materials
Article
New Volatile Perfluorinated Amidine–Carboxylate Copper(II)
Complexes as Promising Precursors in CVD and
FEBID Methods
Katarzyna Madajska
and Iwona Barbara Szymańska *
Department of Chemistry, Nicolaus Copernicus University in Toruń, Gagarina 7, 87-100 Toruń, Poland;
502533@doktorant.umk.pl
* Correspondence: pola@umk.pl; Tel.: +48-56-611-4317
Citation: Madajska, K.; Szymańska,
I.B. New Volatile Perfluorinated
Amidine–Carboxylate Copper(II)
Complexes as Promising Precursors
in CVD and FEBID Methods.
Materials 2021, 14, 3145. https://
Abstract: In the present study, we have synthesised and characterised newly copper(II) complexes
with the general formula [Cu2 (NH2 (NH=)CC2 F5 )2 (µ–O2 CRF )4 ], where RF = CF3 , C2 F5 , C3 F7 , C4 F9 .
Infrared spectroscopy, mass spectrometry with electron ionisation (EI MS), and density-functional theory (DFT) calculations were used to confirm compounds’ composition and structure. The volatility of
the compounds was studied using thermal analysis (TGA), EI MS mass spectrometry, variable temperature infrared spectroscopy (VT IR), and sublimation experiments. Research has revealed that these
compounds are the source of metal carriers in the gas phase. The thermal decomposition mechanism
over reduced pressure was proposed. TGA studies demonstrated that copper transfer to the gaseous
phase occurs even at atmospheric pressure. Two selected complexes [Cu2 (NH2 (NH=)CC2 F5 )2 (µ–
O2 CC2 F5 )4 ] and [Cu2 (NH2 (NH=)CC2 F5 )2 (µ–O2 CC3 F7 )4 ] were successful used as chemical vapour
deposition precursors. Copper films were deposited with an evaporation temperature of 393 K
and 453 K, respectively, and a decomposition temperature in the range of 573–633 K without
the use of hydrogen. The microscopic observations made to investigate the interaction of the
[Cu2 (NH2 (NH=)CC2 F5 )2 (µ–O2 CC2 F5 )4 ] with the electron beam showed that the ligands are completely lost under transmission electron microscopy analysis conditions (200 keV), and the final
product is copper(II) fluoride. In contrast, the beam energy in scanning electron microscopy (20 keV)
was insufficient to break all coordination bonds. It was shown that the Cu-O bond is more sensitive
to the electron beam than the Cu-N bond.
doi.org/10.3390/ma14123145
Academic Editor: Piotr Smoleński
Keywords: chemical vapour deposition; focused electron beam induced deposition; volatile compounds; volatility study; nanomaterials; electron interactions; copper fluoride; thermal properties
Received: 26 April 2021
Accepted: 2 June 2021
Published: 8 June 2021
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
published maps and institutional affiliations.
Copyright: © 2021 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
1. Introduction
Nowadays, copper compounds are used, for example, to design hybrid organicinorganic materials [1–3], anticancer agents [4,5], and sensors [6,7]. Additionally, copper
complexes are used as model compounds and tested as antimicrobial agents [8–10].
On the other hand, copper nanostructures are used, for example, in electronics [11–13],
plasmonics [14,15], catalysis [16–18], and as antibacterial agents [19,20]. These structures
can be obtained by gas-assisted methods such as chemical vapour deposition (CVD) [21–23]
and focused electron beam induced deposition (FEBID) [24–26]. CVD is a process where
one or more volatile precursors are transported in the vapour carrier gas to the reactor
chamber, where they decompose on a heated substrate and deposit a solid material [27].
In FEBID, the precursor is supplied without the use of carrier gas, then the adsorbed
compound is dissociated in the focus of an electron beam provided by a scanning or
transmission electron microscope (SEM/TEM) [28]. For the FEBID deposition of metals,
CVD precursors are often tested. However, in addition to being sufficiently volatile, the
compound must also decompose under the influence of an electron beam [29]. In copper
FEBID, several complexes previously used as CVD precursors, such as [Cu(hfac)(VTMS)],
Materials 2021, 14, 3145. https://doi.org/10.3390/ma14123145
https://www.mdpi.com/journal/materials
Materials 2021, 14, 3145
2 of 24
[Cu(hfac)(MHY)], and [Cu(hfac)(DMB)] of Cu(I); [Cu(tbaoac)2 ], [Cu(hfac)2 ]·H2 O, and most
recently [Cu2 (µ-O2 CC2 F5 )4 ] of Cu(II), were applied, but the purity (10–26 at.%) of the
formed deposits remains a problem [30–33].
There are no literature data about the synthesis of Cu(II) perfluorinated amidinate and
heteroleptic amidine-carboxylate compounds and research on their volatility to the best
of our knowledge. In the available literature, only silver(I) and mercury(II) compounds
with perfluorinated amidinates are known: [Ag2 ((NH)2 CCF3 )2 ], [Ag2 ((NH)2 CC2 F5 )2 ], and
[Hg((NH)2 CC2 F5 )2 ], which however, have not been tested for use in vapor deposition
methods [34]. Only copper(I) amidinate complexes with non-fluorinated ligands of the
general formula [Cu2 (AMD)2 ], known from the ALD (atomic layer deposition) process [35],
have been tested in the CVD process. For example, when [Cu2 (i-Pr-MeAMD)2 ] was
vaporised at 368 K and hydrogen was used as a reducing gas, pure copper films were
obtained in the temperature range 473–573 K [36]. Using the same precursor, copper films
were deposited on steel, silicon, and SiO2 /Si substrates at 473–623 K, at a total pressure
of 1333 Pa with hydrogen as a reactive gas [37]. On the other hand, in 2019, (N,N′ -di-secbutylacetamidinato)copper(I) was used to produce Cu2 O thin films in a water assistance
process in a low-temperature range of 398−473 K [38]. In all of the above-mentioned
CVD processes, although copper(I) compounds were used, which are often spontaneously
disproportionate to copper(0), it was necessary to use hydrogen as a reducing agent.
Furthermore, both the synthesis and the manipulation of the used compounds should be
carried out under inert gas [39].
It is known that fluorinated ligands favour thermal stability and vapour pressure [40].
Copper CVD precursors containing perfluorinated carboxylates and N-donor ligands are
known. [Cu2 (t BuNH2 )2 (µ-O2 CRF )4 ] (RF = Cn F2n+1 , n = 1−6) were the source of volatile
copper species: [Cu2 (t BuNH2 )2 (µ-O2 CRF )4 ] and [Cu2 (O2 CRF )2 ], which were transported in
the gas phase in the temperature range of 473–533 K. Metallic copper films were obtained
between 623 K and 733 K without an additional reducing agent [41,42].
On the other hand, in their 2008 paper, Brown et al. reported the amidinate–carboxylate
complexes of dimolybdenium(II) and ditungsten(II) [M2 (O2 CMe)2 ((Ni Pr)2 CR)2 ] obtained
from the reaction of the [M2 (O2 CMe)4 ] and the lithium amidinates Li{(Ni Pr)2 CR}, where
M = Mo, W and R = Me or M = Mo and R = –C≡Ct Bu, –C≡CPh, and –C≡C–Fc (Fc =
1-ferrocenyl) [43]. In other studies [Mo(amidinate)(O2 CR)3 ] or [W2 (amidinate)2 (O2 CR)2 ]
complexes with bridging amidinate were obtained in reaction of [M2 (O2 CR)4 ] (M = Mo, W,
R = CF3 ) with one and two equivalents of Li{(N-2,6-i Pr2 C6 H3 )2 CH}(THF)2 , respectively [44].
However, in the mechanochemical reaction of unsymmetrical N,N′ -diarylformamidine ligands (Ar-amidine), for example, N-(2-methoxyphenyl)-N′ -2,6-dichorophenyl)-formamidine
with copper(II) acetate and zinc(II) acetate, paddle wheel complexes of general formula
[Zn2 (Ar-amidine)2 (OAc)4 ] and [Cu2 (Ar-amidine)2 (OAc)4 ] were synthesised. In the obtained complexes, the N-coordinated ligand in the axial position remained protonated [45].
We decided to synthesise compounds containing a perfluorinated carboxylate and a
perfluorinated amidine as a secondary ligand in the axial position based on the previously
mentioned literature results. We wanted to investigate how the volatility and thermal stability
of the compound would be affected by introducing the perfluorinated amidine instead of the
previously used amine with bulky group. In this paper, we discuss the synthesis of novel
copper(II) complexes with the general formula [Cu2 (NH2 (NH=)CC2 F5 )2 (O2 CRF )4 ], where
RF = CF3 , C2 F5 , C3 F7 , C4 F9 . For simplicity, the abbreviation AMDH for C2 F5 C(=NH) NH2 is
used in this work. We also describe the structural characteristics and thermal properties of
the new compounds. Importantly, we examined the volatility of the compounds and their
interactions with an electron beam. We selected the most promising compounds and used
them in preliminary chemical vapour deposition experiments based on volatility studies.
2. Materials and Methods
CuCO3 ·Cu(OH)2 (>95%), anhydrous acetonitrile (99.8%), CF3 COOH (99%), and
RCOOH (R = Cn F2n+1 ; n = 2–4; 97–98%) were purchased from Sigma Aldrich (Saint Louis,
Materials 2021, 14, 3145
3 of 24
MO, USA) and used as received. Pentafluoropropylamidine C2 F5 C(=NH)NH2 (AMDH)
(98.7%) was purchased from Apollo Scientific (Stockport, UK) and used as received. Copper(II) carboxylates [Cu2 (µ-O2 CRF )4 ], where RF = Cn F2n+1 , n = 1–4, were prepared as earlier
reported [46].
2.1. Instrumentation
Mass spectra using a Finnigan MAT 95 mass spectrometer, Waltham, MA, USA and
electron ionisation (EI) method were registered over the temperature range of 303–623
K. The Cu content was determined with an iCE3300 FL atomic absorption spectrometer
Thermo Fisher Scientific, Waltham, MA, USA. The content of C was found out with a
Vario MACRO CHN ELEMENTAR Analysensysteme GmbH, Langenselbold, Germany. IR
spectra were measured on a Vertex 70V spectrometer (Bruker Optik, Leipzig, Germany)
using a reflective–single reflection diamond ATR unit (400–4000 cm−1 ). Thermal studies
(TGA/DTA) were performed on an SDT 2960 TA analyser (New Castle, DE, USA) (dry N2 ;
heating rate 2.5 K/min, heating range up to 598–692 K; sample mass 7–13 mg). Variable
temperature infrared spectra (VT IR) were registered using a PerkinElmer Spectrum 2000
spectrometer (Waltham, MA, USA) over the range 400–4000 cm−1 with a medium slit
width and a peak resolution of 2.0 cm−1 . The glass vessel with the precursor sample
(~100 mg) was placed in the homemade reactor tube and heated (from 303 to 753 K)
under a dynamic vacuum (p = 10−1 mbar). IR spectra of vapours collected at selected
temperatures were registered. Preliminary deposition experiments were carried out using
the homemade horizontal hot-wall CVD reactor. Copper films were grown on Si(111) in
argon within 60 min. The vaporisation temperatures TV were 393 K and 453 K, whereas the
deposition temperature values TD were 593 K and 633 K. The morphology and composition
studies of obtained materials were performed using scanning electron microscopy: SEM–
LEO 1430VP, Ltd., Cambridge, UK (operating voltage 20 kV) equipped with an energy
dispersive X-ray spectrometer (EDS) Quantax 200 with detector XFlash 4010 (Bruker AXS
microanalysis GmbH, Berlin, Germany). XRD data were collected with a Philips X’PERT
diffractometer with an X’Celerator Scientific detector (Malvern Panalytical Ltd., Malvern,
UK). Diffractogramms were measured in the 10–100◦ 2θ range, using Cu Kα irradiation
and sample spinning (step size—0.0167◦ , program. res. slit—0.5, measuring time—30 s
per point). SEM images were registered using Quanta 3D FEG (FEI, Hillsboro, OR, USA).
Transmission electron microscopy (TEM G2 F20X-Twin 200 kVFEI, Hillsboro, OR, USA) was
used to determine the composition of the thermal analysis residue, confirm the composition
of the deposited materials, and test the sensitivity of compounds to the high-energy electron
beam. A sample for TEM testing was prepared by dissolving in anhydrous ethanol (99.8%),
afterwards applying a drop on carbon-coated copper or nickel (CVD deposits) mesh with
holes (Lacey type 400 mesh) and evaporating the solvent at room temperature. To identify
the chemical elements, energy-dispersive X-ray spectroscopy (EDX, RTEM model SN9577,
134eV, EDAX, FEI, Hillsboro, OR, USA) spectra and selected area (electron) diffraction
pattern were recorded. Atomic Force Microscopy (AFM) analysis of films was performed
using a Veeco microscope (Veeco, Plainview, NY, USA).
2.2. DFT Calculations
The theoretical structures of complexes were optimised using density functional
theory (DFT). All DFT calculations were performed using the Gaussian 16 (Wallingford,
CT, USA) [47] available in PLGrid Infrastructure and the following sets of functionals
and functional databases: B3LYP/6-31G, B3LYP/SDD(Cu) + 6-31G, RBW91/LANL2DZ,
RBW91/LANL2DZ, MN15/6-31G, M06/6-31G, M06/6-311G(d,p), and M06/LANL2DZ.
The lowest energy of the system was obtained when using the hybrid functional of Truhlar
and Zhao M06 [48] and 6-311G(d,p) basis set. GaussView 5.0 (Wallingford, CT, USA) was
used to visualise optimised structures.
Materials 2021, 14, 3145
4 of 24
2.3. Software
All graphical data were further processed with OriginPro 9.1 (Northampton, MA,
USA). Diagrams of thermal and electron compounds decomposition were drawn using
ChemDraw Ultra 12.0 (Mid-Cambridge, MA, USA). TA Universal Analysis (New Castle,
DE, USA) was used to analyse thermograms. AFM images were prepared using the
Gwyddion program (Brno, Czech Republic), while the roughness was determined using
NanoScope Analysis (Bruker, Billerica, MA, USA). Grain size in the layers was determined
by ImageJ (LOCI, University of Wisconsin, Madison, WI, USA). To assign the decomposition
products in VT IR experiments, the registered spectra were compared with the spectra of
molecules in the gas phase collected in SpectraBase [49].
2.4. Synthesis of [Cu2 (NH2 (NH=)CC2 F5 ))2 (µ-O2 CRF )4 ]) ((1)–(4))
Synthesis was carried out using Schlenk glassware and a glove box in an argon
atmosphere. [Cu2 (µ-O2 CR)4 ] (0.5 mmol, R = CF3 , C2 F5 , C3 F7 , C4 F9 ) was placed in a
round-bottomed flask, and 20 cm3 of anhydrous acetonitrile was added, followed by the
introduction of pentafluoropropylamidine (AMDH, 1 mmol). The resulting dark-blue
solution was stirred for 24 h. After this time, the solvent was removed under reduced
pressure (1.3 × 10−3 –1.3 × 10−4 mbar). The obtained compounds were green-blue solids,
stable in air. Yields ranged between 66% and 83%. No crystals suitable for an X-ray
structure determination were obtained.
[Cu2 (AMDH)2 (µ-O2 CCF3 )4 ] (1) Cu2 O8 N4 H6 C14 F22 (calc./found) % Cu 14.1/15.1, C
18.6/18.3; EI MS T = 396 K (m/z, RI %) [AMDH]+. (162, 36); [Cu(AMDH)]+ (225, 100);
[Cu2 (O2 CCF3 )]+ (239, 27); [Cu(AMDH)2 ]+ (387, 71); [Cu2 (AMD)2 ]+ (448, 29); [Cu2 (AMDH)2
(O2 CCF3 )3 ]+ (789, 67); IR (3519, 3380, 3323, 3258, 3191, 3009, 1703, 1670, 1649, 1597, 1507,
1443, 1376, 1334, 1276, 1189, 1141, 1029, 974, 926, 846, 795, 772, 727, 686, 647, 637, 618, 599,
557, 522, 490, 429, 403 cm−1 ).
[Cu2 (AMDH)2 (µ-O2 CC2 F5 )4 ] (2) Cu2 O8 N4 H6 C18 F30 (calc./found) % Cu 11.5/11.5,
C 19.6/18.4; EI MS T = 353 K (m/z, RI %) [AMDH]+. (162, 39), [Cu(AMDH)]+ (225, 23),
[Cu2 (O2 CC2 F5 )]+ (289, 5), [Cu(AMDH)2 ]+ (387, 14), [Cu2 (AMD)3 ]+ (609, 7), [Cu2 (AMDH)2
(O2 CC2 F5 )3 ]+ (939, 24); IR (3573, 3516, 3390, 3240, 3194, 2955, 1707, 1657, 1603, 1555, 1510,
1424, 1324, 1209, 1157, 1032, 974, 897, 822, 772, 737, 686, 650, 612, 589, 554, 544, 516,
432 cm−1 ).
[Cu2 (AMDH)2 (µ-O2 CC3 F7 )4 ] (3) Cu2 O8 N4 H6 C22 F38 (calc./found) % Cu 9.8/10.4, C
20.2/20.1; EI MS T = 383 K (m/z, RI %) [AMDH]+. (162, 13), [Cu(AMDH)]+ (225, 4),
[Cu2 (O2 CC3 F7 )]+ (339, <1), [Cu(AMDH)2 ]+ (387, 1), [Cu2 (AMDH)(O2 CC3 F7 )]+ (501, 1),
[Cu2 (AMDH)2 (O2 CC3 F7 )3 ]+ (1089, 2); IR (3516, 3387, 3317, 3252, 3198, 3059, 3012, 1710,
1675, 1606, 1504, 1411, 1337, 1270, 1209, 1183, 1166, 1119, 1083, 1029, 971, 936, 820, 772, 743,
721, 682, 650, 637, 602, 564, 532, 483, 429 cm−1 ).
[Cu2 (AMDH)2 (µ-O2 CC4 F9 )4 ] (4) Cu2 O8 N4 H6 C26 F46 (calc./found) % Cu 8.5/8.6, C
20.8/21.3; EI MS T = 396 K (m/z, RI %) [AMDH]+. (162, 23), [Cu(AMDH)]+ (225, 54),
[Cu(AMDH)2 ]+ (387, 25), [Cu2 (O2 CC4 F9 )]+ (389, 21), [Cu2 (AMDH)(O2 CC4 F9 )]+ (551, 9),
[Cu(AMDH)2 (O2 CC4 F9 )]+ (650, 47), [Cu2 (AMDH)2 (O2 CC4 F9 )3 ]+ (1239, 38); IR (3525, 3390,
3294, 3202, 2970, 1704, 1672, 1608, 1577, 1502, 1413, 1353, 1330, 1202, 1160, 1132, 1061, 1032,
974, 936, 913, 884, 874, 817, 772, 743, 714, 679, 650, 612, 554, 528, 426 cm−1 ).
3. Results and Discussion
In the acetonitrile environment, two amidine molecules are coordinated by dinuclear
copper(II) carboxylate and complexes of formula [Cu2 (AMDH)2 (µ-O2 CR)4 ], where R =
CF3 , C2 F5 , C3 F7 , C4 F9 , were formed:
[ Cu2 (µ − O2 CRF )4 ] + 2AMDH → [ Cu2 (AMDH)2 (µ − O2 CRF )4 ]
The obtained compounds (1)–(4) were blue-green air-stable solids.
Materials 2021, 14, 3145
5 of 24
3.1. Infrared Spectra Analysis
In the infrared spectra (Supplementary Materials, Figures S1–S4), characteristic bands
for asymmetric stretching vibrations of the carboxylate group were found over the range of
1649–1675 cm−1 (Table 1, Figure 1). Moreover, the value for symmetric stretching vibration
was observed over the range of 1411–1443 cm−1 . The calculated parameter ∆ν = νasCOO −
νsCOO , which is used for proposing of carboxylate coordination mode, was 206 cm−1 for
the compound [Cu2 (AMDH)2 (µ-O2 CCF3 )4 ] (1), which was close to the ∆ν(RCOONa) value
for sodium trifluoroacetate (223 cm−1 ). This suggests the bridging coordination mode
of the carboxylate ligand. A similar relationship ∆ν ∼
= ∆ν(RCOONa) occurred for the next
compounds (2)–(4) and the formation of the carboxylate bridges was proposed.
Table 1. Selected IR absorption bands (cm−1 ) of the studied compounds.
Compound
νasNH2
ν=NH
δNH2
νN=C–N
νasCOO
νsCOO
∆COO
[Cu2 (AMDH)2 (µ-O2 CCF3 )4 ] (1)
[Cu2 (AMDH)2 (µ-O2 CC2 F5 )4 ] (2)
[Cu2 (AMDH)2 (µ-O2 CC3 F7 )4 ] (3)
[Cu2 (AMDH)2 (µ-O2 CC4 F9 )4 ] (4)
AMDH
3380
3390
3387
3390
3362
3009
3240
3012
3202
3130
1597
1603
1606
1608
1593
1507
1510
1504
1502
1450
1649
1657
1675
1672
-
1443
1418
1411
1413
-
206
239
264
259
-
∆ν(CF3 CO2 Na) = 223 cm−1 , ∆ν(C2 F5 CO2 Na) = 268 cm−1 , ∆ν(Cn F2n+1 CO2 Na; n = 3–4) = 272 cm−1 .
Figure 1. Infrared spectra for the reactants and the reaction product [Cu2 (AMDH)2 (µ-O2 CC2 F5 )4 ] (2).
Bands characteristic of the amidine molecule were also observed in the spectrum:
stretching vibration of the C=N bond (1703–1710 cm−1 ), stretching vibration of a =NH
group (3009–3240 cm−1 ), deformation vibrations of the NH2 amino group (1597–1608 cm−1 ),
and stretching of N=C–N group (1502–1510 cm−1 ). The coordination shifts of the signals
mentioned above (relative to the free ligand) were in the range of 39–46 cm−1 , 72–121 cm−1 ,
4–15 cm−1 , and 52–60 cm−1 , respectively (Table 1). The obtained results suggest the amidine coordination through the =NH group because the coordination shift for the vibrations
of this group has the most significant value.
Materials 2021, 14, 3145
6 of 24
Interestingly, the band characteristic for the stretching vibrations of the =NH group
shifts towards lower values for compounds (1) and (3) with an odd number of carbon
atoms in the carboxylate chain and higher values for compounds (2) and (4) with an even
number of carbon atoms in the carboxylate chain. There is also a difference in the area of
the stretching bonds of the =NH and NH2 groups (Figure 1 and Supplementary Materials,
Figures S1–S4). It seems that, for compounds (1) and (3), the amidine arrangement in the
structure is different than for compounds (2) and (4).
3.2. DFT Calculations
To confirm the way of amidine coordination, we decided to optimise the geometry
using computational chemistry methods. For the complex [Cu2 (AMDH)2 (µ–O2 CCF3 )4 ]
(1), calculations were started for the two structures in which the amidine coordination
occurred through the =NH group (Figure 2a) or –NH2 (Figure 2b). The calculations showed
that the lower energy structure is that in which the amidine was bound via the = NH
group (Figure 2). Therefore, in the case of complex (2), calculations were made only for the
molecule with an amidine coordinated by the =NH bond (Figure 2c) owing to the size of
the molecule, and thus the long computation time.
Figure 2. Optimised compounds structures by the density functional theory (DFT) M06/6-311G(d,p)
method: (a) [Cu2 (AMDH)2 (µ–O2 CCF3 )4 ] (1), =NH bonded; (b) [Cu2 (AMDH)2 (µ–O2 CCF3 )4 ] (1),
–NH2 bonded; (c) [Cu2 (AMDH)2 (µ–O2 CC2 F5 )4 ] (2), =NH bonded.
It should be noted that the calculated structures are asymmetric, there is a distortion
of the coordinated carboxylate, and the Cu-O bonds are not equal (Figure 2). There are also
differences in the spatial arrangement of amidines. The optimised structures confirm the
different arrangement of the amidines in the structure for the compound (1) and (2). One
of the amidine groups H2 N–C=NH is perpendicular to the copper atoms, while the other
is parallel.
Materials 2021, 14, 3145
7 of 24
Infrared spectra (Figure 3) were also calculated for the two alternative optimised
structures to confirm the assigned bands in the experimental spectra of the complexes
(Table 1). In the area of stretching bonds characteristic of the –NH2 and =NH groups for
compound (1), in the case of amidine coordination by the =NH group (Figure 2a), a slight
shift only in the position of the asymmetric stretching band of the –NH2 group vibrations
in relation to the free ligand is calculated− (3716 cm−1 , AMDH; 3722 cm−1 , 3700 cm−1 ,
complex (1)). On the other hand, the band characteristic for the stretching vibrations
of the =NH group shifts towards lower values (3540 cm−1 → 3498 cm−1 , 3485 cm−1 ).
In the case of compound (1) alternative structure (Figure 2b) in which the amidine was
coordinated by the –NH2 group, a significant shift of the NH2 asymmetric stretching
ν
vibrations towards lower values was observed (3716 cm−1 → 3591 cm−1 ). Comparing the
theoretical and experimental infrared spectra for complex (1), it was confirmed that the
amidine coordination occurs through the =NH group, because, in the experimental spectra,
no significant shift of the band characteristic for the –NH2 group was detected (Table 1).
Figure 3. Theoretical infrared spectrum calculated by DFT M06/6-311G(d,p): (a) [Cu2 (AMDH)2 (µ-O2 CCF3 )4 ] (1), =NH
bonded amidine, (b) [Cu2 (AMDH)2 (µ-O2 CCF3 )4 ] (1), –NH2 bonded amidine; (c) [Cu2 (AMDH)2 (µ-O2 CC2 F5 )4 ] (1), =NH
bonded (blue line) and AMDH (black line).
Interestingly, in the theoretical spectra, a difference was also observed in the shift
of the stretching =NH band for complex (1) and (2) as in the experimental spectra. In
−1
−1
compound (2), this band shifts towards
ν higher values (3540 cm → 3563 cm ). Moreover,
in −the
there is a double
set of bands for asymmetric
− spectrumδfor complex (1),
− →
−
→theoretical
and symmetric stretching vibrations of the NH2 group. However, in the case of (2), only for
–NH2 symmetrical stretching vibrations, a double set of bands was observed. As a result,
more bands in the region >3000 cm−1 were observed in the spectrum for complex (1) than
for complex (2), and it was also visible in the experimental spectra.
In the case of the stretching vibrations characteristic for the C=N group, no coordination shift was observed in the theoretical spectra (Figure 3). It should be noted that the
νC=N bands overlap with the asymmetric stretching vibrations of the carboxylic groups.
Because the theoretical spectra were calculated for molecules in the gas phase, the individ-
Materials 2021, 14, 3145
8 of 24
ual bands are better separated. In the solid-phase spectrum (Figure 1), one band with the
shoulder was observed, making it difficult to distinguish individual vibrations.
The theoretical spectra also confirmed the proposed assignment of bands characteristic
for deformation vibrations of the –NH2 group and stretching vibrations of the N=C–N
group (Section 3.1). The νas N=C–N shifts more towards higher values (1442 cm−1 →
1485 cm−1 ) than the δNH2 (1602 cm−1 → 1623 cm−1 ).
In conclusion, both the calculated energies and the band coordination shifts in the
experimental and theoretical spectrum confirmed the coordination of the amidine by the
=NH group.
3.3. Thermal Analysis
Thermal analysis was performed to examine the behavior of the complexes during
heating under a nitrogen atmosphere and to determine the final decomposition temperature
and the degradation product. Only for complex (1), two decomposition stages can be
distinguished (Figure 4a). For the other compounds ((2)–(4)), the thermal decomposition
consists of several overlapping steps particularly visible in the derivative graph (Figure 4b).
This may be because, for complexes (2), (3), and (4), a small amount of residue was
observed (Table 2) after the thermal analysis, which means that the complex can be partially
evaporated under atmospheric pressure. However, under these conditions, in addition to
evaporation, compound decomposition also occurs.
Figure 4. (a) Thermogravimetric analysis (TGA) for the complexes: (1), (2), (3), (4); (b) derivative thermogravimetry (DTG)
and TGA for the complex (2) showing the complexity of the decomposition process; (c) transmission electron microscope
(TEM) diffraction pattern for the residue from TGA of (2); (d) energy-dispersive X-ray spectroscopy (EDX) spectrum for the
residue after TGA of (2).
Table 2. Thermal analysis results.
Complex
[Cu2 (AMDH)2 (µ-O2 CCF3 )4 ] (1)
Temperature (K)
Residue Cu2 O (%)
Ti
Tm
Tf
Found
Calc.
391
544
471
573
492
586
13.0
15.8
Materials 2021, 14, 3145
9 of 24
To confirm the composition of the final product of thermal analysis, transmission
electron microscope (TEM) was used owing to the small amount left in the crucible. TEM
diffraction pattern and EDX spectra (Figure 4c,d) revealed that the decomposition product
under the conditions of thermal analysis (atmospheric pressure, nitrogen atmosphere) is
copper(I) oxide.
The temperature of the decomposition process onset changed over the range of
337–391 K, and the temperature of the final product formation varied from 551 K (3)
to 586 K (1) (Table 2). For comparison, in the case of complexes mentioned in the Introduction with amines [Cu2 (t BuNH2 )2 (µ-O2 CRF )4 ], these values were 318–396 K and 458–668 K,
respectively. Interestingly, in the case of complexes with CF3 and C2 F5 substituents, lower
Ti and Tf temperatures were observed for the complexes with tert-butylamine in the axial
position. However, for the C3 F7 and C4 F9 chains, lower temperatures were registered for
the complexes with amidine [41].
3.4. Mass Spectra Analysis
Mass spectra (EI-MS) of the studied compounds were registered between 303 and
623 K and applied to determine the composition of the complex and identify metallated
fragments. The benefit of this approach is that the compound is first evaporated and then
exposed to electrons (70 eV).
For all compounds, the pseudomolecular ion of the [Cu2 (AMDH)2 (O2 CRF )3 ]+ formula, where RF = Cn F2n+1 , n = 1–4, formed by the detachment of one carboxylate ligand, was visible in EI MS spectra (Table 3 and Tables S1–S3). Pseudomolecular ions
of this type were previously observed for [Cu2 (t BuNH2 )2 (µ-O2 CRF )4 ] complexes with
tert-butylamine in the axial position [41]. For compounds (1) and (2), this ion is already
present at low temperatures: 313 K (RI = 4%) and 327 K (RI = 8%), respectively. The
[Cu2 (AMDH)2 (O2 CRF )3 ]+ ion reaches its highest intensity over the temperature range of
353–396 K with only 2% RI for compound (3) and 67% RI for compound (1). The low
intensity of the pseudomolecular ion for compound (3) may result from its high sensitivity
to the electron beam, which caused its rapid decomposition. This is also evidenced by the
high intensity of the [Cu2 F]+ ion, which is the end product of the decomposition (Figure
5) and for the compound (3) reached 97% RI at 550 K (Table S2). Generally, there are also
other detected dinuclear and mononuclear metallated fragments containing both ligands,
such as [Cu2 (AMDH)2 (O2 CRF )2 ]+ , [Cu2 (AMDH)(O2 CRF )3 ]+ , [Cu(AMDH)2 (O2 CRF )]+ , and
[Cu(AMDH)(O2 CRF )]+ . This fact confirms the coordination of the entire amidine molecule
to the perfluorinated copper(II) carboxylate. Moreover, the intact molecule of the tested
compound passes into the gas phase. Interestingly, ions containing Cu(I) detection, such as
[Cu2 (AMDH)2 (O2 CRF )]+ and [Cu2 (AMDH)(O2 CRF )]+ , suggest Cu(II) reduction during the
thermolysis of the compounds. In the case of compounds (3) and (4), even trinuclear and
tetranuclear ions [Cu3 (AMDH)(O2 CC4 F9 )5 ]+ (RImax = 3, 428 K), [Cu3 (AMDH)(O2 CRF )4 ]+
(RImax = 17, 460 K), and [Cu4 (NH2 )2 (AMDH)(AMD)2 (O2 CC4 F9 )2 ]+ (RImax = 4, 460 K) were
observed. This may indicate the polymeric structure of these complexes.
Materials 2021, 14, 3145
10 of 24
Table 3. Mass spectrometry with electron ionisation (EI MS) results for [Cu2 (AMDH)2 (µ-O2 C2 F5 )4 ] (2).
Temperature [K]
Fragments
m/z
]+
[Cu2 (AMDH)2 (O2 CC2 F5 )3
[Cu2 (AMDH)(HN=C=NH)(O2 CC2 F5 )3 ]+
[Cu2 (AMDH)2 (O2 CC2 F5 )2 ]+
[Cu2 (AMD)2 (O2 CC2 F5 )2 ]+.
[Cu2 (AMD)3 ]+
[Cu(AMDH)(AMD)(O2 CC2 F5 )]+
[Cu2 (O2 CC2 F5 )2 ]+.
[Cu2 (AMDH)(O2 CC2 F5 )]+
[Cu2 (AMD)(O2 CC2 F5 )]+
[Cu(AMDH)2 ]+
[Cu(AMD)2 ]+.
[Cu2 (O2 CC2 F5 )]+
[Cu(AMDH)(HN=C=NH)]+
[Cu(AMD)(HN=C=NH)]+
[AMDH]+.
[NH2 CC2 F5 ]+
[Cu2 F]+
[C2 F4 CN]+
[C2 F5 ]+
[CF2 CN]+
[Cu]+
[CO2 H]+
[CO2 ]+.
[HN=C=NH]+.
[F2 ]+.
939
819
776
774
609
549
452
451
450
387
385
289
267
266
162
147
145
126
119
76
63
45
44
42
38
327
353
398
433
529
8
<1
1
<1
2
9
1
2
2
4
6
2
4
1
9
9
62
86
80
40
100
5
9
24
2
3
1
7
30
4
5
4
14
18
5
10
3
39
34
<1
100
1
1
1
1
75
7
3
7
<1
4
1
2
10
3
3
4
9
7
5
6
1
7
11
1
66
100
85
2
45
76
2
3
3
7
1
4
3
4
8
10
3
1
8
23
2
75
1
100
2
90
90
2
2
18
<1
<1
50
12
13
51
100
67
9
24
1
2
1
Figure 5. Fragmentation scheme for metallated fragments of [Cu2 (AMDH)2 (µ-O2 CRF )4 ].
Ions containing deprotonated amidine (i.e., amidinate) having the following composition:
[Cu2 (AMD)(O2 CRF )3 ]+. , [Cu2 (AMD)2 (O2 CRF )2 ]+. , [Cu2 (AMD)2 (O2 CRF )-H]+ , [Cu2 (AMD)
(O2 CRF )2 -H]+ , and [Cu2 (I,II)(AMD)(O2 CRF )]+ /[Cu2 (I)(AMD)(O2 CRF )]+. (the highest in-
Materials 2021, 14, 3145
11 of 24
tensities they achieved in the temperature range of 398–486 K) were detected in the spectra.
The obtained results exhibited the deprotonation of the coordinated amidine ligand during
heating (vide infra, Section 3.5) under measurement conditions.
It is also worth noting the formation of ions containing coordinated ligands that were
formed as a result of amidine decomposition, for example, [Cu2 (NH2 )(AMD)2 (O2 CCF3 )]+ ,
[Cu(AMDH)(HN=C=NH)]+ , and [Cu(AMD)(HCN)]+ . This may indicate that the formed
Cu-N bond is strong. Therefore, the bonds in the ligand molecule were broken, and the
amidine/amidinate was not completely detached.
Interestingly, at low temperatures (313–371 K), copper(I) and copper(II) fragments
containing only N-donor ligands are also formed, such as [Cu2 (AMD)3 ]+ [Cu2 (AMD)2 ]+. ,
[Cu(AMDH)2 ]+ , [Cu(AMD)2 ]+. , and [Cu(AMDH)]+ . Moreover, in the case of compound (1),
the [Cu(AMDH)]+. fragment reaches 100% RI at 393 K, while for compound (4), it achieves
99% RI at 428 K. On the other hand, ions containing only the carboxylate ligand reach
their highest intensity at higher temperatures. In the case of [Cu2 (O2 CRF )2 ]+. ion, the
relative intensity varies from 18 to 96% in the temperature range of 505–550 K (Table 3 and
Tables S1–S3). In turn, the ion [Cu2 (O2 CRF )]+ reaches the highest intensity (50–75% RI)
over the 428–538 K temperature range. The obtained results may indicate that the metal
species with the amidinate ligand are more volatile than those with the carboxylate.
At the beginning of heating, fragments such as [AMDH]+ , [C2 F5 ]+ , [CO2 ]+. , [CO2 H]+ ,
[HN=C=NH]+. , and [F2 ]+. appear, which indicate the degradation of the compounds. However, it is difficult to distinguish which ions result from thermal decomposition and which
are the result of interactions with electrons. The appearance of carbodiimide [HN=C=NH]+
suggests carbodiimide detachment as one of the mechanisms of degradation [50]. However,
there is also release of amidine as a decomposition factor. The presence of [CO2 H]+ is
evidence of amidine deprotonation by carboxylate and the formation of carboxylic acid
and amidinate.
[Cu2 F]+ fragment was also detected in MS EI spectra for all studied compounds. It
reaches an intensity of 13 to 97% RI at high temperatures (505–550 K) (Table 3 and Tables
S1–S3). In the case of compound (2), [Cu]+ ion was also observed. The formation of
this type of ions indicates that electrons can cause the decomposition of compounds into
copper(I) fluoride or copper (vide infra). This is a desirable property for the potential
precursors used in the FEBID method.
3.5. Variable Temperature Infrared Spectroscopy (VT IR)
To examine the composition of the gas phase formed during the heating of the complexes, variable temperature infrared spectroscopy was carried out. The measurement
conditions are similar to the deposition parameters during the CVD experiments.
In the first step of decomposition (stage I), the band characteristics of the amidine
molecule (3565 cm−1 ; νas NH2 , 3527 cm−1 ; νs NH2 , 3446 cm−1 ; ν=NH, 1776 cm−1 ; νCN,
1590 cm−1 , δNH) were observed for compounds (1), (3), and (4) in the temperature range of
333–413 K (Figures S1–S3). In the case of compound (2), in addition to the bands mentioned
above, there are the following observed signals: 3191 cm−1 (νNH2 ), 1734 cm−1 (νCN +
νas COO), and 1683 cm−1 (νas COO) characteristic for complex (2), and signals 3515 cm−1
(νNH2 ) and 1705 cm−1 (νCN) assigned to the coordinated amidinate (Cu–AMD) species. It
is difficult to propose a mode of amidinate coordination owing to the lack of spectroscopic
data in the literature for copper compounds with perfluorinated AMD. These species occur
in the gas phase over the temperature range of 353–533 K (Figure 6 and Figure S8).
Materials 2021, 14, 3145
12 of 24
Figure 6. VT IR spectra of [Cu2 (AMDH)2 (µ-O2 CC2 F5 )4 ] (2) in the temperature range of 353–753 K.
Similarly, the second decomposition step (stage II) is comparable for compounds (1),
(3), and (4). HAMD’s characteristic bands are still being observed. The entire molecules
of complexes (1), (3), and (4) are also present at this stage in the temperature range of
433–453 K (1), (3), and 433–473 K (4). Moreover, in the gas phase, Cu–AMD species occur
(Figures S5–S7). In the case of complex (2), in addition to the molecules mentioned above,
signals,
1824 cm−1 (νas COO)
and 1440 cm−1 (νs COO) bands were observed in the gas
− ν
− ν
phase at 553 K, characteristic for carboxylic acid (Figure 6) [51].
At higher temperatures, there are more differences in the way compounds are decomposed. Generally, compounds (1), (3), and (4) have more decomposition stages than
complex (2).
For compound (2) in the temperature range of 573–653 K (stage III), there are characteristic bands for Cu–AMD species, Cu2 (O2 CRF )4 (1651 cm−1 ; νas COO) [42], HAMD,
−
RF COOH [51], CF3 CFO (1889 cm−1 , 736 cm−1 , 676 νcm−1 ) [52], CF2 O (1957 cm−1 , 1936
−
−
−
−
−
cm−1 ) [52], CO2 (2322 cm−1 ) [53], and CO (2174 cm−1 , 2126 cm−1 ) [54]. In turn, in the
−
−
−
673–753 K temperature range (stage IV), there are no more metalled species. Only decomposition products mentioned above such as HAMD, RF COOH, CF3 CFO, CF2 O, CO2 , and
CO occurred (Figure 6).
In the case of compound (1) in the third step of decomposition (stage III), there are
characteristic bands for the entire complex (1) molecules, Cu–AMD, HAMD, and the
carboxylic acid, at the temperature range of 473–533 K. The fourth stage of decomposition
(stage IV) occurs in the temperature range of 553–593 K. The gas-phase contained the
following species: Cu–AMD, Cu2 (O2 CCF3 )4 , HAMD, CF3 COOH [55], CF3 CFO, CO2 , and
CO. The next stage (stage V, 613–633 K) differs from the previous one by the lack of metal
carriers. In the last tested temperature range (stage VI, 653–753 K) in the gas phase, HAMD,
CF3 COOH, CO2 , and CO (Figure S5) were detected.
Materials 2021, 14, 3145
13 of 24
Compared with compound (1), in the third stage of complex (3) decomposition (stage
−
III, 473–493 K), an additional band (2272 cm−1 ) was observed in the VT IR spectra. This
band is characteristic of a gaseous C2 F5 CN. [56] At 513 K (stage IV), the gas phase consists
of the following species: complex (3), Cu–AMD, HAMD, C3 F7 COOH [57], C2 F5 CN, and
CO2 . In turn, in the temperature range of 533–553 K (stage V), in addition to the aforementioned degradation products, CO molecule was detected. In the next step of decomposition
−
(stage VI, 573–593 K), copper(II) perfluorobutyrate (1658 cm−1 ) was the only metal
carrier.
−
−
1
−
1
Additionally, the bands 1795 cm (C=C stretching vibrations) and 1396 cm (asymmetric
C-F− stretching mode in the CF3 group) were recorded in the VT IR spectra characteristic
of the CF2 =CF–CF3 molecule. [58] RF COOH, CF3 CFO, C2 F5 CN, CO2 , and CO were also
observed in the gas phase at these temperatures. With increasing temperature (613–653 K,
stage VI), the composition of the gas phase changes slightly compared with stage V, and
there are no CF3 CFO molecules (Figure S6).
In the case of compound (4), the molecules Cu–AMD, HAMD, and C4 F9 COOH occur
in the third phase of degradation (stage III, 493–533 K). In the next stage of decomposition
(stage IV, 553–573 K), the gas phase is composed of Cu–AMD, HAMD, C4 F9 COOH [59],
CF3 CFO, C2 F5 CN, CO2 , and CO. At higher temperatures (stage V, 593–613 K), the amount
of gaseous products decreased, and Cu–AMD, HAMD, carboxylic acid, and CF3 CFO were
observed. Finally, in the last stage of decomposition (stage VI, 633–753 K), the gas phase
consists only of CF2 =CF–CF3 and C4 F9 COOH (Figure S7).
Based on the VT IR spectra and the ions observed in the gas phase during the EI MS
experiments (Section 3.4), the thermal decomposition mechanism of compounds (1)–(4)
was proposed, as shown in Figure 7.
Figure 7. Proposed mechanism of [Cu2 (AMDH)2 (µ-O2 CRF )4 ] thermal decomposition based on VT IR and EI MS studies.
Based on the results described above, it can be concluded that complex (2) has the
highest volatility. It was observed in the gas phase in the broadest temperature range.
Moreover, the bands characteristic of this compound have a higher intensity than the bands
assigned to degradation products in the temperature range of 433–533 K (Figure 8b). The
inverse relationship was observed for compounds (1), (3), and (4) (Figure 8a,c,d). This
means that complex (2) demonstrates the highest concentration in the gas phase. It is also
worth adding that the concentration of other metal carriers is equally high. For this reason,
Materials 2021, 14, 3145
14 of 24
this compound was selected as the best of the newly synthesised potential CVD precursors.
Another compound ensuring a satisfactory amount of metal carriers in the gas phase was
complex (3), which was also selected for testing in the CVD process.
Figure 8. Enlargement of the area of occurrence of metal carriers (green) and degradation products’ (red) characteristic
bands: (a) [Cu2 (AMDH)2 (µ-O2 CCF3 )4 ] (1), (b) [Cu2 (AMDH)2 (µ-O2 CC2 F5 )4 ] (2), (c) [Cu2 (AMDH)2 (µ-O2 CC3 F7 )4 ] (3),
(d) [Cu2 (AMDH)2 (µ-O2 CC4 F9 )4 ] (4).
The obtained complexes (1)–(4) have higher volatility than analogues with tertbutylamine, because the complexes [Cu2 (AMDH)2 (µ-O2 CRF )4 ] go into the gas phase in the
temperature range of 353–553 K, while the [Cu2 (NH2 t Bu)2 (µ-O2 CRF )4 ] compounds are a
source of metal carriers over the temperature range of 373–553 K [42].
It is also worth noting that relatively small changes in the compound structure (the
carbon chain length increased in the carboxylate) affect the mechanism of compound
decomposition. One interesting observation is the lack of a simple correlation between
the carbon chain length in the carboxylate and complex volatility. Both complex (1) and
complex (4) have a low concentration in the gas phase.
Compared with the mass spectrometry results, no characteristic signals for carbodiimide or perfluorinated hydrocarbons
(for compounds (1) and (2)) were detected in the
−
VT IR spectra. It can be concluded that these products are formed only as a result of
electron-induced decomposition rather than thermal decomposition. This observation
confirms that VT IR spectra are an effective method for testing CVD precursors because the
measurement conditions (10−1 mbar) are close to those in the used CVD reactor.
3.6. Sublimation Experiments
To further investigate the volatile species, sublimation experiments at 393 K and
1.3 × 10−4 mbar were performed for compounds (1) and (2). Then, IR spectra were taken
for the compounds deposited on the cooling finger. As can be seen for compound (2), the
Materials 2021, 14, 3145
15 of 24
−
spectra before and after sublimation are similar (Figure 9 bottom). There are only slight
differences in the positions and intensities of the bands assigned to the N-H stretching
and deformation vibration. Therefore, it can be concluded that compound (2) evaporates
without decomposition. In the case of compound (1), the changes in the IR spectra before
and after sublimation are much more noticeable (Figure 9 top). It is worth mentioning
that there are bands in the spectrum of the sublimation product characteristic for both
coordinated ligands (3155 cm−1 , νN-H ; 1724 cm−1 , νC=N ; 1666 cm−1 , νC=O ; 1543 cm−1 ,
− ν
− ν
− ν
−
δNH ). This proves that compound
(1) is a source
of the metal carrier
in the gasδphase. It
is likely that, during heating, a rearrangement in the structure of the compound occurs.
The reason may be the breaking of hydrogen bonds between the molecules of the heated
compound [60].
Figure 9. Infrared spectra for the compound [Cu2 (AMDH)2 (µ-O2 CCF3 )4 ] (1) (top) and [Cu2 (AMDH)2 (µ-O2 CC2 F5 )4 ] (2)
(bottom) before sublimation (gray) and after sublimation (black).
3.7. TEM and SEM Observations
To initially assess the sensitivity of compound (2) to the high-energy electron beam,
observations were made using a scanning electron microscope (SEM, 20 keV) and a transmission electron microscope (TEM, 200 keV). Complex (2) was selected for testing because
it demonstrates the best volatility. For SEM analysis, the compound was prepared by
depositing its layer on a silicon wafer using the earlier described sublimation method
(Section 3.6). For TEM analysis, the compound was dissolved in ethanol, and the resulting
solution was applied to a mesh and allowed to solvent evaporate.
EDX spectra were taken during the SEM analysis, each of them recorded for 30 s. During the experiment, the scanning area was reduced, which increases the dose of electrons
(Figure 10). With the increasing dose of electrons, the oxygen content decreases, the carbon
content is slightly reduced, and the copper content increases. The contents of the other
elements were almost the same.
Materials 2021, 14, 3145
16 of 24
Figure 10. Examined scan areas’ EDX spectra for [Cu2 (AMDH)2 (O2 CC2 F5 )4 ] (2) deposited on a
silicon wafer.
Based on the obtained results and the mass spectra (Section 3.4), the following mechanism of the compound decomposition under SEM conditions was proposed. In the
first stage, a proton transfer from the amidine molecule to the carboxylate probably took
place, resulting in the formation of a coordinated deprotonated amidine and C2 F5 COOH
(Figure 11). Ions with deprotonated amidine were observed in EI MS spectra (Table 3).
Acid was decomposed by electrons, forming CO2 and C2 F5 H. This is confirmed by the EI
MS spectra of the complexes (Section 3.4) and C2 F5 COOH [61], in which no molecular ion
derived from the acid was observed. The unchanged amount of fluorine before compound
interaction with electrons may indicate that C2 F5 H has not been released from the layer as
a gas. Stoffels et al. proved by electron attachment mass spectrometry experiment that, in
radio frequency fluorocarbon plasmas of CF4 , C2 F6 , and C4 F8 , polymerisation occurred
readily, and molecules containing up to ten carbon atoms have been found [62]. Therefore,
it is possible that, during the interaction of C2 F5 H with the electron beam, polymerisation
has occurred, and a carbon-fluorine matrix was formed.
Figure 11. Proposed mechanism of [Cu2 (AMDH)2 (O2 CC2 F5 )4 ] (2) decomposition under the influence of an electron beam
in the SEM microscope.
In the second stage of decomposition, decarboxylation occurred, and the perfluorinated groups were still bound to the central atom. Formation of [Cu(CF3 )2 ]− ion, with
Materials 2021, 14, 3145
17 of 24
fluorocarbon group bonded to the copper atom, has been proved by Rijs et al. in their
research on the mechanism of decomposition of [Cu2 (µ-O2 CCF3 )4 ] using mass spectrometry [63]. Moreover, in the electron collision experiments with [Cu2 (µ-O2 CC2 F5 )4 ], the
[Cu(CF3 )2 ]− ion was formed [64].
It is seen that, during the interaction of the electron beam with a solid-phase compound, the amount of energy under SEM conditions is insufficient to remove the coordinated ligands effectively. The obtained results showed that the Cu-O bond is the weakest
in the complex. The proposed mechanism of compound decomposition under the influence of the SEM electron beam is presented in Figure 11. It still seems necessary to
conduct FEBID experiments because, in this method, the precursor is deposited from the
gas phase and, as evidenced by the EI MS spectrum (Section 3.4), in which the precursor is
in gaseous form and interact with low-energy electrons, it is possible to decompose it into
copper(I) fluoride.
During TEM imaging, structural changes of the sample were visible from the beginning of the measurement. After a few seconds, only the image of the sample was
stabilised, and crystallites were observed (Figure 12). The obtained diffraction confirmed
the crystallinity of the formed substance with diffraction rings matching the cubic CuF2 .
Figure 12. Image of the sample after a few seconds of [Cu2 (AMDH)2 (O2 CC2 F5 )4 ] (2) interaction
with electrons and diffraction pattern, which confirms the crystallinity of the deposit.
The EDX spectra of the TEM obtained material showed the following atomic contents
of the elements: 11.6 at.% Cu, 41.2 at.% F, and 47.2 at.% C. Such values suggest that, apart
from copper(II) fluoride formation, after the compound (2) interaction with the high-energy
electron beam, a matrix consisting of fluorine and carbon is formed. However, it should
be taken into account that the TEM grid was covered with carbon, which affects the semiquantitative measurement. It is worth noting that the resulting material does not contain
oxygen and nitrogen, which were in the pristine compound (2). This may indicate the
breaking of Cu-N and Cu-O bonds under the conditions of TEM observation. The proposed
mechanism of compound decomposition under the influence of the TEM electron beam is
presented in Figure 13.
The microscopic observations showed that compound (2) is sensitive to the highenergy electron beam. It is visible that the decomposition of the compound strongly
depends on the beam power and electron dose. When high electron energy (TEM) is used,
CuF2 is the final decomposition product, but in SEM, CO2 detachment and organometallic
species formation were suggested, while electrons have low energy (EI MS, Section 3.4),
and Cu2 F2 is formed.
Materials 2021, 14, 3145
18 of 24
Figure 13. Proposed mechanism of [Cu2 (AMDH)2 (O2 CC2 F5 )4 ] (2) decomposition under the influence of an electron beam
under the TEM analysis conditions.
3.8. CVD Preliminary Experiments
Based on the volatility studies (TA, EI MS, VT IR–Sections 3.3–3.5), complexes (2) and
(3) were selected for the hot-wall CVD experiments (Table 4). In the case of complex (2), the
evaporation temperature of 393 K was sufficient to deposit the layer, while for precursor
(3), it was necessary to increase the temperature to 453 K. This is consistent with the results
of the VT IR analysis (Section 3.5), because for complex (2), the bands characteristic for the
copper carriers appeared at lower temperatures (353 K) than for the complex (3) (433 K),
and additionally they had higher intensities, which indicates a higher concentration of
these metal carriers in the gas phase. The process was carried out using decomposition
temperatures from 573 K to 633 K. The best results were obtained using 633 K as the
deposition temperature TD .
Table 4. Summary of chemical vapour deposition (CVD) conditions for the compounds (2) and (3).
Precursor
Precursor mass (m) [mg]
Vaporization temperature (TV ) [K]
Deposition temperature (TD ) [K]
Carrier gas
Substrates
Deposition time (t) [min]
100
393 (2), 453 (2), (3)
573–633
Ar
Si(111)
60
The deposited layer was smooth (Figures 14b and 15b), and no impurity signals
(C, N, O) were observed in the EDX spectrum (Figure 16a, red line).
In the case of compound (2), nanometric layers with a thickness of 69–360 nm were
produced (Figure 14). It is seen that the higher evaporation temperature of the precursor
results in a greater layer thickness (Figure 14a). This is consistent with the spectra of
infrared spectroscopy because, at 393 K (Figure 8b), the concentration of the metal carriers
in the gas phase was much lower than at 453 K (Figure 8b). Using the higher vaporization
temperature (TV = 453 K) and lower deposition temperature (TD = 593 K) (Figure 14a), the
grains (51–254 nm, Figure 15a) were more dispersed than at TV = 393 K and TD = 633 K
(Figure 14b–d). The calculated roughness parameter Ra for the layer shown in Figure 14a
was 51.0 nm, while for the films shown in Figure 14b–d, it was 11.1 nm, 11.1 nm, and
9.4 nm, respectively. This confirms that the layers deposited at lower vaporization and
higher deposition temperature were smoother. The grain size for the films deposited at
the higher temperatures was in the range of 36–201 nm (Figure 15c,d). Moreover, using
a shorter transport way of the metal carriers, densely packed single grains with a size of
138–445 nm were obtained (Figure 15b).
In the layer with a thickness of 69 nm deposited with complex (3), single, densely
packed, round shape grains were also formed (Figure 17a). The continuity of the layer
is also confirmed by the low value of the roughness parameter (Ra = 9.9 nm). It is worth
noting that using a higher evaporation temperature TV from 551 K (3) to 586 K for complex
(3) leads to obtaining layers of similar thickness than when using a lower temperature for
precursor (2). This also confirms the lower volatility of compound (3).
Materials 2021, 14, 3145
19 of 24
It is worth adding that both the precursor evaporation and decomposition temperatures were lower for the complex with perfluorinated amidine than for the previously used
complex with t BuNH2 [Cu2 (t BuNH2 )2 (µ–O2 CRf )4 ], which was 393 K, 453 K (amidine);
435–473 K (t BuNH2 ) and 573–633 K (amidine); 623–733 K (t BuNH2 ), respectively [42]. This
shows that the introduction of a secondary perfluorinated ligand in the axial position
increases the volatility of the compounds.
The EDX (SEM analysis) and XRD spectra of the layers showed that they contain
metallic copper (Figure 16a,b and Figure 17b,c). The formation of metallic copper during
the CVD process was also confirmed by the diffraction patterns and EDX spectra obtained
by TEM analysis (Figures 16c and 17d,e). The great advantage of compound (2) is that it
was possible to obtain pure copper films from the air-stable copper(II) complex in an argon
atmosphere without using an additional reducing agent. However, in the case of complex
(3), apart from metallic copper, copper(I) oxide and a small amount of carbon (2.9 at.%)
were also detected. In contrast, the layers deposited with the use of [Cu2 (NH2 t Bu)2 (µO2 CRF )4 ] precursors contained metallic copper and a low content of fluorine (1.5 at.%) [42].
Interestingly, despite the high content of fluorine in compounds (2) and (3), it is not present
in the obtained deposits.
Figure 14.
SEM and AFM results of preliminary chemical vapour deposition (CVD) experiment using
[Cu2 (AMDH)2 (O2 CC2 F5 )4 ] (2): (a) TV = 453 K, TD = 593 K; (b–d) TV = 393 K, TD = 633 K, in the case of (b), the transport
way of vapours was shorter.
Materials 2021, 14, 3145
20 of 24
Figure 15. Histograms of a grain-size distribution: (a) [Cu2 (AMDH)2 (O2 CC2 F5 )4 ] (2), TV = 453 K, TD = 593 K;
(b–d) [Cu2 (AMDH)2 (O2 CC2 F5 )4 ] (2), TV = 393 K, TD = 633 K, in the case of (b), the transport way of vapours was shorter;
(e) [Cu2 (AMDH)2 (µ-O2 CC3 F7 )4 ] (3), TV = 453 K, TD = 633 K; (f) [Cu2 (AMDH)2 (µ-O2 CC3 F7 )4 ] (3) TV = 393 K, TD = 573 K.
Figure 16. Composition analysis of the obtained layers using precursor (2) [Cu2 (AMDH)2 (µ-O2 CC2 F5 )4 ]: (a) EDX spectra:
black line: TV = 453 K, TD = 593 K; red, blue and pink line: TV = 393 K, TD = 633 K, in the case of the red line, the transport
way of vapours was shorter; (b) XRD analysis; (c) TEM image; (d) diffraction pattern; (e) EDX spectra from TEM.
Materials 2021, 14, 3145
21 of 24
Figure 17. (a) SEM and AFM results of preliminary CVD experiments for the [Cu2 (AMDH)2 (µ-O2 CC3 F7 )4 ] precursor (3)
TV = 453 K, TD = 633 K; (b) SEM results, [Cu2 (AMDH)2 (µ-O2 CC3 F7 )4 ] (3), TV = 393 K, TD = 573 K; (c) EDX spectra from
SEM; (d) XRD spectra; (e) TEM image and diffraction pattern; (f) EDX spectra from TEM.
4. Conclusions
Copper(II) carboxylate compounds with perfluorinated amidine of the general formula
[Cu2 (AMDH)2 (µ-O2 CRF )4 ], where RF = CF3 , C2 F5 , C3 F7 , C4 F9 , were obtained by the
reaction of copper(II) perfluorinated carboxylates with perfluorinated amidine. The IR
spectra analysis suggested the bridging coordination mode of carboxylates. =NH-bonding
of amidine was proposed from infrared spectra analysis and confirmed by DFT calculations.
The structure of the obtained compounds was also confirmed by the results of the EI MS
analysis. The following pseudomolecular ions [Cu2 (AMDH)2 (µ-O2 CRF )3 ]+ proved the
dimeric structure of the studied complexes with bridging carboxylates forming the “paddlewheel” structure and axially N-coordinated amidines. The results of mass spectrometry
(MS-EI) and variable temperature infrared spectroscopy demonstrated that the synthesised
compounds are a source of copper carriers in the gas phase. Studies have shown that the
decomposition process of the compounds is complicated. However, two main mechanisms
can be distinguished: amidine deprotonation and formation of volatile copper amidinates
and a carboxylic acid, and amidine detachment followed by decarboxylation of copper
carboxylate. Moreover, the thermal analysis showed that copper transfer to the gaseous
phase occurred even at atmospheric pressure.
The preliminary CVD experiments confirmed that compounds (2) and (3) are the
source of metal carriers in the gas phase. Complex (2) was also shown to be the most
volatile. Copper films were successfully deposited using precursor [Cu2 (AMDH)2 (µO2 CC2 F5 )4 ] (2) by the CVD method in the Ar atmosphere. The great advantage of the
[Cu2 (AMDH)2 (µ-O2 CC2 F5 )4 ] (2) is that a metallic film can be obtained from the air-stable
copper(II) compound without the need to use any additional reducing agent such as H2 .
The best results (smooth and pure structures) were achieved when the temperature of the
Materials 2021, 14, 3145
22 of 24
precursor evaporation was 393 K, while the decomposition temperature was 633 K. Thanks
to our research, we have obtained a promising compound that can be used as a precursor
in CVD methods. Moreover, based on the study of the complexes’ interactions with
high-energy electrons, it can be concluded that the energy and dose influence the deposit
composition and compound [Cu2 (AMDH)2 (µ-O2 CC2 F5 )4 ] (2) may also be a precursor in
the FEBID method for the preparation of nanostructures composed of copper(II) fluoride.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/
10.3390/ma14123145/s1, Figure S1. Infrared spectra for the complex [Cu2 (AMDH)2 (µ-O2 CF3 )4 ]
(1); Figure S2. Infrared spectra for the complex [Cu2 (AMDH)2 (µ-O2 C2 F5 )4 ] (2); Figure S3. Infrared
spectra for the complex [Cu2 (AMDH)2 (µ-O2 C3 F7 )4 ] (3); Figure S4. Infrared spectra for the complex
[Cu2 (AMDH)2 (µ-O2 C4 F9 )4 ] (4); Figure S5. VT IR spectra of [Cu2 (AMDH)2 (µ-O2 CCF3 )4 ] (1) in the
temperature range 333–753 K; Figure S6. VT IR spectra of [Cu2 (AMDH)2 (µ-O2 CC3 F7 )4 ] (3) in the
temperature range 333–613 K; Figure S7. VT IR spectra of [Cu2 (AMDH)2 (µ-O2 CC4 F9 )4 ] (4) in the
temperature range 333–633 K; Figure S8. IR spectrum in the gas phase at 353 K for the complex
[Cu2 (AMDH)2 (µ-O2 CC2 F5 )4 ] (2); Table S1. EI MS results for the complex [Cu2 (AMDH)2 (µ-O2 CF3 )4 ]
(1); Table S2. EI MS results for the complex [Cu2 (AMDH)2 (µ-O2 C3 F7 )4 ] (3); Table S3. EI MS results
for the complex [Cu2 (AMDH)2 (µ-O2 C4 F9 )4 ] (4).
Author Contributions: Conceptualisation, K.M. and I.B.S.; investigation, K.M.; writing—original
draft preparation, K.M.; writing—review and editing, I.B.S.; visualisation, K.M.; supervision, I.B.S.;
funding acquisition, K.M. and I.B.S. All authors have read and agreed to the published version of
the manuscript.
Funding: The financing of this work was by Nicolaus Copernicus University in Torun (PDB no.103).
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: The data presented in this study are available on request from the
corresponding author.
Acknowledgments: This research was supported in part by PL-Grid Infrastructure.
Conflicts of Interest: The authors declare no conflict of interest.
References
1.
2.
3.
4.
5.
6.
7.
8.
Padnya, P.; Shibaeva, K.; Arsenyev, M.; Baryshnikova, S.; Terenteva, O.; Shiabiev, I.; Khannanov, A.; Boldyrev, A.; Gerasimov, A.;
Grishaev, D.; et al. Catechol-Containing Schiff Bases on Thiacalixarene: Synthesis, Copper (II) Recognition, and Formation of
Organic-Inorganic Copper-Based Materials. Molecules 2021, 26, 2334. [CrossRef] [PubMed]
Li, X.; Zhong, X.; Hu, Y.; Li, B.; Sheng, Y.; Zhang, Y.; Weng, C.; Feng, M.; Han, H.; Wang, J. Organic-Inorganic Copper(II)-Based
Material: A Low-Toxic, Highly Stable Light Absorber for Photovoltaic Application. J. Phys. Chem. Lett. 2017, 8, 1804–1809.
[CrossRef]
Wang, L.; Sun, H.; Sun, C.; Xu, D.; Tao, J.; Wei, T.; Zhang, Z.; Zhang, Y.; Wang, Z.; Bi, W. Lead-free, stable orange-red-emitting
hybrid copper based organic-inorganic compounds. Dalton Trans. 2021, 50, 2766–2773. [CrossRef]
Doniz Kettenmann, S.; Nossol, Y.; Louka, F.R.; Legrande, J.R.; Marine, E.; Fischer, R.C.; Mautner, F.A.; Hergl, V.; Kulak, N.;
Massoud, S.S. Copper(II) Complexes with Tetradentate Piperazine-Based Ligands: DNA Cleavage and Cytotoxicity. Inorganics
2021, 9, 12. [CrossRef]
Hussain, A.; AlAjmi, M.F.; Rehman, M.T.; Amir, S.; Husain, F.M.; Alsalme, A.; Siddiqui, M.A.; AlKhedhairy, A.A.; Khan, R.A.
Copper(II) complexes as potential anticancer and Nonsteroidal anti-inflammatory agents: In vitro and in vivo studies. Sci. Rep.
2019, 9, 1–17. [CrossRef]
Cao, M.G.; Wang, J.J.; Wang, Y.; Wang, X.F.; Li, J.X.; Chen, J.X.; Hu, B.Q.; Hu, D.D. Two cobalt(II) and copper(II) complexes with
2,4,5-tri(4-pyridyl)-imidazole and 5-hydroxyisophthalic acid as turn-on luminescence sensors for Mg2+, Ca2+ and SCN−ions. Z.
Anorg. Allg. Chem. 2021, 647, 714–721. [CrossRef]
Kanso, H.; Ben Jrad, A.; Inguimbert, N.; Rammal, W.; Philouze, C.; Thomas, F.; Noguer, T.; Calas-Blanchard, C. Synthesis and
Characterization of Bis-1,2,3-Triazole Ligand and its Corresponding Copper Complex for the Development of Electrochemical
Affinity Biosensors. Chem. A Eur. J. 2021, 27, 1–10. [CrossRef]
Guo, R.-F.; Yan, H.-T.; Liu, R.-X.; Li, H.-C.; Liu, Y.-C.; Chen, Z.-F.; Liang, H. Structural characterization and pharmacological
assessment: In vitro/in vivo of a new copper(ii)-based derivative of enrofloxacin. Metallomics 2020, 12, 2145–2160. [CrossRef]
[PubMed]
Materials 2021, 14, 3145
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
23 of 24
Andrejević, T.P.; Aleksic, I.; Počkaj, M.; Kljun, J.; Milivojevic, D.; Stevanović, N.L.; Nikodinovic-Runic, J.; Turel, I.; Djuran, M.I.;
Glišić, B. Tailoring copper(ii) complexes with pyridine-4,5-dicarboxylate esters for anti-: Candida activity. Dalton Trans. 2021, 50,
2627–2638. [CrossRef] [PubMed]
Velo-Gala, I.; Barceló-Oliver, M.; Gil, D.M.; González-Pérez, J.M.; Castiñeiras, A.; Domínguez-Martín, A. Deciphering the
H-Bonding Preference on Nucleoside Molecular Recognition through Model Copper(II) Compounds. Pharmaceuticals 2021, 14,
244. [CrossRef]
Yang, Y.; Huang, Q.; Payne, G.F.; Sun, R.; Wang, X. A highly conductive, pliable and foldable Cu/cellulose paper electrode
enabled by controlled deposition of copper nanoparticles. Nanoscale 2019, 11, 725–732. [CrossRef] [PubMed]
Kwon, Y.T.; Kim, Y.S.; Lee, Y.; Kwon, S.; Lim, M.; Song, Y.; Choa, Y.H.; Yeo, W.H. Ultrahigh Conductivity and Superior Interfacial
Adhesion of a Nanostructured, Photonic-Sintered Copper Membrane for Printed Flexible Hybrid Electronics. ACS Appl. Mater.
Interfaces 2018, 10, 44071–44079. [CrossRef] [PubMed]
Zhang, Y.; Zhang, L.; Cui, K.; Ge, S.; Cheng, X.; Yan, M.; Yu, J.; Liu, H. Flexible electronics based on micro/nanostructured paper.
Adv. Mater. 2018, 30, 1801588. [CrossRef]
Chan, G.H.; Zhao, J.; Hicks, E.M.; Schatz, G.C.; Van Duyne, R.P. Plasmonic properties of copper nanoparticles fabricated by
nanosphere lithography. Nano Lett. 2007, 7, 1947–1952. [CrossRef]
Huang, C.-L.; Kumar, G.; Sharma, G.D.; Chen, F.-C. Plasmonic effects of copper nanoparticles in polymer photovoltaic devices for
outdoor and indoor applications. Appl. Phys. Lett. 2020, 116, 253302. [CrossRef]
Gu, Z.; Shen, H.; Shang, L.; Lv, X.; Qian, L.; Zheng, G. Nanostructured Copper-Based Electrocatalysts for CO 2 Reduction. Small
Methods 2018, 2, 1800121. [CrossRef]
Luc, W.; Fu, X.; Shi, J.; Lv, J.J.; Jouny, M.; Ko, B.H.; Xu, Y.; Tu, Q.; Hu, X.; Wu, J.; et al. Two-dimensional copper nanosheets for
electrochemical reduction of carbon monoxide to acetate. Nat. Catal. 2019, 2, 423–430. [CrossRef]
Kuroda, K.; Keller, P.; Kawasaki, H. Mild synthesis of single-nanosized plasmonic copper nanoparticles and their catalytic
reduction of methylene blue. Colloids Interface Sci. Commun. 2019, 31, 100187. [CrossRef]
Gao, F.; Pang, H.; Xu, S.; Lu, Q. Copper-based nanostructures: Promising antibacterial agents and photocatalysts. Chem. Commun.
2009, 3571–3573. [CrossRef] [PubMed]
Mahmoodi, S.; Elmi, A.; Hallaj-Nezhadi, S. Copper Nanoparticles as Antibacterial Agents. J. Mol. Pharm. Org. Process. Res. 2018,
6, 1–7. [CrossRef]
Jain, A.; Chi, K.M.; Kodas, T.T.; Hampden-Smith, M.J.; Farr, J.D.; Paffett, M.F. Chemical Vapor Deposition of Copper from
(Hexafluoroacetylacetonato)(alkyne)-copper(I) Complexes via Disproportionation. Chem. Mater. 1991, 3, 995–997. [CrossRef]
Chi, Y.; Hsu, P.-F.; Liu, C.-S.; Ching, W.-L.; Chou, T.-Y.; Carty, A.J.; Peng, S.-M.; Lee, G.-H.; Chuang, S.-H. Fluorinated aminoalkoxide CuII complexes: New CVD precursors for deposition of copper metal. J. Mater. Chem. 2002, 12, 3541–3550. [CrossRef]
Toh, B.H.W.; McNeill, D.W.; Gamble, H.S. Investigation of copper layers deposited by CVD using Cu(I)hfac(TMVS) precursor. J.
Mater. Sci. Mater. Electron. 2005, 16, 437–443. [CrossRef]
Córdoba, R.; Sesé, J.; De Teresa, J.M.M.; Ibarra, M.R.R.; Cordoba, R.; Sese, J. High-purity cobalt nanostructures grown by
focused-electron-beam-induced deposition at low current. Microelectron. Eng. 2010, 87, 1550–1553. [CrossRef]
Córdoba, R.; Sharma, N.; Kölling, S.; Koenraad, P.M.; Koopmans, B. High-purity 3D nano-objects grown by focused-electron-beam
induced deposition. Nanotechnology 2016, 27, 355301. [CrossRef] [PubMed]
Shawrav, M.M.; Taus, P.; Wanzenboeck, H.D.; Schinnerl, M.; Stöger-Pollach, M.; Schwarz, S.; Steiger-Thirsfeld, A.; Bertagnolli, E.
Highly conductive and pure gold nanostructures grown by electron beam induced deposition. Sci. Rep. 2016, 6, 34003. [CrossRef]
Kodas, T.T.; Hampden-Smith, M.J. The Chemistry of Metal CVD; Wiley-VCH: Weinheim, Germany, 1994.
Huth, M.; Porrati, F.; Schwalb, C.; Winhold, M.; Sachser, R.; Dukic, M.; Adams, J.; Fantner, G. Focused electron beam induced
deposition: A perspective. Beilstein J. Nanotechnol. 2012, 3, 597–619. [CrossRef] [PubMed]
Utke, I.; Hoffmann, P.; Melngailis, J. Gas-assisted focused electron beam and ion beam processing and fabrication. J. Vac. Sci.
Technol. B 2008, 26, 1197–1276. [CrossRef]
Luisier, A.; Utke, I.; Bret, T.; Cicoira, F.; Hauert, R.; Rhee, S.-W.; Doppelt, P.; Hoffmann, P. Comparative Study of Cu Precursors for
3D Focused Electron Beam Induced Deposition. J. Electrochem. Soc. 2004, 151, C535–C537. [CrossRef]
Utke, I.; Luisier, A.; Hoffmann, P.; Laub, D.; Buffat, P.A. Focused-electron-beam-induced deposition of freestanding threedimensional nanostructures of pure coalesced copper crystals. Appl. Phys. Lett. 2002, 81, 3245–3247. [CrossRef]
Haverkamp, C.; Sarau, G.; Polyakov, M.N.; Utke, I.; dos Santos, M.V.P.; Christiansen, S.; Höflich, K. A novel copper precursor for
electron beam induced deposition. Beilstein J. Nanotechnol. 2018, 9, 1220–1227. [CrossRef]
Berger, L.; Jurczyk, J.; Madajska, K.; Edwards, T.E.J.; Szymańska, I.; Hoffmann, P.; Utke, I. High-Purity Copper Structures from a
Perfluorinated Copper Carboxylate Using Focused Electron Beam Induced Deposition and Post-Purification. ACS Appl. Electron.
Mater. 2020, 2, 1989–1996. [CrossRef]
Reilly, W.L.; Brown, H.C. Reactions of Perfluoronitriles. I. Synthesis of Derivatives of Perfluoroamidines, N-Substituted Perfluoroamidines and Perfluorothioamides. J. Am. Chem. Soc. 1956, 78, 6032–6034. [CrossRef]
Hagen, D.J.; Pemble, M.E.; Karppinen, M. Atomic layer deposition of metals: Precursors and film growth. Appl. Phys. Rev. 2019,
6, 041309. [CrossRef]
Krisyuk, V.V.; Aloui, L.; Prud’homme, N.; Sarapata, B.; Senocq, F.; Samélor, D.; Vahlas, C. CVD of pure copper films from a novel
amidinate precursor. ECS Trans. 2009, 25, 581–586. [CrossRef]
Materials 2021, 14, 3145
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
24 of 24
Krisyuk, V.; Aloui, L.; Prud’homme, N.; Sysoev, S.; Senocq, F.; Samélor, D.; Vahlas, C. CVD of pure copper films from amidinate
precursor. Electrochem. Solid State Lett. 2011, 14, D26–D29. [CrossRef]
Chua, D.; Kim, S.B.; Li, K.; Gordon, R. Low temperature chemical vapor deposition of cuprous oxide thin films using a copper(i)
amidinate precursor. ACS Appl. Energy Mater. 2019, 2, 7750–7756. [CrossRef]
Li, Z.; Barry, S.T.; Gordon, R.G. Synthesis and characterization of copper(I) amidinates as precursors for atomic layer deposition
(ALD) of copper metal. Inorg. Chem. 2005, 44, 1728–1735. [CrossRef] [PubMed]
Mishra, S.; Daniele, S. Metal-Organic Derivatives with Fluorinated Ligands as Precursors for Inorganic Nanomaterials. Chem. Rev.
2015, 115, 8379–8448. [CrossRef]
Szymańska, I.B. Gaseous phase studies of new copper(II) carboxylate complexes with tert-butylamine as potential precursors for
chemical vapor deposition (CVD). Polyhedron 2013, 50, 200–207. [CrossRef]
Piszczek, P.; Szymańska, I.B.; Talik, E.; Heimann, J. Deposition of thin copper layers using copper(II) carboxylate complexes with
tert-butylamine as new CVD precursors. Chem. Vap. Depos. 2013, 19, 251–259. [CrossRef]
Brown, D.J.; Chisholm, M.H.; Gallucci, J.C. Amidinate–carboxylate complexes of dimolybdenum and ditungsten: M2(O2CR)
2((NiPr)2CR’)2. Preparations, molecular and electronic structures and reactions. J. Chem. Soc. Dalton Trans. 2008, 1615–1624.
[CrossRef] [PubMed]
Mendoza, I.; Curado, N.; Carrasco, M.; Álvarez, E.; Peloso, R.; Rodríguez, A.; Carmona, E. Synthesis and structure of mixed
carboxylate-aminopyridinate and -amidinate complexes of dimolybdenum and ditungsten. Inorganica Chim. Acta 2015, 424,
120–128. [CrossRef]
Munzeiwa, W.A.; Nyamori, V.O.; Omondi, B. Zn(II) and Cu(II) unsymmetrical formamidine complexes as effective initiators for
ring-opening polymerization of cyclic esters. Appl. Organomet. Chem. 2018, 32, e4247. [CrossRef]
Szłyk, E.; Szymańska, I. Studies of new volatile copper(I) complexes with triphenylphosphite and perfluorinated carboxylates.
Polyhedron 1999, 18, 2941–2948. [CrossRef]
Frisch, M.J.; Trucks, G.W.; Schlegel, H.B.; Scuseria, G.E.; Robb, M.A.; Cheeseman, J.R.; Scalmani, G.; Barone, V.; Petersson, G.A.;
Nakatsuji, H.; et al. G16_C01; Gaussian, Inc.: Wallingford, CT, USA, 2016.
Zhao, Y.; Truhlar, D.G. The M06 suite of density functionals for main group thermochemistry, thermochemical kinetics, noncovalent interactions, excited states, and transition elements: Two new functionals and systematic testing of four M06-class functionals
and 12 other function. Theor. Chem. Acc. 2008, 120, 215–241. [CrossRef]
John Wiley & Sons. SpectraBase. Available online: https://spectrabase.com/ (accessed on 29 May 2021).
Barry, S.T. Amidinates, guanidinates and iminopyrrolidinates: Understanding precursor thermolysis to design a better ligand.
Coord. Chem. Rev. 2013, 257, 3192–3201. [CrossRef]
John Wiley & Sons. SpectraBase Compound ID=JSNx6EfxlY2 SpectraBase Spectrum ID=3ZkQWtxcvyC. Available online:
https://spectrabase.com/spectrum/3ZkQWtxcvyC (accessed on 29 May 2021).
Majima, T.; Sugita, K.; Nakao, A.; Matsunuma, S.; Abe, K. Synthesis of thin layer of fluorinated polymer by laser CVD—direct
synthesis of perfluoropolyether thin film on surface by oxidative polymerization of a perfluoropropene-oxygen mixture using an
ArF excimer laser. J. Photopolym. Sci. Technol. 1993, 6, 421–428. [CrossRef]
John Wiley & Sons. SpectraBase Compound ID=8A1dEqurmmH SpectraBase Spectrum ID=G4CBuKNSKtQ. Available online:
https://spectrabase.com/spectrum/G4CBuKNSKtQ (accessed on 29 May 2021).
John Wiley & Sons. SpectraBase Compound ID=n8DC1v4dvL SpectraBase Spectrum ID=Big2BLmUPf2. Available online:
https://spectrabase.com/spectrum/Big2BLmUPf2 (accessed on 29 May 2021).
John Wiley & Sons. SpectraBase Compound ID=4G3cv2HmDfG SpectraBase Spectrum ID=89U74D3zedG. Available online:
https://spectrabase.com/spectrum/89U74D3zedG (accessed on 29 May 2021).
Shurvell, H.F.; Bulmer, J.T. The infrared and Raman spectra of pentafluoropropionitrile. J. Fluor. Chem. 1972, 1, 391–406. [CrossRef]
John Wiley & Sons. SpectraBase Compound ID=COBj6FpC1Cb SpectraBase Spectrum ID=4dr5ESO3mVP. Available online:
https://spectrabase.com/spectrum/4dr5ESO3mVP (accessed on 29 May 2021).
Run Nielsen, J.; Claassen, H.H.; Smith, D.C. Infrared and Raman Spectra of Fluorinated Ethylenes. V. Hexafluoropropene. J.
Chem. Phys. 1952, 20, 1916–1919. [CrossRef]
U.S. Secretary of Commerce on Behalf of the United States of America. Perfluoro-n-Pentanoic Acid.. Available online: https:
//webbook.nist.gov/cgi/cbook.cgi?ID=C2706903&Mask=80 (accessed on 29 May 2021).
Yuan, Q.; Zhou, T.; Li, L.; Zhang, J.; Liu, X.; Ke, X.; Zhang, A. Hydrogen bond breaking of TPU upon heating: Understanding
from the viewpoints of molecular movements and enthalpy. RSC Adv. 2015, 5, 31153–31165. [CrossRef]
John Wiley & Sons. SpectraBase Compound ID=JSNx6EfxlY2 SpectraBase Spectrum ID=C0XmNpe0v68. Available online:
https://spectrabase.com/spectrum/C0XmNpe0v68 (accessed on 29 May 2021).
Stoffels, W.W.; Stoffels, E.; Tachibana, K. Polymerization of fluorocarbons in reactive ion etching plasmas. J. Vac. Sci. Technol. A
Vac. Surf. Films 1998, 16, 87–95. [CrossRef]
Rijs, N.J.; O’Hair, R.A.J.J. Forming trifluoromethylmetallates: Competition between decarboxylation and C-F bond activation of
group 11 trifluoroacetate complexes, [CF 3CO 2ML]. Dalton Trans. 2012, 41, 3395–3406. [CrossRef] [PubMed]
Lacko, M.; Papp, P.; Szymańska, I.B.; Szłyk, E.; Matejčík, Š. Electron interaction with copper(II) carboxylate compounds. Beilstein
J. Nanotechnol. 2018, 9, 384–398. [CrossRef]