Synthesis, characterization, thermal properties of silicon(IV) compounds containing guanidinato ligands and their potential as CVD precursors

Abstract

The title compounds of the type (Me₃Si)₂N–C(=N′R)(–N′′RSiMe₃) (with R = iPr or Cy) as potential CVD precursors have been synthesized and characterized by X-ray diffraction, ¹H NMR, ¹³C NMR, ²⁹Si NMR and elemental analysis where necessary. Among these characterizations, solid-/liquid-state ²⁹Si NMR were accomplished to study their behavior in the solid and solution. Thermal properties including stability, volatility, transport behavior and vapour pressure were evaluated by thermogravimetric analysis (TGA) to confirm that they are suitable for the CVD procedure. Deposition was accomplished in a hot wall CVD reactor system, which preliminarily verified the ability of these compounds as CVD precursors.

---

Introduction

Guanidines/amidinates are N-donor bidentate chelating ligands, which demonstrate a great diversity by the variation of substituents on the conjugated N–C–N backbone.¹ The steric and electronic properties are easily modified to meet the needs of different metal centers.¹ Corresponding compounds have shown enormous potential as catalysts for the oligomerization and polymerization of olefins, hydroamination, intramolecular hydroamination/cyclization and hydrosilylation.²

Meanwhile, the numerous features and widespread applications have made guanidinato/amidinate an important ligand for precursors used in chemical vapor deposition (CVD)/atomic layer deposition (ALD) manufacturing process. For example, Anjana Devi and coworkers obtained nitrogen-doped CVD-grown TiO₂ thin films using [Ti(NMe₂)₃(guan)] (guan = N,N′-diisopropyl-2-dimethylamidoguanidinato) as precursor.³ Tianniu Chen introduced chelating guanidinate ligands improving thermal stability and maintaining the volatility of precursors for CVD tantalum nitride thin films.⁴ Copper(I)amidinate [Cu(i-Pr-Me-AMD)]₂ was investigated to produce copper films in conventional low pressure chemical vapor deposition (LPCVD) using hydrogen as reducing gas-reagent.⁵

In contrast, the chemistry of silicon compounds containing guanidinato/amidinate ligands is still a largely unexplored field. To date, a few silicon compounds containing guanidinato/amidinate ligands have been reported (Scheme 1).⁶ All of these researches were merely focused on synthetic methodology or molecular geometry. Few literatures, however, involved the application of these compounds as CVD/ALD precursors and relevant properties.

Scheme 1 Reported silicon compounds with guanidinato/amidinate ligands.

As known, many silicon compounds have been investigated to grow silicon-based thin films such as silicon nitride (SiN_x),⁷ amorphous silicon (α-Si),⁸ silicon carbide (SiC),⁹ polycrystalline silicon¹⁰ and silicon carbonitride (SiCN),¹¹ which have shown a wide range of properties and have various applications in microelectromechanical systems (MEMS) technology. Apart from the industrialized silanes (SiH₄ and Si₂H₆), many hydrosilicons linked with alkyl and/or amino ligands have been reported, such as SiH₃CH₃, Et₃SiH, SiH₂(CH₃)₂, SiH(NMe₂)₃, SiMe₂(NMe₂)₂ and (SiHMe₂)₂NH.¹² Insufficiently, the introduction of gaseous reactant Me₂NH or Si–H units has increased the synthetic difficulty, thus hindering their further application. Furthermore, the toxicity and transportation difficulties of commonly-used SiH₄/Si₂H₆ have always been a big trouble for the users.¹³

The search for new silicon precursors is still a great challenge not only for the above-mentioned reasons but also because the chemical and physical properties of the Si source can be adjusted by various ligands/substituents in order to obtain films with many particular properties and thereby satisfy different application requirements.¹⁴ Thus, we have recently been intrigued by the question as to whether such silicon compounds containing guanidinato/amidinate ligands are capable for a robust CVD/ALD process.

On the basic of these studies, herein, we report the synthesis of two new silicon compounds (Scheme 2) with guanidinato ligands, their thermal properties and further potential as CVD precursors.

Scheme 2 Synthesis of compounds 1 and 2.

Compared with the existing CVD precursors mentioned above, many foreseeable advantages can be obtained through the design and synthesis of these compounds as CVD precursors: (i) the synthesis method is well known as a salt elimination approach that is easily accomplished; (ii) the specific –SiMe₃ groups are beneficial to enhance the volatility of these compounds;¹⁵ (iii) many guanidinato/amidinate ligands are either easily sublimated solid or volatile liquid that will facilitate the desorption of the byproducts from the surface;¹⁶ and (iv) they are nontoxic, non-explosive and non-pyrophoric and are thus readily applicable to a robust CVD system.

Results and discussion

Synthesis and characterization

Compounds 1 and 2 were obtained by reaction of lithium intermediates, which originated from the reaction of corresponding carbodiimides and LiN(SiMe₃)₂, with equivalent SiMe₃Cl according to a general procedure. The synthetic approach involved the formation of lithium intermediates and LiCl elimination, which proceeded easily under benign conditions in diethyl ether.

Between these two precursors, compound 1 is low melting solid (mp 48–50 °C), while 2 melts at a higher temperature (mp 150–153 °C). Both of them were recrystallized out of n-hexane at −30 °C for structural determination and purified further through sublimation under vacuum. Furthermore, both elemental analysis of 1 and 2 are consistent with the proposed monomeric structures.

One ingenious idea of the introduction of –N(SiMe₃)₂ groups is for convenient and unambiguous learning of coordination states of these two central silicon atoms (Si3 for 1 and Si1 for 2).

The single-crystal X-ray diffraction results (Fig. 1 and 2) confirm the monomeric nature of them. Characteristic crystal parameters of these compounds 1 and 2 are listed in Table 1.

Fig. 1 ORTEP diagram of 1 showing 40% thermal ellipsoids. Selected bond lengths [Å] and angles [deg]: C1–N1 1.451(4), C1–C5 1.519(4), C2–N1 1.276(3), C2–N2 1.395(4), C2–N3 1.433(3), C3–N2 1.478(4), C6–Si1 1.844(4), C9–Si2 1.841(4), C12–Si3 1.876(4), N2–Si3 1.777(3), N3–Si2 1.754(2), N3–Si1 1.754(2); N1–C2–N2 114.6(3), N1–C2–N3 125.7(3), N2–C2–N3 119.7(2), C2–N1–C1 123.7(3), C2–N2–C3 121.1(2), C2–N2–Si3 111.79(19), C3–N2–Si3 127.1(2), C2–N3–Si2 117.36(18), C2–N3–Si1 117.43(17), Si2–N3–Si1 125.05(14), N3–Si1–C7 110.54(17), N3–Si2–C9 109.69(15), N2–Si3–C12 109.78(16), C14–Si3–C12 103.41(19).

Fig. 2 ORTEP diagram of 2 showing 40% thermal ellipsoids. Selected bond lengths [Å] and angles [deg]: C1–N1 1.457(3), C1–C2 1.508(4), C7–N1 1.278(3), C7–N2 1.396(3), C7–N3 1.436(3), C8–N2 1.475(3), C14–Si1 1.851(4), C17–Si3 1.854(4), C20–Si2 1.847(4), N2–Si1 1.782(2), N3–Si3 1.752(2), N3–Si2 1.759(2); N1–C7–N2 114.7(2), N1–C7–N3 125.9(2), N2–C7–N3 119.37(19), C7–N1–C1 122.9(2), C7–N2–C8 121.13(19), C7–N2–Si1 111.24(15), C8–N2–Si1 127.35(16), C7–N3–Si3 117.79(15), C7–N3–Si2 117.01(15), Si3–N3–Si2 124.99(12), N2–Si1–C14 111.26(15), N3–Si2–C20 112.56(16), –N3–Si2–C21 111.30(16), N3–Si3–C19 109.89(13).

Table 1 Crystal parameters of 1 and 2

	1	2
Empirical formula	C16H41N3Si3	C22H49N3Si3
Molecular weight	359.79	439.91
Temperature (K)	296	296
Wavelength Mo Kα (Å)	0.71073	0.71073
Crystal size (mm)	0.27 × 0.30 × 0.32	0.22 × 0.24 × 0.26
Crystal system, space group	Triclinic, P1̄	Triclinic, P1̄
a (Å)	9.331(4)	9.0667(12)
b (Å)	9.361(4)	11.4557(15)
c (Å)	14.996(7)	14.1193(19)
α (deg)	99.880(11)	100.930(3)
β (deg)	93.805(12)	102.475(2)
γ (deg)	109.819(10)	91.188(2)
Cell volume (Å^3)/Z	1203.0(9)/2	1403.0(3)/2
Density δcalcd (g cm^−3)	0.993	1.041
Absorption coefficient μ (mm^−1)	0.199	0.181
F(000)	400	488
θ range for data collection (deg)	1.4–26.3	1.5–25.0
Completeness to θ = full [%]	98.3	98.6
Index ranges	−11 ≤ h ≤ 10, −10 ≤ k ≤ 11, −18 ≤ l ≤ 18	−10 ≤ h ≤ 10, −12 ≤ k ≤ 13, −16 ≤ l ≤ 15
Reflections collected/unique [R(int)]	9120/4807/0.043	9737/4885/0.028
Refinement method	Full-matrix least-squares on F^2	Full-matrix least-squares on F^2
Data/restraints/parameters	4807/0/212	4885/0/262
Goodness-of-fit on F^2 (GOF)	1.038	1.089
Observed data [I > 2.0 × sigma(I)]	2592	3867
R, wR2, S	0.0607, 0.1834, 1.04	0.0574, 0.2258, 1.09
Largest difference peak and hole (e Å^−3)	−0.30, 0.21	−0.43, 0.42
CCDC number	1400899	1033427

---

For every compound, each focal silicon center (Si3 for 1 and Si1 for 2) is chelated with one nitrogen atom from the amidinate units. The coordination environment around these two silicon centers is approximately tetrahedral (∠N–Si–C and ∠C–Si–C values range from 101.4 to 113.2°). Moreover, they are almost coplanar with NCN atoms of respective amidinate units (dihedral angle N1C2N2Si3 0.8° and N1C7N2Si1 1.4°).

The significant difference between the two N–C distances in both of the amidinate units [1.39 vs. 1.27 Å] indicates the higher degree of localization of the N=C double bond compared to the bidentate guanidinato ligands (1.33 vs. 1.34 Å).^6d Meanwhile, Si3/Si1–N2 distance (1.777 and 1.782 Å) is significantly shorter than those of compounds with bidentate guanidinato ligands reported by Reinhold Tacke^6d (all approximately 1.9 Å), clearly reflecting the different coordination modes. This Si3/Si1–N2 distance, however, is slightly longer than that of –N(SiMe₃)₂ (approximately 1.75 Å of 1 and 2) and the monodentate ligand (1.76 Å), which implies the interaction between focal silicon center and the ligand. Furthermore, the medium-strong Si–N interactions are emphasized by the Si3–N1/Si1–N1 distances (2.6324 Å for 1 and 2.6242 Å for 2), which are under the sum of the Si and N van der Waals radii of 3.48 Å.¹⁷

Besides, we recorded solid-state ²⁹Si CP/MAS NMR spectrums of precipitated powder of them, which show one single peak at −0.489 ppm (1) and −0.543 ppm (2), respectively. These are in agreement with the solid-state structures observed by X-ray diffraction well, where all Si atoms possess almost the same chemical environment.

To elucidate the solution structure in the studied compounds, the conventional NMR parameters (chemical shifts of ¹H, ¹³C, ²⁹Si nuclei) were obtained in CDCl₃. For compound 1, the ¹H and ¹³C NMR spectra exhibited chemical shift equivalent N2C2N1 ligand resonances at room temperature. Though ¹H NMR analysis of 2 was not distinguishable, ¹³C NMR spectra of it confirmed its equivalent resonances for N2C7N1 ligand with four singlets of cyclohexyl units. Thus, we concluded that both 1 and 2 behaved as symmetric structures in CDCl₃, which disagreed with their solid state formation.

To further verify this conclusion, ²⁹Si NMR was applied. Analogously, two separated signal for both of them (1: 3.5 and −7.4 ppm; 2: 3.8 and −7.4 ppm; 25 °C) indicated the variation in the chemical environment of focal silicon center (Si3 for 1 and Si1 for 2) and corresponding referential silicon in –N(SiMe₃)₂ groups. Moreover, in comparison to the tetra-coordinated referential silicon, the fact that resonances of focal silicon were shifted towards higher-field (at least to a small extend) revealed the intramolecular donor-coordination (N1 → Si3 for 1 and N1 → Si1 for 2).¹⁸

Based on the similarity of these two compounds, only ²⁹Si NMR solvent effect of 2 was investigated. Compared with the result in CDCl₃, there was no obvious difference when d8-toluene was used (3.6 and −7.0 ppm, 25 °C). An unexpected lower-field signal appeared in CD₂Cl₂, however, with the chemical shift of 7.4 ppm (another two: 3.8 and −7.6 ppm, 25 °C), which might suggest the dynamic equilibrium of the N1 → Si1 interaction mentioned above. Subsequently, VT-²⁹Si NMR of compound 2 was accomplished in d8-toluene. At high temperature (40 °C) compound 2 had an extra silicon chemical shift of 7.0 ppm (another two: 3.6 and −7.0 ppm), suggesting a weakening of the N1 → Si1 coordination. That is, high temperature weakens the interaction which has also been found by other researchers.¹⁹ And when given a low temperature of −60 °C, both intrinsic ²⁹Si signal tended to be closer to each other in a slight extent with shifts of 3.4 and −6.7 ppm, which implied the trend of them to merge as solid-state ²⁹Si CP/MAS signal with the decrease of the temperature. Regrettably, lower temperature ²⁹Si NMR failed to obtain because of instrument limit.

Combining all tests above, we conclude that compound 1 and 2 exhibit different nuclear magnetic resonance behaviors by solid state and in solution, which has been widely studied.²⁰ Besides, a dynamic equilibrium exists revealing the intensity change of the interaction between Si and corresponding N.

Thermal properties of compounds 1 and 2

The study of the basic parameters of precursors has not only scientific but also of great practical importance for understanding the chemistry of precursor preparation and their long-term stability as well as for the calculation of proper CVD precursor supply into the reactor for the preparation of the appropriate device structure. In particular, volatility data and a detailed vapour pressure equation are essential for controlled precursor dosimetry.

Both of these two compounds are volatile and thermally stable enough to be purified by vacuum sublimation and their thermal characteristics were first studied and measured by simultaneous thermal analyses (STA) including thermogravimetric analysis (TGA) and differential scanning calorimetry (DSC).

As seen in Fig. 3, both of compounds 1 and 2 are stable up to their melting temperature and exhibit monotonic weight loss with negligible residual mass remaining (0.22% and 0.79% respectively), in a single step. This demonstrates that compounds 1 and 2 have sufficient thermal stability for vaporization without significant decomposition. Besides, the entire weight loss of compounds 1 and 2 finish at 215 and 240 °C respectively, implying the comparable volatility of them.

Fig. 3 Thermogravimetric curves of 1 and 2.

The temperature of 50% mass loss derived from TGA data (T₅₀) has been demonstrated to correlate with the volatility of the sample and is used to compare relative volatilities (Table 2).²¹ The combination of lower T₅₀ with lower onset temperature derived from TGA curve is usually a strong indication of good volatility of a precursor. Clearly, compound 1 possesses lower T₅₀ (197 °C) and onset temperature (179 °C), which suggests its better volatility. Both temperatures increase as the molecular mass increases (1 < 2), which agrees with previous reports by other researchers. For example, the nearly linear relationship between T₂₀ values and molecular weigh in tBu derivatives of gallium chalcogenide cubanes were reported by Barron.²² And the consistent result was obtained by T. Chen, in tantalum amido compounds.⁴

Table 2 STA data of compounds 1 and 2^a

	Molecular weight (g mol^−1)	T50/°C	Onset temperature/°C	Residual/%
a Mass residuals were measured at 300 °C.
1	359.79	197	179	0.79
2	439.91	221	203	0.22

---

To further investigate the transport properties of them, isothermal TGA experiments were carried out, in which the temperature of the STA furnace was held at 150 °C for 600 minutes and the rates of the linear mass loss, presumably by evaporation instead of decomposition, were calculated from the TGA curves shown in Fig. 4. The fact that both DSC curves remain constant for these samples tested verifies the assumption that only evaporation contributes to the mass loss reflected by isothermal TGA curves. Compound 1 showed faster mass loss rate of 97.2 μg min⁻¹ (22.9 μg min⁻¹, 2) at 150 °C and purged by 60 sccm Ar, which coincided with its lower T₅₀ and onset temperature and confirmed its better volatility between these two compounds. Significantly, the higher volatility of 1 over the other precursor most likely arises from its lower molecular mass and suggests the future strategy for designing new volatile CVD/ALD precursors.

Fig. 4 Isothermal TGA of 1 and 2.

Vapour pressure data are very important for industrial process control and in manufacturing processes. Especially, the film growth rate can be limited by the precursor vapour pressure. If the vapour pressure of precursor compounds is known, then it would be easy to compare their volatilities and find the right evaporation temperature for each precursor so that the precursor flux would be maintained in the CVD reactor at the right level.

Herein, vapour pressure–temperature plots were obtained by thermogravimetry, which is considered a rapid and convenient technique for the determination of vapour pressure curves. The theoretical basis of the TG procedure is the Langmuir and Antoine equation, where benzoic acid was chosen as a standard.²³

As seen from the P–T curves (Fig. 5) for 1 and 2, compound 1 possesses higher vapor pressure at relatively low temperatures with almost the same range (90 °C). The plot of ln P against 1/T as straight lines presents a more comparable view. Compound 1 is shown to be more volatile at lower temperatures, and to have 1 Torr of vapor pressure at 121 °C. Compound 2 has a slightly lower volatility, and gives 1 Torr of vapor pressure at 156 °C. This also verifies the T₅₀ results are reliable. However, both of these compounds can be expected to give reasonably high vapor pressures at normal bubbler temperatures, and the similarity of the slopes of their evaporation curves suggests that they experience similar intermolecular attraction during evaporation. Moreover, the above temperatures cover the interval important for the CVD process.²⁴

Fig. 5 Vapor pressure–temperature curves of 1 and 2.

On the basis of the above results, we consider both of them to be qualified precursor for a robust CVD process. And a similar assumption could be valid under the real CVD conditions.

Potential as CVD precursors

The straightforward way to qualitatively verify the potential of these compounds as CVD precursors is making them undergo a deposition process. Thus compound 1 was chosen as the representative to undergo a single-precursor deposition process at 900 °C. The as-grown film, with a thickness about 400 nm, showed comparable morphology without obvious agglomeration as displayed in ESI.†

Besides, Fig. 6 displayed the XP survey spectra acquired from the surface of the as-grown film. The relative peak positions for the Si 2p and Si 2s are 103.7 and 154.08 eV respectively, being consistent with earlier data reported²⁵ and thereby indicating the participation of the precursor in the film deposition. The oxygen observed here was assumed to be largely due to surface oxidation or contamination. Thus, we qualitatively conclude that compounds 1 and 2 with guanidinato ligands have the potential to be CVD precursors. Certainly, further work should be done in regard to the deposition process such as optimization of conditions, quantitative analysis of the film component and mechanism. All these will be presented in the subsequent report.

Fig. 6 XP spectra of the as-grown film.

Conclusions

In this study, we have synthetized and characterized two silicon(IV) compounds containing guanidinato ligands of the type (Me₃Si)₂N–C(=N′R)(–N′′RSiMe₃) (with R = iPr or Cy) as CVD precursors. They were obtained by reaction of lithium intermediates, which originated from the reaction of corresponding carbodiimides and LiN(SiMe₃)₂, with equivalent SiMe₃Cl according to a general procedure in high yields. Thermal stability, volatility, transport behavior and vapor pressure were evaluated by simultaneous thermal analyses (STA) to confirm that they are suitable for CVD process. Deposition which was accomplished in a simple hot wall CVD reactor system, qualitatively demonstrated the ability of them as CVD precursors for fabricating silicon containing films.

Moreover, further work with respect to the deposition process such as optimization of conditions, quantitative analysis of the film component, mechanism and deposition along with other auxiliary source (e.g. NH₃) to form different kinds of films will be presented in the subsequent report elsewhere.

Experiment section

All manipulations were carried out using standard Schlenk techniques or in a glove box under a nitrogen atmosphere. Prior to use, diethyl ether and n-hexane was freshly distilled from Na. Other chemicals were purchased from Aldrich and used as received. X-ray data was collected with a Bruker SMART APEX II CCD area detector using monochromated Mo-Kα radiation (λ = 0.71073 Å). NMR spectrum was collected on a Bruker ACF-400 spectrometer tetramethylsilane as internal standard (0.03% content as purchased). The C, H, N analyses were performed with a Carlo Erba model EA 1108 microanalyzer. LiN(SiMe₃)₂ was prepared according to the reported procedure.²⁶ The TG curve was obtained with an STA 449 F3 analyzer in argon at a heating rate of 10 °C min⁻¹ from 30 to 800 °C. The surface morphology and X-ray photoelectron spectroscopy (XPS) analysis were characterized by a Hitachi S-4800 scanning electron microscopy (SEM) and a Thermo ESCALAB 250Xi equipment with a residual pressure of 10⁻⁷ Pa.

Synthesis of 1 and 2

Due to the similarities of preparing these two compounds, only a detailed synthetic procedure for 1 is described here. To a 100 mL Schlenk flask charged with LiN(SiMe₃)₂ (1.673 g, 10 mmol) in 40 mL diethyl ether, diisopropylcarbodiimide (1.262 g, 10.0 mmol) was added. After a stir for 24 h, SiMe₃Cl (1.086 g, 10.0 mmol) was added at −78 °C. A white precipitate appeared after the addition started. The reaction mixture was warmed to room temperature and stirred overnight. After removing diethyl ether by vacuum, the residue was extracted by n-hexane (30 mL × 3). Then the filtrate was concentrated and recrystallized out of n-hexane at −30 °C for a few hours to yield colourless crystal 1 (3.274 g, 91% yield, mp 48–50 °C): ¹H NMR (400 MHz, CDCl₃) δ 3.92–3.82 (m, 2H, –CH(CH₃)₂), 1.12–1.10 (d, J = 8 Hz, 12H, –CH(CH₃)₂), 0.24 (s, 9H, –Si(CH₃)₃), 0.19 (s, 18H, –N[Si(CH₃)₃]₂); ¹³C NMR (101 MHz, CDCl₃) δ 153.68, 45.98, 24.73, 5.06, 2.20; ²⁹Si NMR (79 MHz, CDCl₃, 25 °C) δ 3.54, −7.40; ²⁹Si CP/MAS NMR δ −0.489; anal. calcd for C₁₆H₄₁N₃Si₃: C, 53.41; H, 11.49; N, 11.68; found: C, 53.39; H, 11.58; N, 11.65.

2 (colourless crystal, 83% yield, mp 150–153 °C): ¹H NMR (400 MHz, CDCl₃) δ 3.44–3.37 (m, 2H, N–CH), 1.74–1.58 (m, 8H, Cy), 1.51–1.41 (m, 4H, Cy), 1.28–1.05 (m, 9H, Cy), 0.24 (s, 9H, –Si(CH₃)₃), 0.20 (s, 18H, –N[Si(CH₃)₃]₂); ¹³C NMR (101 MHz, CDCl₃) δ 153.92, 55.41, 34.64, 26.17, 25.86, 5.33, 2.34; ²⁹Si NMR (79 MHz, CDCl₃, 25 °C) δ 3.77, −7.36; ²⁹Si CP/MAS NMR δ −0.543; ²⁹Si NMR (79 MHz, CD₂Cl₂, 25 °C) δ 7.41, 3.81, −7.61; ²⁹Si NMR (79 MHz, d8-toluene, 25 °C) δ 3.59, −7.03; ²⁹Si NMR (79 MHz, d8-toluene, 40 °C) δ 7.04, 3.65, −6.99; ²⁹Si NMR (79 MHz, d8-toluene, −60 °C) δ 3.43, −6.67; anal. calcd for C₂₂H₄₉N₃Si₃: C, 60.07; H, 11.23; N, 9.55; found: C, 60.11; H, 10.98; N, 9.71.

Thermal analysis and CVD deposition

The TG curve was obtained with an STA 449 F3 analyzer in argon at a heating rate of 10 °C min⁻¹ from 30 to 800 °C. Subsequently, chemical vapor deposition was accomplished in a simple system shown schematically in ESI,† to further demonstrate the ability of them as CVD precursors, where 1 was chosen as the representative.

The furnace consists of a hot-wall tubular quartz reactor with a large (≈60 cm) isothermal (±5 °C) zone. The substrates used were 1 cm × 1 cm N-type Si (100) & (110) & (110) wafers doped with P with a resistivity of 0.01–0.02 Ω cm. Prior to the deposition, the wafers were treated by a H₂SO₄ + H₂O₂ (v/v = 7/3) mixture followed by HF etching, thoroughly rinsed with ultrapure water and then dried with N₂. The growth parameters used during the process were: total pressure: 2.1 × 10³ Pa, 10% H₂/N₂ flow: 120 mL min⁻¹, N₂ flow: 120 mL min⁻¹, deposition time: 180 min and temperature 900 °C respectively. A 150 °C heating temperature of 1 was kept during the whole deposition. After deposition, samples were cooled to room temperature at a rate of 5 °C min⁻¹ in a N₂ (flow: 120 mL min⁻¹) atmosphere.

The surface morphology and X-ray photoelectron spectroscopy (XPS) analysis were characterized by a Hitachi S-4800 scanning electron microscopy (SEM) and a Thermo ESCALAB 250Xi equipment with a residual pressure of 10⁻⁷ Pa.

Acknowledgements

We gratefully acknowledge financial support of this work by the National Natural Science Foundation of China (No. 21371080) and “333 Talent Project” of Jiangsu Province (BRA2012165).

Notes and references

1. F. T. Edelmann, in Advances in the Coordination Chemistry of Amidinate, Guanidinate Ligands, ed. A. F. Hill and M. J. Fink, Elsevier Academic Press Inc, San Diego, CA, 2008, vol. 57, pp. 183–352.
2. (a) E. Smolensky and M. S. Eisen, Dalton Trans., 2007, 5623; (b) B. D. Ward, H. Risler, K. Weitershaus, S. Bellemin-Laponnaz, H. Wadepohl and L. H. Gade, Inorg. Chem., 2006, 45, 7777; (c) S. Ge, A. Meetsma and B. Hessen, Organometallics, 2008, 27, 3131; (d) W. Zhang and L. R. Sita, J. Am. Chem. Soc., 2008, 130, 442; (e) S. Collins, Coord. Chem. Rev., 2011, 255, 118; (f) F. T. Edelmann, Chem. Soc. Rev., 2009, 38, 2253; (g) M. P. Coles, Dalton Trans., 2006, 985; (h) F. T. Edelmann, Chem. Soc. Rev., 2012, 41, 7657; (i) C. Alonso-Moreno, A. Antiñolo, F. Carrillo-Hermosilla and A. Oterob, Chem. Soc. Rev., 2014, 43, 3406; (j) P. J. Bailey and S. Pace, Coord. Chem. Rev., 2001, 214, 91.
3. S. J. Kim, K. Xu, H. Parala, R. Beranek, M. Bledowski, K. Sliozberg, H. W. Becker, D. Rogalla, D. Barreca, C. Maccato, C. Sada, W. Schuhmann, R. A. Fischer and A. Devi, Chem. Vap. Deposition, 2013, 19, 45.
4. T. Chen, C. Xu, T. H. Baum, G. T. Stauf, J. F. Roeder, A. G. DiPasquale and A. L. Rheingold, Chem. Mater., 2010, 22, 27.
5. V. Krisyuk, L. Aloui, N. Prud'homme, S. Sysoev, F. Senocq, D. Samélor and C. Vahlas, Electrochem. Solid-State Lett., 2011, 14, D26.
6. (a) K. Junold, J. A. Baus, C. Burschka, D. Auerhammer and R. Tacke, Chem.–Eur. J., 2012, 18, 16288; (b) K. Junold, C. Burschka, R. Bertermann and R. Tacke, Dalton Trans., 2010, 9401; (c) K. Junold, J. A. Baus, C. Burschka, T. Vent-Schmidt, S. Riedel and R. Tacke, Inorg. Chem., 2013, 52, 11593; (d) F. M. Mück, K. Junold, J. A. Baus, C. Burschka and R. Tacke, Eur. J. Inorg. Chem., 2013, 5821; (e) K. Junold, C. Burschka, R. Bertermann and R. Tacke, Dalton Trans., 2011, 9844; (f) K. Junold, J. A. Baus, C. Burschka and R. Tacke, Angew. Chem., Int. Ed., 2012, 51, 7020.
7. (a) F. L. Riley, J. Am. Ceram. Soc., 2000, 83, 245; (b) H. Klemm, J. Am. Ceram. Soc., 2010, 93, 1501; (c) Y. Q. Chen, X. N. Zhang, Q. Zhao, L. He, C. K. Huang and Z. P. Xie, Chem. Commun., 2011, 47, 6398.
8. (a) H. Sirringhaus, Adv. Mater., 2014, 26, 1319; (b) B. Rech and H. Wagner, Appl. Phys. A, 1999, 69, 155.
9. (a) N. G. Wright, A. B. Horsfall and K. Vassilevski, Mater. Today, 2008, 11, 17; (b) H. P. Phan, D. V. Dao, P. Tanner, J. S. Han, N. T. Nguyen, S. Dimitrijev, G. Walker, L. Wang and Y. Zhu, J. Mater. Chem. C, 2014, 2, 7176.
10. (a) J. K. Rath, Sol. Energy Mater. Sol. Cells, 2003, 76, 43; (b) A. F. B. Braga, S. P. Moreira, P. R. Zampieri, J. M. G. Bacchin and P. R. Mei, Sol. Energy Mater. Sol. Cells, 2002, 82, 418; (c) S. D. Brotherton, Semicond. Sci. Technol., 1995, 10, 721.
11. (a) O. Flores, R. K. Bordia, D. Nestler, W. Krenkel and G. Motz, Adv. Eng. Mater., 2014, 1; (b) A. Badzian, T. Badzian, R. Roy and W. Drawl, Thin Solid Films, 1999, 354, 148; (c) A. Badzian, T. Badzian, W. D. Drawl and R. Roy, Diamond Relat. Mater., 1998, 7, 1519.
12. (a) I. Golecki, F. Reidingger and J. Marti, Appl. Phys. Lett., 1992, 60, 1703; (b) A. M. Wrobel, A. Walkiewicz-Pietrzykowska, M. Ahola, I. J. Vayrynen, F. J. Ferrer-Fernandez and A. R. Gonzalez-Elip, Chem. Vap. Deposition, 2009, 15, 39; (c) A. M. Wrobel, A. Walkiewicz-Pietrzykowska, M. Stasiak, T. Aoki, Y. Hatanaka and J. Szumilewicz, J. Electrochem. Soc., 1998, 145, 1060; (d) A. M. Wrobel and A. Walkiewicz-Pietrzykowska, Chem. Vap. Deposition, 1998, 4, 133; (e) N. Fainer, A. Golubenko, Y. M. Rumyantsev and E. Maximovskii, Glass Phys. Chem., 2009, 35, 274; (f) A. M. Wrobel, I. Blaszczyk-Lezak, P. Uznanski and B. Glebocki, Chem. Vap. Deposition, 2010, 16, 211; (g) R. D. Mundo, M. Ricci, R. d'Agostino, F. Fracassi and F. Palumbo, Plasma Processes Polym., 2007, 4, S21; (h) G. Suchaneck, V. Norkus and G. Gerlach, Surf. Coat. Technol., 2001, 142, 808; (i) A. M. Wrobel and I. Blaszczyk-Lezak, Chem. Vap. Deposition, 2007, 13, 595; (j) N. Fainer, A. Golubenko, Y. M. Rumyantsev and E. Maximovskii, Glass Phys. Chem., 2009, 35, 274.
13. (a) H. Gleskova, S. Wagner, V. Gašparık and P. Kováč, Appl. Surf. Sci., 2001, 175, 12; (b) A. Kshirsagar, P. Nyaupane, D. Bodas, S. Duttagupta and S. Gangal, Appl. Surf. Sci., 2011, 257(11), 5052; (c) H. Sato, A. Izumi, A. Masuda and H. Matsumura, Thin Solid Films, 2001, 395(1), 280; (d) F. V. Assche, W. Kessels, R. Vangheluwe, W. Mischke, M. Evers and M. van de Sanden, Thin Solid Films, 2005, 484(1), 46; (e) A. Santoni, J. Lancok, S. Loreti, I. Menicucci, C. Minarini, F. Fabbri and D. Della Sala, J. Cryst. Growth, 2003, 258, 272; (f) J. Song, G. S. Lee and P. K. Ajmera, Thin Solid Films, 1995, 270, 512.
14. M. Asay, C. Jones and M. Driess, Chem. Rev., 2011, 111, 354.
15. (a) P. Chadha, D. J. H. Emslie and H. A. Jenkins, Organometallics, 2014, 33, 1467; (b) I. Giebelhaus, R. Müller, W. Tyrra, I. Pantenburg, T. Fischer and S. Mathur, Inorg. Chim. Acta, 2011, 372, 340; (c) H. C. Aspinall, J. F. Bickley, J. M. Gaskell, A. C. Jones and G. Labat, Inorg. Chem., 2007, 46, 5852; (d) R. Anwander, F. C. Munck, T. Priermeier and W. Scherer, Inorg. Chem., 1997, 36, 3545.
16. S. T. Barry, Coord. Chem. Rev., 2013, 257, 3192.
17. (a) Y. V. Zefirov, Russ. J. Inorg. Chem., 2000, 45, 1552; (b) S. S. Batsanov, Russ. J. Inorg. Chem., 1991, 36, 1694; (c) S. S. Batsanov, Inorg. Mater., 2001, 37, 871; (d) P. M. Zorkii and A. A. Stukalin, Crystallogr. Rep., 2005, 50, 522; (e) A. Bondi, J. Phys. Chem., 1964, 68, 441.
18. U. H. Berlekamp, A. Mix, B. Neumann, H. G. Stammler and P. Jutzi, J. Organomet. Chem., 2003, 667, 167.
19. B. Gostevskii, K. Adear, A. Sivaramakrishna, G. Silbert, D. Stalke, N. Kocher, I. Kalikhman and D. Kost, Chem. Commun., 2004, 1644.
20. (a) J. Wagler, U. Bohme, E. Brendler and G. Roewer, Organometallics, 2005, 24, 1348; (b) T. Cheisson and A. Auffrant, Dalton Trans., 2014, 13399; (c) B. M. Kraft and W. W. Brennessel, Organometallics, 2014, 33, 158; (d) M. Sohail, R. Panisch, A. Bowden, A. R. Bassindale, P. G. Taylor, A. A. Korlyukov, D. E. Arkhipov, L. Male, S. Callear, S. J. Coles, M. B. Hursthouse, R. W. Harringtond and W. Clegg, Dalton Trans., 2013, 10971.
21. (a) M. F. Richardson and R. E. Sievers, Inorg. Chem., 1971, 10, 498; (b) Z. Li, A. Rahtu and R. G. Gordon, J. Electrochem. Soc., 2006, 11, C787.
22. E. G. Gillan, S. G. Bott and A. R. Barron, Chem. Mater., 1997, 9, 796.
23. S. F. Wright, D. Dollimore, J. G. Dunn and K. Alexander, Thermochim. Acta, 2004, 421, 25.
24. (a) E. G. Gillan, S. G. Bott and A. R. Barron, Chem. Mater., 1997, 9, 796; (b) T. Chen, C. Xu, T. H. Baum, G. T. Stauf, J. F. Roeder, A. G. DiPasquale and A. L. Rheingold, Chem. Mater., 2010, 22, 271; (c) W. Hunks, P. S. Chen, T. Chen, M. Stender, G. T. Stauf, L. Maylott, C. Xu and J. F. Roeder, Mater. Res. Soc. Symp. Proc., 2008, 1071, F09.
25. (a) J. H. Boo, S. B. Lee, K. S. Yu, M. M. Sung and Y. Kim, Surf. Coat. Technol., 2000, 131, 147; (b) M. B. J. Wijesundara, G. Valente, W. R. Ashurst, R. T. Howe, A. P. Pisano, C. Carraro and R. Maboudiana, J. Electrochem. Soc., 2004, 151, C210.
26. I. V. Magedov, Y. I. Smushkevich, D. N. Plutitskii and N. N. Suvorov, Zh. Obshch. Khim., 1988, 58, 1934.

---

Footnote

† Electronic supplementary information (ESI) available: ¹H NMR, ¹³C NMR, ²⁹Si NMR spectra for 1 and 2; STA curves; SEM images; deposition furnace. CCDC 1400899 and 1033427. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c5ra09755j

---