Effect of Free Radicals on Irradiation Chemistry of a Double-Coordination Organotin (Sn 4 ) Photoresist by Adjusting Alkyl Ligands

Open AccessCCS ChemistryRESEARCH ARTICLES24 Jan 2024Effect of Free Radicals on Irradiation Chemistry of a Double-Coordination Organotin (Sn4) Photoresist by Adjusting Alkyl Ligands Hao Chen, Yifeng Peng, Haichao Fu, Fuping Han, Guangyue Shi, Feng Luo, Jun Zhao, Danhong Zhou, Pengzhong Chen and Xiaojun Peng Hao Chen State Key Laboratory of Fine Chemicals, Frontiers Science Center for Smart Materials, School of Chemical Engineering, Dalian University of Technology, Dalian 116024 , Yifeng Peng State Key Laboratory of Fine Chemicals, Frontiers Science Center for Smart Materials, School of Chemical Engineering, Dalian University of Technology, Dalian 116024 , Haichao Fu Research and Development Center of Valiant Co., Ltd., YEDA, Yantai 264006 , Fuping Han State Key Laboratory of Fine Chemicals, Frontiers Science Center for Smart Materials, School of Chemical Engineering, Dalian University of Technology, Dalian 116024 , Guangyue Shi School of Materials Science and Engineering, Nankai University, Tianjin 300350 , Feng Luo School of Materials Science and Engineering, Nankai University, Tianjin 300350 , Jun Zhao Shanghai Advanced Research Institute, Chinese Academy of Sciences, Shanghai 201203 , Danhong Zhou State Key Laboratory of Fine Chemicals, Frontiers Science Center for Smart Materials, School of Chemical Engineering, Dalian University of Technology, Dalian 116024 , Pengzhong Chen *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] State Key Laboratory of Fine Chemicals, Frontiers Science Center for Smart Materials, School of Chemical Engineering, Dalian University of Technology, Dalian 116024 and Xiaojun Peng *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] State Key Laboratory of Fine Chemicals, Frontiers Science Center for Smart Materials, School of Chemical Engineering, Dalian University of Technology, Dalian 116024 State Key Laboratory of Fine Chemicals, College of Materials Science and Engineering, Shenzhen University, Shenzhen 518060 https://doi.org/10.31635/ccschem.023.202303616 SectionsSupplemental MaterialAboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack Citations ShareFacebookTwitterLinked InEmail Metal oxide cluster (MOC) photoresists are highly promising materials for the next generation of extreme ultraviolet lithography (EUVL). The consecutive exploration of novel MOC materials and their structural irradiation chemistry are the major concerns associated with EUVL. Herein, we report two bicoordinated tin-oxo clusters (TOCs), the organic ligands of which contain both adamantane carboxylic acids and alkyl groups (methyl: Sn4–Me–C10; butyl: Sn4–Bu–C10). We explore the correlation between the structures of the TOCs and their patterning properties by adjusting the alkyl groups coordinated to the Sn atom. The structural variation causes different irradiation chemistry, with Sn4–Me–C10 exhibiting improved resolution and Sn4–Bu–C10 demonstrating higher sensitivity. These differences are attributed to the bonding energies of the Sn-methyl and Sn-butyl groups, the size of the resulting alkyl radicals, and their reaction probabilities. Both clusters occur in the reactions of Sn–C bond cleavage and the decarboxylation of adamantane carboxylic acids upon irradiation. However, the entire process exhibits distinct characteristics. Based on the electron-beam lithography and other experiments, we proposed irradiation-induced reaction mechanisms for both clusters. The Sn4–Bu–C10 cluster predominantly undergoes alkane chain linkage, whereas the Sn4–Me–C10 cluster mainly follows the adamantanes linkage pathway. Download figure Download PowerPoint Introduction Over the past two decades, advancements in photolithographic patterning tools and corresponding photoresist materials have played an important role in downscaling device feature sizes in integrated circuits (ICs).1,2 Currently, extreme ultraviolet (EUV) lithography (EUVL) is the most prevalent technique and gradually replacing deep ultraviolet (DUV; 193 nm) lithography for high-volume IC manufacturing.3–5 EUVL enables the achievement of a theoretical limit for sub-10 nm patterning resolution by employing a short exposure wavelength (13.5 nm). This high resolution is heavily dependent on photoresist materials with tailored properties compatible with the high energy of EUV photons (92.5 eV).6 However, the use of conventional chemically amplified resists (CARs) in EUVL faces limitations such as low EUV sensitivity, large component sizes, and poor etching resistance, causing them to fall short of the manufacturing requirements for high-resolution patterning.7 Therefore, the development of novel photoresist materials that exhibit optimal EUV response is crucial with respect to the long-term success of EUVL.8,9 Unlike the photochemistry observed in DUV lithography, where the photoacid generator is selectively excited by UV light, the highly energetic EUV photons far surpass the ionization threshold and can ionize valence and inner-valence electrons.10–13 The resulting photoemitted electrons scatter through the material and transfer their energy to initiate reactions in the surrounding material.14–16 Considering the strong resonance to EUV light for some elements, such as Sn,17–20 Zn,21,22 and Hf,23,24 incorporating these inorganic components into organic resist formulations is a widely accepted strategy for developing novel EUV photoresist materials.25,26 In particular, tin-oxo clusters (TOCs) have attracted considerable attention because of their most promising application potential in the industry.27,28 TOCs comprise a tin–oxygen inner core coated with organic ligands on its surface and can be synthesized with precise atomic-scale control through the hydrolysis and condensation of alkyl-tin precursors. TOCs exhibit superior characteristics and figures of merit than conventional CARs. These include well-defined molecular structures, metallic centers with high EUV absorption density, good etching resistance, and improved critical dimensions (CDs) resulting from small and homogeneous geometrical sizes. Recent studies on TOCs have mainly focused on two types of complexes: butyl-Sn1229,30 and butyl-Sn13 (butyl-tin Keggin);31,32 these complexes patterning performance under electron irradiation have been well studied. Furthermore, these TOCs undergo ligand cleavage from the inner core upon EUV irradiation, leading to the generation of reactive radicals that form organometallic polymers with a solubility change compared with the unexposed photoresist films.33–36 Previously proposed mechanisms elucidate the basic performance of organotin resists. However, the effect of ligand characteristics and the tin–oxygen core configuration on the aforementioned irradiation-induced chemistry remains to be fully elucidated. To further gain an in-depth understanding of their structure–function relationships, it is imperative to diversify relevant studies and investigate the irradiation-induced chemistry of TOC materials. Herein, we report two dimeric organotin derivatives, namely Sn4–Me–C10 and Sn4–Bu–C10, and investigate their patterning performance under EUVL and electron-beam lithography (EBL) irradiation. Both complexes contain four Sn atoms bound to four adamantane carboxylic acids through Sn–O bonds. In addition, each Sn atom is bound to two methyl groups in Sn4–Me–C10 and two butyl groups in Sn4–Bu–C10. Single-crystal X-ray diffraction analysis reveals a small size of approximately 2 nm for these TOCs. Density function theory (DFT) calculations show that the Sn–C bond energy in Sn4–Me–C10 is higher than that in Sn4–Bu–C10. Therefore, the former exhibits considerably lower irradiation sensitivity than the latter as well as different patterning behaviors in the EBL and EUVL experiments. Through electron paramagnetic resonance (EPR), thermogravimetry analysis-mass spectrometry (TGA-MS), and X-ray photoelectron spectroscopy (XPS), we elucidate the mechanism of the irradiation-induced reaction. Sn4–Me–C10 and Sn4–Bu–C10 undergo distinct reaction processes, resulting in different compositions in the final film matrix. This mechanism will contribute to the understanding and rational design of such EUV photoresist materials. Experimental Methods General procedure for the synthesis of Sn4–Me–C10 and Sn4–Bu–C10 The organotin oxide, R2SnO (6.05 mmol; 1.00 g for R = Me; 1.50 g for R = Bu), and 1-adamantane carboxylic acid (6.05 mmol; 1.05 g) were suspended in 150 mL of toluene in a round-bottom flask equipped with a Dean–Starck trap and a water-cooled condenser.37 The reaction mixture was refluxed for 10 h until complete dissolution of the R2SnO precursor. The clear reaction mixture was then evaporated under reduced pressure, and the resulting residue was recrystallized from toluene. After 2 days at 25 °C, colorless crystals were obtained. Sn4–Me–C10 ([Me2Sn(C11H15O2)]2O)2: White solid (1.28 g; 83%; m.p. 276–280 °C). As shown in Supporting Information Figure S1, 1H NMR (CDCl3) δH (ppm): 1.99 (s, 3H), 1.80 (s, 6H), 1.68 (m, 6H), 0.78 and 0.70 (s, 6H). 13C NMR (CDCl3) δC (ppm): 184.95, 41.18, 39.46, 36.67, 28.20, 9.65, 6.88. 119Sn NMR (CDCl3) δSn (ppm): −190.96 and −194.35. Elemental analysis: Anal. Calcd for C52H84O10Sn4: C, 46.47; H, 6.30; O, 11.90; Sn, 35.33. Found: C, 46.69; H, 6.23; O, 12.03; Sn, 35.13. Sn4–Bu–C10 ([Bu2Sn(C11H15O2)]2O)2: White solid (1.52 g, 79%; m.p. 243–246 °C). As shown in Supporting Information Figure S2, 1H NMR (C6D6) δH (ppm): 1.99 (s, 3H), 1.84 (s, 6H), 1.70 (m, 6H), 4.68-1.51 (m, 6H), 1.42–1.28 (m, 6H), 0.92 and 0.90 (s, 6H). 13C NMR (C6D6) δC (ppm): 184.75, 41.94, 40.25, 37.03, 28.75, 28.13, 27.70, 27.42, 14.10, 13.88. 119Sn NMR (C6D6) δSn (ppm): −219.66 and −223.86. Elemental analysis: Anal. Calcd for C76H132O10Sn4: C, 54.31; H, 7.92; O, 9.52; Sn, 28.25. Found: C, 54.15; H, 7.95; O, 9.65; O, 12.03; Sn, 28.11. Spin-coating procedures The aforementioned reagents (10 mg/mL) were dissolved in chloroform via ultrasonication for 10 min. The precursor solutions were filtered through a 0.22-μm poly(tetrafluoroethylene) syringe filter and then spin-coated onto 1 × 1 cm2 polished wafers at 6000 rpm for 30 s, followed by baking at 70 °C for 1 min in air. Contrast curves The prepared films were patterned using a eLINE Plus EBL from Germany Raith company. Arrays of 8 × 8 squares were exposed at an accelerating voltage of 1 kV using minimum and maximum doses of 10 and 640 μC/cm2, respectively. After exposure, the samples were subjected to post-exposure baking at 100 °C for 1 min, followed by development in a solution of isopropyl alcohol: deionized water (DIW) = 3:1 for 20 s and rinsing in DIW for 5 s. Finally, these samples were hard baked at 100 °C for 1 min. The resulting thickness of each square was measured using atomic force microscopy (AFM) and plotted as a function of the dose to determine the contrast curves. A fitted line was drawn between the normalized thickness values of 0.2 and 0.8, which was extrapolated to determine the two important values D0 and D100. D0 is the value-fitted line crossing the x-axis, representing the highest dose at which the material can still be completely washed away during development, while D100 is the value-fitted line crossing the y-value of 1, representing the lowest dose at which all the material becomes completely insoluble after development. The contrast γ is calculated using eq 1. γ = ( log 10 D 100 D 0 ) − 1 (1) The contrast γ describes the transition from the exposed area to the unexposed area. The higher the contrast of the photoresist, the steeper the side walls and the higher the aspect ratio of the patterns after development. Patterned lines L/3S dense lines (100 nm) were obtained using EBL with doses ranging from 40 to 640 μC/cm2 for Sn4–Me–C10 and Sn4–Bu–C10. Line/space (L/S) lines were also patterned by EUVL (13.5 nm) at the Shanghai Synchrotron Radiation Facility (SSRF), with exposure doses ranging from 3.7 to 149.85 mJ/cm2. The same development procedure as that described in the "Contrast Curves" section was performed. The CD was measured using scanning electron microscopy (SEM), and the line edge roughness (LER) (3σ) was calculated using SEM-measured image lines estimator (SMILE).38,39 Results and Discussion TOCs have been widely studied in the field of EUVL over the two past decades.17,27,28 However, the EUV performance of bicoordination Sn4 clusters has seldom been reported. These clusters can be prepared via hydrolysis and condensation of organotin oxide (R2SnO) precursors. In this study, we synthesized atomically precise Sn4–Me–C10 and Sn4–Bu–C10 to investigate the influence of different Sn-alkyl bonds on the chemistry induced by irradiation (Figure 1). Furthermore, the use of an adamantane framework with a high C/H ratio implied good dry-etching resistance,40 and the steric effect of this framework facilitated the formation of smooth spin-coated films. The single-crystal structure of Sn4–Me–C10 revealed a small size of approximately 2 nm for such Sn4 complexes ( Supporting Information Figure S3). These structural characteristics were beneficial for the formation of uniform films with a thickness of approximately 40 nm and roughness (Rq) of 0.3 nm for Sn4–Me–C10 and 0.6 nm for Sn4–Bu–C10 ( Supporting Information Figure S4). Figure 1 | Schematic structure of Sn4–Me–C10 and Sn4–Bu–C10. Download figure Download PowerPoint Contrast curves and patterned periodic lines The butyl-liganded Sn4–Bu–C10 exhibited high sensitivity with a D0 of 4.05 μC/cm2 and D100 of 123.94 μC/cm2 (Figure 2a and Supporting Information Figure S5). Its irradiation sensitivity was considerably higher than that of most reported single-butyl-liganded Sn clusters.17,32,34 The methyl-liganded Sn4–Me–C10 exhibited higher D0 and D100 values of 77.76 and 392.93 μC/cm2, respectively (Figure 2b). DFT calculations showed that the bond energies of the Sn-methyl bond and the Sn-butyl bond were 52 and 40 kcal/mol, respectively ( Supporting Information Figure S6). Thus, the Sn-butyl bond was susceptible to cleavage under irradiation, thereby generating highly reactive species and inducing further photolithographic reactions. However, this high irradiation sensitivity for Sn4–Bu–C10 resulted in low contrast γ values, with 1.42 for Sn4–Me–C10 and only 0.67 for Sn4–Bu–C10. In addition, the small structural difference between the two complexes leads to different patterning performances. Figure 2 | Contrast curve of (a) Sn4–Bu–C10 and (b) Sn4–Me–C10. (c) SEM and AFM images of pattern lines of Sn4–Me–C10 and Sn4–Bu–C10 with or without benzoquinone at a dose of 640 μC/cm2. (d) CD of pattern lines of Sn4–Bu–C10 and Sn4–Me–C10 rose in the wake of the increasing dose. (e) LER (3σ) of pattern lines of Sn4–Me–C10 with or without benzoquinone decreased with increasing dose. Download figure Download PowerPoint An L/3S pattern of 100-nm density was used as a template in EBL at 1 kV, with exposure doses ranging from 40 to 640 μC/cm2 for Sn4–Me–C10 and Sn4–Bu–C10 (Figure 2c and Supporting Information Figures S7 and S8). The line CD for both clusters increased with increasing dose (Figure 2d). In particular, the lines of Sn4–Bu–C10 considerably broadened from 202 nm at 200 μC/cm2 to 308 nm at 640 μC/cm2 because of the fragile Sn-butyl bond that readily generated free radicals and triggered exposure reactions. The low-energy electrons diffusing in the film could readily cause Sn4–Bu–C10 to excite and react, thereby broadening the pattern lines. Conversely, the pattern lines of Sn4–Me–C10 barely broadened with increasing dose, measuring 99 and 114 nm at 480 and 640 μC/cm2, respectively. This can be attributed to the strong Sn-methyl bond, which was less likely to generate free radicals; further, the low-energy electrons in the film were difficult to induce excess exposure reactions. The LER (3σ) of Sn4–Me–C10 pattern lines was also determined using SMILE. The LER (3σ) decreased with increasing dose, measuring 7.26 and 13.6 nm at 640 and 480 μC/cm2, respectively (Figure 2e and Supporting Information Figures S7). This trend is analogous to the trends observed in the case of CARs, where CD increases and LER decreases with increasing dose. The decrease in LER can be attributed to the free radicals generated via the breakage of the Sn–C bond moving through the film, which is similar to H+ in the case of CARs. The increasing migration of free radicals with increasing dose compensates for the inhomogeneity of the pattern line edges but simultaneously increases the CD. However, the LER (3σ) for Sn4–Bu–C10 could not be calculated because of the excessively large CD. Benzoquinone (BQ), a common inhibitor, can be combined with radicals to inhibit free radical reactions. The addition of 1% BQ to the photoresist formulation increased D100 and contrast γ for Sn4–Me–C10 ( Supporting Information Figure S9). The CD of Sn4–Me–C10 remained almost constant, while LER (3σ) increased from 7.26 to 11.9 nm when BQ was added at 640 μC/cm2 (Figure 2c and Supporting Information Figure S10). Interestingly, for Sn4–Bu–C10, the addition of BQ enhanced its reactivity and further expanded the pattern lines to a mass (Figure 2c and Supporting Information Figure S11). Sn4–Me–C10 and Sn4–Bu–C10 underwent different reaction processes in the presence of BQ. Therefore, the relationship between the structure and patterning properties of photoresists was subtle; even tiny structural variations could produce considerably different photolithography patterns. The EUVL (13.5 nm) patterning performance of Sn4–Me–C10 and Sn4–Bu–C10 was further investigated at SSRF. A 100-mL precursor solution was dropped onto 2-in. wafers at the same rotational speed and time to obtain ultrasmooth and uniformly thick films ( Supporting Information Figure S12). Similar to the observation in the case of EBL, the pattern lines for Sn4–Bu–C10 broadened with increasing EUV exposure doses (Figure 3a,b). At a minimum dose of 40.7 mJ/cm2, 51.9 nm L/S pattern lines were obtained with an LER of 7.67 nm. As the dose increased to 98.1 mJ/cm2, the L/S pattern lines reached 65.9 nm. For Sn4–Me–C10, distinct lines could not be obtained, possibly owing to the higher bond energy of the Sn-methyl bond, which passivated the exposure reaction under the experimental conditions. Figure 3 | SEM images of Sn4–Bu–C10 after EUV exposure at doses of (a) 40.7 mJ/cm2 and (b) 98.1 mJ/cm2. Download figure Download PowerPoint Composition and structure of the exposure product EPR measurements demonstrated the production of alkyl radicals for both Sn4–Me–C10 and Sn4–Bu–C10 under irradiation. DMPO (5,5-dimethyl-1-pyrroline N-oxide) was used to trap alkyl radicals, forming DMPO-R. For Sn4–Me–C10, the EPR signal emerged as a combination of two sets of peaks (Figure 4a). The simulation results of the signals closely matched the experimental results. The DMPO-R signal was identified with α N = 20.37 G and α H β = 15.24 G, while the DMPOX signal was attributed to the DMPO oxidation by free radicals. Furthermore, no EPR signal was observed in the absence of Sn4–Bu–C10 (Figure 4b). When a methanol solution of Sn4–Bu–C10 was irradiated for 30 min, the observed radical adducts were assigned to DMPO-OR. We speculated that the ·OR free radicals were generated from methanol, which was oxidized by the more oxidizing and reactive butyl radicals. These results confirmed the formation of alkyl radicals through homogeneous cleavage of the Sn–C bond upon irradiation. Figure 4 | EPR signal of the mixture containing cluster resist and DMPO after UV radiation of (a) Sn4–Me–C10 and (b) Sn4–Bu–C10. TGA-MS of (c) Sn4–Me–C10 and (d) Sn4–Bu–C10. The gray dashed line represents the change in sample mass with temperature. Download figure Download PowerPoint TGA-MS was conducted to investigate the thermal stability of Sn4–Me–C10 and Sn4–Bu–C10 as well as the species generated from the irradiation-induced reaction; this investigation was helpful for determining the exposure mechanism. TGA results showed that Sn4–Me–C10 and Sn4–Bu–C10 remained thermally stable at 250 °C (Figure 4c,d). As the temperature increased further, the sample weight began to decrease because of the dissociation of organic ligands. At 550 °C, approximately 20 wt % of the sample was retained, probably derived from the Sn–O nucleus. Sn4–Me–C10 exhibited a CH4 peak at a mass-to-charge ratio (m/z) of 16, resulting from the thermally induced cleavage of the Sn–C bond accompanied by hydrogen addition to the methyl group, leading to the formation of CH4. Sn4–Bu–C10 showed an m/z of 56, corresponding to 1-butene, which resulted from hydrogen abstraction from the butyl radical. At 280 °C and 390 °C, both clusters exhibited m/z = 135 and 136, respectively, which were assigned to adamantane. In addition, a peak was observed at m/z = 44, which could be attributed to CO2. These results suggest that the ligand 1-adamantane carboxylic acids can dissociate from the Sn–O nucleus, accompanied by a decarboxylation process to generate CO2 and adamantane with reactive bridgehead carbon. To gain further insights regarding the exposure mechanisms under irradiation, C 1s, O 1s, and Sn 3d XPS of Sn4–Me–C10 and Sn4–Bu–C10 were investigated before and after X-ray irradiation (Figure 5 and Supporting Information Figure S13). XPS analysis of the films with BQ was also performed. Furthermore, the atomic concentrations of C and O, and the atomic percentages relative to Sn were calculated based on the integral area for comparison ( Supporting Information Tables S1 and S2), with the Sn content kept constant throughout the reaction. The initial film atom compositions determined via XPS were consistent with the molecular formula, while the C/Sn and O/Sn ratios decreased after irradiation because of the desorption of the alkyl groups and 1-adamantane carboxylic acid. Figure 5 | XPS for C 1s of (a) Sn4–Me–C10 and (b) Sn4–Bu–C10; XPS for O 1s of (c) Sn4–Me–C10 and (d) Sn4–Bu–C10 before X-ray exposure (initial film) and after 1 h of irradiation (after exposure) and spectra with 1% BQ after irradiation (after exposure with BQ). The experimental data correspond to black circle markers, and the fitted data correspond to black solid lines. The numbers in the graphs were obtained by integrating each peak relative to the Sn content. Download figure Download PowerPoint The C 1s spectra (Figure 5a,b) were fitted using three components: Eb = 284.8 eV (red line) for aliphatic carbon (C–H), 286.2 eV (green line) for Tin-carbon (Sn–C), and 288.6 eV (blue line) for carbonyl groups (C=O). For the initial Sn4–Me–C10 film, the C–H peak was attributed to the carbon in adamantane. After irradiation, the integrals of the C–H peak decreased from 11.5 to 7.63, whereas those of the Sn–C peak increased from 1.20 to 1.85, implying that the cleavage of C–H in adamantane is initiated by methyl radicals and the formation of carbon-free radicals. Furthermore, the reduction in the integrals of the C=O peak signified that the carboxylic group was stripped off to produce carbon dioxide and form new active Sn sites. These new active sites of adamantane reacted with the contiguous active Sn atoms, generating new Sn–C bonds. Thus, a portion of adamantane participated in the construction of the main framework of the polymer, thereby enhancing the film's etching resistance. Compared with the methyl group, the end-group H of butyl group in Sn4–Bu–C10 was more easily shed to form new radicals that cross-linked with the surrounding Sn free radical. Therefore, the Sn–C signal increased from 1.48 to 1.73. However, the addition of 1% BQ to the Sn4–Me–C10 film attenuated the enhancement of the Sn–C peak signal, probably because BQ combined with adamantane radicals.41,42 For the O 1s spectra shown in Figure 5c,d, two peaks were used to fit at Eb = 531.7 eV (light green) for Sn–O–C and Eb = 530.4 eV (pale purple) for Sn–O–Sn. After irradiation, the integrals of the Sn–O–C peak decreased from 1.99 to 1.19 for Sn4–Me–C10 and from 1.82 to 1.54 for Sn4–Bu–C10, indicating the dissociation of partial 1-adamantane carboxylic acid from the tin-oxo inner core. Accordingly, Sn active sites were generated, some of which further reacted with carbon-free radicals to form new Sn–C bonds. The remaining active sites interacted with the adjacent cluster nucleus, leading to a slight increase in the Sn–O–Sn peak from 0.445 to 0.517 for Sn4–Me–C10. The addition of BQ increased the integrals of the Sn–O–Sn peak, probably because BQ inhibited the recombination of the Sn–C bonds, causing Sn active sites to combine with adjacent cluster nuclei to form a Sn–O–Sn network. Overall, the changes in the Sn–C signal were more substantial than those for the Sn–O–Sn network and consistent with the variation in the photoresist formulation reactivity. Therefore, the reorganization and crosslinking of the Sn–C bonds played a dominant role during exposure. Based on the aforementioned experimental results, an exposure mechanism was proposed to explain the different irradiation behaviors of Sn4–Me–C10 and Sn4–Bu–C10 (Figure 6). First, irradiation caused the cleavage of precarious Sn–C bonds, resulting in the generation of alkyl free radicals and Sn active sites, which constituted the primary exposure reaction. Radical hydrogen abstraction occurred, leading to generation of new alkyl radicals. These radicals directly reacted with uncoordinated active Sn sites, yielding an Sn-alkyl network bridging the clusters, which is known as alkane chain linkage. In addition, these alkyl radicals decarboxylated the adamantine acid to form CO2 and adamantine radicals on the bridgehead carbon. The adamantine radicals readily combined with contiguous active Sn sites to form a Sn–adamantine network, referred to as the adamantane linkage. The two pathways competed with each other. The alkane chain linkage played a minor role in the solubility change in Sn4–Me–C10 because of the unfavorable steric hindrance caused by the short methyl group bound to the lateral active Sn sites. Therefore, the adamantane linkage predominated in Sn4–Me–C10, and the immobility of bulky adamantane free radicals prevented the spreading of lines and enhanced resolution. The addition of BQ specifically bound to the adamantane free radical inhibited the adamantane linkage and attenuated the diffusion of adamantane free radicals in the films. Therefore, the addition of BQ decreased the sensitivity of Sn4–Me–C10, decreased the line CD, and increased the LER. For Sn4–Bu–C10, the butyl free radicals generated from the hydrogen extraction reaction easily reacted with the surrounding Sn active sites because of the proper steric hindrance. Although BQ inhibited the adamantane linkage, it did not affect the sensitivity of Sn4–Bu–C10 but rather reduced the initial butyl radical consumption and enhanced the alkyl linkage. This explains why the addition of BQ decreased the sensitivity of Sn4–Me–C10 but enhanced that of Sn4–Bu–C10. Figure 6 | Proposed chemical mechanism for the full lithographic process for Sn4–Me–C10 and Sn4–Bu–C10. Download figure Download PowerPoint Etching resistance The selectivity of etching between the substrate and photoresist was crucial for achieving well-defined micropatterns. Therefore, the etching rates of the two cluster photoresists and silicon substrates were evaluated using deep silicon etching (DSE), reactive ion etching (RIE), and inductively coupled plasma etching (ICPE) ( Supporting Information Figures S14–S17 and Table S3). Sn4–Me–C10 and Sn4–Bu–C10 exhibited remarkable etching resistance in DSE, with rates of 0.91 and 0.93 nm/s for Sn4–Me–C10 and Sn4–Bu–C10, respectively (Table 1). Etch selectivity in the case of silicon reached as high as 1:4. The two cluster photoresists also demonstrated reasonable etching resistance during RIE, with rates of 1.45 and 1.49 nm/s for Sn4–Me–C10 and Sn4–Bu–C10, respectively. However, both cluster photoresists performed poorly during inductively coupled plasma (ICP) etching. Table 1 | Three Etching Rates of Silicon and Two Cluster Photoresists under DSE, RIE, and ICPE DSE (nm/s) RIE ICPE Si 3.94 2.92 1.87 Sn4–Me–C10 0.91 1.45 4.81 Sn4–Bu–C10 0.93 1.49 5.27 Conclusion In this study, we have described two Sn-containing clusters, Sn4–Me–C10 and Sn4–Bu–C10, and investigated their performance in EBL and EUVL for the first time. These two clusters were simply pr