Science - USA (2022-03-04)

the symmetry of the supramolecular Vernier
complex. The 1:3 ratio of environments also rules
out alternative structures, such as a cyclic tetra-
trefoil shield knot (supplementary materials,
section 12) or link-like assemblies in which the
pdc unit of a strand used to bind the bridging
second metal ion is one of the internal sites.
The open-knot Vernier complexes were also
characterized by circular dichroism (Fig. 3E).
The overall appearance of the CD spectrum of
the open triskelion complex ((R) 8 - L2) 3 • [Yb] 4
(blue) closely resembles that of theL-open
trefoil complex (R) 6 - L3•[Yb] (green), and the
spectra are almost identical when normalized
for the fourfold number of chromophores in
((R) 8 - L2) 3 • [Yb] 4. The small differences between
thetwocanbeattributedtothedifferentchi-
ral expression of the central entangled feature
of ((R) 8 - L2) 3 • [Yb] 4 , which lacks the folded loops
of the outer trefoil tangles ( 48 ).
Multiple tethered pdc groups cannot assem-
ble around a single lanthanide (III) ion with
tethered pdc groups of opposing stereochem-
istry because of steric clashes (supplementary
materials, section 5). Because strand-crossing
stereochemistry is governed by the point chi-
rality in the pdc groups, this offers the pos-
sibility of a new element of selection and
control in assembling Vernier complexes. To
explore this, we prepared a diastereomeric
ligand strand of (R) 8 - L2in which one ter-
minal pdc unit had the (S,S)-stereochemistry,
namely (S) 2 (R) 6 - L2(supplementary materials,
section 4). Treatment of (S) 2 (R) 6 - L2with
Lu(CF 3 SO 3 ) 3 or Yb(CF 3 SO 3 ) 3 in a 3:4 strand:
metal ion ratio at 80°C in MeCN generated
the corresponding Vernier open inverted core
triskelion complexes ((S) 2 (R) 8 - L2) 3 • [Lu] 4 and
((S) 2 (R) 8 - L2) 3 • [Yb] 4 over 7 days (Fig. 3D). Mass
spectrometry confirmed the 3:4 ligand-to-metal
ratio of the resulting supramolecular structures
(fig. S25).
The CD spectrum of ((S) 2 (R) 6 - L2) 3 • [Yb] 4
(Fig. 3E, red) is consistent with the structure
being the inverted core triskelion knot with
three outerLentanglements and a centralD
entanglement. The inverted exciton coupling
of the centralDentanglement compensates
for the intensity of the coupling of one of the
outerLentanglements, leading to an intensity
for the inverted core triskelion of approximate-
ly half that of the all-Lopen triskelion complex
((R) 8 - L2) 3 • [Yb] 4 (Fig. 3E, red).
Closure of ((R) 8 - L2) 3 • [Lu] 4 by RCM, sub-
sequent demetallation by Et 4 NF, and size-
exclusion chromatography gave the metal-free
organic knot (L 3 ,L)- 2 in 6% isolated yield over
three steps (Fig. 4A) as a single species as
determined by DOSY (fig. S70). The broadness
of the^1 H NMR spectrum of (L 3 ,L)- 2 (Fig. 4C)
only reduced noticeably at 348 K (fig. S73),
reflecting the degree and congestion of strand
entanglement. The most abundant ion in the
MALDI-TOF spectrum atm/z7535 corresponds

to the organic knot (fig. S88). The isolated or-

ganic knot could be remetalated by treatment

with an excess of Lu(CF 3 SO 3 ) 3 over 48 hours in

MeCN/CDCl 3 (4:1, v/v) to form (L 3 ,L)- 2 • [Lu] 4

in 92% yield (supplementary materials, sec-

tion 7.3). ESI-MS of the regenerated (L 3 ,L)-

2 • [Lu] 4 shows a pristine spectrum with

molecular ion charges ranging from 12+ to 5+

(Fig. 4B). The corresponding inverted core

triskelion complex ((S) 2 (R) 6 - L2) 3 • [Lu] 4 was

also closed by RCM and demetalated by

Et 4 NF to give (L 3 ,D)- 2 (Fig. 4D). The normal-

ized CD spectrum of closed knot (L 3 ,D)- 2 • [Lu] 4

shows the expected reduction in intensity com-

pared with (L 3 ,L)- 2 • [Lu] 4 as a result of the chi-

rality of the centralD-handed fragment (Fig. 4E

and supplementary materials, section 11).

Although the triskelion knot (L 3 ,L)- 2 and

inverted core triskelion knot (L 3 ,D)- 2 both have

12 crossings, their topologies are distinctly dif-

ferent. Triskelion knot (L 3 ,L)- 2 has 12 alter-

nating crossings, meaning that for each crossing,

the thread must formally traverse from one

face of the loop to the other and back again;

however, (L 3 ,D)- 2 has six alternating and six

nonalternating crossings, the latter crossings

achievable by the strand lying across two ad-

jacent sections of the loop. This greatly affects

the intrinsic tightness of the isomeric knots, as

reflected in their approximate hydrodynamic

radii measured by DOSY ((L 3 ,L)- 2 2.03 nm,

(L 3 ,D)- 2 2.73 nm, CDCl 3 , 298 K; supplemen-

tary materials, section 9.2, and table S3).

TheC 3 -symmetric triskelion conformation of

(L 3 ,L)- 2 • [Lu] 4 was calculated to be ~150 kcal/

mol more stable than other tangled conformers

(supplementary materials, section 14), a sym-

metry consistent with the 1:3 signal intensities

observed for the HDprotons in the^1 H NMR

spectrum of ((R) 8 - L2) 3 • [Yb] 4. The stability was

estimated by condensed-phase hybrid density-

functional theory calculations [PCM-B3LYP-

D3(BJ)/def2-SV(P), with effective core potentials

for Lu, using ORCA 4.2.1] on low-energy struc-

tures identified by an extensive simulated

annealing-based structural search protocol ( 28 )

(supplementary materials, section 14). Other

than being flexible about the chains that link

the trefoil tangles to the central core, the metal-

coordinated triskelion knot adopts a well-defined

2D conformation despite being composed of

a single continuous self-entangled 1D strand.

We note a fundamental similarity to the struc-

ture of knitted materials ( 54 ), which arise from

the systematic self-entanglement of a single 1D

strand in a pattern that results in a 2D layer

that can adopt complex 3D curvature (like a

knitted hat or sweater). This differs from weav-

ing ( 32 – 36 ), which results in the periodic mu-

tual entanglement of multiple 1D strands.

In summary, Vernier template synthesis is

a powerful tool for constructing large molecu-

lar structures with complex systematic net-

works of entanglements. Granny knots (six

alternating crossings) were accessed from a

3:2 Vernier complex, and a triskelion knot with

12 alternating crossings and an inverted-core

triskelion knot with six alternating and six

nonalternating crossings was accessed from

3:4 Vernier complexes. The latter, 378-atom-

long closed-loop knots (~40 nm strand length)

are among the largest discrete macrocycles

synthesized to date, with the added complex-

ity of defined sequences of stereocontrolled

strand entanglements, the precise topology

affecting the structure and properties of the

resulting knots. Both types of triskelion knot

topologies are common motifs in extended

Hiberno-Saxon knotwork and interlace ( 55 ).

The ability to synthesize large, hierarchically

knotted molecular architectures with precise

control of entanglements, crossing stereochem-

istry, and overall symmetry presents opportu-

nities and research directions for topological

molecules and materials.

REFERENCESANDNOTES

1. S. D. P. Fielden, D. A. Leigh, S. L. Woltering,Angew. Chem. Int. Ed.
   56 , 11166–11194 (2017).


2. J. F. Stoddart,Nano Lett. 20 , 5597–5600 (2020).


3. D. Meluzzi, D. E. Smith, G. Arya,Annu. Rev. Biophys. 39 ,
   349 – 366 (2010).


4. M. Zhao, M. T. Woodside,Nat. Chem. Biol. 17 , 975–981 (2021).


5. N. C. H. Lim, S. E. Jackson,J. Phys. Condens. Matter 27 ,
   354101 (2015).


6. J.-F. Aymeet al.,J. Am. Chem. Soc. 137 , 9812–9815 (2015).


7. R. A. Bilbeisiet al.,Chem. Sci. 7 , 2524–2531 (2016).


8. D. P. Augustet al.,J. Am. Chem. Soc. 142 , 18859–18865 (2020).


9. V. Marcoset al.,Science 352 , 1555–1559 (2016).


10. G. Gil-Ramírezet al.,J. Am. Chem. Soc. 138 , 13159–13162 (2016).


11. N. Katsoniset al.,Nat. Chem. 12 , 939–944 (2020).


12. F. Benyettouet al.,Chem. Sci. 10 , 5884–5892 (2019).


13. D. A. Leigh, L. Pirvu, F. Schaufelberger, D. J. Tetlow, L. Zhang,
    Angew. Chem. Int. Ed. 57 , 10484–10488 (2018).


14. C. O. Dietrich-Buchecker, J.-P. Sauvage,Angew. Chem. Int. Ed.
    28 , 189–192 (1989).


15. J.-P. Sauvage,Angew. Chem. Int. Ed. 56 , 11080–11093 (2017).


16. J.-F. Aymeet al.,Nat. Chem. 4 , 15–20 (2011).


17. J. J. Danonet al.,Science 355 , 159–162 (2017).


18. J. P. Carpenteret al.,Chem 7 , 1534–1543 (2021).


19. Y. Inomata, T. Sawada, M. Fujita,Chem 6 , 294–303 (2020).


20. Y. Inomata, T. Sawada, M. Fujita,J. Am. Chem. Soc. 143 ,
    16734 – 16739 (2021).


21. H.-N. Zhang, W.-X. Gao, Y.-J. Lin, G.-X. Jin,J. Am. Chem. Soc.
    141 , 16057–16063 (2019).


22. L.-L. Dang, H.-J. Feng, Y.-J. Lin, G.-X. Jin,J. Am. Chem. Soc. 142 ,
    18946 – 18954 (2020).


23. Z. Cui, Y. Lu, X. Gao, H.-J. Feng, G.-X. Jin,J. Am. Chem. Soc.
    142 , 13667–13671 (2020).


24. D. A. Leighet al.,Nat. Chem. 13 , 117–122 (2021).


25. Y. Segawaet al.,Science 365 , 272–276 (2019).


26. H. Adamset al.,Nature 411 , 763 (2001).


27. P. E. Barranet al.,Angew. Chem. Int. Ed. 50 , 12280–12284 (2011).


28. D. A. Leighet al.,Nature 584 , 562–568 (2020).


29. J. Brüggemannet al.,Angew. Chem. Int. Ed. 46 , 254–259 (2007).


30. N. Ponnuswamy, F. B. L. Cougnon, J. M. Clough, G. D. Pantoş,
    J. K. M. Sanders,Science 338 , 783–785 (2012).


31. F. B. L. Cougnon, K. Caprice, M. Pupier, A. Bauzá, A. Frontera,
    J. Am. Chem. Soc. 140 , 12442–12450 (2018).


32. U. Lewandowskaet al.,Nat. Chem. 9 , 1068–1072 (2017).


33. Z. Wanget al.,Nat. Commun. 8 , 14442 (2017).


34. D. P. Augustet al.,Nature 588 , 429–435 (2020).


35. Y. Liuet al.,Science 351 , 365–369 (2016).


36. Y. Liuet al.,J. Am. Chem. Soc. 140 , 16015–16019 (2018).


37. Q. Wuet al.,Science 358 , 1434–1439 (2017).


38. J. S. Lindsey,New J. Chem. 15 , 153–180 (1991).


39. T. R. Kelly, R. L. Xie, C. Kraebel Weinreb, T. Bregant,
    Tetrahedron Lett. 39 , 3675–3678 (1998).


40. C. A. Hunter, S. Tomas,J. Am. Chem. Soc. 128 , 8975– 8979
    (2006).


41. M. C. O’Sullivanet al.,Nature 469 , 72–75 (2011).

1040 4 MARCH 2022•VOL 375 ISSUE 6584 science.orgSCIENCE

RESEARCH | REPORTS