2.2. Transition Metal Complexes
Some of the alkali metal guanidinates described above were synthesized or prepared in situ as starting materials for reactions with transition metal halides and related precursors. Although the guanidinato derivatives of transition metals had already been described [
38], Richeson’s group opened up the organometallic chemistry of metals of the early groups with this type of ligands. The alkylation of a symmetrically substituted guanidinato derivative of zirconium by metathesis reaction with PhCH
2MgCl generates complex
42 (
Figure 25). As with analogous classic complexes with cyclopentadienyl ligands, this compound exhibits a η
2-type bond interaction of a benzyl ligand [
96].
Figure 25. Zirconium complex with a η2-type bond interaction of a benzyl ligand.
Alternatively, the same complex could be accessed via a toluene elimination reaction between neutral guanidine and tetrabenzylzirconium. This reaction can also be extended to other alkyltitanium derivatives [
97,
98,
99].
There are far fewer examples of complexes containing a guanidinato (2−) ligand than the monoanionic ligands mentioned above [
100,
101]. For example, the reaction of [Cp*ZrCl
3] with the dianion [{(PhN)
3C}Li
2] gave the zwitterionic species [Cp*Zr{C(NPh)
3}Cl
2Li(Et
2O)(THF)], a complex active in ethylene polymerization [
100]. As for their monoanionic analogues, these doubly negatively charged ligands also exhibit chelate-type coordination. Fluxional behavior was found in solution, and this involves a
syn/anti isomerization process of the guanidinato (2−) ligand [
101].
The formation of mixed cyclopentadienyl zirconium guanidinato derivatives can also be achieved by a formal [2 + 2] cycloaddition reaction of carbodiimides to imides. These azaallyl-like ligands still exhibit the κ
2-chelate coordination described above and not the possible κ
3-coordination similar to that of the allyl ligands, due to the orbitals involved in both cases (
Figure 26). These complexes enable exchange reactions with other carbodiimides under mild conditions, which extends the family of this type of derivatives [
102,
103].
Figure 26. Dianionic zirconium guanidinato complex.
This type of complex still presents an uncoordinated imino group with a lone pair that could participate in the coordination to Lewis acid centers. However, the more than likely delocalization of these electrons along the ligand prevents such reactivity [
104].
Whereas the alkylation with MeLi of the derivative Cp(C(C(NiPr)
2(NMe
2))ZrCl
2 leads to a metathesis process to form the metallocene derivative and the bis-guanidinate, the use of the Grignard reagent MeMgCl or, more conveniently, the protonation reaction of the derivative (C
5H
3(SiMe
3)
2)ZrMe
3 with triisopropylguanidine allows the obtention of dialkyl derivatives with guanidinato ligands [
102,
103,
105].
Following the similarity with the chemistry of metallocene derivatives, these neutral alkyl zirconium complexes with guanidinato ligands can be transformed into cationic derivatives by reaction with reagents, such as tris(pentafluorophenyl)borane, [Ph3C][B(C6F5]3]4 or Me3SiOTf (Figure 27).
Figure 27. Cationic derivatives of a bis-guanidinato zirconium compound.
The great influence of the geometry and the donor capacity of the guanidinato ligands on the reactivity of the complexes they form can be exemplified by studying the reactions of dialkyl derivatives with aryl isonitriles. The titanium or zirconium complexes of this class form intermediates, sometimes isolated, with η
2-iminoacyl ligands for which subsequent transformations lead to vinylamido-type species, in the case of linked guanidinato ligands [
106], while their unlinked analogues evolve towards imido or enediamido species (
Figure 28) [
107,
108,
109].
Figure 28. Aryl isonitriles insertion and the consecutive evolution of the iminoacyl groups.
Occasionally, the lithiated guanidine derivative does not generate the expected chelate complex observed in most of the compounds described. Instead, in complexes such as Cp
2TiCl(k
1-N=C(NMe
2)
2)
50, the guanidinate assumes an unusual binding mode, in the form of a monodentate ligand [
110,
111].
The metallacarborane complexes of zirconium and titanium can be obtained either by protonolysis with neutral guanidine or carbodiimide insertion in amido precursors. Titanium complex
51 is an effective catalyst for the addition of primary and secondary aliphatic and aromatic amines to carbodiimides with good functional group tolerance (
Figure 29) [
112,
113].
Figure 29. Metallocarborane complexes of titanium.
Using chiral modified 2-aminooxazolines gave rise to chiral half-sandwich complexes where the amino-oxazolinato group had undergone a ring opening and a migratory insertion of a dimethylamido ligand, providing a configurationally stable chiral-at-metal system (
Figure 30) [
114,
115].
Figure 30. Chiral half-sandwich complexes with an oxazolinato group.
Although, as with metallocene derivatives, their application in catalytic processes makes the guanidinato complexes of group 4 metals very attractive compounds, these complexes have received interest as precursors to metal nitride or oxide thin films due to their nitrogen or carbon content, their potential to increase the volatility of the compound and their ability to stabilize the metal center due to their electronic flexibility. For example, thin films of carbonitrides or oxides of titanium, zirconium or hafnium can be obtained from guanidinato or mixed cyclopentadienyl guanidinato precursors. The above properties of these molecules satisfy the important requirements for their application in chemical deposition techniques, such as ALD or CVD [
116,
117,
118,
119].
As mentioned above, guanidinato ligands allow the stabilization of less common oxidation states for transition metals. Some mixed halide complexes of zirconium and hafnium can be reduced in the presence of N
2 to form bimetallic complexes with a side-on-bridged dinitrogen molecule and proved to be very active toward both hydrosilylation and hydrogenation (
Figure 31) [
120].
Figure 31. Coordination and activation of dinitrogen with group 4 guanidinato complexes.
The titanium (II) complex
54 (
Figure 32), obtained by a two-electron reduction from a dichloride precursor, exhibits an arene–M bond interaction similar to that previously described for group 1 elements, such as lithium or potassium. This type of complexes allows catalytic hydrogenation by hydrogen transfer to monocyclic and polycyclic arenes, mimicking the reactivity of late transition metals, as well as demonstrating a great ability to reductively activate a wide range of substrates, such as ketones, azides, N
2O or C–F bonds [
121,
122,
123].
Figure 32. Arene–metal bond interaction in a titanium complex.
There is some parallelism between the organometallic chemistry of guanidinato complexes of groups 4 and 5. For example, the alkyl derivatives of niobium of the type
55 (
Figure 33) show a η
2-benzyl group in the solid state, and this is in rapid exchange with the other η
1-benzyl ligand in solution. While the migratory insertion of tBuNC into the niobium–alkyl bonds gives rise to the corresponding bis(iminoacyl) derivatives, the insertion of an aryl isonitrile, XyNC (Xy = 2,6-Me
2C
6H
3), into a niobium–benzyl bond results, as with some zirconium complexes [
107], in the isomerization of an η
2-iminoacyl group to a vinylamido-type one via a 1,2-hydrogen shift. The influence of the guanidinato ligand on this process is clear, since the same reaction does not take place starting from the precursor compound of the guanidinato complexes, the trialkylimido [NbBz
3(NtBu)] (Bz = benzyl) [
124,
125].
Figure 33. Asymmetric coordination of a guanidinato ligand in niobium complexes.
These dibenzyl complexes with guanidinato ligands derived from diisopropylcarbodiimide and an aromatic amine show an asymmetric coordination of the ligand via an alkylamino nitrogen atom and the arylimino nitrogen atom. Computational studies confirm this preference, and the results suggest that electronic factors prevail over steric factors. In fact, when a guanidine with three alkyl substituents, one of them different, was used for the synthesis, a mixture of isomeric complexes was formed with the coordinated ligands without any selectivity for any of the nitrogen atoms. In addition, these complexes were proposed as intermediates in the mechanism of the catalytic guanylation of anilines, using the complex [NbBz
3(NtBu)] as a precatalyst [
126].
Another example of the remarkable influence of the coordination of a guanidinato ligand on the reactivity on the coordination sphere of a metal is the process, in which an alkyl guanidine reacts with an iminocarbamoyl complex, which undergoes an easy cleavage of a C–N bond at room temperature (
Figure 34) [
127].
Figure 34. C–N activation mediated by a guanidine in an iminocarbamoyl compound.
This process contrasts with that observed when migratory insertion of aryl isonitriles into the Ta-NMe
2 bonds of a previously formed guanidinato complex was undertaken [
128]. In this case, the iminocarbamoyl species [Ta(NMe
2)[C(NiPr)
3][η
2-(Me
2N)C=N(2,6-Me
2C
6H
3)]
2]
56, which is stable, was formed.
As was the case for group 4 metals [
120], guanidinato ligands stabilize organometallic tantalum species in a medium or low oxidation state [
129,
130,
131]. For example, mixed cyclopentadienyl-guanidinato species of Ta(IV) give rise to N≡N bond cleavage to provide the bimetallic bis(μ-nitrido) complex, {Cp*Ta[N(iPr)C(NMe
2)N(iPr)](μ-N)}
2 57. The presence of the uncoordinated NMe
2 group in the guanidinato ligand can serve to modulate the magnitude of the free energy barrier for N≡N bond cleavage.
Very recently, the participation of tantalum alkyl guanidinato derivatives of the type [Ta(hpp)
2(CH
2SiMe
3)
3] (hpp = 1,5,7-triazabicyclo [4.4.0]dec-5-ene)
58 has been studied as probable intermediates in the process of hydroaminoalkylation of terminal alkenes with a variety of secondary amine substrates to give substituted secondary amine products [
132].
While the organometallic chemistry of group 6 derivatives is limited to a molybdenum complex, [(ArN)
2MoMe{N(Cy)C[N(iPr)
2]N(Cy)}] (Ar = 2,6-iPr
2C
6H
3)
59 [
133], and that of group 7 is non-existent, group 8 organometallic complexes with guanidine-derived ligands offer interesting structural and reactivity aspects.
Iron guanidinato (2−) complexes
60 and
61 (
Figure 35) were obtained via a formal [2 + 2] cycloaddition of carbodiimide to imido precursors in high or low oxidation states [
134,
135].
Figure 35. Iron dianionic guanidinato compounds.
Ruthenium witnesses the first reference of a complex in which the guanidinato ligand acts as a chelate. Indeed, complex
62 contains an η
6-bonded aromatic ligand, a terminal chloride and the chelating triphenylguanidine anion. The structure of this compound shows, for the first time, a feature that is repeated in many guanidinato complexes: the angles around the central carbon of the guanidine ligand total 360° indicating the strict planarity of the CN
3 core (
Figure 36) [
136].
Figure 36. First example of a chelate guanidinato compound.
This coordination mode is present in other similar ruthenium complexes with both monoanionic and dianionic ligands [
137,
138,
139,
140,
141].
It is also worth noting that, despite the prominent role of ruthenium complexes in organic synthesis, there are only a few examples of guanidinato complexes reported to date that have been explored as catalysts. Complexes similar to those described are excellent catalysts for the redox isomerization of allylic alcohols in the absence of base (
Figure 37) [
142].
Figure 37. Ruthenium guanidinato complex acting as catalyst for allylic alcohol isomerization.
This catalytic process also takes place using ruthenium (IV) guanidinato complexes, namely [RuCl{κ
2-C(NR)(NiPr)NHiPr}(η
3:η
3-C
10H
16)]
64 (
Figure 38). This type of compounds are excellent precursors for the preparation of octahedral ruthenium (II) guanidinato complexes,
mer-[RuCl{κ
2-C(NR)(NiPr)-NHiPr}(CN-2,6-C
6H
3Me
2)
3]
65, through the reductive elimination of the ligand 2,7-dimethylocta-2,6-diene-1,8-diyl with high stereoselectivity [
143].
Figure 38. Ruthenium (IV) guanidinato complex [RuCl{κ2-C(NR)(NiPr)NHiPr}(η3:η3-C10H16)].
In contrast, less attention has been paid to osmium guanidinato complexes [
137,
144], although some examples were able to transform a wide variety of aldoximes selectively and catalytically into the corresponding nitriles (
Figure 39).
Figure 39. Osmium guanidinato complexes acting as aldoxime dehydration catalysts.
Interestingly, in the search for new coordination modes, guanidines have been modified with other donor groups, such as phosphanes. In this way, ruthenium (II) (
Figure 40) and osmium (II) complexes were obtained [
145,
146].
Figure 40. Phosphane-modified guanidine as ligand in a ruthenium complex.
This type of multidentate ligand allows the preparation of the iridium and rhodium guanidinato complexes as
68, which behave as a frustrated Lewis pair (FLP) capable of reversibly activating polar bonds, such as those of water, and non-polar bonds, such as those of dihydrogen (
Figure 41) [
147,
148].
Figure 41. Frustrated Lewis pair based in rhodium phosphanoguanidinato complexes.
Although the usual ligand arrangement in many rhodium complexes is the κ
2-
N,
N chelate [
137,
149,
150], examples are known where the guanidinate acts as a bridge between two metal centers [
99], or even a complex where the anionic ligand coordinates the rhodium center in an unusual η
5-cyclohexadienyl mode, before isomerizing to the conventional chelate form (
Figure 42) [
151].
Figure 42. Bridging and arene coordination of guanidinato ligands in rhodium complexes.
Not surprisingly, as well as several examples of rhodium complexes with guanidinato ligands are known, the organometallic chemistry of analogous iridium compounds is wide. In addition to reactivities parallel to those described for rhodium [
137,
148,
150], there are interesting examples of the influence of these ligands on the behavior of the metal. For example, the presence of the lone pair of electrons on the uncoordinated NR
2 group increases the electron density of the ligand available for coordination to the metal. This obviously makes these ligands stronger donors than the related amidinato ligands. Consequently, if they coordinate to a metal in a low oxidation state, the possibility of reaching higher oxidation states increase. Thus, in the reactions of iridium guanidinato complexes with O
2, the observed reactivity trends correlate well with both the electronic and steric properties of the substituents of the guanidinato ligands, allowing the stabilization and characterization of complexes with peroxo groups (
Figure 43) [
152,
153,
154,
155].
Figure 43. Dioxygen iridium complex.
This same donor capacity of the guanidinato ligand also lies at the basis of several catalytic processes involving organometallic iridium species, which allow the metathesis of carbodiimides and isocyanates, the carbodiimide cleavage or the cross-dimerization of terminal alkynes [
156,
157,
158].
Organometallic iridium (III) complexes are also useful as photophysical active materials in a wide variety of applications, e.g., as lighting devices, as dyes in solar panels, as bio-imaging agents or as photosensitizers for the production of hydrogen from water by photocatalysis [
159]. It has been shown that these properties can be altered practically on a whim by modifying the ligands in these complexes. This is why, given the high tunability provided by the guanidinato ligands, they appear as excellent precursors of phosphorescent iridium cyclometalated complexes, becoming prominent in the preparation of Organic Light Emitting Devices (OLEDs), where the proper selection of ligand substituents allows high efficiencies and high variability of the emission color (
Figure 44) [
160,
161,
162,
163].
Figure 44. Luminescent iridium guanidinato compound.
The analogy between bulky guanidinates and β-diketiminates has been used to achieve the stabilization of Co(I) toluene adducts, as precursors of dimeric compounds with short Co–Co bonds [
164].
As with other platinum group metals, the first reference of group 10 organometallic complexes with guanidinato ligands presents species bearing dianionic ligands with chelate coordination [
137,
138,
139,
140]. Since then, it is worth noting the studies carried out by Thirupathi’s group, who have extensively studied aspects of the coordination chemistry of triarylguanidines towards palladium or platinum, in which the presence of Me or Ome substituents on the aromatic rings of the guanidines, together with their inherent basic properties, allowed the formation of interesting six-membered cyclometalated complexes, in which the guanidinato ligand shows a κ
2-
C,
N coordination mode (
Figure 45) [
165,
166,
167,
168,
169,
170,
171,
172,
173,
174,
175].
Figure 45. Cyclometalated guanidinato palladium and platinum complexes.
To conclude this section, the organometallic derivatives of group 11 are limited to an example of a strongly phosphorescent copper (I) complex with an (alkyl)(amino)carbene ligand (
Figure 46) [
176].
Figure 46. Phosphorescent carbene copper guanidinato compound.
2.3. Group 3 and the Lanthanoid Complexes
The group 3 elements and the lanthanoids represent a family of metals with very interesting properties in terms of chemical reactivity, luminescence or magnetism [
177,
178]. In particular, the guanidinato derivatives of these metals stand out for their potential use in catalytic processes or as molecular precursors of new ceramic materials. Although there are excellent reviews by Edelmann and Trifonov on rare earth amidinates and guanidinates [
43,
44,
45], in this work we have selected the references focused on organometallic complexes, highlighting their usefulness and the possible differential nuances with the complexes of other groups of metals previously described.
As with some of these examples, the metathesis of a lithium guanidinate and a metal halide is one of the main routes to rare-earth metal derivatives [
179,
180,
181]. This is the case for
75, the first reported organolanthanide complex supported by such a ligand [
182]. After the formation of the halide derivative, the reaction with LiCH(SiMe
3)
2 yields a mononuclear organometallic compound (
Figure 47).
Figure 47. First reported organolanthanide complex with guanidinato ligands.
Stable low-coordinate organometallic complexes, using bulky guanidinato ligands, allow agostic interactions between the yttrium atom and the tert-butyl group in the solid state for [{(Me
3Si)
2NC(NCy)
2}
2YtBu]
76 [
183] or with two methyl carbon atoms of SiMe
3 groups in [{(Me
3Si)
2NC(NCy)
2}
2Y(μ-CH
2SiMe
3)
2Li]
77 [
184].
Alternatively, the insertion reaction of carbodiimides into M–N bonds of amido derivatives allows a very efficient synthesis of organometallic complexes with symmetric and, mainly, asymmetric guanidinates [
185,
186].
The presence of NH groups on the ligands assists the insertion and can provide an interesting isomerization of these coordinated fragments (
Figure 48) [
187,
188].
Figure 48. NH-assisted insertion of carbodiimides in amido ytterbium compounds.
In contrast to the usual lack of reactivity observed in guanidinates of other metals, certain organolanthanides allow examples of the insertion of small unsaturated molecules into the metal–guanidinate bond. Complexes [(C
5H
5)
2Ln(μ-η
1:η
2-N=C(NMe
2)
2)]
2 (Ln = Y
79, Gd
80, Er
81) can mono-insert phenyl isocyanate into the Ln–N bond to yield [(C
5H
5)
2Ln(μ-η
1:η
2-OC(N=C(NMe
2)
2)NPh)]
2 without further insertion or the usual isocyanate cyclotrimerization [
110,
189]. The presence of reactive alkyl groups, such as benzyl, offers the possibility to study alternative insertion processes [
190,
191,
192]. Complex
82 inserts nitriles and isocyanates, generating binuclear complexes (
Figure 49) [
191].
Figure 49. Nitrile insertion in alkyl samarium complex.
A third route is the protonolysis reaction of alkyl derivatives with neutral guanidines [
193]. In this way, an alkyl yttrium compound has been obtained with a guanidine with elevated steric demand and with the adequate control of the electronic deficiency of the metal. Compound
84 was highly efficient in the hydrosilylation of alkenes via hydride intermediates with excellent anti-Markovnikov selectivity (
Figure 50) [
193].
Figure 50. Alkene hydrosilylation with an yttrium guanidinato compound as catalyst.
The catalytic activity of rare-earth metal guanidinates also extends to olefin polymerization processes [
194,
195,
196]. For example, the coordination of two guanidinate ligands allows the stabilization of hydride complexes. The appropriate steric bulk of the substituents on the nitrogen atoms provides high solubility of the lanthanide derivatives and enables Lewis base-free hydrides to be obtained from alkyl derivatives and silanes with an extremely low coordination index on the metal atom. Species such as compound
86 (
Figure 51) have shown high catalytic activity in ethylene polymerization [
194].
Figure 51. Synthesis of an yttrium hydride stabilized by guanidinato ligands.
Such alkyl or hydride complexes stabilized by guanidinato ligands also allow ring-opening or polar monomer polymerizations [
197,
198].
To conclude this section, guanidinato ligands have met the needs for rare-earth metal molecular precursors for ALD or CVD techniques: the bidentate coordination mode allows to saturate coordinatively even large cations; the electronic delocalization across the ligand stabilizes the whole molecular structure; and the easy modification of the substituents allows to modulate the physical properties, all contributing to expand the library of potential precursors of new high quality materials obtained by these deposition techniques [
199].
3. Conclusions
In conclusion, a great deal of work has been carried out in this field, but there is still scope for the development of new organometallic complexes, in which the choice of metal and the design of the ligands provide greater control of their chemical and physical properties. Guanidines provide easy access to such control, as they can be obtained by very straightforward and efficient syntheses from widely available reagents. The ease of formation of coordination and organometallic complexes, even in unconventional or unstable oxidation states, can be enhanced by potentially modifying these ligands to increase the number and type of donor atoms. A further step towards a firm foothold among the most successful ligands is the possibility to design and synthesize new chiral derivatives for enantioselective catalysis applications. So, again, an appropriate choice of precursors and methods to obtain these ligands, if possible, in a sustainable manner, opens up a whole range of possibilities for organometallic chemists.