Ion Beam Modification of Poly(methyl methacrylate) (PMMA)

Abstract

In this chapter, we review fundamental issues related to the damaging processes of PMMA films induced by high-energy ions with kinetic energies from a few keV to a few GeV, covering the regimes of energy deposition dominated by nuclear collisions and by electronic excitation. Emphasis is given to present an overview of the bond-breaking processes, the changes in the polymer chemical structure, and the corresponding modifications in selected macroscopic physical properties (optical, mechanical, and electrical).

1 Introduction

Poly(methyl methacrylate) (PMMA) is one of the most important polymeric materials available commercially. Due to its excellent mechanical properties, chemical resistance, and optical behavior similar to glass, it has been widely utilized across several industries, mainly as shatterproof windows, illuminated signs, and optical parts, but also as a structural material for different electronic, construction, and household goods [1]. The high biological compatibility made PMMA a common material for biomedical applications in ophthalmology, prosthetics, orthopedics, and dentistry, to name a few [2]. Resins based on PMMA have also been widely used in microlithography with UV photons, X-rays, or electron beams [3,4,5,6], or in direct micro-structuring processes such as proton-beam writing [7, 8]. More recently, in the growing field of proton radiotherapy [9, 10], PMMA has been used as a material for phantoms and beam-monitoring devices [11]. Therefore, the interaction of various forms of radiation with PMMA has been a persistent topic of investigation for more than four decades [3, 12,13,14]. A deep understanding of the radiation effects in PMMA is important not only to evaluate the stability and behavior of this material under various high-energy radiation fields, but also to optimize procedures in modern ion-beam-based processing techniques, where PMMA plays an important role [5,6,7,8, 15].

In this chapter, we review fundamental issues related to damaging processes of PMMA films induced by high-energy ions with kinetic energies from a few keV to a few GeV, covering the regimes of energy deposition dominated by either nuclear collisions or electronic excitation. Emphasis is given to present an overview of the bond-breaking processes and the corresponding changes in macroscopic physical properties, and their dependence on the energy loss of the ions.

2 Chemical Modification of PMMA by High-Energy Ions

Polymer degradation encompasses many reactions that change the material’s structure, such as irreversible cleavage of bonds in the main chain, intermolecular crosslinking, and formation of unsaturated bonds and of volatile products [16,17,18]. Both crosslinking and scission may occur simultaneously during irradiation of polymers, but the relative magnitude of crosslinking to scission events depends on the structure of the polymer [1], on the ion stopping power, and ion fluence [19, 20]. In this section, we review the degradation pathways of PMMA by high-energy ions, starting with a discussion of the balance between scission and crosslinking and how it is affected by the stopping power, dE/dx. We then review the radiolysis of PMMA, the main degradation compounds formed, and the structure of highly irradiated PMMA.

2.1 Chain Scission and Crosslinking

The chemical structure of the poly(methyl methacrylate) (PMMA) repeating unit is shown in Fig. 1. For the majority of irradiation conditions, ion-bombarded PMMA undergoes main chain scission, which leads to a decrease of the average molecular weight [19, 21,22,23,24]. Chain scission is also dominant in the irradiation of PMMA with γ-rays, electrons, and UV photons [25,26,27]. The propensity of degradation has been attributed to a steric hindrance due to the methyl ester (CH₃OOC–) groups attached to the PMMA backbone structure [19, 28]. Large pendant groups restrict chain mobility and thus hamper crosslinking. The presence of a tetra-substituted carbon atom in the main chain also favors degradation [29, 30].

Fig. 1

Chemical structure of the repeating unit of PMMA

Figure 2 shows the molecular weight distribution extracted from PMMA films irradiated with 200 keV He⁺ [31]. Even at low fluences (2.5 × 10¹² ions/cm²), the molecular weight distribution (MWD) of bombarded PMMA exhibits a pronounced tail at the low-mass side due to scission products. For higher fluences, the MWD continues to shift toward lower molar masses, but a small fraction of high molecular weight crosslinked molecules is also seen (Fig. 2a).

Fig. 2

Adapted from Ref. [31]

a Molecular weight distribution of 24 kDa PMMA irradiated with 200 keV He⁺. b $\left( {1/M_{n} - 1/M_{n}^{0} } \right)$ as a function of dose for samples of 95 and 24 kDa PMMA irradiated with 200 keV He⁺. A fluence of 6 × 10¹³ ions/cm² of 200 keV He⁺ corresponds to a dose of ~23 eV/g. M_n is the number-averaged molecular weight.

The solubility of irradiated PMMA increases with fluence (because of chain scission) only until a certain maximum dose (Fig. 2b). For larger fluences, solubility decreases progressively until an insoluble gel is eventually formed. Thus, even for an easily degradable polymer as PMMA, crosslinking of the chains will be extensive at large fluences. The change to a crosslinking dominant behavior occurs in the first place due to the higher concentration of macroradicals at high fluences, which favors bonding in between chains. In addition, at high fluences, PMMA is transformed to a material resembling a sort of disordered polyethylene [32], a polymer for which the probability for crosslinking is higher than for scission [21]. In general, the larger the dE/dx or LET of the ion beam, the lower is the fluence required for the onset of crosslinking in irradiated PMMA. Table 1 illustrates this for beams of H, He, and O ions of 1.5 MeV. The ability of PMMA to crosslink at high doses is also seen in electron beam irradiated PMMA, which changes from positive to negative resist behavior (insoluble in the developer) at sufficiently large exposures to the e-beam [33].

Table 1 Regions of dominance of scission and crosslinking for ion beams of various LET values

Several studies have been conducted to investigate the effect of dE/dx on the yield of chain scission and crosslinking [19, 22, 24]. This is usually quantified by the respective G-values (the number of crosslinks G(x) or chain scissions G(s) per 100 eV of absorbed energy). For irradiation with ⁶⁰Co gamma rays (dE/dx = 0.2 eV/nm), for example, G(s) ~ 1.2 [24]. Similar G(s) values were observed for low LET ion beams (Fig. 3). However, G(s) starts to decrease for beams with LET larger than about 15 eV/nm [19, 22]. G(s) values as low as 0.3 was found at LET = 577 eV/nm (90 MeV O⁴⁺) [24]. The lower yield of scission at high LET were attributed to the reduced distance between radical pairs or dangling bonds, which increases the probability for two neighboring radicals to be close enough for crosslinking.

Fig. 3

Reprinted with permission from [19]

G-value of chain scission, G(s), for PMMA as a function of LET. Irradiations were performed with different ions and velocities.

An interesting effect related to the degradation of PMMA is ion-beam-induced depolymerization (unzipping) [34]. Unzipping is the main mechanism of thermal degradation of PMMA [35] and is characterized by sequential scissioning of the chain ends, resulting in the preferential emission of the monomer methyl methacrylate [36]. The monomer is found to be the most abundant product (yield > 85%) in the pyrolysis of PMMA [21]. This phenomenon starts at the so-called ceiling temperature (around 220 °C) [34] in the conventional thermal degradation of PMMA. However, when the polymer was exposed to a beam of 200 keV He⁺ while heated, unzipping initiated at temperatures as low as 105 °C, slightly above the glass transition temperature (T_g) [34, 37,38,39,40]. Thus, irradiation of PMMA at temperatures above T_g may result in a stronger degradation rate and a pronounced reduction in the polymer molecular weight. Up to now, irradiation-induced depolymerization of PMMA has not been detected at room temperature, or using swift heavy ions.

2.2 Radiolysis, Volatiles, and Changes in the Chemical Structure

The radiolytic decomposition mechanisms of ion-beam-irradiated PMMA have been extensively investigated, using beams in a wide range of energies and stopping powers [19, 40,41,42]. While several details have yet to be clarified, especially at the very large dE/dx of swift heavy ions, the general degradation scheme has been established and is not very different from those obtained from “conventional” forms of radiation such as X-rays, electrons, and γ-rays.

A simplified scheme of the mechanism of PMMA degradation is shown in Fig. 4. High-energy irradiation is very efficient for inducing backbone chain scission (Fig. 4-path II) compared to photoexcitation process (path I). Both processes can provoke hydrogen abstraction from the polymer main chain (Fig. 4-path VI) and main chain scission (Fig. 4-path III). Besides main chain scission (and the consequent formation of unsaturated bonds), rupture and decomposition of the side-chain methyl ester groups CH₃COO– are the dominant events (Fig. 4-path IV–V). The main volatiles formed from the decomposition of the pendent group are CO, CO_2, and CH₄ [19, 32, 40, 43]. The amount of such gases will vary depending on the stability of the fragments generated, and on the type of beam.

Fig. 4

Reprinted with permission from [40]

Radiolysis pathway of PMMA and main volatile products formed.

Figure 5 shows the G-values for CO, H₂, CO₂, and CH₄ production as a function of the track average LET. Again, at low LET, the G-values for ion beams are similar to gamma rays. The G-values start to increase strongly with LET above a LET around 10 eV/nm. Since most of the analyzed gases have the same precursor (·COOCH₃), the data of Fig. 5 indicate that the amount of the methyl ester radical must increase with LET [44]. Studies of PMMA irradiated with gamma rays show that the amount of CO, CO₂, and CH₄ are also temperature-dependent [44]. CO and CO₂ yields at T = 40 °C are 2.5–3 times larger than the values found at room temperature. On the other hand, molecular hydrogen is produced from radicals that are not thermally dependent.

Fig. 5

Reprinted with permission from [44]

G-values for the production of various volatile molecules as a function of track average LET for beams of γ-rays, H, He, and C ions.

Many other small molecules originating from the breakage of the chains or from the subsequent decomposition of intermediate species are usually observed by in situ mass spectrometry. Figure 6 shows an example of an in situ residual gas analysis (RGA) of a PMMA film irradiated by 4.5 MeV/u Au ions [45], illustrating the multiplicity of emitted species. This includes H₂, CH₂, CH₃, CH₃O∙, HCOOCH_3, and unsaturated hydrocarbons, such as H₂C=CH₂, ·CH=C=CH₂, CH₂=C=CH₂, and CH₃CH=CH [40, 45, 46]. The abundant emission of oxygen-containing small molecules is a clear indication that a significant fraction of methyl ester groups is broken and easily eliminated by outgassing. The emission of numerous unsaturated hydrocarbons supports the observation of significant scissioning of the macromolecular chains [26, 32].

Fig. 6

Reprinted with permission from [45]

Outgassing spectra of PMMA irradiated with 4.5 MeV/u Au ions. Mass spectra recorded before and after irradiation.

The decomposition scheme described above is also supported by several spectroscopic studies [20, 31, 32, 40, 47,48,49,50]. In particular, infrared vibrational spectroscopy (FTIR) was widely used to probe bond breaking of irradiated PMMA. An example is given in Fig. 7, where FTIR spectra of PMMA bombarded by ions of low (500 keV He) and high dE/dx (~886 MeV Au) are shown. The reduction of the peaks associated to the vibrational modes at 1150 and 1190 cm⁻¹ (C–O stretching vibrations) and the carbonyl (C=O) band at 1720 cm⁻¹ are consistent with the gradual elimination of the pendent methyl ester groups (Fig. 7) [20, 26, 47, 48, 51]. The diminishing of carbon–oxygen bonds (C=O and C–O–C) is also shown in the XPS spectra of irradiated PMMA [27, 32, 52, 53] or in measurements of the fraction of oxygen atoms derived from RBS data [26] (Figs. 8 and 9). It is interesting to note that the ratio of oxygen to carbon in a sample irradiated with a beam of low dE/dx ions (2 MeV H⁺, dE/dx ~ 0.019 keV/nm) is smaller by ~20% than in a film irradiated with a high dE/dx ion (2.2 GeV Bi, dE/dx ~ 14 keV/nm) at a similar dose (in the case of Fig. 8 close to 100 eV/nm³) [52]. This suggests a greater importance of particle ejection unzipping by PMMA chains at high dE/dx, which would keep the O/C ratio closer to the pristine value.

Fig. 7

Reprinted with permission from [45, 54]

FTIR spectra of PMMA films bombarded by a 500 keV He and b 886 MeV Au ions.

Fig. 8

Reprinted with permission from [52]

C1s XPS spectra of PMMA films irradiated by 2 MeV H⁺ and 2.1 GeV Bi at similar deposited energy densities. The peaks associated to C atoms directly bound to oxygen are labeled in the figure. These are C₃ (O–CH₃) and C₄ (C=O). All spectra were normalized to the peak intensity of the C₁ line (C–H) at 285 eV.

Fig. 9

Reprinted with permission from [26]

Change of the relative contents of carbon and oxygen of PMMA irradiated by 350 keV H as a function of fluence, measured by Rutherford backscattering spectrometry.

FTIR [45, 47, 49], Raman [27], and UV spectroscopy [45, 49] reveal the formation of conjugated C=C bonds in the irradiated polymer. In FTIR, this is seen by the appearance of a new absorption band at 1600–1670 cm⁻¹. In one study, the number of C=C bonds generated in the films was similar to that of C=O groups removed by the same incident dose, suggesting that removal of an ester group generates also a C=C bond by main chain scission and H abstraction [40]. For low dE/dx, the efficiency of C=C bond formation is smaller than that found at high dE/dx (Fig. 7).

At high irradiation fluences, the pristine polymer chemical structure is virtually lost and FTIR spectra are almost featureless due to generalized bond breaking [21, 31, 32, 47, 49]. Carbonization is the fate of all heavily irradiated polymers, but the type of carbonaceous material that is eventually formed depends on the monomer structure and dE/dx of the ions. Most of the studies indicate that PMMA films eventually turn into a form of hydrogenated amorphous carbon material a-C:H [27, 49]. The Raman spectra of PMMA irradiated by 400 keV Cr⁺ ions (Fig. 10) show clearly the development of the characteristics broad peaks corresponding to the D and G band of disordered carbon.

Fig. 10

Reprinted with permission from [55]

Raman spectra of pristine and irradiated PMMA samples. The irradiation was performed using 400 keV Cr⁺ ions with fluences up to 5 × 10¹⁵ cm⁻². The solid lines indicate the measured Raman spectrum, while the dashed lines indicate the two Gaussian-fitted D and G bands.

Although the general scheme presented above for the radiolysis of PMMA is observed for irradiations with ions in a wide range of stopping powers, there are some effects that are only observed at very high dE/dx due to swift heavy ions. For example, GeV heavy ion irradiation of PMMA leads to the formation of various aromatic compounds [56, 57], including large mass pure carbon clusters [58] that are not observed at low dE/dx [51]. It was suggested that in the core of swift heavy ion tracks, as a result of the very high transient temperatures a gas of electrons and free carbon ions is formed. Carbon clusters and aromatic rings would form during the subsequent expansion and cooling of this highly excited region [58]. The proposed scheme is shown in Fig. 11.

Fig. 11

Reprinted with permission from [58]

Suggested mechanisms for the decomposition of PMMA and carbon cluster formation in the core of swift heavy ion tracks.

2.3 Damage Cross Sections

In the previous section, the changes introduced in the chemical structure of irradiated PMMA were reviewed. Here, we discuss the problem of the degradation efficiency and the rates of destruction (or formation) of various chemical bonds. The damaging efficiency is usually quantified extracting the damage cross section $\sigma$ for a certain process of interest. $\sigma$ is a probability of occurrence, which is interpreted as an effective area around the ion path where a certain modification takes place. For the quasi-continuous cylindrical damage track of a swift heavy ion, the geometrical interpretation of $\sigma$ is intuitive. The value of $\sigma$ can be obtained by monitoring the changes in the intensity $I\left( \phi \right)$ of a given signal (a FTIR band, UV absorbance, XRD peaks, etc.) as a function of fluence. In most of the situations, an exponential behavior is observed experimentally, which allows the extraction of a damage cross section by fitting the data with the expression:

$$ I\left( \phi \right) = I_{0} {\text{e}}^{ - \sigma \phi } $$

(1)

where $I_{0}$ is the intensity of the investigated signal for the pristine sample. If radiation induces the creation of new structures, then the signal intensity can usually be expressed as:

$$ I\left( \phi \right) = I_{\text{sat}} \left( {1 - {\text{e}}^{ - \sigma \phi } } \right), $$

(2)

where I_sat is the signal obtained at high fluences.

Table 2 provides a compilation of cross-sectional values obtained from the literature. Some of them were already extracted in the original references, while others were obtained by us from fittings of the available data. Cross sections for the processes of chain scission, loss of various functional groups, decrease in oxygen and hydrogen content, and formation of conjugated double bonds (chromophores) are shown. The magnitude of the cross sections depends on the technique used to probe the damage (e.g., chemical composition, optical properties, secondary yields) [59], and on the dE/dx and ion velocity. Hence, a direct comparison between cross sections of different processes is only meaningful when similar irradiation conditions are used in the experiments. Despite the large amount of papers on ion beam modification of PMMA, there is, unfortunately, no systematic study comparing cross sections of different processes under similar irradiation conditions, similar to those found, for example, for PET [60], PPS [61], or PC [62]. Yet, some general considerations can be drawn.

Table 2 Ion energy, velocity, electronic stopping power (dE/dX)_e, nuclear stopping power (dE/dX)_n, and cross sections (σ) extracted from ion bombardment of PMMA films

The largest damage cross-sectional values are found for processes related to changes in the polymer crystallinity and average molecular weight [60]. For these processes, either the activation energies or the required number of hits by secondary electrons to induce an observable effect is small. However, PMMA is usually in an amorphous configuration (or semi-crystalline structure with high amorphous content). Hence, there are only few studies investigating the changes induced in the PMMA crystallinity and cross sections were not extracted [50, 63]. The rates of amorphization nevertheless appear to be similar to those found for molecular weight changes described below. Fragalà et al. [21] investigated the number-averaged molecular weight M_n of PMMA irradiated by 200 keV He⁺. Figure 12 shows the values of $1/M_{n}$ as a function of fluence. The fitting (dashed line) gives a cross section of 9 × 10⁻¹⁴ cm². In another study, also with 200 keV He bombardment, a cross section of ~3 × 10⁻¹³ cm² was extracted [64]. The major difference between those two measurements was the procedure used to quantify the signal. While in Ref. [64] only the low-mass portion of the MWD chromatogram was used; Fragalà et al. [21] used the average over the complete molecular weight distribution. This example illustrates how easily cross sections may be influenced by the procedure of data quantification. In any case, both cross sections are relatively large, considering the low stopping power of the He beam. We found no study on damage cross sections for chain scission of PMMA using swift heavy ions in the literature to compare with.

Fig. 12

Reprinted with permission from [21]

Reciprocal number-averaged molecular weight $(1/M_{n} )$ of bombarded PMMA as a function of ion fluence.

Cross sections for bond breaking and formation of new bonds in irradiated PMMA are the most abundant in the literature. They typically range from ~10⁻¹⁶ cm² for light ions up to ~10⁻¹¹ cm² for swift heavy ion projectiles (see Table 2). By monitoring the changes in the intensity of the FTIR bands of PMMA irradiated with MeV light ions (5 MeV He and Li and 2.5 MeV B), Fink et al. [20] found damage cross section of the order of 10⁻¹⁶–10⁻¹⁴ cm²). This corresponds to effective track radii of chemical damage of about 0.4 Å for He or 3 Å for B. On the other hand, the damage cross section of 887 MeV Au ions is roughly three orders of magnitude larger [45]. Figure 13 shows the evolution of three different FTIR bands at 2994, 2842, and 1641 cm⁻¹ of PMMA irradiated with 887 MeV Au ions. Both processes of bond destruction and new bond formation are shown. The loss of CH₂ and CH₃ groups (bands at 2994 and 2842 cm⁻¹, respectively) occurs with a cross section of $\sigma_{{2842\,{\text{cm}}^{ - 1} }} = 7.8 \times 10^{ - 13} \,{\text{cm}}^{2}$ and $\sigma_{{2994\,{\text{cm}}^{ - 1} }} = 1.5 \times 10^{ - 12} \,{\text{cm}}^{2}$. The very large cross section obtained for the new band appearing at 1641 cm⁻¹ $(\sigma_{{1641\,{\text{cm}}^{ - 1} }} = 1.3 \times 10^{ - 11} \,{\text{cm}}^{2} )$ demonstrates the efficiency of such high dE/dx ions to create carbon–carbon double bonds, which was larger than the breaking of C–H bonds.

Fig. 13

Reprinted with permission from [45]

Peak intensity of absorption bands at 2994, 2842, and 1641 cm⁻¹ as a function of the fluence for PMMA films irradiated with 887 MeV Au ions.

Another process investigated is chromophore formation by the conjugation of unsaturated carbon bonds. This is usually extracted from UV–Vis spectra, as shown in Fig. 14. The absorbance of irradiated samples in the UVA and visible region increases with irradiation fluence. Figure 14b shows the increase in absorbance at 375 nm after irradiation with 100 MeV Si⁸⁺ ions, depicting an exponential growth typical of Eq. 2. The resulting cross section was about 5.5 × 10⁻¹³ cm².

Fig. 14

Reprinted with permission from [65]

a Optical absorption spectra of pristine PMMA and PMMA after irradiation with 100 MeV Si⁸⁺ ion beam. b Peak intensity of absorption at 375 nm of the spectra shown in (a) and its resulting damage cross section.

Overall, the various damage cross sections extracted for PMMA are comparable to those observed in other polymers, with differences in magnitude that stem from the degree of radiation stability of each material. For example, cross sections for emission of volatiles in polystyrene (a more radiation-resistant polymer) bombarded by 500 keV He, were more than 50% smaller than in PMMA, under the same experimental conditions [49]. On the other hand, heteroatom emission in ion-bombarded PVC, a highly degradable polymer, is even larger than those found for PMMA [66].

Although cross sections are important parameters quantifying the sensitivity of a polymer to radiation damage, care must be taken when comparing data probed by different techniques, because of their specific spatial sensitivity and probing signal. XPS, for example, tends to give smaller cross sections than FTIR, although both techniques are probing the amount of chemical bonds. This is related to the fact that XPS is very little affected by long-range changes in the chemical environment and gives information of modification occurring mostly close to the ion path where bond disruption is severe. In addition, the signal from spectroscopic techniques such as UV–Vis and FTIR are proportional to the optical path of the probing beam in the sample. This means that the signal may change not only because specific bonds are broken, but also due to reductions in the sample thickness. As discussed in the next section, PMMA thinning by ion irradiation is very efficient [67]. Therefore, many of the cross sections reported in the literature may be the convolution of chemical damage with effects that cause thickness reductions (sputtering, densification, etc.).

2.4 Changes in Physicochemical Properties

Bond breaking, formation of new bonds, and preferential emission of oxygen and hydrogen in bombarded PMMA result in irreversible changes in its macroscopic physicochemical properties. In this section, the main changes observed in mechanical, optical, and electrical properties of ion-bombarded PMMA are reviewed.

2.4.1 Density Enhancement and Compaction

PMMA films and foils irradiated by ion beams become remarkably thinner after ion bombardment [26, 43, 70,71,72]. The thickness decrease was mainly attributed to outgassing of volatile products [73, 74]. Sputtering is a small component, especially for light ions and low dE/dx, but cannot be neglected for swift heavy ions. The crater volume of a single 600 MeV Au ion is approximately a half sphere of 10 nm radius [75, 76]. In addition, main chain scission results in shorter and more mobile molecules, what facilitate spatial rearrangements and local conformational changes. This eventually leads to the compaction of the layers and increase in density [77, 78].

Figure 15 presents the compaction induced in 1-mm-thick PMMA foils irradiated with 350 keV H⁺ (the ion range is 4.2 μm) as a function of the fluence. The compaction reached values as high as few micrometers at high fluences. The shrinking rate was not steady, but decreased with fluence, until a saturation level is reached at very high fluences [67, 71, 78]. This occurs because of the cumulative effects of track overlap. At large fluences, the probability of ions hitting regions already modified by the ions increases, reducing the gas yield per ion impact.

Fig. 15

Reprinted with permission from [26]

Compaction of a 1-mm-thick PMMA foil as a function of fluence of 350 keV H⁺ ions.

Swelling, however, may precede the shrinking process. Due to the long projected range of swift ions, gaseous products are formed in deep regions of the polymer and cannot escape from the material as quick as they are produced. As a consequence, a high pressure of gases is built up in the polymer, resulting in a transiently swollen surface [74]. In a foil of PMMA irradiated through a micro-patterned mask with 1.8 MeV H⁺ (ϕ = 6 × 10¹³ ions/cm²), within a time of 12 min after the irradiation the surface swollen about 70 nm with respect to the non-irradiated region (Fig. 16a). Only later, 200 min after the irradiation, profilometry of the same region revealed that the surface retracted to more than 200 nm below the original surface level (Fig. 16b). It was noted by the authors that if the ion current was low enough, swelling was not observed.

Fig. 16

Reprinted with permission from [74]

Surface profiles of the irradiated sample measured by profilometry, a 12 min after the irradiation and b after 200 min.

A discrepancy between compaction values is found in the literature, even when similar irradiation conditions are used [43, 74]. These differences can be explained by the distinct times after irradiation that the depth profile was measured. The shrinking due to outgassing is diffusion-limited and may continue over several weeks if the effective diffusion coefficient, D_eff is low (Fig. 17). At room temperature, the D_eff extracted for PMMA after irradiation with 1.8 MeV H⁺ ions to a fluence of 6 × 10¹³ cm⁻² was 8.3 × 10⁻¹¹ cm²/s. Gas diffusion is accelerated at higher temperatures, increasing the shrinking rate. At temperatures just below the glass transition, T_g (~105 °C), the effective diffusion constant found by Schrempel et al., is about six orders of magnitude higher than the value found at room temperature [74]. At 130 °C, thermal degradation also plays an important role, being gas diffusion no longer the predominant effect.

Fig. 17

Reprinted with permission from [74] by Publisher

Shrinking of the surface (z_l) of 1-mm-thick PMMA foils bombarded with 1.8 MeV H⁺ ions as a function of time t after irradiation to a fluence of 6 × 10¹³ ions/cm². Measurements were performed at different temperatures T_T ranging from 23 to 130 °C. The solid lines are fittings with a diffusion model.

2.4.2 Optical Properties

There has been a great practical interest on the investigation of optical properties of irradiated PMMA, because of the wide range of potential photonic applications and the widespread use of PMMA in optical components. Proton-beam writing has been used for the fabrication of microphotonic devices (waveguides, gratings, microlens arrays, etc., Fig. 18). For such applications, knowledge on ion-induced modifications in the absorption coefficient and refractive index are of utmost importance.

Fig. 18

Adapted from [79]

a Optical image of waveguides fabricated in 3-mm-thick PMMA. Electron microscopy images of a b surface relief gratings fabricated in 800 nm layer of PMMA spin coated on a Si wafer and c a microlens array.

As discussed in the previous section, the damage produced by ion irradiation leads to the formation of absorption centers in PMMA at the UV and visible regions of the spectra due to the increase in conjugated C=C double bonds, which provides more optically active electrons. This process changes the polymer color gradually from transparent to yellowish-brown. The color change is usually irreversible, but there are reports of transparency recovery depending on the beam conditions and storage environment [28].

Accompanying the changes in UV absorption, there is an increase in the refractive index with ion fluence (Fig. 19) [47, 71, 72, 80, 81]. The refractive index enhancement was attributed as well to the formation of unsaturated bonds along the irradiated layer [28]. Compaction and densification of the irradiated material are also associated to the increase in the refractive index [77]. It is remarkable in the study of Kallweit and Biersack [80] that $\Delta n$ increases steadily with fluence, with no saturation yet visible (Fig. 20). The relatively large changes in $\Delta n$ of up to 0.3 is an interesting feature for waveguide fabrication.

Fig. 19

Reprinted with permission from [80]

Change in refractive index, $\Delta n$, of PMMA irradiated by N ions with kinetic energies between 100 and 350 keV. a $\Delta n$ versus ion fluence for 300 keV N⁺. b $\Delta n$ versus ion energy (constant fluence of 2 × 10¹⁴ ions/cm²).

Fig. 20

Reprinted with permission from [81]

Change in refractive index at 633 nm of PMMA irradiated by 2 MeV H⁺ as a function of penetration depth. For comparison, the energy-loss profile was also plotted as a solid line.

Several experiments were performed to investigate the correlation between the ion range and the changes of $\Delta n$ with depth. When 2 MeV H⁺ was used to bombard a block of PMMA (Fig. 21), the highest increase of the refractive index occurs at the end of range, i.e., at the Bragg peak of the energy-loss curve [81].

Fig. 21

Reprinted with permission from [70]

Intensity loss at 632.8 nm of a PMMA waveguide fabricated using 2 MeV H⁺ at fluence of ~50 nC/mm².

The use of irradiated PMMA in waveguides was also tested. The main problem was the optical losses. The measured intensity attenuation of waveguides fabricated using a 3-mm-thick foil irradiated by 2 MeV H⁺ at fluence of ~50 nC/mm² is shown in Fig. 22. In this case, the mean attenuation coefficient was found to be 1.4 ± 0.2 dB/cm [70]. Propagation losses are, additionally, very dependent on the ion dose [72]. For PMMA blocks irradiated with Li⁺ ions in the energy range of 100–130 keV with fluences up to 10¹⁴ cm⁻², waveguides with total loss values of <2 dB/cm could be fabricated [47]. At higher fluences, because of ion-beam-induced damage, the propagation loss increases by more than a factor of two. Optical losses are also a function of the beam energy. Other studies show that attenuation measurements of waveguides obtained after 1–2 MeV proton irradiation showed loss values as high as 5 dB/cm up to 8 dB/cm, depending on the beam conditions [71].

Fig. 22

Reprinted with permission from [47]

Loss value of irradiated PMMA waveguides for different fluences of Li⁺ ions in the energy range of 100–130 keV.

Another complication of using ion irradiation to prepare waveguides is aging. In one study, long-term stability of n was investigated in PMMA bombarded by 1 MeV He⁺ ions and stored at 70 °C in vacuum or at ambient conditions (Fig. 23). Both samples show a decrease of the initial index of refraction during the first 60 days and become stable after 3 months. The aging effect was more pronounced for the samples stored at ambient conditions [71].

Fig. 23

Reprinted with permission from [71]

Long-term stability of the refraction index of PMMA samples irradiated with 1 MeV He⁺ ions.

2.4.3 Mechanical Properties

Figure 24 shows the results of the changes of flexural strength of 3-mm-thick sheets of PMMA irradiated by 45 and 30 MeV proton beams and γ-rays as a function of the absorbed dose. The decay in flexural strength is similar for all types of such low LET radiations (LET ~ 1 eV/nm for the three beams). It is well known that such a decrease of the strength is induced by chain scission of PMMA [82].

Fig. 24

Reprinted with permission from [82]

Flexural strength of PMMA as a function of dose (○ gamma ray; □ 45 MeV proton; ■ 30 MeV proton).

Hardness, scratch resistance, and elastic modulus of the polymer are also affected by crosslink and chain scission [23, 83]. Figure 25 shows averaged values of indentation elastic modulus versus maximum indentation depth for bulk PMMA implanted by 40 keV B⁺ ions. The maximum penetration depth of B⁺-ions into PMMA is about 400 nm. However, implanted PMMA showed an increased elastic modulus with respect to the pristine material at much larger depths (up to 1100 nm [84]).

Fig. 25

Reprinted with permission from [84]

Indentation elastic modulus versus maximum indentation depth for un-implanted PMMA (squares) and PMMA implanted by 40 keV B⁺ ions (circles) samples.

Figure 26 illustrates the changes in hardness of PMMA irradiated by 2 MeV Ar ions (LET = 1134 eV/nm) and 2 MeV He ions (LET = 242 eV/nm) as a function of ion fluence. Hardness increased significantly with increasing fluence, especially for the larger dE/dx Ar beam, where it jumped from the pristine value of 0.45 GPa to over 10 GPa after irradiation to a fluence of 1 × 10¹⁵ cm⁻². The improvement in hardness was greater at larger dE/dx [19]. In fact, under low LET radiations, such as e-beam and γ-rays, the material becomes extremely brittle, but there is no appreciable change in hardness [19]. However, it has been shown that at high electron doses (7.5 × 10¹⁷ cm⁻²), the surface hardness of the irradiated PMMA increased to 2.8 GPa, and the elastic modulus from 5.6 to 22.9 GPa [27].

Fig. 26

Reprinted with permission from [19]

Hardness changes as a function of fluence for PMMA irradiated by 2 MeV He⁺ and Ar⁺ ions.

As hardness of polymers increases with increasing crosslinking density [83], this difference in mechanical behavior of PMMA irradiated by ions of low and high LET reinforces that scission is dominant for low LET and low fluences, whereas crosslinking becomes more important with increasing LET and at high ion fluences.

2.4.4 Changes in Electrical Properties

In general, polymers are highly insulating materials with wide band gaps [28]. Ion irradiation or implantation can be used to increase the polymer conductivity by several orders of magnitude. The possibility of preparing conducting regions in a highly insulating polymeric matrix was one of the first exciting effects of ion bombardment of polymers to be explored [85]. Figure 27 presents data on the electrical conductivity of irradiated PMMA as a function of 400 keV Cr⁺ ions. An increase in electrical conductivity from ~2 × 10⁻¹⁰ S/cm (pristine) to ~7 × 10⁻⁶ S/cm (at 5 × 10¹⁵ ions/cm²) was observed. In another study, using beams of 150–200 keV Si⁺ ions at much larger fluences (3.2 × 10¹⁶ Si⁺/cm²), the conductivity reached values of almost 11 orders of magnitude higher than that of pristine PMMA [86].

Fig. 27

Reprinted with permission from [55]

Electrical conductivity of 400 keV Cr⁺-implanted PMMA versus ion fluence. The line was drawn to guide the eye.

The conductivity increase reflects the carbonization process discussed in the previous sections. Conductivity rises sharply only after the material has been extensively damaged by the beam and significant amounts of carbon clusters are formed in the material. The formation of a network of carbon clusters provides a charge transportation system in implanted PMMA [55, 85]. In the early work at the Bell Laboratories [85, 87], 2 MeV Ar⁺ was used to irradiate several nonpolymeric and polymeric organic compounds, including PMMA. The temperature dependence of the resistivity $\rho \left( T \right)$ of the irradiated material was found to follow $\rho \left( T \right) \propto { \exp }\left[ { + \left( {T_{0} /T} \right)^{1/2} } \right]$ over a wide range of fluences and temperatures. Based on this dependence, it was proposed that conduction of electricity at intermediate irradiation fluences involves hopping of carriers between isolated conducting islands. At very high fluences, (>10¹⁶ cm⁻²), conducting islands interconnect and the resistivity becomes temperature-independent. In general, ion bombardment of organic thin films leads to the production of a carbon-enriched material with some properties similar to those found for amorphous carbon, but with electrical characteristics that are unique [87]. Conductivity also depends on the initial monomer structure. The value found for heavily irradiated PMMA (9.5 × 10⁻⁴ S/cm, after 10¹⁶ ions/cm² of 1 MeV Xe⁺) was smaller than for other polymers like PS (1 × 10⁻⁴ S/cm) and PE (1.5 × 10⁻⁴ S/cm) irradiated under similar conditions [54].

3 Concluding Remarks

In this chapter, fundamental issues related to damaging processes of PMMA induced by high-energy ions were reviewed. The different steps in the polymer modification, from the identification of the main bond-breaking pathways and chemical changes to the corresponding modifications in the macroscopic physical properties are reasonably well established, and are not very different from those obtained from “conventional” forms of radiation, for which the available data is much more abundant. The similarities are more evident for low LET particles, such as fast H or He ions. In this regime, G-values for X-rays, γ-rays, electrons, and ions are similar and the mean dose is the main parameter determining the level of the final effect (chain scission, hydrogen release, etc.). However, several details of the complex chain of events induced by high-energy ions in PMMA have yet to be clarified, especially at the very large dE/dx of swift heavy ions (SHI).

In fact, already at intermediate dE/dx (e.g., of 1 MeV Ar, ~800 eV/nm), the dose/response curve might become nonlinear. Certain degradation products are clearly produced at much higher yields at large dE/dx. This includes the formation of triple bond carbon clusters, and sputtering of macromolecular fragments. Thus, although carbonization is the fate of all heavily irradiated polymers, the type and properties of the carbonaceous material that is eventually formed will depend on the dE/dx of the ions. The balance between scission and crosslinking and ion-beam-induced unzipping are also expected to alter at large dE/dx. Systematic studies aiming a better understanding of transformations specifically induced by energetic heavy ions are necessary. Even the basic effect of the changes in M_w distribution of PMMA after SHI irradiation is lacking in the literature. Of the major physical properties of materials, thermal property is one of the least investigated after ion irradiation of polymers, including PMMA. Studies on the changes in thermal properties like thermal expansion or heat conductivity are practically absent in the literature and need to be conducted. There are scattered data on damage cross sections for bond-breaking of PMMA in the literature. However, there is no systematic study comparing cross sections of different modification processes (chain scission, bond breaking, chromophore formation, carbon cluster formation, etc…) under the same irradiation conditions. This will allow a detailed view of the spatial (radial) distribution of different types of structures/defects created around the ion track.

The effect of irradiation temperature is also worthy of further investigation. We have shown examples where temperature alters the effects produced by ion irradiation of PMMA (e.g., in irradiation-induced depolymerization or in the kinetics of volatile emission). However, it is unclear at present how temperature alters the profile of intermediate products produced by ion irradiation or the final properties of the bombarded material. It is expected that, near and above T_g, all processes which are diffusion-controlled will be enhanced. The temperature effects may also be different for ions of high and low dE/dx, as recent studies on the surface morphology of ion-irradiated PMMA thin films suggest [88]. Irradiations at low temperatures with subsequent in situ annealing and spectroscopic characterization will allow obtaining valuable information on the creation and stability of different chemical groups formed by irradiation. This has been explored in a few polymers, like polyethylene, but not in PMMA.

The vast majority of irradiation experiments are performed in vacuum. It would be interesting to study the effect of ion irradiation in different atmospheres, including oxidative environments, and how they affect the yield and type of degradation products.

Additionally, the influence of the sample conditions themselves needs to be carefully addressed. For example, there is no systematic investigation comparing irradiation effects on samples prepared from monodisperse standards in a broad range of molecular weights. How much the molar mass influence radiation chemistry and properties changes induced by energetic ions? PMMA is also used in the form of composites with various fillers or polymer blends. A systematic investigation of the ion beam modification of such systems is also scarce and needs to be examined in closer detail.

Overall, in spite of the importance of PMMA as an engineering material, the volume of data available on ion modification of PMMA is still relatively limited, compared to other important polymers such as polystyrene, polyethylene, or poly (ethylene terephthalate). Thus, PMMA needs for sure an expanded and reliable database, built from “targeted” experiments with well-controlled irradiation and sample conditions to help answering some of the open issues raised above.