記住我
In recent years, gold nanostructures have received much attention owing to their dielectric properties , their biocompatibility , and their electrical properties , which enable a multitude of exciting applications in the field of nanotechnology. These include, but are not limited to electronic interconnects , metamaterials , growth substrates for nanowires and nanotubes , and complex plasmonic structures . For the latter application, mastery over the shape as well as accurate control of the distribution of the nanostructures are critical for the enhancement of absorption and controlled scattering of light . Focused-electron-beam-induced deposition (FEBID) is a direct writing method for controlled deposition/fabrication of nanostructures on either flat or nonflat surfaces. It offers excellent shape control and thus the potential to widen the scope of applicable nanomaterials. In FEBID, a focused electron beam is directed onto the surface of a substrate in close proximity to a gas inlet, through which a precursor compound is supplied to deliver the material for the nanostructures to be built. For metallic structures, these precursor molecules are commonly organometallics that adsorb on the substrate and are decomposed by the electron beam irradiation. Ideally, a pure metal is deposited while fragmented volatile ligands are pumped away .
Several parameters affect the FEBID process, including the electron beam energy and current, the substrate material, the environment inside the deposition chamber, and the composition of the precursor . Heretofore, various chemical vapor deposition (CVD) precursors have been applied for FEBID depositions. For gold nanostructures, these include, for example, dimethyl(acetylacetonate)gold(III) (Au(acac)(CH3)2), dimethyl(trifluoroacetylacetonate)gold(III) (Au(tfac)(CH3)2), and dimethyl(hexafluoroacetylacetonate)gold(III) (Au(hfac)(CH3)2) . Although these precursors have proven suitable for CVD, in FEBID their application has mainly resulted in amorphous matrixes of carbon with embedded metal crystallites and a gold content of 2–3 atom % , 10–40 atom % , and 8–20 atom % , respectively. This is most likely due to the fact that the CVD process is thermally driven, while in FEBID, the precursor fragmentation is primarily electron driven. This may partly explain the insufficient metal content achieved when using CVD precursors in FEBID . In this context, efforts have been made to design gold precursors optimized for FEBID. Arguably, the most noticeable of these are chloro(carbonyl)gold(I) (AuICl(CO)) and chloro(trifluorophosphine)gold(I) (AuICl(PF3)) . These precursors have enabled depositions of ≈95 atom % Au and a resistivity of Au grains as low as 22 µΩ, respectively. However, the short lifetime of both precursors, which results from their moisture sensitivity and thermal instability, has limited their applicability.
In FEBID, the irradiation of the substrate with a high-energy focused electron beam results in elastic and inelastic electron scattering, including ionizing events. The latter leads to the production of numerous reactive, low-energy scattered and secondary electrons. These play a significant role in the precursor decomposition and thus in the deposit formation . Hence, the decomposition of the precursor molecules is not only effectuated by the primary electron beam. In fact, the reactivity of these low-energy electrons may even determine the fragmentation of the precursor molecules, which in turn is critical to the resulting purity of the FEBID deposits. In general, electron-induced fragmentation processes are categorized as dissociative ionization (DI), dissociative electron attachment (DEA), dipolar dissociation (DD), and neutral dissociation (ND) . To fully comprehend the electron–molecule interactions in FEBID, it is critical to understand the extent and nature of these processes and how they are reflected in the deposit formation from individual precursors or specific ligand structures. A very interesting approach in this direction was recently introduced by Jurczyk et al. under the term focused-electron-beam-induced mass spectrometry (FEBiMS). In this approach, ion-extraction mass spectrometry, in close proximity to the forming FEBID structure, is used to analyze the charged, desorbing ligand fragments. Another approach in this direction is to combine ultrahigh-vacuum (UHV) surface science studies and mass spectrometry in high-vacuum (HV) gas-phase investigations . In this context, surface science experiments allow for electron-dose-dependent studies of the elemental composition of the deposit, and desorbing ligands may be monitored by means of mass spectrometry. On the other hand, gas-phase studies using controllable, quasi-monoenergetic electron beams under single collision conditions, provide information on the electron energy dependence and extent of the individual fragmentation processes . A number of such comparative gas-phase and surface science studies have been carried out in the past using a 500 eV flood gun in the surface studies , and also in combination with higher energy FEBID studies . In a recent study , we took a similar approach and investigated (CH3)AuP(CH3)3 as a potential gold precursor for FEBID. We used gas-phase experiments under single-collision conditions and quantum mechanical calculations for data interpretation, in combination with FEBID in an UHV setup. The results of this study demonstrated that at 5 keV electron energy, FEBID deposits with 31–34 atom % Au content were attainable with this precursor in the UHV setup. A close-to-complete phosphorous removal was observed and the Au/C ratio of the deposit was found to be 1:2. This corresponds to the average carbon loss per incident beam found in the DI gas-phase experiment, and was consistent with the dominating reaction pathways as determined by the quantum chemical calculations. However, in one specific channel in the DI gas-phase study, a significant retention of the phosphorous at the gold precursor was found indicating significant surface effects.
In the current study, we extended this approach to investigate the deposition of Au using [Au(CH3)2Cl]2 as a potential FEBID precursor. The FEBID in an UHV setup was conducted, in conjunction with a gas-phase study on the electron energy dependence of the fragmentation of this compound through DI and DEA. Moreover, quantum chemical calculations were used to determine the dominating reaction pathways. The UHV FEBID results are discussed in the context of the observed DI and DEA fragmentation processes, and also in the context of a previous FEBID study of this precursor under HV, conducted by van Dorp et al. In that study, a promising Au content of 29–41 atom % was achieved without additional substrate purifications. In the current study, we found the Au content to be further improved to reach about 50 atom % in the UHV setup without pretreatment of the substrate surface. With a pretreated surface, a gold content of 61 atom % was reached.
Results and Discussion FEBID on SiO2 (500 nm)/Si(111)In this experiment, 4 × 4 µm2 FEBID structures were written with [Au(CH3)2Cl]2 as the precursor using an acceleration voltage of 5 keV and a beam current of 1.5 nA. The fabricated structures were examined with scanning electron microscopy (SEM) and Auger electron spectroscopy (AES). Figure 1a depicts an SEM image of the FEBID deposit created with an electron exposure of 7.80 C/cm2. The position of the corresponding AES analysis is marked in Figure 1a by a green-colored star. The AES spectra acquired on the bare substrate and the deposit are shown in Figure 1b. On the pristine SiO2 (500 nm)/Si(111) substrate (black spectrum), only two main AES signals are visible: The peak at 272 eV is attributed to CKLL Auger transitions of carbon , and the peaks at 468, 483, and 503 eV to OKLL Auger transitions of SiO2.
![[2190-4286-14-98-1]](https://www.beilstein-journals.org/bjnano/content/figures/2190-4286-14-98-1.png?scale=2.0&max-width=1024&background=FFFFFF)
Figure 1: (a) An SEM image of a 4 × 4 µm2 FEBID structure deposited on SiO2 from [Au(CH3)2Cl]2 with an electron dose of 7.80 C/cm2 using the electron beam parameters of 5 keV and 1.5 nA. (b) An AES plot of the SiO2 substrate prior to deposition (black line) and from the FEBID structure (green line); the green-colored star in (a) indicates the position where the spectrum was acquired. (c) Magnified image from the area within the red-colored square shown in (a). (d) The same image as shown in (c) after the background subtraction process was applied using the ImageJ program .
After deposition, AES signals at 69, 181, 272, and 430 eV are present. These are assigned to the AuNOO, ClLMM, CKLL, and SnMNN Auger transitions , respectively (Figure 1b, green spectrum). The broad and small peak at approximately 367 eV is attributed to an Sn signal . The contamination with Sn is from the synthesis process of the [Au(CH3)2Cl]2 precursor, which involves SnMe4 as a methylation agent . The elemental composition of the FEBID structure was calculated using the relative sensitivity factors (S) , that is, SAu: 0.21; SCl: 0.69; SC: 0.08; SSn: 0.53. From these, the atomic concentration of the deposit was found to be 51 atom % Au, 2 atom % Cl, 42 atom % C, and 5 atom % Sn. Considering the ratio of the SnMNN signal to that of the remaining OKLL signal from the deposit, compared to that expected for stannic oxide SnO2, it is clear that the Sn impurity is predominantly elemental rather than on the oxidized form. A magnification of a selected area of the SEM image shown in Figure 1a is depicted in Figure 1c, where nanoparticles in the deposition are noticeable, although the picture is somewhat blurry. To better visualize the observed nanoparticles, a background subtraction was performed with the image enhancement program ImageJ . The image after the background subtraction is shown in Figure 1d, where the particles can be more clearly distinguished. After background subtraction, some of the deposited nanoparticles appear facetted; however, the majority are spherical.
HAADF-STEM on FEBID (SiO2 (500 nm)/Si(111))As a next step, high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) was performed to analyze the morphology of the deposited nanoparticles. For this purpose, several FEBID structures were prepared on the SiO2 substrate with the size of 4 × 4 µm2 and an electron dose of 7.80 C/cm2. For the HAADF-STEM measurements, a lamella was prepared with a thickness of approx. 100 nm and a width of approx. 4 µm (Supporting Information File 1, Figure S1). In Figure 2a, the HAADF-STEM image of deposited nanoparticles is shown, revealing a nearly uniform spatial distribution of nanoparticles with a size lower than 5 nm. Nonuniformly distributed nanoparticles with larger sizes (≈15–20 nm) were also observed. The magnified image of a selected larger nanoparticle from Figure 2a is shown in Figure 2b.
![[2190-4286-14-98-2]](https://www.beilstein-journals.org/bjnano/content/figures/2190-4286-14-98-2.png?scale=2.0&max-width=1024&background=FFFFFF)
Figure 2: (a) A HAADF-STEM image of a FEBID gold nanoparticle. (b) Enlarged image of the area depicted with red-dashed lines in (a), showing the interplanar distance of 0.23 nm between the planes of the FCC lattice. (c) A SAED pattern of the FEBID gold nanoparticles, compared with the lattice spacings (d-spacings) of FCC gold.
The determined fringe spacing of that particle is ≈0.23 nm, which is consistent with the spacing between the (111) planes of a face-centered cubic (FCC) gold nanoparticle . The crystallinity of the gold nanoparticles was also investigated by using selected area electron diffraction (SAED) pattern shown in Figure 2c. The relatively bright circular patterns indicate polycrystallinity of the deposits. For comparison, the lattice spacings (d-spacings) of 2.30, 2.07, 1.42, 1.23, and 1.17 Å , corresponding to the (111), (200), (220), (133), and (222) growth planes, respectively, of the FCC lattice of crystalline gold is also shown in Figure 2d.
FEBID on SiO2 (500 nm)/Si(111) at different beam currentsThe FEBID deposits were also prepared with [Au(CH3)2Cl]2 using beam currents of 0.4 nA (deposit size: 2 × 2 μm2), 1.5 nA (deposit size: 4 × 4 μm2), and 3 nA (deposit size: 4 × 4 μm2). The other deposition parameters (electron dose: 7.80 C/cm2 and acceleration voltage: 5 keV) were the same in all three experiments. The FEBID structures were investigated by SEM and noncontact atomic force microscopy (AFM). Figure 3a shows the SEM images of the deposits along with the respective deposition parameters. Magnified sections from these SEM images are shown in Figure 3b. Auger electron spectroscopy was performed on these structures to determine their composition and to better understand the effect of different beam currents on the compositions. The respective spectra are shown in Figure 3c.
![[2190-4286-14-98-3]](https://www.beilstein-journals.org/bjnano/content/figures/2190-4286-14-98-3.png?scale=2.0&max-width=1024&background=FFFFFF)
Figure 3: (a) An SEM image of FEBID structures deposited on SiO2 from [Au(CH3)2Cl]2 with an electron dose of 7.80 C/cm2 using a 5 keV electron beam and different beam currents of 0.4 nA, 1.5 nA, and 3 nA. (b) Magnified images of FEBID structures from (a). (c) An AES plot of the FEBID structures deposited with 0.4 pA, 1.5 nA, and 3 nA depicted with blue, green, and purple lines, respectively.
A careful investigation of the SEM images reveals that the particle size gets smaller when the current is increased, as is clearly discernible from Figure 3b when comparing deposition at a beam current of 1.5 and 3 nA. For detailed particle analysis, the ImageJ software was used to obtain the numbers of nanoparticles and their mean diameter. As mentioned before, the observed gold nanoparticles (Figure 1d) have irregular shapes. Therefore, the mean Feret diameter, which gives the average value over all possible orientations was used (see Supporting Information File 1, Figure S2). The average particle sizes determined from the SEM images were found to be similar at beam currents of 0.4 and 1.5 nA (i.e., 9.8 and 10.1 nm, respectively). At 3 nA, on the other hand, a clear size reduction to 8.2 nm is observed in the SEM images. This size reduction with increasing deposition current is even clearer from the AFM images as discussed in the following section. From the AES data shown in Figure 3c, the atomic concentrations of the structures were calculated: At 0.4 nA, the composition was found to be 45 atom % Au, 1 atom % Cl, 49 atom % C, 5 atom % Sn; at 1.5 nA it was found to be 50 atom % Au, 2 atom % Cl, 42 atom % C, 6 atom % Sn; and at 3 nA, it was found to be 52 atom % Au, 2 atom % Cl, 38 atom % C, 8 atom % Sn. For ease of comparison, the respective elemental compositions are also reported in Table 1 along with the composition of a deposit on thermally cleaned Si(111) at a 1.5 nA beam current, as discussed in the next section.
Table 1: Elemental composition (atom %) of depositions on SiO2 (500 nm)/Si(111) at different beam currents (nA). Also shown is the elemental composition of a deposition on thermally cleaned Si(111).
SiO2 (500 nm)/Si(111) Current Au C Cl Sn 0.4 45 49 1 5 1.5 50 42 2 6 3.0 52 38 2 8 Thermally cleaned Si(111) Current Au C Cl Sn 1.5 61 35 1 3Clearly, the increase in gold content with increasing deposition current is correlated with the decrease in carbon content, which is also reflected in the proportionally increasing Sn contaminations showing the same trend as that of gold. This is more evident from the reduction of the carbon peak areas in the AES, which is ≈36% when comparing the depositions at 0.4 and 3 nA, and ≈14% when comparing the depositions at 0.4 and 1.5 nA. We thus attribute the observed size reduction of the deposited gold particles with increasing deposition current to the decrease in carbon content. A similar size reduction of gold nanoparticles has been reported for post-deposition oxidative purification of FEBID deposits, where carbon removal led to ≈18% height reduction of the respective nanoparticles . Notwithstanding, changes in the deposition time and in the associated different volume of the deposited material may also contribute to the observed particle size reduction.
AFM of FEBID on (SiO2 (500 nm)/Si(111)) at different beam currentsIn order to obtain complementary information on the structures deposited with different beam currents, noncontact AFM was used to investigate the height of the deposits and their particle size. Figure 4a and Figure 4b depict the 2D AFM images and magnified sections of these, respectively. The corresponding height profiles are shown in Figure 4c. The magnified sections of the 2D AFM images (Figure 4b) show the same trend as observed in the SEM images shown in Figure 3b. The size of the gold nanoparticles is approximately the same for the FEBID structures written with 0.4 and 1.5 nA, while they are smaller in the deposit written with 3 nA beam current. The average particle sizes obtained from the AFM images are approx. 10.4 nm for 0.4 nA, 9.5 nm for 1.5 nA, and 7.0 nm for 3 nA (Supporting Information File 1, Figure S3b). These values are in good agreement with the values obtained from the SEM images (9.8 nm – 0.4 nA; 10.1 nm – 1.5 nA; 8.2 nm – 3 nA). Notably, the line profiles in Figure 4b for the structures created with 0.4 and 1.5 nA reveal thicknesses of the deposits of ≈17 nm, while the thickness of the deposit written with 3 nA is only ≈9 nm. As aforementioned, we attribute the size reduction, at least in part, to a more efficient carbon removal at higher currents. The same applies to the observed reduction in thickness with increasing beam current. However, the reduction of volume growth rate per dose at higher currents and the thickness of the deposits may also affect the observed particle size.
![[2190-4286-14-98-4]](https://www.beilstein-journals.org/bjnano/content/figures/2190-4286-14-98-4.jpg?scale=2.0&max-width=1024&background=FFFFFF)
Figure 4: (a) A set of 2D AFM images and magnified AFM images (red-dashed squares). (b) The corresponding line profiles for the FEBID structures produced with an electron dose of 7.80 C/cm2 using the beam currents of 0.4 nA (blue line), 1.5 nA (green line), and 3 nA (purple line).
Interestingly, the height profiles of depositions also change according to the applied beam current (Figure 4b). For example (most significant in the height profile of the 0.4 nA deposition, Figure 4b, blue line), there is a negative dip at the edge of the deposits, indicated by the orange dashed lines. It is important to note that this negative dip is also observed for other line profiles throughout the deposit. Therefore, the negative dips are present in the entire structure (Supporting Information File 1, Figure S4). This negative dip can also be seen for the depositions created with 1.5 and 3 nA beam currents, depicted in Figure 4b with green and purple lines, respectively. However, the depth of the dip decreases with increasing applied beam current. This indicates that an etching process occurs simultaneously with the deposition process, wherein the etching effect is less pronounced than that of the deposition for all beam currents. Similar etching effects were observed with other halogenated precursors, where it was reported that one of the expected effects when working with halogen-based precursors is the observation of etching as well as deposition . In these studies, the release of halogen ligands was indicated as the main reason for the etching process.
FEBID on thermally cleaned Si(111)In several UHV-FEBID studies it has been shown that an UHV setup alone is not sufficient to produce FEBID structures with relatively low organic contaminations. In addition, a comparably clean and well-defined substrate also helps to increase the metal content. Thus, the Si(111) substrate was sputtered using Ar+ for 45 minutes (V![[Graphic 1]](https://www.beilstein-journals.org/bjnano/content/inline/2190-4286-14-98-i1.svg?max-width=637&scale=1.18182)
= 1 eV, P![[Graphic 2]](https://www.beilstein-journals.org/bjnano/content/inline/2190-4286-14-98-i2.svg?max-width=637&scale=1.18182)
= 4 × 10−6 mbar) and subsequently annealed up to 823 K under an oxygen atmosphere for 90 minutes to demonstrate the effect of surface preparation (reduction of C and O contaminants) on the quality of deposition. After preparation, AES was performed to check the surface cleanliness and to compare with the uncleaned surface (Supporting Information File 1, Figure S5). The Supporting Information File 1, Figure S5, clearly shows that the carbon (CKLL at 272 eV) and oxygen peaks (OKLL at 508 eV) were reduced (by ≈17% for C, ≈67% for O), and thus the SiLMM peak at 92 eV became observable. Using the cleaned sample, a FEBID experiment with the [Au(CH3)2Cl]2 precursor was performed to create 4 × 4 µm2 structures using the same parameters as were used for the deposits depicted in Figure 1. The results were analyzed with SEM and AES, see Figure 5. The red-colored star indicates the AES measured point.
![[2190-4286-14-98-5]](https://www.beilstein-journals.org/bjnano/content/figures/2190-4286-14-98-5.png?scale=2.0&max-width=1024&background=FFFFFF)
Figure 5: (a) An SEM image of a 4 × 4 µm2 FEBID structure deposited on Si(111) from [Au(CH3)2Cl]2 with an electron dose of 7.80 C/cm2 using the electron beam parameters of 5 keV and 1.5 nA. (b) An AES plot of the Si(111) substrate prior to deposition (black line) and the result from the FEBID structure (red line). The red-colored star in (a) indicates the position where the spectrum was acquired.
The AES plot (red-colored line) shows the following peaks: AuNOO at 69 eV, ClLMM at 181 eV, CKLL at 272 eV, and SnMNN at 430 eV , yielding atomic concentrations of 61 atom % Au, 1 atom % Cl, 35 atom % C, and 3 atom % Sn, respectively. A comparison to Figure 1b reveals an increase of Au content by 10%, while the C content decreases by 10%. In a previous study, the same precursor (i.e., [Au(CH3)2Cl]2) was used to create FEBID deposits on a SiO2 substrate by using 5 keV and 0.1/0.4 nA in an HV atmosphere . The composition of the structures was checked via energy dispersion X-ray (EDX) mapping and reported to be 29–41 atom % Au, 2–6 atom % Cl, and 53–68 atom % C. The SEM images of the FEBID deposits also revealed grainy structures. These concentrations, reported in the reference study, support the idea of complete Cl ligand desorption and incorporation of both CH3 ligands into the deposit. The main difference between the work at hand and the aforementioned study in HV, is that this study was carried out in UHV with a higher electron beam current of 1.5 nA. However, it should be mentioned that the AES used here is surface sensitive as compared to EDX, which is bulk sensitive. As the synthesis and purification routes in both HV and the current UHV studies are apparently the same, it is surprising that no Sn impurities were reported in the deposits made under HV. As the information on the synthesis route is limited in the HV study reported, we can only speculate that a different methylation agent may have been used (i.e., one that did not contain tin). An alternative explanation may lie in the usage of different characterization tools (i.e., EDX and AES). From both the UHV and HV FEBID results, one can conclude that the Cl ligands are completely removed and desorbed under the impact of the electron beam. The ease of Cl ligand desorption during electron beam deposition has also been addressed in several previous studies . Notably, the UHV-FEBID results yield 10–20 atom % higher Au content than those reported in the HV study. Therefore, the reaction pathway of [Au(CH3)2Cl]2 can be suggested as:
![[Graphic 3]](https://www.beilstein-journals.org/bjnano/content/inline/2190-4286-14-98-i3.svg?max-width=637&scale=1.18182)
where Au2(CH3)x is the deposited material, while 2Cl and (4 − x)(CH3) are desorbed from each molecule. We note that these may be desorbed as Cl2, CH3Cl, or CH3CH3 as discussed in the next section. Further, we expect the final deposit to rather result from electron-induced secondary and tertiary reaction than from a single electron precursor interaction. According to the AES depicted in Figure 1b and Figure 5b, x can be inferred to be 1–2.
Gas-phase studiesFigure 6a shows a positive ion mass spectrum of [Au(CH3)2Cl]2, recorded for the m/z range from 10 to 550 at a 50 eV electron impact energy. A rich fragmentation pattern, characterized by progressive loss of methyl groups, is observed.
![[2190-4286-14-98-6]](https://www.beilstein-journals.org/bjnano/content/figures/2190-4286-14-98-6.png?scale=2.0&max-width=1024&background=FFFFFF)
Figure 6: Positive ion mass spectrum of electron impact ionization and dissociation of [Au(CH3)2Cl]2 recorded at an incident electron energy of 50 eV.
The first progression is that of the molecular cation at m/z 524 followed by a sequential loss of methyl ligands, appearing at m/z 509, 494, 479, and 464, with the most significant contribution being from the loss of all four methyl ligands at m/z 464. The second progression shows the loss of one chlorine atom and two, three, and four methyl ligands at m/z 458, 444, and 429, respectively. From these, m/z 458 has lost an additional hydrogen and m/z 444 overlaps with lesser contributions from m/z 443, which we also attribute to additional hydrogen loss. Similar to the preceding progression, the loss of all four methyl ligands, m/z 429, is also the dominating contribution here. The third progression shows the loss of both chlorines in combination with the sequential loss of two, three, and four methyl ligands at m/z 422, 408, and 394, respectively. Here, the contributions are of similar intensity, although the loss of three methyl ligands, m/z 408, is slightly more apparent. Lesser contribution is also observed at m/z 407 and is attributed to additional hydrogen loss as compared to that of m/z 408. The last progression is from the loss of both chlorine atoms along with one gold atom, and two and three methyl groups and is observed at m/z 227 and 225, 212 and 197, respectively. From these, m/z 225 is ascribed to the loss of two methyl groups and two additional hydrogens, and 197 represents Au+ (i.e., the elemental gold). Additionally, m/z 247 is observed with fair intensity and we attribute this fragment to the loss of three methyl ligands in combination with the loss of one chlorine and one gold atom (i.e., [Au(CH3)Cl]+). There are some broad low-intensity impurity contributions in the EI MS in the m/z range of SnCl (150–160) and SnCl(CH3) (200–210). However, these are low intensity and cannot be unambiguously assigned from their isotope distribution. The most significant low m/z contributions are around m/z 28 and 15. The contributions at and around m/z 28 are predominantly from the background gas in the chamber, including N2, but may also contain C2Hn contributions from rearrangement reactions of [Au(CH3)2Cl]2 upon ionization. Similarly, m/z 15 is in part attributed to CH3+ resulting from DI of [Au(CH3)2Cl]2.
For most of the observed m/z ratios, the assignment of the underlying fragmentation process is not straightforward as the neutral fragments, complementary to the m/z ratios observed, may be assigned to more than one composition. Thus, to better understand the underlying fragmentation process, the respective appearance energies (AEs) are determined using a Wannier-type threshold function (see the Methods section) and compared to calculated threshold energies for a variety of potential reaction pathways. A full list of all optimized geometries (Cartesian coordinates) of the parent and positively charged ions at the PBE0-TZVP level of theory are provided in Supporting Information File 1, Table S1. Figure 7 shows the fitted onset region of representative ion yield curves for the individual fragments along with their average AEs determined from fits to 3–4 ion yield curves recorded on different days. Also shown are the respective confidence limits and the structures of the respective positive ions optimized at the PBE0-TZVP level of theory.
![[2190-4286-14-98-7]](https://www.beilstein-journals.org/bjnano/content/figures/2190-4286-14-98-7.png?scale=2.0&max-width=1024&background=FFFFFF)
Figure 7: Representative fits to the onset region of the DI ion yield curves for the parent cation and the most dominant positively charged fragments from [Au(CH3)2Cl]2. The appearance energies and their confidence limits for each ion are shown along with the respective chemical structure optimized at the PBE0-TZVP level of theory.
Table 2 compares the individual AEs with calculated thresholds for different potential reactions leading to the respective fragments. Here the values for single bond ruptures without new bond formations are shown along with the best matches of the AEs with the respective threshold values. For comparison, threshold values for selected processes that are next to the assigned processes are also shown in Table 2. A complete set of all calculated threshold values, and the respective processes are shown in Supporting Information File 1, Table S2. The thresholds are calculated at both the PBE0-TZVP and DLPNO-CCSD(T)-TZVP levels of theory, as discussed in the Methods section. The assigned fragmentation reactions shown in Table 2 are in bold. In the assignment we primarily compare to the DLPNO-CCSD(T)-TZVP values. We note that as activation barriers may shift the AEs to higher values as compared to the respective thermochemical thresholds, the true thermochemical threshold may in some cases be lower than the respective AE. Where the current comparison does not allow the assignment to one combination of neutral fragments, the closest matches are in bold in Table 2.
Table 2: Comparison of the experimental AE values with the respective calculated threshold values. The threshold energies are calculated for homolytic bond ruptures and where no new bonds are formed the neutral fragments are the radical species.
m/z Products AE (eV) PBE0-TZVP (eV) DLPNO-CCSD(T)-TZVP (eV) 524 [Au(CH3)2Cl]2+ 9.4 ± 0.3 9.23 9.92 494 [Au2Cl2(CH3)2]+ + 2CH3For the appearance energy of the parent cation [Au(CH3)2Cl]2+, (ionization energy of [Au(CH3)2Cl]2) we determine a value of 9.4 ± 0.3 eV. Within the confidence limit, this agrees well with the calculated threshold of 9.23 eV found at the PBE0-TZVP level of theory. However, at the DLPNO-CCSD(T)-TZVP levels of theory, we calculate a threshold of 9.92 eV, which is ≈0.2 eV above the upper confidence limit for the experimental AE.
For the loss of one methyl group, m/z 509, we find the intensity too low to determine the AE. However, for the loss of two methyl groups, m/z 494, we find an AE of 9.7 ± 0.3 eV. Considering only single bond ruptures (i.e., the formation of two CH3 radicals in this process) it results in threshold values of 13.51 and 14.06 eV at the PBE0-TZVP and DLPNO-CCSD(T)-TZVP levels of theory, respectively. These are ≈4 eV abov
留言 (0)