Trimethylphosphine(hexafluoro-2,4-pentanedionate)copper (I)

• CAS: 135707-05-0
• Molecular Formula: C8H11CuF6O2P
• Molecular Weight: 347.684
• SMILES: CP(C)C.C(=C(C(=O)C(F)(F)F)O)C(F)(F)F.[Cu]
• InChI: InChI=1S/C5H2F6O2.C3H9P.Cu/c6-4(7,8)1-2(12)3(13)5(9,10)11;1-4(2)3;/h1,12H;1-3H3;/b2-1-
• InChIKey: GVSUUFKMQJJKKJ-UAIGNFCESA-N

Trimethylphosphine(hexafluoro-2,4-pentanedionate)copper(I), commonly abbreviated as (hfac)Cu(PMe₃), is a copper(I) β-diketonate Lewis-base adduct belonging to the (hfac)Cu(I)L family of volatile organometallic precursors developed for chemical vapor deposition (CVD) and atomic layer deposition (ALD) of copper thin films. The compound comprises a hexafluoroacetylacetonate (hfac) chelating ligand and a trimethylphosphine (PMe₃) neutral donor.

Trimethylphosphine(hexafluoro-2,4-pentanedionate)copper(I) – A Differentiated Cu(I) CVD/ALD Precursor for Electronic-Grade Copper Thin Films

Trimethylphosphine(hexafluoro-2,4-pentanedionate)copper(I), commonly abbreviated as (hfac)Cu(PMe₃), is a copper(I) β-diketonate Lewis-base adduct belonging to the (hfac)Cu(I)L family of volatile organometallic precursors developed for chemical vapor deposition (CVD) and atomic layer deposition (ALD) of copper thin films [1]. The compound comprises a hexafluoroacetylacetonate (hfac) chelating ligand and a trimethylphosphine (PMe₃) neutral donor. It deposits high-purity metallic copper via a thermally induced disproportionation reaction without requiring an external reducing agent [2]. Within the broader class of Cu(I) hfac precursors—where the neutral ligand L can be systematically varied to tune volatility, thermal stability, and surface reactivity—(hfac)Cu(PMe₃) occupies a distinct position defined by quantifiable differences in bond strength, thermal robustness, and deposition behavior relative to the more widely commercialized analogs (hfac)Cu(VTMS) and (hfac)Cu(1,5-COD) [3].

• [1] Shin, H. K.; Chi, K. M.; Farkas, J.; Hampden-Smith, M. J.; Kodas, T. T.; Duesler, E. N. Chemistry of Copper(I) β-Diketonate Complexes. 2. Synthesis, Characterization, and Physical Properties of (β-Diketonato)copper(I) Trimethylphosphine and Bis(trimethylphosphine) Complexes. Inorg. Chem. 1992, 31, 424–431. View Source
• [2] Choi, H.; Hwang, S. Copper(I) tert-Butyl 3-Oxobutanoate Complexes as Precursors for Chemical Vapor Deposition of Copper. Chem. Mater. 1998, 10 (9), 2326–2328. View Source
• [3] Cavallotti, C.; Gupta, V.; Sieber, C.; Jensen, K. F. Dissociation Reactions of CuI(hfac)L Compounds Relevant to the Chemical Vapor Deposition of Copper. Phys. Chem. Chem. Phys. 2003, 5 (13), 2818–2827. View Source

Why (hfac)Cu(PMe₃) Cannot Be Interchanged with (hfac)Cu(VTMS), (hfac)Cu(COD), or Other Cu(I) hfac Analogs

The (hfac)Cu(I)L precursor family exhibits steep property cliffs driven by the identity of the neutral ligand L. Although all share the same hfac chelate and the same formal Cu(I) oxidation state, swapping L alters the Cu–L bond dissociation energy, thermal stability window, vapor pressure, and surface reaction kinetics by margins sufficient to change deposition outcomes from selective, high-purity films to non-selective growth, premature decomposition in the delivery line, or parasitic gas-phase reactions [1]. Three quantitative parameters define the substitution risk: (i) the Cu–L bond energy, which governs the rate-limiting ligand-dissociation step during film growth [2]; (ii) the thermal stability at the sublimation or delivery temperature, which determines shelf-life and source reproducibility [3]; and (iii) the propensity for parasitic side reactions during co-deposition with other metals [4]. These properties are not linearly correlated with simple molecular descriptors such as ligand molecular weight or boiling point, making empirical or computational head-to-head comparison the only reliable basis for precursor selection.

• [1] Cavallotti, C.; Gupta, V.; Sieber, C.; Jensen, K. F. Dissociation Reactions of CuI(hfac)L Compounds Relevant to the Chemical Vapor Deposition of Copper. Phys. Chem. Chem. Phys. 2003, 5 (13), 2818–2827. View Source
• [2] NIST Chemical Kinetics Database. Calculated Bond Dissociation Energies for Cu(hfac)L → Cu(hfac) + L. Data derived from Cavallotti et al., Phys. Chem. Chem. Phys. 2003, 5, 2818. View Source
• [3] Shin, H. K.; Chi, K. M.; Farkas, J.; Hampden-Smith, M. J.; Kodas, T. T.; Duesler, E. N. Chemistry of Copper(I) β-Diketonate Complexes. 2. Inorg. Chem. 1992, 31, 424–431. View Source
• [4] CVD and Characterization of Al–Cu Metallization Thin Films. MRS Online Proceedings Library 1992, 260, 119–124. View Source

Quantitative Differentiation Evidence for (hfac)Cu(PMe₃) vs. Closest Analogs

Cu–L Bond Dissociation Energy: (hfac)Cu(PMe₃) Exhibits the Highest Computed Bond Strength Among Four hfac-Based Cu(I) CVD Precursors

Density functional theory (DFT) calculations at the B3LYP level with 6-311+G(d) basis sets yielded gas-phase Cu–L heterolytic bond dissociation energies for the reaction Cu(hfac)L → Cu(hfac) + L. (hfac)Cu(PMe₃) displayed the highest value, 38.4 kcal mol⁻¹, compared with 33.6 kcal mol⁻¹ for (hfac)Cu(VTMS), 35.6 kcal mol⁻¹ for (hfac)Cu(η⁴-COD), and 32.1 kcal mol⁻¹ for (hfac)Cu(2-butyne) [1]. The NIST Chemical Kinetics Database reports an independently fitted value of 39.4 kcal mol⁻¹ for the PMe₃ adduct, corroborating the trend [2]. This represents a 14% increase in Cu–L bond strength relative to the most widely used industrial precursor (hfac)Cu(VTMS).

Evidence Dimension	Cu–L heterolytic bond dissociation energy (kcal mol⁻¹)
Target Compound Data	38.4 kcal mol⁻¹ (DFT); 39.4 kcal mol⁻¹ (NIST fit)
Comparator Or Baseline	(hfac)Cu(VTMS): 33.6; (hfac)Cu(η⁴-COD): 35.6; (hfac)Cu(2-butyne): 32.1 kcal mol⁻¹
Quantified Difference	+4.8 kcal mol⁻¹ vs. (hfac)Cu(VTMS); +2.8 kcal mol⁻¹ vs. (hfac)Cu(COD); +6.3 kcal mol⁻¹ vs. (hfac)Cu(2-butyne)
Conditions	B3LYP/6-311+G(d) DFT calculations, gas phase, 0 K, no solvation or surface effects

Why This Matters

Higher Cu–L bond energy translates to a higher thermal budget before premature ligand dissociation, enabling wider process windows for delivery-line temperature control and reducing the risk of gas-phase nucleation that fouls reactor walls and degrades film uniformity.

• [1] Cavallotti, C.; Gupta, V.; Sieber, C.; Jensen, K. F. Dissociation Reactions of CuI(hfac)L Compounds Relevant to the Chemical Vapor Deposition of Copper. Phys. Chem. Chem. Phys. 2003, 5 (13), 2818–2827. View Source
• [2] NIST Chemical Kinetics Database. Calculated Bond Dissociation Energies for Cu(hfac)L → Cu(hfac) + L (entries for VTMS: 33.6 kcal mol⁻¹; PMe₃: 39.4 kcal mol⁻¹; 2-Butyne: 32.1 kcal mol⁻¹; COD: 35.6 kcal mol⁻¹). View Source

Thermal Stability at Melting Point: (hfac)Cu(PMe₃) Survives ≥4 Days at 67 °C, Whereas (acac)Cu(PMe₃) and (tfac)Cu(PMe₃) Decompose Rapidly

In a systematic study of (β-diketonate)Cu(PMe₃) compounds, Shin et al. quantified thermal stability by isothermal hold at the melting point. (hfac)Cu(PMe₃) (mp = 67 °C) remained unchanged by ¹H NMR and visual inspection for at least 4 days at 67 °C [1]. In contrast, the acetylacetonate analog (acac)Cu(PMe₃) and the trifluoroacetylacetonate analog (tfac)Cu(PMe₃) exhibited only limited thermal stability under identical conditions, undergoing visible decomposition [1]. This difference is attributed to the electron-withdrawing effect of the fully fluorinated hfac ligand, which stabilizes the Cu(I) oxidation state against disproportionation or ligand redistribution.

Evidence Dimension	Isothermal thermal stability at melting point
Target Compound Data	Stable for ≥4 days at 67 °C (mp)
Comparator Or Baseline	(acac)Cu(PMe₃): limited thermal stability at mp; (tfac)Cu(PMe₃): limited thermal stability at mp
Quantified Difference	(hfac)Cu(PMe₃) > 4 days stable; (acac)Cu(PMe₃) and (tfac)Cu(PMe₃) decompose within <1 day (not quantified beyond 'limited stability')
Conditions	Isothermal hold at respective melting point; monitored by ¹H NMR spectroscopy and visual inspection; neat compound under inert atmosphere

Why This Matters

The ability to withstand prolonged heating at the delivery temperature (typically 35–60 °C for bubbler-based CVD) without decomposition ensures reproducible precursor delivery rates over multi-day deposition campaigns, reducing source changeover frequency and lot-to-lot process drift.

• [1] Shin, H. K.; Chi, K. M.; Farkas, J.; Hampden-Smith, M. J.; Kodas, T. T.; Duesler, E. N. Chemistry of Copper(I) β-Diketonate Complexes. 2. Synthesis, Characterization, and Physical Properties of (β-Diketonato)copper(I) Trimethylphosphine and Bis(trimethylphosphine) Complexes. Inorg. Chem. 1992, 31, 424–431. View Source

Vapor Pressure and Volatility: (hfac)Cu(PMe₃) Delivers 100 mTorr at 60 °C with a Sublimation Temperature of 35 °C

The vapor pressure of (hfac)Cu(PMe₃) was measured as a function of temperature, yielding 100 mTorr (0.133 hPa) at 60 °C [1]. From the Clausius–Clapeyron plot, an enthalpy of vaporization (ΔHvap) of approximately 10 kcal mol⁻¹ was derived [1]. The sublimation temperature at reduced pressure is reported as 35 °C [2]. By comparison, (hfac)Cu(VTMS) exhibits a vapor pressure of approximately 0.4 hPa (300 mTorr) at 40 °C (313 K) [3], indicating roughly 3× higher volatility at equivalent temperatures. However, this higher volatility of (hfac)Cu(VTMS) comes at the cost of lower thermal stability, as noted in multiple comparative studies [4].

Evidence Dimension	Vapor pressure and sublimation behavior
Target Compound Data	100 mTorr at 60 °C; ΔHvap ≈ 10 kcal mol⁻¹; sublimation temperature ~35 °C
Comparator Or Baseline	(hfac)Cu(VTMS): ~0.4 hPa (≈300 mTorr) at 40 °C (313 K); (hfac)Cu(COD): comparable vapor pressure, different thermal stability profile
Quantified Difference	(hfac)Cu(VTMS) approximately 3× more volatile at 40 °C, but with documented thermal stability limitations during evaporation [4]
Conditions	Vapor pressure measured by static or saturation method; (hfac)Cu(PMe₃) data from Shin et al. 1992; (hfac)Cu(VTMS) data from modeling studies

Why This Matters

The moderate volatility of (hfac)Cu(PMe₃) is adequate for standard bubbler-based delivery at 40–60 °C, and its combination with superior thermal stability means that the vapor-phase composition remains chemically homogeneous over extended delivery periods, unlike higher-volatility analogs that may partially decompose in the bubbler.

• [1] Shin, H. K.; Chi, K. M.; Farkas, J.; Hampden-Smith, M. J.; Kodas, T. T.; Duesler, E. N. Inorg. Chem. 1992, 31, 424–431. View Source
• [2] Electronic Structures of Copper(I) and Silver(I) β-Diketonate Complexes. Reported sublimation temperatures: (hfac)Cu(PMe₃) 35 °C; (hfac)Cu(CNMe) 60 °C; (hfac)Ag(PMe₃) 91 °C. View Source
• [3] Modeling of the Kinetics in Copper Chemical Vapor Deposition from Cu(hfac)VTMS. Reports vapor pressure of 0.4 hPa at 313 K for (hfac)Cu(VTMS). View Source
• [4] Patent EP 0 466 345 A2 / related prior art. Notes that (hfac)Cu(VTMS) has poor thermal stability and Cu(hfac)₂ requires excessively high decomposition temperatures. View Source

Parasitic Reaction Suppression: Cu(I) PMe₃ Adducts Dramatically Reduce Uncontrolled Gas-Phase Reactivity vs. Cu(II) hfac Sources During Al–Cu Co-Deposition

In a comparative CVD study of Al–Cu metallization thin films using trimethylamine alane as the aluminum source, three copper sources were evaluated side by side: bis-(hexafluoroacetylacetonato)copper(II), Cu(hfac)₂; (cyclopentadienyl)copper(I) triethylphosphine, CpCu(PEt₃); and (hfac)Cu(PMe₃) [1]. Cu(hfac)₂ exhibited a 'massive parasitic reaction' that made control of the copper content in the co-deposited film very difficult [1]. In contrast, both Cu(I) complexes—including (hfac)Cu(PMe₃)—showed 'very minor parasitic reactions,' enabling straightforward control of film composition up to approximately 5 wt% Cu by simply varying the Cu source temperature [1].

Evidence Dimension	Severity of parasitic gas-phase reaction during Al–Cu co-deposition
Target Compound Data	Very minor parasitic reaction; Cu content controllable up to ~5 wt% by source temperature variation
Comparator Or Baseline	Cu(hfac)₂ (Cu(II) source): massive parasitic reaction, very difficult composition control; CpCu(PEt₃): very minor parasitic reaction (comparable to PMe₃ adduct)
Quantified Difference	Qualitative ranking: Cu(hfac)₂ ≫ (hfac)Cu(PMe₃) ≈ CpCu(PEt₃) in parasitic reaction severity; quantitative Cu content control achievable only with Cu(I) precursors
Conditions	Vertical low-pressure cold-wall CVD reactor; Al source: trimethylamine alane at 20 °C; substrates: Si, SiO₂, TiN; deposition temperature ~350 °C

Why This Matters

For multi-metal CVD processes (e.g., Al–Cu or Cu–Mn seed layers), a precursor that minimizes parasitic gas-phase reactions prevents uncontrolled film stoichiometry, reactor fouling, and particulate generation—directly impacting manufacturing yield and tool uptime.

• [1] CVD and Characterization of Al–Cu Metallization Thin Films. MRS Online Proceedings Library 1992, 260, 119–124. View Source

Deposited Copper Film Resistivity: (hfac)Cu(PMe₃) Yields Films with 4 μΩ·cm Resistivity on Non-Irradiated PTFE at 200 °C

Selective chemical vapor deposition of copper onto chemically etched poly(tetrafluoroethylene) (PTFE) using (hfac)Cu(PMe₃) at a substrate temperature of 200 °C produced continuous Cu films with a resistivity of 4 μΩ·cm [1]. X-ray photoelectron spectroscopy (XPS) confirmed the deposited films were pure copper with only surface impurities attributed to air exposure during sample transfer [1]. For comparison, (hfac)Cu(1,5-COD) deposited under hot- and cold-wall CVD conditions yielded film resistivities in the range of 1.9–5.7 μΩ·cm depending on process conditions [2], while (hfac)Cu(VTMS) films exhibited a strong dependence of resistivity on precursor partial pressure, with optimal values approaching bulk copper resistivity only within narrow pressure windows [3].

Evidence Dimension	Electrical resistivity of as-deposited copper films
Target Compound Data	4 μΩ·cm (selective CVD on PTFE, 200 °C)
Comparator Or Baseline	(hfac)Cu(1,5-COD): 1.9–5.7 μΩ·cm (range across multiple conditions); (hfac)Cu(VTMS): strong pressure dependence; bulk Cu: 1.67 μΩ·cm
Quantified Difference	(hfac)Cu(PMe₃) resistivity ~2.4× bulk Cu resistivity; within the same range as (hfac)Cu(COD) best values but achieved on a challenging low-surface-energy substrate (PTFE)
Conditions	Selective CVD at 200 °C on chemically etched PTFE; three-step patterned deposition process involving x-ray/e-beam crosslinking, chemical etching, and CVD

Why This Matters

Demonstrating low-resistivity copper films on a challenging substrate (PTFE) indicates that (hfac)Cu(PMe₃) can deliver electronic-grade film quality without requiring high substrate temperatures or aggressive surface pre-treatments, broadening the range of compatible device architectures.

• [1] Rye, R. R.; Chi, K.-M.; Hampden-Smith, M.; Kodas, T. T. Patterned Deposition of Copper on Poly(tetrafluoroethylene). J. Electrochem. Soc. 1992, 139 (6), L60–L61. View Source
• [2] Chemical Vapor Deposition of Copper via Disproportionation of Hexafluoroacetylacetonato(1,5-cyclooctadiene)copper(I), (hfac)Cu(1,5-COD). J. Mater. Res. 1992; reports resistivities of 1.9–5.7 μΩ·cm with no detectable impurities by AES. View Source
• [3] Jain, A.; Chi, K.-M.; Kodas, T. T.; Hampden-Smith, M. J. Chemical Vapor Deposition of Copper from (hfac)Cu(VTMS): Deposition Rates, Mechanism, Selectivity, Morphology, and Resistivity as a Function of Temperature and Pressure. J. Electrochem. Soc. 1993, 140 (5), 1434–1439. View Source

Selective Deposition on W vs. SiO₂: (hfac)Cu(PMe₃) Enables Intrinsic Surface Selectivity Without Ex Situ Surface Modification Steps

A series of (hfac)CuL precursors—including L = PMe₃, 1,5-COD, VTMS, and 2-butyne—were evaluated for selective copper deposition onto tungsten surfaces in the presence of unmodified SiO₂ [1]. The patent explicitly identifies (hfac)Cu(PMe₃) as one of the four precursor candidates capable of selective deposition, where the inherent differential reactivity between W and SiO₂ surfaces eliminates the need for ex situ SiO₂ passivation with chlorosilanes [1]. While (hfac)Cu(VTMS) required SiO₂ surface pre-treatment with chlorotrimethylsilane or dimethyldichlorosilane to achieve selectivity [1], (hfac)Cu(PMe₃) and (hfac)Cu(2-butyne) exhibited intrinsic selectivity under the tested conditions [2].

Evidence Dimension	Intrinsic selectivity for W vs. unmodified SiO₂ surfaces
Target Compound Data	Selective deposition on W in presence of unmodified SiO₂ demonstrated
Comparator Or Baseline	(hfac)Cu(VTMS): requires SiO₂ pre-treatment with chlorosilanes for selectivity; (hfac)Cu(2-butyne) and (hfac)Cu(1,5-COD): also show selectivity under certain conditions
Quantified Difference	Binary selectivity achieved without ex situ SiO₂ passivation step for (hfac)Cu(PMe₃); (hfac)Cu(VTMS) requires additional processing step
Conditions	Test wafers with patterned W and SiO₂ features; deposition conditions as specified in US 5,358,743

Why This Matters

In semiconductor interconnect fabrication, eliminating the SiO₂ passivation step from the process flow reduces cycle time, chemical consumption, and the risk of surface contamination—directly impacting manufacturing cost and yield for selective copper via-fill applications.

• [1] Jain, A.; Farkas, J.; Kodas, T. T.; Chi, K.-M.; Hampden-Smith, M. J. US Patent 5,358,743. Selective and Blanket Chemical Vapor Deposition of Cu from (β-Diketonate)Cu(L)n by Silica Surface Modification. Filed November 23, 1992; issued October 25, 1994. View Source
• [2] Jain, A.; Chi, K.-M.; Hampden-Smith, M. J.; Kodas, T. T.; Paffett, M. F.; Farr, J. D. Control of Selectivity During Chemical Vapor Deposition of Copper from Copper(I) Compounds via Silicon Dioxide Surface Modification. Chem. Mater. 1991, 3 (6), 995–997. View Source

Procurement-Relevant Application Scenarios Where (hfac)Cu(PMe₃) Differentiation Drives Selection Decisions

Al–Cu or Multi-Metal CVD Co-Deposition Requiring Minimal Parasitic Gas-Phase Reactivity

When co-depositing copper with aluminum or other metals for interconnect metallization, the choice of copper precursor directly determines the severity of uncontrolled gas-phase reactions. The MRS head-to-head comparison demonstrated that Cu(II) sources such as Cu(hfac)₂ generate massive parasitic reactions that render film composition control nearly impossible, whereas (hfac)Cu(PMe₃) exhibits only very minor parasitic reactions, enabling precise composition tuning up to ~5 wt% Cu via simple source-temperature adjustment [1]. For process engineers developing Al–Cu or Cu–Mn seed layers, selecting (hfac)Cu(PMe₃) eliminates the need for complex gas-phase separation strategies or post-deposition composition correction steps.

Selective Copper Via-Fill on Tungsten Contacts Without Ex Situ SiO₂ Passivation

In back-end-of-line (BEOL) interconnect fabrication, selective copper deposition onto tungsten vias or contacts while leaving surrounding SiO₂ dielectric uncoated is a high-value process capability. The patent literature confirms that (hfac)Cu(PMe₃) can achieve this selectivity intrinsically, without requiring the ex situ SiO₂ surface passivation step (chlorosilane treatment) that is necessary for the more volatile but less surface-discriminating (hfac)Cu(VTMS) [1]. This translates to a simplified process flow with fewer chemical handling steps, reduced queue time between surface preparation and deposition, and lower risk of chloride contamination in the final device.

Extended-Duration CVD Campaigns Requiring Stable Precursor Delivery Without Source Degradation

For high-volume manufacturing or multi-wafer research campaigns where the precursor source must remain at elevated delivery temperature (40–60 °C) for days without degradation, (hfac)Cu(PMe₃) offers a uniquely quantified stability advantage. The experimental demonstration of ≥4-day thermal stability at the melting point (67 °C) [1] contrasts sharply with (acac)Cu(PMe₃) and (tfac)Cu(PMe₃), which decompose under identical conditions, and with (hfac)Cu(VTMS), whose lower thermal stability is a known limitation for reproducible selective growth [2]. Procurement of (hfac)Cu(PMe₃) for long-duration deposition workflows reduces the risk of mid-campaign source replacement, minimizes lot-to-lot variability, and preserves the calibration of mass-flow-controlled delivery systems.

Copper Metallization on Temperature-Sensitive or Low-Surface-Energy Substrates

The demonstration of continuous, 4 μΩ·cm copper films deposited on PTFE at 200 °C using (hfac)Cu(PMe₃) [1] indicates that this precursor can nucleate and grow high-quality copper on substrates that are challenging for other Cu(I) precursors. For applications involving flexible electronics, polymer substrates, or MEMS devices where substrate temperature is constrained and surface energy is low, the combination of moderate delivery temperature (sublimation at ~35 °C), sufficient thermal stability to avoid premature decomposition, and demonstrated film quality on PTFE makes (hfac)Cu(PMe₃) a scientifically justified candidate for process development.

• [1] Shin, H. K.; Chi, K. M.; Farkas, J.; Hampden-Smith, M. J.; Kodas, T. T.; Duesler, E. N. Chemistry of Copper(I) β-Diketonate Complexes. 2. Inorg. Chem. 1992, 31, 424–431. View Source
• [2] Choi, H.; Hwang, S. Copper(I) tert-Butyl 3-Oxobutanoate Complexes as Precursors for Chemical Vapor Deposition of Copper. Chem. Mater. 1998, 10 (9), 2326–2328. Notes that (hfac)Cu(TMVS) stability is 'not satisfactory for the selective growth of copper films with reproducibility.' View Source