Experimental evidence for cobalt(III)-carbene radicals: Key intermediates in cobalt(II)-based metalloradical cyclopropanation

Article abstract of DOI:10.1021/ja203434c

New and conclusive evidence has been obtained for the existence of cobalt(III)-carbene radicals that have been previously proposed as the key intermediates in the underlying mechanism of metalloradical cyclopropanation by cobalt(II) complexes of porphyrins. In the absence of olefin substrates, reaction of [Co(TPP)] with ethyl styryldiazoacetate was found to generate the corresponding cobalt(III)-vinylcarbene radical that subsequently dimerizes via its γ-radical allylic resonance form to afford a dinuclear cobalt(III) porphyrin complex. X-ray structural analysis reveals a highly compact dimeric structure wherein the two metalloporphyrin units are arranged in a face-to-face fashion through a tetrasubstituted 1,5-hexadiene C₆-bridge between the two Co(III) centers. The γ-radical allylic resonance form of the cobalt(III)-vinylcarbene radical intermediate could be effectively trapped by TEMPO via C-O bond formation to give a mononuclear cobalt(III) complex instead of the dimeric product. The allylic radical nature and related reactivity profile of the cobalt(III)-carbene radical, including its inability to abstract hydrogen atoms from toluene solvent, were established by DFT calculations.

Full text of DOI:10.1021/ja203434c

COMMUNICATION
pubs.acs.org/JACS
Experimental Evidence for Cobalt(III)-Carbene Radicals: Key
Intermediates in Cobalt(II)-Based Metalloradical Cyclopropanation
Hongjian Lu,† Wojciech I. Dzik,‡ Xue Xu,† Lukasz Wojtas,† Bas de Bruin,*,‡ and X. Peter Zhang*,†
†Department of Chemistry, University of South Florida, Tampa, Florida 33620-5250, United States
‡Department of Homogeneous and Supramolecular Catalysis, Van't Hoff Institute for Molecular Sciences,
University of Amsterdam, Science Park 904, 1098 XH Amsterdam, The Netherlands
S Supporting Information
b
electron-deficient olefins has proven to be challenging. The
ABSTRACT: New and conclusive evidence has been
obtained for the existence of cobalt(III)-carbene radicals
that have been previously proposed as the key intermediates
in the underlying mechanism of metalloradical cyclopropa-
nation by cobalt(II) complexes of porphyrins. In the ab-
sence of olefin substrates, reaction of [Co(TPP)] with ethyl
styryldiazoacetate was found to generate the corresponding
cobalt(III)-vinylcarbene radical that subsequently dimerizes
via its γ-radical allylic resonance form to afford a dinuclear
cobalt(III) porphyrin complex. X-ray structural analysis
reveals a highly compact dimeric structure wherein the
two metalloporphyrin units are arranged in a face-to-face
fashion through a tetrasubstituted 1,5-hexadiene C6-bridge
between the two Co(III) centers. The γ-radical allylic
resonance form of the cobalt(III)-vinylcarbene radical inter-
mediate could be effectively trapped by TEMPO via CꢀO
bond formation to give a mononuclear cobalt(III) complex
instead of the dimeric product. The allylic radical nature and
related reactivity profile of the cobalt(III)-carbene radical,
including its inability to abstract hydrogen atoms from
toluene solvent, were established by DFT calculations.
electrophilic nature of the metallocarbene intermediates is also
responsible for the notorious side reaction of carbene dimeriza-
tion, a problem in cyclopropanation that often has to be addressed
through the slow addition of diazo reagents and/or by the use of
excess olefins.
Different from the Rh(II)2-based and other closed-shell metal
catalysts, cobalt(II) complexes of porphyrins, [Co(Por)], which
exist as stable metalloradicals with a well-defined low-spin
1
(dxy)2(dxz,yz)4(dz2
)
electronic configuration, have recently
been developed into a unique class of open-shell catalysts for
cyclopropanation.5 Supported by a toolbox of D2-symmetric
chiral porphyrins with tunable electronic, steric, and chiral
environments,5a,6 we have demonstrated that Co(II)-based chiral
metalloradical catalysts [Co(D2-Por*)] are highly effective for
asymmetric cyclopropanation of a broad scope of olefin substrates
in excellent diastereoselectivity and enantioselectivity.5ꢀ9 Parti-
cularly, the Co(II)-based metalloradical system permits for the
first time efficient asymmetric cyclopropanation of electron-
deficient olefins.7bꢀd,f,g Furthermore, as a practical attribute that
is atypical for other metal-based catalytic systems, Co(II)-cata-
lyzed cyclopropanation reactions can be operated in a one-time
protocol without the need of slow addition of diazo reagents and
with the alkene as the limiting reagent owing to the absence
of the competitive carbene dimerization side reaction. It is evident
that the cyclopropanation reactivity profile of the Co(II)-based
open-shell metalloradical system is distinctly different from
those of the widely studied Rh2 and other closed-shell systems,
suggesting the involvement of catalytic intermediates that are
nonelectrophilic.5b,7a,10
Our recent EPR spectroscopic investigations and systematic
DFT calculations have led to the disclosure of an unprecedented
reaction mechanism of the Co(II)-catalyzed cyclopropanation
(Scheme 1) that involves an unusual Co(III)-carbene radical
(B) as the key intermediate undergoing a stepwise radical addi-
tionꢀsubstitution pathway.11 Different from the common electro-
philic Fischer-type carbene intermediates, the new radical carbene
intermediate B is effectively an [Co(Por)]-supported carbon-
based radical,12 which is capable of undergoing common radical
reactions (such as radical addition and radical substitution)13 but is
not “free”, as its reactivity is well controlled by the electronic, steric,
and chiral environments of the covalently attached [Co(Por)]
element. While the proposed Co(III)-carbene radical intermediate
yclopropanation of alkenes via carbene transfer from diazo
C
reagents represents an important class of organic transfor-
mations. The resulting three-membered cyclopropanes, espe-
cially in their optically active forms, have found a wide range of
fundamental and practical applications.1 To enhance reactivity
and control stereoselectivities, complexes of different transition
metals supported by various chiral ligands have been developed
to catalyze the stereoselective construction of the smallest
carbocyclic structures.2 Represented by the most recognized
dirhodium(II) complexes of chiral carboxamidates and carbox-
ylates, several classes of chiral metal catalysts have been demon-
strated to enable successful asymmetric cyclopropanation
reactions for certain combinations of alkenes and diazo reagents,
affording the corresponding cyclopropane derivatives in high
yields with excellent diastereoselectivity and enantioselectivity.2,3
Mechanistically, it has been shown that most existing catalytic
systems share a general cyclopropanation mechanism that in-
volves electrophilic (Fischer-type) metallocarbenes as the key
intermediates.2ꢀ4 Because of the electrophilic nature of the
metallocarbene intermediates, highly enhanced reactivity has
been achieved for cyclopropanation of aromatic and other
electron-rich olefins. For the same reason, cyclopropanation of
Received: April 14, 2011
Published: May 12, 2011
r
2011 American Chemical Society
8518
dx.doi.org/10.1021/ja203434c J. Am. Chem. Soc. 2011, 133, 8518–8521
|

Journal of the American Chemical Society
COMMUNICATION
Scheme 1. Mechanism of Metalloradical Cyclopropanation
Scheme 3. Generation of Cobalt(III)-Vinylcarbene Radical
from Reaction of [Co(TPP)] with Ethyl Styryldiazoacetate
and Subsequent Homodimerization via CꢀC Bond
Formation
Scheme 2. Stoichiometric Reactions of [Co(TPP)] with
Diazo Reagents and Subsequent H-Atom Abstraction Process
atom abstraction process (from solvents or some other sources in
the system), Cenini and co-workers proposed a different Co(II)
intermediate for its formation.5c In light of this ambiguity,5c it is
clear that observation of 2 could not be taken as convincing
evidence for the existence of the Co(III)-carbene radicals.
In a search for more definite experimental evidence for the key
Co(III)-carbene radical intermediate, we then turned our atten-
tion to the use of ethyl styryldiazoacetate as the carbene source
for the stoichiometric reaction with [Co(TPP)] (Scheme 3). It
was envisioned that the corresponding carbene R-radical 3 would
be stabilized by its γ-radical resonance form 30 due to its allylic
nature (Scheme 3). Different from our previous experiments
with other diazo reagents, the reaction of [Co(TPP)] with 4
equiv of ethyl styryldiazoacetate in toluene at 80 °C was found to
be pleasingly clean and generated only one major product (4),
which could be isolated in 90% yield after 3 h of reaction. The
data obtained from NMR and HRMS measurements (see
Supporting Information) show 4 to be a dimeric cobalt(III)
porphyrin complex. Particularly, proton resonances with unusual
negative chemical shifts appeared in the high-field region of its 1H
NMR spectrum, which is indicative of the presence of hydrogen-
containing ligands in the axial position of 4. The detailed
structure of 4 (Scheme 3) was finally unveiled by single-crystal
X-ray diffraction, which is consistent with all the spectroscopic
data. As shown in Figure 1, the two cobalt porphyrins in 4 are
arranged in a face-to-face fashion through a tetrasubstituted 1,
5-hexadiene C6-bridge between the two Co(III) centers; the
distance between the two cobalt ions is 9.02 Å. It is instantly
recognizable that the dinuclear complex 4 was formed from the
radical dimerization of the initially generated Co(III)-carbene
R-radical 3 via its γ-radical allylic resonance form 30 (Scheme 3).13
In an effort to probe the electronic structure of the cobalt(III)-
vinylcarbene radical, DFT calculations were performed on
(Por)Co-C(CO2Me)(CHCHPh) (5, Figure 2), which serves
as a model complex for 3 (Scheme 3). As illustrated by Figure 2,
the calculated spin density is mainly distributed over the “car-
bene” moiety with little spin density left at the metal center,
indicating effective spin transfer from the Co(II) ion of
[Co(TPP)] to the “carbene” moiety of the product after the
reaction. Moreover, the spin density in the “carbene” moiety is
found to be concentrated mostly on the R and γ atoms with
similar values (R-C, 46.8; γ-C, 49.9), which supports a strong
B was experimentally detected from a reaction mixture by EPR
spectroscopy and ESI mass spectrometry,11,14 our early efforts
toward direct characterizations of the intermediates in isolated
forms have been met with formidable challenges due to their high
reactivity, resulting in multiple decomposition products. With the
use of ethyl styryldiazoacetate as a unique carbene source, herein
we wish to report our recent success in obtaining conclusive
chemical evidence to support the existence of the Co(III)-carbene
radicals. Takingadvantage ofallylic radicalresonance, weshowthat
the resulting cobalt(III)-vinylcarbene radical, generated in a stoi-
chiometric reaction of ethyl styryldiazoacetate with [Co(TPP)] in
the absence of olefin substrates, cleanly dimerizes and can be
effectively trapped by TEMPO in its allylic γ-radical resonance
form to afford the respective CꢀC or CꢀO radical coupling
products in high yields. Based on DFT calculations, we give an
account on the spin density distribution of the cobalt(III)-vinyl-
carbene radical and provide both kinetic and thermodynamic
grounds for its stability toward H-atom abstraction reactions.
To characterize the Co(III)-carbene radicals, a series of diazo
reagents with different types of R substituents, including diazo-
acetates, aryldiazoacetates, and diazomalonates, wereemployed as
carbene sources to react with [Co(TPP)] in the absence of olefin
substrates (Scheme 2). Our attempts from these experiments
failed to provide concrete evidence for the formation of carbene
radicals 1, as these stoichiometric reactions typically gave a
complex mixture of products. Among several organic and orga-
nometallic species in the reaction mixtures, the corresponding
Co(III)-alkyl complex 2 (Scheme 2, R1 = H, R2 = CO2Et) could
be positively identified by NMR spectroscopy from the reaction
with ethyl diazoacetate. The single-crystal X-ray structure of this
compound was previously reported by Cenini and co-workers.5c
Chan and co-workers recently reported an analogous reaction
with [Rh(II)(Por)].15 While it seems reasonable to assume that 2
would be formed directly from the Co(III) carbon-based radical
intermediate 1 (Scheme 2, R1 = H, R2 = CO2Et) by a hydrogen
8519
dx.doi.org/10.1021/ja203434c |J. Am. Chem. Soc. 2011, 133, 8518–8521

Journal of the American Chemical Society
COMMUNICATION
Scheme 4. DFT-Calculated Kinetic and Thermodynamic
Parameters for Hydrogen Atom Transfer from Toluene to
Cobalt(III)-Vinylcarbene Radical and Its Allylic Resonance
Form
the hydrogen atom abstraction process were often observed in
the reactions with ethyl diazoacetate and other diazo reagents
(Scheme 2). To understand this unique reactivity profile of ethyl
styryldiazoacetate as the carbene source, the hypothetical hydro-
gen atom abstraction reaction of cobalt(III)-vinylcarbene
R-radical 5 and its allylic resonance form γ-radical 50 from toluene
was investigated by DFT calculations (Scheme 4). According to
DFT, the hydrogen atom transfer (HAT) from toluene to the
R-carbon position (reaction via the R-radical resonance form 5)
and the γ-carbon position (reaction via the γ-radical resonance
form 50) is uphill by þ5 and þ16 kcal molꢀ1, respectively (see
Supporting Information). The high value of the energy asso-
ciated with HAT from toluene to γ-radical 50 explains well the
absence of the corresponding hydrogen atom abstraction pro-
duct 60 (Scheme 4). Since the value of þ5 kcal molꢀ1 associated
with the HAT process to R-radical 5 is not high enough to explain
the absence of the corresponding hydrogen atom abstraction
product 6, the activation barrier for the process was calculated.
Hydrogen atom transfer from toluene to R-radical 5 at the
sterically least hindered side (i.e., the toluene facing the side of
the carbonyl group; see Supporting Information) has a calculated
transition state barrier of þ20.8 kcal molꢀ1 (relative to the
isolated vinyl carbene radical and toluene). This barrier is high
enough to hamper any HAT, since the competing CꢀC coupling
of γ-radical 5 is essentially barrierless. Hence, we conclude that
the absence of 6 is kinetic in origin, while the formation of 60 is
thermodynamically unfavorable.
Figure 1. Single-crystal X-ray structure of 4 with thermal ellipsoids
drawn at 30% probability. Hydrogen atoms and chloroform solvent
molecule are omitted for clarity. Selected bond distances (Å): Co1ꢀC1,
1.917(4); C1ꢀC2, 1.325(5); C2ꢀC3, 1.508(6); C3ꢀC4, 1.574(6);
C4ꢀC5, 1.521(6); C5ꢀC6, 1.305(5); C6ꢀCo2, 1.931(4); Co1--Co2,
9.0170(9).
Figure 2. Plot of the calculated spin density distribution for (Por)Co-
C(CO2Me)(CHCHPh) model complex 5, showing substantial deloca-
lization of the unpaired electron over the R and γ carbon atoms. The
numbers represent the calculated spin densities on the respective atoms.
Although the homocoupling of γ-radical 30 is a very facile
process (Scheme 3), it was shown that the radical could be
effectively captured by radical traps. For example, when the
reaction of [Co(TPP)] with 4 equiv of ethyl styryldiazoacetate
was performed in toluene at 80 °C in the presence of an excess of
the stable radical 2,2,6,6-tetramethylpiperidine N-oxide (TEMPO)
(10 equiv), heterocoupling between γ-radical 30 and TEMPO
was found to be the dominant process, affording the heterodimer
7 via CꢀO bond formation in 74% isolated yield, with almost no
formation of the homocoupling product 4 (Scheme 5).16,17 The
isolation and characterization of 7 provide further convincing
evidence for the existence of cobalt(III)-vinylcarbene R-radical 3
and its allylic resonance form γ-radical 30.
contribution of the allylic radical resonance structures 3 (carbene
R-radical) and 30 (carbene γ-radical) to the electronic structure of
3 (Scheme 3). Further DFT calculations using the broken-
symmetry methodology (see Supporting Information) indicate
that the dimerization of carbene γ-radical 30 via radical CꢀC
coupling is a low-barrier process, explaining the facile formation of
4 (Scheme 3). The absence of the alternative dimerization process
through the R-radical resonance form 3 is probably steric in origin.
The radical dimerization (Scheme 3) proceeded very cleanly
and was found to be free of products from hydrogen atom
abstraction, even though the reaction was performed with
toluene as the solvent. As previously mentioned, products from
Insummary, ourcombined experimentaland theoretical studies
have shown that the metalloradical [Co(TPP)] is capable of
activating ethyl styryldiazoacetate to generate the corresponding
8520
dx.doi.org/10.1021/ja203434c |J. Am. Chem. Soc. 2011, 133, 8518–8521

Journal of the American Chemical Society
COMMUNICATION
Scheme 5. Generation of Cobalt(III)-Vinylcarbene Radical
from Reaction of [Co(TPP)] with Ethyl Styryldiazoacetate
and Subsequent Trapping by TEMPO via CꢀO Bond
Formation
(4) (a) Doyle, M. P. Acc. Chem. Res. 1986, 19, 348. (b) Nowlan,
D. T.; Gregg, T. M.; Davies, H. M. L.; Singleton, D. A. J. Am. Chem. Soc.
2003, 125, 15902. (c) Hansen, J.; Autschbach, J.; Davies, H. M. L. J. Org.
Chem. 2009, 74, 6555.
(5) (a) Doyle, M. P. Angew. Chem. Int. Ed. 2009, 48, 850. (b) Huang,
L. Y.; Chen, Y.; Gao, G. Y.; Zhang, X. P. J. Org. Chem. 2003, 68, 8179. (c)
Penoni, A.; Wanke, R.; Tollari, S.; Gallo, E.; Musella, D.; Ragaini, F.;
Demartin, F.; Cenini, S. Eur. J. Inorg. Chem. 2003, 1452.
(6) These D2-symmetric chiral porphyrins can be readily prepared
through a modular, three-step synthetic scheme: Chen, Y.; Fields, K. B.;
Zhang, X. P. J. Am. Chem. Soc. 2004, 126, 14718.
(7) (a) Chen, Y.; Zhang, X. P. J. Org. Chem. 2007, 72, 5931. (b)
Chen, Y.; Ruppel, J. V.; Zhang, X. P. J. Am. Chem. Soc. 2007, 129, 12074.
(c) Zhu, S. F.; Ruppel, J. V.; Lu, H. J.; Wojtas, L.; Zhang, X. P. J. Am.
Chem. Soc. 2008, 130, 5042. (d) Zhu, S. F.; Perman, J. A.; Zhang, X. P.
Angew. Chem. Int. Ed. 2008, 47, 8460. (e) Fantauzzi, S.; Gallo, E.; Rose,
E.; Raoul, N.; Caselli, A.; Issa, S.; Ragaini, F.; Cenini, S. Organometallics
2008, 27, 6143. (f) Ruppel, J. V.; Gauthier, T. J.; Snyder, N. L.; Perman,
J. A.; Zhang, X. P. Org. Lett. 2009, 11, 2273. (g) Zhu, S. F.; Xu, X.;
Perman, J. A.; Zhang, X. P. J. Am. Chem. Soc. 2010, 132, 12796.
(8) For a recent report on [Co(D2-Por*)]-catalyzed enantioselective
cyclopropenation of alkynes, see: Cui, X.; Xu, X.; Lu, H.; Zhu, S.; Wojtas,
L.; Zhang, X. P. J. Am. Chem. Soc. 2011, 133, 3304.
(9) For Co(II)-catalyzed asymmetric cyclopropanation using non-
porphyrin-based ligands, see: (a) Nakamura, A.; Konishi, A.; Tatsuno,
Y.; Otsuka, S. J. Am. Chem. Soc. 1978, 100, 3443. (b) Ikeno, T.; Sato, M.;
Sekino, H.; Nishizuka, A.; Yamada, T. Bull. Chem. Soc. Jpn. 2001,
74, 2139. (c) Niimi, T.; Uchida, T.; Irie, R.; Katsuki, T. Adv. Synth.
Catal. 2001, 343, 79.
Co(III)-vinylcarbene radical that exists in R-carbon-based and
γ-carbon-based allylic radical resonance forms. The Co(III)-
supported carbene radical, which is incompetent for H-atom
abstraction, can cleanly self-dimerize via radical CꢀC coupling
or can be effectively trapped by TEMPO via radical CꢀO
coupling. We take these results as new experimental evidence
for the existence of Co(III)-carbene radicals that are key inter-
mediates in the radical-type mechanism of Co(II)-based metallo-
radical catalysis for cyclopropanation.
(10) Chen, Y.; Zhang, X. P. J. Org. Chem. 2004, 69, 2431.
(11) Dzik, W. I.; Xu, X.; Zhang, X. P.; Reek, J. N. H.; de Bruin, B.
J. Am. Chem. Soc. 2010, 132, 10891.
’ ASSOCIATED CONTENT
(12) (a) Dzik, W. I.; Zhang, X. P.; de Bruin, B. Inorg. Chem. 2011,
DOI: 10.1021/ic200043a. (b) Belof, J. L.; Cioce, C. R.; Xu, X.; Zhang,
X. P.; Space, B.; Woodcock, H. L. Organometallics 2001, DOI: 10.1021/
om2001348.
(13) (a) Dzik, W. I.; Reek, J. N. H.; de Bruin, B. Chem. Eur. J. 2008,
14, 7594. (b) Fuchibe, K.; Iwasawa, N. Chem. Eur. J. 2003, 9, 905. (c)
Sierra, M. A.; Ramirez-Lopez, P.; Gomez-Gallego, M.; Lejon, T.;
Mancheno, M. J. Angew. Chem. Int. Ed. 2002, 41, 3442. (d) Geisbauer,
A.; Mihan, S.; Beck, W. J. Organomet. Chem. 1995, 501, 61. (e) Rabier, A.;
Lugan, N.; Mathieu, R.; Geoffroy, G. L. Organometallics 1994, 13, 4676.
(f) Lattuada, L.; Licandro, E.; Maiorana, S.; Molinari, H.; Papagni, A.
Organometallics 1991, 10, 807.
S
Supporting Information. Experimental details and ana-
b
lytical data for all new compounds, details of the DFT computa-
tional study; and all optimized geometries (pdb files). This
material is available free of charge via the Internet at http://
pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
xpzhang@usf.edu; b.debruin@uva.nl
(14) On the basis of IR and DFT investigations, Co(salen) carbene
complexes were proposed to have radical character: (a) Ikeno, T.;
Iwakura, I.; Yamada, T. J. Am. Chem. Soc. 2002, 124, 15152. (b) Iwakura,
I.; Tanaka, H.; Ikeno, T.; Yamada, T. Chem. Lett. 2004, 33, 140.
(15) Zhang, L.; Chan, K. S. Organometallics 2007, 26, 679.
(16) When preformed 4 (via Scheme 3) was reacted with TEMPO
(10 equiv) in toluene at 80 °C for 3 h, no formation of 7 was observed.
(17) When ethyl styryldiazoacetate (4 equiv) was reacted with
TEMPO (10 equiv) first in toluene for 3 h at 80 °C, followed by addition
of [Co(TPP)], no formation of 7 or 4 was observed, even after further
reaction for an additional 3 h at 80 °C.
’ ACKNOWLEDGMENT
We are grateful for financial support by the American Chemical
Society (44286-AC1; X.P.Z.), the National Science Foundation
(CAREER Award CHE-0711024; X.P.Z.), the University of South
Florida (X.P.Z.), Netherlands Organization for Scientific Re-
searchꢀChemical Sciences (NWO-CW VIDI project 700.55.426;
B.d.B.), the European Research Council (Grant Agreement
202886; B.d.B.), and the University of Amsterdam (B.d.B.).
’ REFERENCES
(1) (a) Carson, C. A.; Kerr, M. A. Chem. Soc. Rev. 2009, 38, 3051. (b)
Reichelt, A.; Martin, S. F. Acc. Chem. Res. 2006, 39, 433. (c) Yu, M.;
Pagenkopf, B. L. Tetrahedron 2005, 61, 321. (d) Reissig, H. U.; Zimmer,
R. Chem. Rev. 2003, 103, 1151.
(2) (a) Pellissier, H. Tetrahedron 2008, 64, 7041. (b) Lebel, H.;
Marcoux, J. F.; Molinaro, C.; Charette, A. B. Chem. Rev. 2003, 103, 977.
(c) Davies, H. M. L.; Antoulinakis, E. G. Org. React. 2001, 57, 1. (d)
Doyle, M. P.; Forbes, D. C. Chem. Rev. 1998, 98, 911.
(3) (a) Hansen, J.; Davies, H. M. L. Coord. Chem. Rev. 2008, 252,
545. (b) Doyle, M. P. J. Org. Chem. 2006, 71, 9253. (c) Doyle, M. P.; Ren,
T. Prog. Inorg. Chem. 2001, 49, 113.
8521
dx.doi.org/10.1021/ja203434c |J. Am. Chem. Soc. 2011, 133, 8518–8521