Photoactivated silicon-oxygen and silicon-nitrogen heterodehydrocoupling with a commercially available iron compound

Article abstract of DOI:10.1039/c9dt04870g

Silicon-oxygen and silicon-nitrogen heterodehydrocoupling catalyzed by the commercially available cyclopentadienyl dicarbonyl iron dimer [CpFe(CO)₂]₂ (1) under photochemical conditions is reported. Reactions between alcohols and PhSi

Full text of DOI:10.1039/c9dt04870g

Dalton
Transactions
View Article Online
View Journal
PAPER
Photoactivated silicon–oxygen and
silicon–nitrogen heterodehydrocoupling
Cite this: DOI: 10.1039/c9dt04870g
with a commercially available iron compound†
Matthew B. Reuter, Michael P. Cibuzar, James Hammerton and Rory Waterman
*
Silicon–oxygen and silicon–nitrogen heterodehydrocoupling catalyzed by the commercially available
cyclopentadienyl dicarbonyl iron dimer [CpFe(CO)2]2 (1) under photochemical conditions is reported.
Reactions between alcohols and PhSiH3 with catalytic 1 under visible-light irradiation produced silyl
ethers quantitively. Reactions between either secondary or tertiary silanes and alcohols also produced silyl
ethers, however, these reactions were marked by their longer reaction times and lower conversions.
Reactions of either primary or secondary amines and silanes with catalytic 1 demonstrated mixed
efficiency, featuring conversions of 20–100%. Mechanistic study indicates that an iron silyl compound is
unimportant in the bond-formation step and argues for either a nucleophilic alkoxide or amide intermedi-
ate. Most important, mechanistic study reveals that the most immediate hurdle in the catalysis is the poor
activation of 1, demonstrating the necessity to fully activate the catalyst to realize the potential of iron in
this reactivity.
Received 23rd December 2019,
Accepted 11th February 2020
DOI: 10.1039/c9dt04870g
rsc.li/dalton
worthy to mention that examples of mild, photoactivated iron
compounds are scarce in comparison to thermally activated
Introduction
The dominance of noble metals in catalysis is, rightly, under catalysts.15,16 This becomes an unfortunate realization, as the
assault. The importance of metals such as palladium, plati- development and improvement of iron-based systems is para-
num, rhodium, and iridium is irrefutable, with some of the mount to inexpensive and green chemical transformations.
more representative transformations including palladium-cata-
Concomitant with the improvement of base metal catalysis,
lyzed C–C or C–N cross-coupling,1,2 platinum-catalyzed hydro- chemists have been challenged with the development of
silylation of olefins,3 rhodium-catalyzed hydrogenation and greener, efficient synthetic pathways.17 Heterodehydrocoupling
hydroformylation,4,5 and iridium-catalyzed C–H activation.6 has gained momentum in this aspect, due to the atom-econ-
Despite their high utility to both academia and industry, there omical formation of element–element bonds. The evolution of
has been a shift away from these noble metals due to their H2 as the sole byproduct is also attractive, providing an
cost, toxicity, and most importantly, increasing scarcity.7 In excellent driving force and simplifying the purification of
their stead, a plethora of transformations have emerged, products. It is also important to recognize that dehydro-
including C–C cross-coupling,8 hydrosilylation of olefins and coupling is often symmetry forbidden, necessitating the use
aldehydes,9,10 and C–H activation,11,12 by base metals includ- of either main group or transition-metal compounds.18
ing iron, manganese, and cobalt. Iron is particularly attractive Consequentially, the identification and development of hetero-
in catalysis due to its high abundance and access to a range of dehydrocoupling catalysts is extremely attractive for green,
oxidation states.13,14 However, a variety of factors limit base catalytic transformations.
metal-catalyzed transformations, such as high catalyst load-
The commercially available iron dimer [CpFe(CO)2]2 (1) is a
ings, significant heating, or other forcing conditions to achieve rare example of a mild, photoactivated iron compound.
conversions comparable to those with noble metal catalysts.15 Heterodehydrocoupling via compound 1 has already been
Iron is no exception to these limitations, and it is also note- demonstrated on amine-borane substrates by Manners and co-
workers as well as between dimethylformamide and PhMe2SiH
by Waterman and co-workers.19,20 Furthermore, compound 1
University of Vermont, Department of Chemistry, Discovery Hall, Burlington,
is known to photoactivate under either ultraviolet or visible-
VT 05401, USA. E-mail: rory.waterman@uvm.edu
light irradiation to produce two equiv. of the 17-electron
†Electronic supplementary information (ESI) available. See DOI: 10.1039/
compound, 3, via the all terminal carbonyl intermediate 2
c9dt04870g
This journal is © The Royal Society of Chemistry 2020
Dalton Trans.

View Article Online
Paper
Dalton Transactions
silanes and alcohols. An equimolar amount of nPrOH and
PhSiH3 in the presence of 1 mol% of 1 in a benzene-d6 solu-
tion was irradiated under visible-light from a commercial LED
bulb. After 24 h, the mixture showed 32% conversion to
PhSiH2(OnPr) and 43% conversion to PhSiH(OnPr)2 as
measured by 1H NMR spectroscopy. Comparable conversions
were not achieved under ultraviolet irradiation, which was
attributed to a known, unproductive activation pathway of 1
involving the loss of CO.21 The molar equivalences of alcohol
and silane were varied in an effort to generate the third
addition silyl ether product PhSi(OnPr)3. Four-fold excess of
silane to one equivalent of alcohol showed little effect on silyl
ether generation. However, increasing the concentration of
alcohol four-fold and the catalyst loading to 2 mol% of 1 gen-
erated PhSi(OnPr)3 in quantitative conversion after 24 h accord-
ing to 1H NMR spectroscopy (Table 1, entry 1).28 These reac-
tion conditions were uniformly applied to other substrates
(eqn (1)).
Scheme 1 Photoactivation pathway of compound 1 under either ultra-
violet or visible-light irradiation.17
(Scheme 1).21 Thus, the photoirradiation of compound 1 may
provide a green and facile method to forming other element–
element bonds in the main group.
Molecules and materials containing Si–O and Si–N possess
diverse applications due to their unique chemical character-
istics. Silyl ethers, or small molecules containing Si–O bonds,
are of importance in the protection of alcohols.22 Poly(silyl
ethers) are appealing compounds due their degradability in
acidic and basic medium.23 Molecules containing Si–N bonds
such as silamines are well established as bases and silylating
agents in organic syntheses,24 while poly(silazanes) are sought
after for their potential as ceramic precursors.25 Although Si–O
heterodehydrocoupling using 1 has not been reported, deriva-
tives of 1 catalyzing similar transformations are known, most
notably by Nakazawa and co-workers in Si–O and Si–S heterode-
hydrocoupling via Cp(CO)2FeMe.26,27 Herein, we report 1 as a
heterodehydrocoupling catalyst in the formation of Si–O and Si–
N bonds. Mechanistic study of the reaction indicates nucleophi-
lic attack of a silane by an intermediate iron-alkoxide or -amide,
but more germane to the further development of iron, complete
activation of 1 was not achieved in these reactions, which
suggests that full activation of iron catalyst precursors is an
important pursuit in developing base metal catalysis.
Catalytic Si–O heterodehydrocoupling
ð1Þ
i
Coupling of alcohols such as BnOH (Bn = CH2Ph) and PrOH
with PhSiH3 was also accomplished with 1. Reaction of BnOH
and PhSiH3 in a 4 : 1 ratio generated PhSi(OBn)3 after 6 h, as
determined by 1H and 29Si{1H} NMR spectroscopy (Table 1,
Results and discussion
Condition optimization
This study sought to expand the scope of heterodehydrocou- entry 2).29,30 Analogously, Parkin and co-workers have reported
pling by 1,19,20 initially investigating coupling of primary Si–O heterodehydrocoupling between PhSiH3 and BnOH via
Table 1 Catalytic conditions for the coupling of alcohols and silanesa
Entry
Silane
Alcohol
Equiv.b
Product
Conversionc (%)
1
PhSiH3
PhSiH3
PhSiH3
PhMeSiH2
PhMeSiH2
PhMeSiH2
nPrOH
BnOH
iPrOH
nPrOH
BnOH
iPrOH
4.0
4.0
5.0
4.0
4.0
5.0
PhSi(OnPr)3
PhSi(OBn)3
100
100
100
100
100
9
2d
3
PhSi(OiPr)3
PhMeSi(OnPr)2
PhMeSi(OBn)2
e
4
5
6
PhMeSiH(OiPr)e
PhMeSi(OiPr)2
91
e
Ph2Si(OnPr)2
100
100
100
93
100
93
e
7
8
9
10
11
12
Ph2SiH2
Ph2SiH2
Ph2SiH2
PhMe2SiH
PhMe2SiH
PhMe2SiH
nPrOH
BnOH
iPrOH
nPrOH
BnOH
iPrOH
4.0
4.0
5.0
5.0
5.0
6.0
Ph2Si(OBn)2
Ph2SiH(OiPr)
PhMe2Si(OnPr)e
PhMe2Si(OBn)
PhMe2Si(OiPr)
a Conditions: 2.0 mol% of 1 under visible-light irradiation in benzene-d6 solution at ambient temperature for 24 h unless otherwise specified.
Catalyst loading was with respect to silane. Reactions were monitored by 1H and 29Si{1H} NMR spectroscopy. b Refers to mol. of alcohol per mol.
of silane. c Conversions were determined by 1H NMR integration. d Reaction was complete in 6 h. e Literature spectral data of these silyl ethers
have not been previously reported.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Dalton Transactions
Paper
nickel-catalysts, however, to a mixture of PhSiH(OBn)2 and silane Et3SiH produced no change in 1H NMR spectroscopy
PhSi(OBn)3 after 24 h at 80 °C.31 Reactions between iPrOH and after 24 h.
PhSiH3 at similar alcohol/silane ratio proceeded to incomplete
conversion from PhSiH3 after 24 h, which prompted an
Catalytic Si–N heterodehydrocoupling
increase in the alcohol/silane ratio. Reaction of a 5 : 1 mixture
i
of PrOH and PhSiH3 completely converted from PhSiH3 by
24 h to PhSi(OiPr)3 (Table 1, entry 3).30,32,33 The silyl ether
PhSi(OiPr)3 has previously been synthesized utilizing a half-
sandwich iron complex by Royo and co-workers; although this
ð2Þ
was accomplished at slightly lower catalyst loadings of 1 mol%
after 8 h, only 66% conversion was afforded at 70 °C.33
Attempts at coupling PhSiH3 with heavily encumbered alco-
hols such as tBuOH with 1 did not produce silyl ethers accord-
ing to 1H NMR spectroscopy.
Heterodehydrocoupling with secondary silanes using com-
pound 1 was also investigated. Reaction of PhMeSiH2 and Compound 1 also proved to be competent at Si–N heterodehy-
nPrOH in a 1 : 4 ratio generated a single peak at δ −18.07 in drocoupling but at higher catalyst loadings (eqn (2)). Silamines
29Si{1H} NMR spectroscopy after 24 h under irradiation, con- were produced less efficiently than were silyl ethers, as evident
sistent with PhMeSi(OnPr)2 (Table 1, entry 4). The final by the longer reaction times and mixture of silamine products.
resonance generated at δ 3.89 in 1H NMR spectroscopy
Treatment of nPrNH2 with PhSiH3 in a 6 : 1 amine/silane
indicated 100% conversion to PhMeSi(OnPr)2.
A
similar ratio produced PhSiH2(HNnPr) in only 13% conversion after
i
strategy was applied to reactions of PrOH and PhSiH3 in a 4 h by 1H NMR spectroscopy. After 18 h, the reaction produced
5 : 1 ratio, where PhMeSi(OiPr)2 was afforded in 91% conver- PhSiH(HNnPr) in 50% conversion and PhSiH2(HNnPr) in 23%
sion with 9% of PhMeSiH(OiPr) remaining after 24 h (Table 1, conversion (Table 2, entry 1).40 Similar to trends in Si–O coup-
entry 6). Reaction of excess BnOH with PhMeSiH2 produced ling, hard-acid catalysts appear to produce silamines in good
PhMeSi(OBn)2 in 100% conversion after 24 h (Table 1, conversions under heating.41 The analogous reaction with
entry 5).34
tBuNH2 and PhSiH3 produced PhSiH2(HNtBu) in 100% after
Reaction of nPrOH and Ph2SiH2 under visible-light only 4 h according to 1H NMR spectroscopy, and in 24 h,
irradiation in the presence of 1 proceeded slowly according to PhSiH2(HNtBu) and PhSiH(HNtBu) were produced in 89% and
1H NMR spectroscopy, but all starting material was consumed 11% conversions (Table 2, entry 2).42,43 The disparity between
to a single new product. Isolation of pure product from the the two amines indicates that more basic (i.e., nucleophilic)
highly soluble Fp-catalyst remains a challenge, but in compari- amines give greater silamine conversions. This observation
son to similar resonances of known compounds, it is hypoth- was supported by reaction of 4 equiv. of iPrNH2 and PhSiH3 to
esized that Ph2Si(OnPr)2 was generated in 100% conversion furnish PhSiH2(HNiPr) in 100% conversion after 20 h accord-
(Table 1, entry 7). Reactions BnOH and Ph2SiH2 in a 4 : 1 ratio ing to 1H NMR spectroscopy (Table 2, entry 4).44 Moreover,
1
produced Ph2Si(OBn)2 in 100% conversion as measured by H reaction of 4.7 equiv. of Et2NH with PhSiH3 produced
NMR spectroscopy (Table 1, entry 8).30,35 Several examples of PhSiH2(NEt2) and PhSiH(NEt2)2 in 29% and 71% conversions,
hard-acid catalysts such as KN(SiMe3)2 and a Mn(V)-salen respectively, after 24 h (Table 2, entry 5).45 Finally, reaction of
complex have quantitatively afforded Ph2Si(OBn)2 in signifi- 4.6 equiv. of PhNH2 with 9.3 mol% of 1, PhSiH2(HNPh) was
cantly shorter reactions times, however, elevated heating afforded in only 20% conversion after 20 h (Table 2, entry 3).
appears to be a necessary factor in these reactions.31,36,37
Compound 1 was also demonstrated to be a competent het-
i
Interestingly, reacting 5 equiv. of PrOH with Ph2SiH2 exclu- erodehydrocoupling with amines and PhMeSiH2. Treatment of
sively yielded Ph2SiH(OiPr)25 in quantitative conversion with nPrNH2 with PhMeSiH2 in a 5 : 1 amine/silane ratio affords the
no evidence of the fully substituted product Ph2Si(OiPr)2 corresponding silamine PhMeSiH(HNnPr) in 60% conversion
(Table 1, entry 9).30
after 24 h by 1H NMR spectroscopy (Table 2, entry 6).44
n
Reaction of PrOH and PhMe2SiH in a 5 : 1 ratio afforded a Meanwhile, PhMeSiH(HNtBu) was generated in 100% conver-
new product, tentatively assigned to PhMe2Si(OnPr) based on sion by 1H NMR spectroscopy after 24 h (Table 2, entry 7).44
analogy to PhMe2Si(OBn) and PhMe2Si(OiPr), in 93% conver- Furthermore, reacting 4 equiv. of iPrNH2 with PhMeSiH2 quan-
sion as a resonance at δ 6.67 in the 29Si{1H} NMR spectrum titatively produced PhMeSiH(HNiPr) after 24 h according to H
1
(Table 1, entry 10). Reaction of excess BnOH and PhMe2SiH, and H–29Si HSQC NMR spectroscopy (Table 2, entry 8).44 The
1
however, showed complete disappearance of PhMe2SiH in the reaction between Et2NH and PhMeSiH2 in a 6 : 1 ratio quanti-
1H NMR spectrum and generation of PhMe2Si(OBn) after 24 h tatively converted from PhMeSiH2 after 24 h according to 1H
i
(Table 1, entry 11).34,38 Reaction of PrOH and PhMe2SiH in a and 1H–29Si HSQC NMR spectroscopy (Table 2, entry 9).45
6 : 1 ratio showed 93% conversion to PhMe2Si(OiPr) after 24 h Notably, in addition to PhMeSiH(NEt2), a second peak was
1
according to 1H NMR spectroscopy (Table 1, entry 12).39 also discernible H–29Si HSQC NMR spectroscopy. Although it
Finally, attempts at coupling alcohols to the tertiary alkyl was initially believed to be the second addition product
This journal is © The Royal Society of Chemistry 2020
Dalton Trans.

View Article Online
Paper
Dalton Transactions
Table 2 Catalytic conditions for the coupling of amines and silanesa
Entry
1
Silane
Amine
Loadingb
6.0
Equiv.c
3.5
Product
Conversiond (%)
Time (h)
18
PhSiH3
nPrNH2
PhSiH2(HNnPr)
PhSiH(HNnPr)2
PhSiH2(HNtBu)
PhSiH(HNtBu)2
PhSiH2(HNPh)
PhSiH2(HNiPr)
PhSiH2(NEt2)
23
50
89
11
20
100
29
2
PhSiH3
tBuNH2
7.8
6.0
24
3
4
5
PhSiH3
PhSiH3
PhSiH3
PhNH2
iPrNH2
Et2NH
9.3
8.5
8.5
5.0
4.0
6.0
20
20
24
PhSiH(NEt2)2
71
60
100
100
100
74
40
100
22
6
7
8
9
10
11
12
13
PhMeSiH2
PhMeSiH2
PhMeSiH2
PhMeSiH2
Ph2SiH2
Ph2SiH2
Ph2SiH2
nPrNH2
tBuNH2
iPrNH2
Et2NH
9.3
7.8
9.3
10.2
6.8
8.1
7.8
8.5
5.0
5.0
4.0
6.0
3.0
6.0
6.0
7.0
PhMeSiH(HNnPr)
PhMeSiH(HNtBu)
PhMeSiH(HNiPr)
PhMeSiH(NEt2)
Ph2Si(HNnPr)
24
24
24
24
24
24
24
24
nPrNH2
tBuNH2
iPrNH2
Et2NH
Ph2Si(HNtBu)
Ph2Si(HNiPr)
Ph2SiH2
Ph2SiH(NEt2)
a Conditions: visible-light irradiation in benzene-d6 solution at ambient temperature. Reactions were monitored by 1H, 29Si{1H}, and 1H–29Si
HSQC NMR spectroscopy. b Mol% of 1 was with respect to silane. c Refers to mol. of amine per mol. of silane. d Conversions were determined by
1H NMR integration.
PhMeSi(NEt2)2, literature chemical shifts do not agree,45 and
this minor byproduct remains unidentified.
Heterodehydrocoupling reactions with amines and Ph2SiH2
catalyzed by compound 1 were also tested. Reaction between
nPrNH2 and Ph2SiH2 showed 74% conversion to Ph2SiH
ð3Þ
(HNnPr) after 24 h according to 1H NMR spectroscopy (Table 2,
t
entry 10).44 Conversely, reaction between Ph2SiH2 and BuNH2
showed only 50% conversion to Ph2SiH(HNtBu) after 24 h
(Table 2, entry 11).42 These observations indicated that steric
factors can play a more significant role when both the amine
and silane exhibit steric pressure. Of note, steric factors were
more pronounced with the alcohol substrate in silyl ether reac-
tions (vide supra). This supposition is buttressed by the reac-
tion of Et2NH and Ph2SiH2 in which 22% conversion to
Ph2SiH(NEt2) was observed after 24 h, despite 7 equiv. of
amine to silane (Table 2, entry 13).43 The balance can be
That P–H bond activation was also not quantitative, doubt-
lessly related to the kinetics of visible-light activation of 1.21
The known decomposition of 4 to 1 and the possibility of a
process that directly converts 4 to 5 with free PhSiH3 likely con-
tribute to the ∼20% excess of
5
as compared to 4.47
Observation of catalytic reactions with PhSiH3 by 1H NMR
spectroscopy confirm formation of 5 under catalytic conditions
as well as apparently unreacted 1. Apparent Si–H bond acti-
vation products at iron are consistently presented in catalytic
reactions, regardless of substrate.
i
tipped back with amine substitution where reaction of PrNH2
and Ph2SiH2 gave nearly quantitative conversion to Ph2SiH
(HNiPr) with a minor byproduct discernible only in 1H–29Si
HSQC NMR (Table 2, entry 12).38 Finally, attempts at coupling
these amines with tertiary silanes such as PhMe2SiH or Et3SiH
were unsuccessful.
This observation suggests that iron could activate the orga-
nosilane substrate for nucleophilic attack by alcohol. To test
this supposition, a known silyl derivative, Cp(CO)2FeSiMe2Ph
(6) was prepared.48 Treatment of 6 with 1 equiv. of nPrOH
failed to afford the anticipated silyl ether to any detectable
extent by 1H NMR spectroscopy, and variations on the reaction
including 10 equiv. of alcohol, irradiation, or heating failed to
Mechanistic study
n
Treatment of 1 with 1 equiv. of PrOH resulted in no observa- afford silyl ether as well. However, treatment of 4 equiv. of
ble change as monitored by 1H NMR spectroscopy after 24 h of BnOH with PhSiH3 in the presence of 2 mol% of 6 resulted in
visible-light irradiation in a benzene-d6 solution. In contrast, formation of PhSiH(OBn)2 and PhSi(OBn)3 in 28% and 60%
reaction of equimolar 1 and PhSiH3 over 24 h in benzene-d6 conversions, respectively, after 24 h of irradiation. While
under visible-light irradiation resulted in approximately 22% initially promising, comparative reaction with catalytic 1 fully
1
formation of hydride 4 as measured by H NMR spectroscopy converts PhSiH3 and BnOH to PhSi(OBn)3 after only 6 h
(eqn (3)).46 A new iron compound, tentatively assigned as Cp (Table 1, entry 2). Moreover, the lack of reactivity between stoi-
(CO)2FeSiH2Ph (5) based on resonances at δ 5.22 (SiH) and δ chiometric 6 and nPrOH cannot be ignored. These obser-
3.98 (C5H5), was observed in approximately 27% conversion. vations suggest that an iron-silyl species like 6 converts to 4
The activation of an E–H bond under photolysis of 1 to form 4 under catalytic conditions, but speciation is unclear during
has been observed with phosphines.20
catalysis, though silyl derivatives are detectable in trace quan-
Dalton Trans.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Dalton Transactions
Paper
tities by 1H NMR spectroscopy in observations of catalytic reac- and catalytic reactions, it is clear that the activation of 1 is not
tions using 1. complete during the reaction, but irradiation would form two
The prevalence of species such as 4, 5, and 6 in catalysis equiv. of 3, which would activate silane substrate to hydride 4
hints that active compounds are formally 18-electron deriva- and a silyl compound. While direct formation of 4 from 6
tives, Cp(CO)2FeX (X = silyl, hydride, alkoxide, etc.). Testing for under H2 was not observed, trace amounts of 4 are detected
17-election intermediates was inconclusive due to the necess- while using 6 catalytically. It is believed that some pathway
ary of Cp(CO)2Fe• as an intermediate under the photochemical allows for the conversion of iron-silyl to 4, albeit in trace quan-
conditions. However, 18-electron compounds are unavailable tities due to the prevalence of these species during catalysis.
for organometallic steps such as oxidative addition or σ-bond Hydride 4 is known to decompose back to 1, which may also
metathesis due to formal electron count, and ligands such as contribute to the steady state concentration of 1 during cataly-
silyl and hydride are less likely to engage in migratory inser- sis as observed by 1H NMR spectroscopy.47 However, 4 likely
tion with carbonyl ligands to activate those derivatives com- reacts with alcohol or amine to give a highly unstable alkoxide
pared to Cp(CO)2FeMe, for example.49,50 This deduction or amide intermediate with evolution of H2. This intermediate
process leaves nucleophilic attack as the most viable mechan- can then attack a silane substrate to form product and regener-
istic hypothesis with available data. Of course, many metals ate 4.
promote nucleophilicity of ligands.51 While we cannot observe
Although Si–O and Si–S heterodehydrocoupling by
an iron-alkoxide or -amide compound in solution, we cannot Nakazawa and co-workers via Cp(CO)2FeMe is a similar
discount it. Such an intermediate would be more nucleophilic system, starting from 1 and the prevalence of 18-electron com-
than its parent alcohol. Indeed, the relative reactivity of PhNH2 pounds such as 4, 5, and 6 as intermediates shows that a
i
and PrNH2 support nucleophilicity at the coupling partner. different activation step is likely here.26,27 A 17-electron σ-
Moreover, attempts at synthesizing Cp(CO)2FeOtBu or Cp silane intermediate could be a potential intermediate and
(CO)2FeN(SiMe3)2 via metathetical reaction of Cp(CO)2FeBr highly reactive.52 We have been unable to differentiate the
with the corresponding alkoxide or amide salt were unsuccess- influence of 17-electron compounds other than Cp(CO)2Fe•,
ful. These results further substantiate the idea that these Fp- and the presence 4 and 5 challenge such a pathway in the
alkoxide or -amide intermediates are highly reactive, short- absence of better data.
lived species. Indeed, Sadow and coworkers have proposed
Perhaps the most important observation from this mechan-
nucleophilic attack of a magnesium-amide intermediate and istic proposal is not the Si–O or Si–N bond-forming step. There
formation of a magnesium-hydride in Si–N heterodehydrocou- is far less active catalyst in the system than the loading of 1
pling.44 Moreover, while literature on isolated piano-stool iron- would indicate, even if the silyl intermediate were completely
alkoxides or -amides is scarce, Nakazawa and co-workers have inactive under catalytic conditions. This information is a clear
implicated piano-stool iron-alkoxide and iron-thio intermedi- indication that a meager fraction of potential activity is being
ates in catalytic Si–O and Si–S heterodehydrocoupling, realized, which limits catalyst activity. These issues are clearly
respectively.26,27
important as the conversation over utilization of base metals
Based on the stoichiometric reactions and observations of and lower energy reactions continues.
the catalysis (though NMR observations were made in the
absence of active irradiation), an initial proposal for the cata-
lytic cycle can be made (Scheme 2). From both stoichiometric
Conclusions
Commercially available iron compound 1 is efficient at Si–O
heterodehydrocoupling under visible-light irradiation.
Reactions between alcohols and silanes catalyzed by 1 afforded
silyl ethers often in quantitative conversions from starting
silanes. Sterically encumbered silanes generally required
longer reaction times but provided near quantitative conver-
sion from starting silanes. Compound 1 is also a competent
Si–N heterodehydrocoupling catalyst. However, longer reaction
times and higher catalyst loadings were necessary to produce
silamines in good conversions. Furthermore, electron-rich
amines were shown to be the most effective substrates to
convert to silamines. Mechanistic study is consistent with
nucleophilic attack of an intermediate iron-alkoxide or -amide
at the organosilane substrate. More important to future study,
though, is the necessity for complete activation of catalyst to
achieve optimal conversions. The ‘unactivated’ fraction of cata-
lyst may be a significant factor in the disparity between base
and noble metals in catalysis, suggesting an area for deeper
Scheme 2 Proposed mechanism for Si–O and Si–N heterodehydro-
coupling catalysed by 1.
This journal is © The Royal Society of Chemistry 2020
Dalton Trans.

View Article Online
Paper
Dalton Transactions
investigation. More specifically, this work expands upon the
heterodehydrocoupling capabilities of 1,19,20 and represents
one of the few instances of mild, photoactivated iron-based
catalysts.
6 S. Pan and T. Shibata, ACS Catal., 2013, 3, 704–712.
7 M. C. White, Adv. Synth. Catal., 2016, 358, 2364–2365.
8 E. Bisz and M. Szostak, J. Org. Chem., 2019, 84, 1640–1646.
9 B. Cheng, W. Liu and Z. Lu, J. Am. Chem. Soc., 2018, 140,
5014–5017.
10 R. J. Trovitch, Acc. Chem. Res., 2017, 50, 2842–2852.
11 Q. Bu, E. Gońka, K. Kuciński and L. Ackermann, Chem. –
Eur. J., 2019, 25, 2213–2216.
12 N. Barsu, D. Kalsi and B. Sundararaju, Catal. Sci. Technol.,
2018, 8, 5963–5969.
13 F. D. Monica, A. Buonerba and C. Capacchione, Adv. Synth.
Catal., 2019, 361, 265–282.
14 A. Fürstner, ACS Cent. Sci., 2016, 2, 778–789.
15 D. Wei and C. Darcel, Chem. Rev., 2019, 119, 2550–2610.
16 M. Parasram and V. Gevorgyan, Chem. Soc. Rev., 2017, 46,
6227–6240.
Experimental
General information
All reactions were prepared under purified N2 atmosphere in
an M. Braun glovebox. Cyclopentadienyl dicarbonyl iron(II)
dimer 1 was purified by sublimation. Alcohols and amines
were distilled from CaH2. Silanes were received from chemical
vendors and used without further purification. Benzene-d6 was
vacuum transferred from NaK alloy. NMR spectra were
acquired on either a Varian 500 MHz spectrometer or a Bruker
AXR 500 MHz spectrometer. Spectra recorded on both instru-
ments were reported to SiMe4 (δ 0.00).
17 I. Delidovich and R. Palkovits, Green Chem., 2016, 18, 590–
593.
18 R. Waterman, Chem. Soc. Rev., 2013, 42, 5629–5641.
19 J. R. Vance, A. Schäfer, A. P. M. Robertson, K. Lee,
J. Turner, G. R. Whittell and I. Manners, J. Am. Chem. Soc.,
2014, 136, 3048–3064.
20 J. K. Pagano, C. A. Bange, S. E. Farmiloe and R. Waterman,
Organometallics, 2017, 36, 3891–3895.
21 T. E. Bitterwolf, Coord. Chem. Rev., 2000, 419–450.
22 J. Seliger and M. Oestreich, Chem. – Eur. J., 2019, 25, 9358–
9365.
23 J. M. Mabry, PhD thesis, University of Southern California,
2002.
24 C. Bellini, T. Roisnel, J.-F. Carpentier, S. Tobisch and
Y. Sarazin, Chem. – Eur. J., 2016, 22, 15733–15743.
25 D. Fonblanc, D. Lopez-Ferber, M. Wynn, A. Lale,
A. Soleilhavoup, A. Leriche, Y. Iwamoto, F. Rossignol,
C. Gervais and S. Bernard, Dalton Trans., 2018, 47, 14580–
14593.
Catalytic experiment conditions
An oven-dried scintillation vial containing
1 (3.5 mg,
2.0 mol%) was charged with silane, followed by excess alcohol,
0.5 mL benzene-d6, and TMS. A similar method was performed
with amine coupling, however, loading of 1 was determined by
substrates. Mixtures were transferred to
a J-Young type
polytetrafluoroethylene-valved NMR tube and subsequently
placed under visible-light irradiation. Reactions were subjected
to a cycle of freeze–pump–thaw after 1 and 2 h of irradiation.
All reactions were performed at ambient temperature under
irradiation in the visible spectrum using a 40 W LED bulb.
Conflicts of interest
The authors have no conflicts of interest to declare.
26 K. Fukumoto, M. Kasa and H. Nakazawa, Inorg. Chim. Acta,
2015, 431, 219–221.
27 K. Fukumoto, M. Kasa, T. Oya, M. Itazaki and
H. Nakazawa, Organometallics, 2011, 30, 3461–3463.
28 S.-S. Yun, I.-H. Suh, E. H. Kim, B.-J. Choi and S. Lee,
J. Organomet. Chem., 2001, 631, 16–18.
Acknowledgements
This work was funded by the National Science Foundation
through CHE-1565658. The authors thank Dr Monika Ivancic
for assistance with NMR spectra.
29 T. K. Mukhopadhyay, C. L. Rock, M. Hong, D. C. Ashley,
T. L. Groy, M.-H. Baik and R. J. Trovitch, J. Am. Chem. Soc.,
2017, 139, 4901–4915.
30 W. Sattler, S. Ruccolo, M. R. Chaijan, T. N. Allah and
G. Parkin, Organometallics, 2015, 34, 4717–4731.
31 M. C. Neary, P. J. Quinlivan and G. Parkin, Inorg. Chem.,
2018, 57, 374–391.
Notes and references
1 A. Biffis, P. Centomo, A. D. Zotto and M. Zecca, Chem. Rev.,
2018, 118, 2249–2295.
2 P. Ruiz-Castillo and S. L. Buckwald, Chem. Rev., 2016, 116,
12564–12649.
32 D. Gao and C. Cui, Chem. – Eur. J., 2013, 19, 11143–
11147.
3 T. M. Meister, K. Riener, P. Gigler, J. Stohrer,
W. A. Hermann and F. E. Kühn, ACS Catal., 2016, 6, 1274– 33 J. M. S. Cardosa, R. Lopes and B. Royo, J. Organomet.
1284. Chem., 2015, 775, 173–177.
4 P. Etayo and A. Vidal-Ferran, Chem. Soc. Rev., 2013, 42, 728– 34 A. A. Toutov, K. N. Betz, M. C. Haibach, A. M. Romine and
754. R. H. Grubbs, Org. Lett., 2016, 18, 5776–5779.
5 R. Franke, D. Selent and A. Börner, Chem. Rev., 2012, 112, 35 M. Kahnes, H. Görls, L. González and M. Westerhausen,
5475–5732. Organometallics, 2010, 29, 3098–3108.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Dalton Transactions
Paper
36 A. Harinath, J. Bhattacharjee, S. Anga and T. K. Panda, 44 J. F. Dunne, S. R. Neal, J. Engelkemier, A. Ellern and
Aust. J. Chem., 2017, 70, 724–730.
A. D. Sadow, J. Am. Chem. Soc., 2011, 133, 16782–
37 S. Vijjamarri, V. K. Chidara, J. Rousova and G. Du, Catal.
Sci. Technol., 2016, 6, 3886–3892.
16785.
45 M. P. Cibuzar and R. Waterman, Organometallics, 2018, 37,
38 M. Igarashi, T. Matsumoto, K. Sato, W. Ando and
S. Shimada, Chem. Lett., 2014, 43, 429–431.
4395–4401.
46 R. M. Bullock and E. G. Samsel, J. Am. Chem. Soc., 1990,
39 J. Francos, J. Borge, S. Conejero and V. Cadierno,
Eur. J. Inorg. Chem., 2018, 3176–3186.
112, 6886–6898.
47 T. A. Shackleton and M. C. Baird, Organometallics, 1989, 8,
40 J. X. Wang, A. K. Dash, J. C. Berthet, M. Ephritikhine and
M. S. Eisen, J. Organomet. Chem., 2000, 610, 49–57.
41 N. V. Forosenko, I. V. Basalov, A. V. Cherkasov, G. K. Fukin,
2225–2232.
48 K. H. Pannell and C. C. Wu, J. Organomet. Chem., 1980,
186, 85–90.
E. S. Shubina and A. A. Trifonov, Dalton Trans., 2018, 47, 49 N. M. West, A. J. M. Miller, J. A. Labinger and J. E. Bercaw,
12570–12581. Coord. Chem., 2011, 255, 881–898.
42 A. E. Nako, W. Chen, A. J. P. White and M. R. Crimmin, 50 R. Waterman, Organometallics, 2013, 32, 7249–7263.
Organometallics, 2015, 34, 4369–4375. 51 D. S. Glueck, Dalton Trans., 2008, 5276–5286.
43 A. Baishya, T. Peddarao and S. Nembenna, Dalton Trans., 52 E. Matarasso-Tchiroukhine, J. Chem. Soc., Chem. Commun.,
2017, 46, 5880–5887.
1990, 681–682.
This journal is © The Royal Society of Chemistry 2020
Dalton Trans.