + , B + , O + , P + , Cl + , Cu + , Ag + , Xe + , and Au + ions implanted into PMMA using the free of charge software of version SRIM-2013 [35].

Positron annihilation spectroscopy measurements
Positron annihilation spectroscopy (PAS) measurements with slow positron beam spectroscopy (SPBS) based on Doppler broadening of positron annihilation gamma-rays as a function of incident positron energy and positron annihilation lifetime at constant positron energy were performed at the National Institute of Advanced Industrial Science and Technology (AIST, Japan) [36]. Doppler broadening spectra (S-E and W-E) were measured using a slow positron beamline in the range of positron incident energy E from 0 to 30 keV. Positron annihilation lifetime spectra at incident positron energy of 2.15 keV were measured by another slow positron beamline using an electron linear accelerator as an intense source of slow positrons. The details of the SPBS measurements are reported elsewhere [36].
PAS measurements with positron annihilation lifetime spectroscopy (PALS) and Doppler broadening of annihilation line (DBAL) techniques in the temperature range of 50-300 K using helium cryostat (Closed Cycle Refrigerator, Janis Research Company, Inc., USA) and vacuum equipment (Pfeiffer Vacuum, HiCUBE, Germany) were carried out at the Institute of Physics, Slovak Academy of Sciences (IPSAS, Slovakia) [37][38][39]. The samples were measured in the cycles of heating and cooling with step of 20 K and elapsed time of 4-5 hours per point. At the selected temperatures, the elapsed time was extended for better statistics in the lifetime spectra. For these temperatures, the continuous lifetime analysis technique to obtain the distributions of ortho-positronium (o-Ps) lifetime (free-volume voids) was employed using maximum entropy lifetime (MELT) program [40]. The positron annihilation lifetime spectra were taken by the conventional fast-fast coincidence method using plastic scintillators coupled to photomultiplier tubes as detectors. The radioactive 22 Na positron source (1.5 MBq activity) was deposited in an envelope of Kapton foils and then sandwiched between two samples. This source-sample assembly was placed in a vacuum chamber between two detectors to acquire lifetime spectra at different temperatures. The time resolution (FWHM) of positron lifetime spectrometer was 0.32 ns, measured by defect free Al sample as a standard. Analysis of lifetime spectra was carried out using the PATFIT-88/POSITRONFIT [41] software package with proper source corrections. Three component fitting procedure for PALS data treatment was applied and long-lived lifetime component τ 3 and its intensity I 3 , ascribing to the o-Ps pick-off annihilation in free-volume spaces, was finally taken into account for analysis. Simultaneously with PALS measurements, the DBAL spectra were recorded using a high-purity Ge detector with energy resolution of 1.9 keV FWHM at the energy 1274 keV. The annihilation line was deconvoluted by Gold algorithm that allows eliminating a linear instability of the measuring equipment if the 1274 keV 22 Na peak is measured simultaneously with the annihilation peak, for deconvolution of annihilation peak 511 keV; this procedure permits the measurement of small changes of the annihilation peak with high confidence [42]. The Doppler S and W parameters were used for analysis of a shape of deconvoluted annihilation peak 511 keV. Their numerical values were defined as the ratio of the central area to the total area of the photon peak for S parameter and as the ratio of wings area to the total area of the photon peak for W parameter.

Optical spectroscopy measurements
Optical UV-visible spectroscopy measurements were performed using a SHIMADZU-UV 3100PC spectrophotometer in the range of 200-800 nm at AIST, Japan [36].

Raman spectroscopy measurements
Room temperature Raman spectra were recorded using a Renishaw Raman inVia Reflex spectrometer in the 400-3800 cm −1 range with a spectral resolution of ≤1 cm −1 [43]. The 514 nm Modu-Laser Stellar-REN Pro 514/50 Argon Laser and 785 nm Near Infrared Diode Laser lines were applied as the excitation source. In order to avoid heat-induced effects, the laser power was set at 10 mW. A standard calibration of wavenumber scale with Si plate was used. The Raman spectroscopy measurements were performed at The John Paul II Catholic University of Lublin (KUL, Poland).

Electrical measurements
The electrical measuring system was developed at the IPSAS (Slovak Republic) [43]. The measurement of current vs. voltage (I-V) by transient method or DC method was used in this experiment. The method was selected to check simply presence of the conductivity layer in the material studied. The measuring system includes a cryostat with sample, resistor heating, thermocouple, CSP (charge sensitive preamplifier), excitation circuit of sample, preamplifier, power supply, DAQ (data acquisition) card, and PC (personal computer). The NI PCI 6229 DAQ card was applied [44]. The electrical DC measurement setup used and detail description of basic parameters are reported elsewhere [43].

Nanoindentation test
Nanoindentation test was carried out using an ultra nano hardness tester (UNHT) with a diamond Berkovich indenter at the Lublin University of Technology (LUT, Poland) [45,46]. Advantages of the new UNHT design applied for nanoindentation test compared to the conventional nano indenter (or NHT) design, both developed by CSM Instruments (Switzerland) [47], allow us to make measurements with high performance. The UNHT experiment was done in a progressive multicycle mode. The details of the progressive multicycle mode parameters used are reported elsewhere [45]. Figure 1 shows the typical SRIM simulation results for B + -implantation into PMMA at energy of 40 keV [36,45]. Numerical values of SRIM simulation for 40 keV He + , B + , O + , P + , Cl + , Cu + , Ag + , Xe + , and Au + ions into PMMA are gathered in Table 1 [48].  The SRIM data, presented in Table 1, were used for estimation a predicted thickness of implanted layer [48].  Table 1, indicate the possible modification of PMMA surface upon low-energy ion implantation in dependence on the ion mass. These values could also be useful in practice for evaluation of geometrical parameters of ion-implanted layers in PMMA matrix.

Positron annihilation spectroscopy data
The 40 keV B:PMMA polymers with different ion doses have been studied for the first time using PAS techniques such as SPBS [36] and temperature-dependent PALS [37].
SPBS, often called variable-energy PAS, is a powerful experimental tool widely used for evaluation of defects in solids as a function of depth (defect depth profiling) by varying the positron energy in the range of a few eV to tens keV (for review, see Ref. [49]). Thus, SBPS could be a very effective for the detection of defects induced by ion implantation, which are localized near the surface of material. At first, the un-implanted PMMA and implanted B:PMMA were characterized by Doppler broadening of annihilation gamma rays (or DBAL) as a function of incident positron energy in the range of 0-30 keV. And then, using Doppler broadening results, the PAL spectra at incident positron energy of 2.15 keV were measured and analyzed for the investigated samples. Figure 2 shows the typical variable-energy DBAL and PAL results obtained, the explanation of which has been done in [36]. It should only be emphasized here that there are two different processes seen from the S-E and W-E curves in dependence on the ion implantation dose [36]. In particular, (i) S(E) increases and W(E) decreases at lower fluences (6.25 × 10 14 -3.13 × 10 15 ions/cm 2 ), while (ii) S(E) decreases and W(E) increases at higher fluences (2.5 × 10 16 -5.0 × 10 16 ions/cm 2 ) in the range up to 400 nm in a good agreement with the maximum penetration depth R max B = 400 nm after SRIM simulation; positron energy of 2-3 keV, showing the extreme values of S(E) and W(E), corresponds to a mean depth of 100-200 nm in consistent with the maximum damage R p V,B = 220 nm after SRIM simulation (Figure 2, top). A mean penetration depth of positrons (z m ) was estimated as z m = (40/ρ)E n , where z m is presented in nm, E the positron energy in keV, ρ is the density (1.18 g/cm 3 for PMMA), and n = 1.6 [50]. The results of variable-energy DBAL and PAL measurements at incident positron energy of 2.15 keV were found to be in consistence (Figure 2, bottom) [36]. That is, the decreasing suddenly the S(E) values and an absence of any observable o-Ps yield (intensity I 3 ~ 0) for the implanted samples at higher ion doses were detected to be explained due to carbonization effect, taking into account that no o-Ps yield has been observed in carbon-based materials such as, for instance, fullerene C 60 cage [51] and carbon molecular sieve membranes [52].
The results of temperature dependent PALS measurements of the investigated pristine PMMA and implanted B:PMMA samples at lower ion dose (3.13 × 10 15 ions/cm 2 ) and at higher ion dose (3.75 × 10 16 ions/cm 2 ) are demonstrated in Figure 3. The detail description and possible explanation of these data have been presented in [37]. Two structural transitions in the vicinity of ~150 and ~250 K, ascribed to γ and β transitions, respectively, should be noted here, which are observed in the both PMMA and B:PMMA. These structural transitions are found to be in consistent with reference data for PMMA [53,54]. The threshold temperatures nearby ~150 and ~250 K for PMMA are also detected on Doppler parameters S/S 0 vs. W/W 0 at different temperatures, normalized to the S 0 and W 0 values at room temperature, using low-temperature DBAL measurements [39] as shown in Figure 4. The observed correlation in temperature dependences of S and W parameters and o-Ps data (see Figure 3) is found to be in a good agreement with literature PALS and DBAL data for PMMA as well [33]. New results are also obtained from the PALS study of B:PMMA on the free-volume voids distribution at room temperature [37] as shown in Figure 5. Namely, estimating the distribution of o-Ps lifetime (or the distribution of free volume detected by o-Ps) by MELT program similarly as it has been done in work [55], it is found that B + -ion implantation leads to the shortening the lifetime distribution and decreasing molecular weight in PMMA at lower ion dose (3.13 × 10 15 ions/cm 2 ), while at higher ion dose (3.75 × 10 16 ions/cm 2 ) the broadening the lifetime distribution is probably caused by local destruction of PMMA matrix and generation of additional free volumes.

Optical spectroscopy data
Application of UV-visible optical absorption spectroscopy for investigation of ion-implanted polymeric materials has already been reported for the Ag + -implanted PMMA, ORMOCER, and Epoxy resin [31,34,[56][57][58][59] as well as for the C + , N + , and Ar + -implanted PMMA [29,30]. It has been suggested by the authors that ion irradiation creates compact carbonaceous clusters in polymers, which may also be responsible for a narrowing of optical band gap, enhanced electrical conductivity, and increasing optical absorbance (for example, see [31,34,56]). In the case of the investigated B:PMMA, it is found the gradual increase of absorbance at lower fluences (<10 16 ions/cm 2 ) and saturation of absorbance at higher fluences (>10 16 ions/cm 2 ) as shown in Figure 6 [36]. The observed increasing absorbance for the B:PMMA samples in the course of the ion implantation should be also interpreted as the signature on the formation of carbonaceous clusters which are plausibly confirmed by slow positrons [36]. Figure 7 shows the Raman spectra of the investigated samples excited by 514 nm and 785 nm diode laser lines [43]. Identification of the observed Raman bands has been reported in [43]. The ion-irradiation-induced structural changes as revealed from Raman study can be summarized as follows. New C=C and C−C bands in the vicinity of ~1590 and 1322 cm −1 , respectively, are formed for the as-implanted samples at higher fluences (>10 16 ions/cm 2 ). At the same time, the decreasing intensity of CH 2 band at ~815 cm −1 , C=O band at ~1730 cm -1 , O−CH 3 band at ~991 and 2845 cm -1 , and C−H band at ~1455, 2955, and 3001 cm −1 is detected. In particular, Raman spectroscopy data for excitation by the laser wavelength of 785 nm seem to be additional confirmation for the carbonization processes in the B:PMMA. Indeed, new Raman bands at ~1325 and 1590 cm -1 , attributed to C−C and C=C vibrations, respectively, are detected only for the ion-irradiated samples with higher fluences (>10 16 ions/cm 2 ). These two new Raman bands may also be attributed to the so-called D and G peaks in the region of 1300-1600 cm −1 which are the main peaks characteristic for graphite and graphene structure [60][61][62][63][64][65][66]. The intensity ratio of the D and G peaks I D /I G is a measure of the size of the sp 2 phase organized in rings [67]. The sp 2 phase is mainly organized in chains, when I D /I G is negligible [63]. The relation between I D /I G and the size of the sp 2 phase L a is given by the equation of Tuinstra and Koenig [60]: I D /I G = C(λ)/L a . Here, C(λ) is a wavelength dependent factor [29,68]: C(λ) = −126 + 0.033λ, where λ is the excitation wavelength in (Å) at which the Raman spectra were recorded.

Raman spectroscopy data
A correlation between slow positron beam and Raman spectroscopy results for B:PMMA is mentioned in our recent works [69]. Table 2 gives the SPBS data, exemplified by o-Ps lifetimes and intensities at incident positron energy of 2.15 keV and Raman data, exemplified by I D /I G and L a , for the same B:PMMA samples. A good correlation between these data is clearly observed. Thus, the expected carbonization processes or formation of carbon nanostructures at higher fluences (>10 16 ions/cm 2 ) can independently be identified using SPBS and Raman spectroscopy, providing their combination as a powerful experimental tool in the investigation of ion-implanted polymers.  Table 2. o-Ps lifetimes and intensities at incident positron energy of 2.15 keV and the intensity ratio of the D and G peaks, and calculated sizes of sp 2 phase L a [69].

Electrical measurements data
The un-implanted PMMA and as-implanted B:PMMA samples were measured at 300 K and 360 K with DC method [43]. It was supposed that the conductive layer in ion-implanted polymer is dependent on the temperature and at a higher temperature the effect of increasing conductivity with temperature will be more pronounced. In order to avoid a possible structural change in polymer matrix, the maximum temperature at 360 K was selected in the electrical measurement experiment, not exceeding the glass transition temperature T g of PMMA ranging from 358 to 438 K; the range is so wide because of the vast number of commercial compositions which are copolymers with co-monomers other than methyl methacrylate [70]. Similar value T g ≅ 355 ± 18 K of PMMA was obtained using temperature dependent positron annihilation lifetime measurements [53]. Figure 8 shows the obtained I-V dependences, fitted by the linear regression line [43]. As an example, the I-V characteristic of as-implanted B:PMMA (5.0 × 10 16 ions/cm 2 ) sample is presented with noise (as-implanted + noise (II)) and linear regression line (linear fit of asimplanted (II)). While for other samples only, the linear regression lines of the DC measurements are demonstrated. More details of the electrical measurement experiment have been reported in [43]. The numerical values of the I-V characteristics are presented in Table 3 [43]. It is evidently proven that the pristine PMMA does not exhibit any conductivity neither at higher temperature applied. At room temperature, the slope of linear regression line of I-V dependence remains unchanged and has even negative value. In contrary to unimplanted sample, the results of the I-V measurements for the B:PMMA (3.75 × 10 16 and 5.0 × 10 16 ions/cm 2 ) samples revealed the change of slope of the linear regression line of I-V dependence at 360 K. It means that the as-implanted samples have created a very thin conductive layer or conductive joints due to carbonization processes or formation of carbon nanostructures in consistent with the results of slow positron beam and Raman spectroscopy measurements [36,43].

Nanoindentation test data
A first time the results of investigation of the influence of low dose (6.25 × 10 14 ions/cm 2 ) B +ion-irradiation on the mechanical properties (hardness and elastic modulus) of PMMA probed by nanoindentation with UNHT in the range of 300-1100 nm indentation depth have been reported in [45]. It has been established that the hardness and elastic modulus versus maximum indentation depth illustrate the main difference between the un-implanted (pristine) and ionimplanted samples in the range up to about 400 nm. The same value of the maximum penetration depth of B + -ions into PMMA has been found using SPBS and SRIM simulation [36]. As a continuation of the nanoindentation test of the B:PMMA, the averaged values of indentation hardness versus maximum indentation depth for the un-implanted and as-implanted samples with ion doses of 6.25 × 10 14 , 1.25 × 10 16 , 2.5 × 10 16 , and 3.75 × 10 16 ions/cm 2 are plotted in Figure 9 [46]. One may see that the hardness dependence on the maximum indentation depth demonstrates the difference between the PMMA and B:PMMA samples in the entire range studied up to 1100 nm with the largest changes in the vicinity of 300-400 nm in consistence with the maximum penetration depth of B + -ions into PMMA as revealed from SPBS measurements and SRIM simulation [36]. The observed improving of surface-sensitive mechanical properties of B:PMMA by ion beam processing is obviously detected to be more significant as ion dose increases, that may be suitable for hard-materials applications. According to Lee et al. [4] these properties are apparently related to the effectiveness of cross-linking. Besides, the abovementioned formation of carbon nanostructures upon high-dose ion implantation (>10 16 ions/cm 2 ) seems to be also important. But further increase of hardness with deeper penetration of indenter up to 1100 nm for higher fluences (3.75 × 10 16 B + /cm 2 ) was found to be very interesting and not fully understood yet (see Figure 9). Actually, similar results that the hardness of B +implanted polycarbonate increased with increasing ion dose has also been observed in Ref. [2] but only for penetration of indenter up to 400 nm, that is not so far from the implanted layer as in our case. A deeper understanding the low-energy ion-irradiation-induced processes in polymeric materials exemplified by PMMA irradiated by accelerated light, middle, and heavy ions is still required.

Conclusions
The formation of carbon nanostructures has been confirmed for the B:PMMA samples irradiated with higher ion fluences (>10 16 ions/cm 2 ) and the experimental results of the comprehensive study have been found to be in a good agreement with SRIM simulation results. It is expected that the results obtained for low-energy B + -ion implantation into PMMA will have impact on research and development in the fields of nanoscience, nanotechnology, and nanooptoelectronics, in particular, to be potentially of interest for fabrication of organic luminescent devices, backlight components in liquid crystal display systems, diffractive elements and microcomponents for integrated optical circuits, solar cells, waveguides, etc. similarly as it has been foreseen for low-energy C + , N + , and Ar + -ion implantation into PMMA [29,30].