Magnetic Bistability in Naphtho-1,3,2-dithiazolyl: Solid State

Communication pubs.acs.org/JACS

Cite This: J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Magnetic Bistability in Naphtho-1,3,2-dithiazolyl: Solid State Interconversion of a Thiazyl π‑Radical and Its N−N σ‑Bonded Dimer Demetris Bates,† Craig M. Robertson,*,† Alicea A. Leitch,‡ Paul A. Dube,⊥ and Richard T. Oakley*,‡ †

Department of Chemistry, University of Liverpool, Liverpool L69 7ZD, United Kingdom Department of Chemistry, University of Waterloo, Waterloo, Ontario N2L 3G1, Canada ⊥ Brockhouse Institute for Materials Research, McMaster University, Hamilton, Ontario L8S 4M1, Canada ‡

S Supporting Information *

Chart 1. Heterocyclic 1,3,2-Dithiazolyl (DTA) π-Radicals

ABSTRACT: Crystals of the heterocyclic radical naphtho1,3,2-dithiazolyl NDTA display magnetic bistability with a well-defined hysteretic phase transition at Tc↓ = 128(2) K and Tc↑ = 188(2) K. The magnetic signature arises from a radical/dimer interconversion involving one of the two independent π-radicals in the P1̅ unit cell. Variable temperature X-ray crystallography has established that while all the radicals in HT-NDTA serve as paramagnetic (S = 1/2) centers, half of the radicals in LT-NDTA form closed-shell N−N σ-bonded dimers (S = 0) and half retain their S = 1/2 spin state. The wide window of bistability (60 K) may be attributed to the large structural changes that accompany the phase transition.

which ordered as a bulk antiferromagnet at 11 K.10 However, no low temperature structural information was reported. This historical difficulty in obtaining low temperature structural data also affected the interpretation of early magnetic susceptibility (χ) measurements on the naphtho-variant NDTA,5a which indicated a phase change (on warming) near 190 K. While ambient temperature structural analysis5a revealed a triclinic cell, space group P1̅, with two independent radicals (Z′ = 2),11,12 their structural fate on cooling was not explored. Using synchrotron radiation sources we have now performed low temperature crystallographic measurements on NDTA, and have identified the low temperature (LT) phase to which the high temperature (HT) phase evolves. As expected, radical dimerization is observed on cooling, but in the LT-phase only half of the radicals are spin-paired, forming nominally N−N σbonded dimers (Chart 1), a motif that is, to our knowledge, without precedent in the structural chemistry of thiazyl radicals. We have also re-examined the magnetothermal behavior of NDTA, and find that the structural changes accompanying the HT-to-LT phase transition are mirrored in the magnitude of the magnetic response, which is about one-half of that expected for complete radical dimerization. The response is, moreover, hysteretic, giving rise to a “window” of bistability with a width near 60 K. Variable temperature magnetic measurements on NDTA were performed over the range T = 2−300 K in both cooling and heating modes; Figure 1 shows data collected with a static

M

olecular materials whose physical properties can be altered by external stimuli hold potential in the development of electronic switching and memory storage devices.1 While transition-metal-based spin-crossover compounds have played a pivotal role in the design of such materials,2 molecular radicals also hold appeal, as their crystal structures are subject to the influence of heat, light and/or pressure.3 Heterocyclic 1,3,2-dithiazolyl (DTA) π-radicals (Chart 1) have been widely studied,4 as their dimers are disposed to thermally driven dimer (S = 0) to radical (S = 1/2) interconversions. The magnetic response accompanying dimer opening and radical closing usually occurs at the same temperature,5 but in some cases, typified by TDTA, PDTA and TPDTA (Chart 1), the response is hysteretic,6 that is, there is a difference in temperature required for the dimer-to-radical (Tc↑) and radical-to-dimer (Tc↓) processes, yielding a region of magnetic bistability wherein the thermodynamically stable and metastable forms can coexist. The three radicals noted above adopt high temperature crystal structures which are based on π-stacked arrays of radicals (S = 1/2); upon cooling these form cofacial or pancake7 π-dimers (S = 0). The mechanism of these transformations and the origin of the associated magnetic hysteresis has been the subject of much discussion.8 The search for related materials with similar behavior has also been actively pursued. In this regard the benzo-derivative BDTA is notable. Early crystallographic work revealed a trans-antarafacial π-dimer which dissociated near ∼90 °C9 and later magnetic measurements suggested a low temperature metastable radical phase © XXXX American Chemical Society

Received: December 27, 2017

A

DOI: 10.1021/jacs.7b13699 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Communication

Journal of the American Chemical Society field H = 0.1 T. On cooling, the initial linear response in 1/χ versus T (Figure 1a) is well described in terms of Curie−Weiss behavior; a fit over the range T = 150−300 K data yields C = 0.373 emu K mol−1 for HT-NDTA, a value very close to that predicted (0.375 emu K mol−1) for a pure S = 1/2 paramagnet with a nominal g = 2. The corresponding θ-value of −23 K indicates antiferromagnetic (AFM) exchange between the radical spins. The phase change from HT-NDTA to LTNDTA is apparent near 125 K, after which the 1/χ versus T plot again reflects Curie−Weiss behavior down to near 25 K. Upon warming, the 1/χ versus T plot follows the same track up to near 180 K. A Curie−Weiss fit to the T = 50−150 K data for LT-NDTA yields C = 0.235 emu K mol−1, a slightly larger value than expected (0.19 emu K mol−1) for association of one-half of the radicals into diamagnetic dimers. The discrepancy may arise from structural defects generated during the phase transition, which is described below. The smaller but still AFM θ-value of −17 K suggests the active spin sites in LT-NDTA are more isolated.

often violently,6a,e,13 on cooling to between 120 and 90 K (Figure S1), but eventually we were able to characterize the structure of the low temperature phase. Like its high temperature counterpart, LT-NDTA belongs to the centric space group P1̅. Figure 2 provides details of the crystal structures of both phases at the same temperature of 160 K, that is, near the middle of the bistable region (Figure 1). Data collected at other temperatures are available in the SI.

Figure 2. Unit cell views and parameters for (a) HT- and (b) LTNDTA, both space group P1,̅ at 160 K. Head-to-head approaches of DTA rings are highlighted in red.

In the HT-phase, the two independent NDTA molecules (Z′ = 2) each serve as a building block for head-to-tail ribbons that run parallel to the c-axis. Antiparallel ribbon-like networks are also present in the LT-phase, but the c-axis is almost doubled, and there are now four independent molecules per cell (Z′ = 4). As illustrated in Figure 3, which shows the asymmetric unit of the LT-phase, two of these four molecules remain as openshell radicals, but slide together across inversion centers to form pairs linked by long 4-center (S···N′)2 contacts, a commonly observed supramolecular motif.14 The other two associate into closed-shell dimers nominally linked by N−N σ-bonds. This separation of the molecular building blocks into nominally diamagnetic (S = 0) and paramagnetic (S = 1/2) groups holds the key to the magnetic response described above.

Figure 1. (a) Plot of 1/χ versus T for NDTA at H = 0.1 T, with Curie−Weiss fits from 300 to 150 K (cooling) and from 50 to 150 K (warming). (b) Expanded plot of χT versus T cycled over the range 50−250 K, highlighting the bistable region of ∼60 K.

The hysteretic nature of the phase transition is more clearly visualized in Figure 1b, which provides a χT versus T plot cycled over the range 50−250 K. This allows an estimation of the temperatures for the associative (Tc↓ = 128(2) K) and dissociative (Tc↑ = 188(2) K) processes, and the width (near 60 K) of the resulting window of bistability, which is intermediate between those found in PDTA (46 K) and TDTA (92 K). The use of smaller and larger fields (H = 0.01 and 1 T) has little effect on the response profile, and while repeated cycling led to a small reduction in the limiting value of χT at 300 K, as well as small shifts in Tc↓ and Tc↑, the window of bistability changed by less than 5%. To establish the structural changes associated with the phase transition, we carried out a series of single crystal X-ray diffraction experiments at temperatures below (85 K) and within (160 K) the hysteretic region. Most crystals fractured,

Figure 3. Asymmetric unit of LT-NDTA at 160 K, with two subunits based on N−N σ-bonded dimers and 4-center (S···N′)2 radical pairs. Intermolecular distances, dihedral angles τ within DTA rings and sums of angles Σ∠N at N2a and N2b are listed. B

DOI: 10.1021/jacs.7b13699 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Communication

Journal of the American Chemical Society

dimerization modes, we performed a series of calculations on model prototypal DTA dimers, tracking their relative total electronic energy Etot as a function of lateral slippage (λ) of the two rings (Figure 5). When ring planarity is enforced (τ = 0), the trans-antarafacial structure shows a shallow minimum near λ = −0.25 Å, but when the planarity restriction is released (τ optimized), the nitrogen atoms rotate inward with increasing slippage to form a more stable N−N σ-dimer near λ = 1.75 Å.

The most striking feature of the dimers is the presence of an N−N σ-bond. The loss in planarity of the DTA rings, measured in terms of the dihedral angles τ, combined with a small value for the sum of the bond angles at nitrogen Σ(∠N), are consistent with near sp3 hybridization, although the two N−N distances are longer than in hydrazine (1.46 Å) and its tetraphenyl derivative (1.406 Å).15 Like the HT phase,5a the LT structure is characterized by numerous structure-making “tiltedT” CH···C′(arene) interactions16 (Figure S2). To aid understanding of the structural changes that occur during the phase transition, Figure 4 shows the head-to-head packing of centrosymmetric pairs of N1- and N2-based DTA rings in the two phases. In HT-NDTA, these radical pairs straddle inversion centers located in the (001) plane. In LTNDTA, there is a doubling of the unit cell along the c-axis as different radical pairs combine, with preservation of inversion symmetry, into 4-center (S···N′)2 radical pairs and N−N σbonded dimers centered on either the (001) or (002) planes of the LT-phase. There are several pathways by which such a transformation might occur, depending on the choice of radicals and inversion centers. The mechanism illustrated here, which leads to the atom numbering shown in Figure 3, represents just one possibility. Within this scheme, d1-linked (S1···N1′)2 pairs and d2-linked N2···N2′ pairs in the HT-phase (Figure 4a) couple, respectively, into 4-center (S1a···N1a′) pairs and (N2a)2 dimers located in the (002) plane of the LTphase (Figure 4b). The (001) plane of the LT-phase is generated in a similar fashion, as d3-linked (S2···N1′)2 pairs and d4-linked N2···N2′ pairs of the HT-phase contract into 4-center (S2a···N1a′) pairs and (N2b)2 dimers.

Figure 5. B3LYP-D3/6-31G(d,p) relative total energy Etot of DTA dimers in C2h symmetry as a function of lateral slippage λ, with dihedral angle τ fixed at 0° (blue circles) and optimized (red squares), relative to Etot at τ = 0° and λ = 0 Å. Metrics for optimized models at λ = −0.25 and +1.75 Å are illustrated.

The σ-bond parlance used here should be interpreted with caution. Unlike the localized P−P σ-bond in 1,3,2-dithiaphospholyl dimers,20 the net bonding orbital in the N−N σ-bonded DTA dimer (Figure S4), derived from the pairing of two radical SOMOs, remains delocalized with significant S−S π-overlap, in keeping with the short interannular S−S distance (3.23 Å). While more detailed calculations are required to understand fully the energetic and structural changes associated with the phase transition in NDTA, the present results open a new horizon in the structural chemistry of thiazyl radicals. While radical dimerization in these systems has been extensively explored, spin pairing occurs almost invariably by overlap of their π-systems. There are a few examples of σ-dimers, and to date these have involved the use of sulfur or carbon.3h,21 The involvement of N−N σ-interactions in the dimerization of NDTA and the hysteretic nature of the phase transition augur well for the development of other 1,3,2-dithiazolyl radicals with magnetothermal switching properties. Based on the lessons learned here, the pursuit of DTA radicals appended to other oligoacene frameworks seems particularly promising.

Figure 4. Centrosymmetric pairs of DTA rings in the (a) HT- and (b) LT-phases of NDTA, viewed parallel to the c-axis. All pairs in (a) are centered in the (001) plane. In panel (b), pairs on the left and right are centered in the (001) plane, while those in the center are located in the (002) plane. Intermolecular (S···N′)2 and N···N′ separations during the phase change are listed below.

■

ASSOCIATED CONTENT

S Supporting Information *

17

Dispersion corrected DFT calculations at the (U)B3LYPD3/6-31G(d,p) level on the putative trans-antarafacial π-dimer of NDTA (Table S2) afford a slightly exothermic disproportionation enthalpy ΔHdis = −1.9 kcal mol−1, while the N−N σdimer is bound, with ΔHdis = 7.1 kcal mol−1.18,19 The salient metrics (Figure S3) of the π-dimer are reminiscent of BDTA,9 while those of the σ-dimer provide a close match with LTNDTA. To explore the relative energetics of these two

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/jacs.7b13699. Experimental and computational details (PDF) Crystal data at 160 K for HT-NDTA (CIF) Crystal data at 85 K for LT-NDTA (CIF) Crystal data at 160 K for LT-NDTA (CIF) Crystal data at 225 K for recovered HT-NDTA (CIF) C

DOI: 10.1021/jacs.7b13699 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Communication

Journal of the American Chemical Society

■

Rawson, J. M.; Feeder, N.; van Duijn, J.; McInnes, E. J. L.; Novoa, J. J.; Burriel, R.; Palacio, F.; Oliete, P. J. Mater. Chem. 2001, 11, 1992. (d) Brusso, J. L.; Clements, O. P.; Haddon, R. C.; Itkis, M. E.; Leitch, A. A.; Oakley, R. T.; Reed, R. W.; Richardson, J. F. J. Am. Chem. Soc. 2004, 126, 8256. (e) Brusso, J. L.; Clements, O. P.; Haddon, R. C.; Itkis, M. E.; Leitch, A. A.; Oakley, R. T.; Reed, R. W.; Richardson, J. F. J. Am. Chem. Soc. 2004, 126, 14692. (7) (a) Preuss, K. E. Polyhedron 2014, 79, 1. (b) Cui, Z.; Lischka, H.; Beneberu, H. Z.; Kertesz, M. J. Am. Chem. Soc. 2014, 136, 12958. (c) Beneberu, H. Z.; Tian, Y. H.; Kertesz, M. Phys. Chem. Chem. Phys. 2012, 14, 10713. (8) (a) Clarke, C. S.; Jornet-Somoza, J.; Mota, F.; Novoa, J. J.; Deumal, M. J. Am. Chem. Soc. 2010, 132, 17817. (b) Vela, S.; Mota, F.; Deumal, M.; Suizu, R.; Shuku, Y.; Mizuno, A.; Awaga, K.; Shiga, M.; Novoa, J. J.; Ribas-Arino, J. Nat. Commun. 2014, 5, 4411. (c) Vela, S.; Deumal, M.; Shiga, M.; Novoa, J. J.; Ribas-Ariño, J. Chem. Sci. 2015, 6, 2371. (9) Awere, E. G.; Burford, N.; Haddon, R. C.; Parsons, S.; Passmore, J.; Waszczak, J. V.; White, P. S. Inorg. Chem. 1990, 29, 4821. (10) Fujita, W.; Awaga, K.; Nakazawa, Y.; Saito, K.; Sorai, M. Chem. Phys. Lett. 2002, 352, 348. (11) (a) Brock, C. P. Acta Crystallogr., Sect. B: Struct. Sci., Cryst. Eng. Mater. 2016, 72, 807. (b) Steed, K. M.; Steed, J. W. Chem. Rev. 2015, 115, 2895. (12) Boeré, R. T.; Hill, N. D. D. CrystEngComm 2017, 19, 3698. (13) This is a classic example of the thermosalient or “jumping crystal” phenomenon. For details, see: Skoko, Ž .; Zamir, S.; Naumov, P.; Bernstein, J. J. Am. Chem. Soc. 2010, 132, 14191 and references cited therein.. (14) (a) Cozzolino, A. F.; Vargas-Baca, I.; Mansour, S.; Mahmoudkhani, A. H. J. Am. Chem. Soc. 2005, 127, 3184. (b) Tsuzuki, S.; Sato, N. J. Phys. Chem. B 2013, 117, 6849. (15) (a) Collin, R. L.; Lipscomb, W. N. Acta Crystallogr. 1951, 4, 10. (b) Hoekstra, A.; Vos, A.; Braun, P. B.; Hornstra, J. Acta Crystallogr., Sect. B: Struct. Crystallogr. Cryst. Chem. 1975, 31, 1708. (16) (a) Hobza, P.; Selzle, H. L.; Schlag, E. W. J. Am. Chem. Soc. 1994, 116, 3500. (b) Sinnokrot, M. O.; Sherrill, C. D. J. Phys. Chem. A 2004, 108, 10200. (17) (a) Grimme, S. Chem. - Eur. J. 2012, 18, 9955. (b) Grimme, S.; Antony, J.; Ehrlich, S.; Krieg, H. J. Chem. Phys. 2010, 132, 154104. (18) For early theoretical work on N−N bonded DTA dimers, see: Brownridge, S.; Du, H.; Fairhurst, S. A.; Haddon, R. C.; Oberhammer, H.; Parsons, S.; Passmore, J.; Schriver, M. J.; Sutcliffe, L. H.; Westwood, N. P. C. J. Chem. Soc., Dalton Trans. 2000, 3365. (19) The dissociation of simple hydrazines into aminyl radicals is highly endothermic, as the resulting spin is not free to delocalize. See: Bohna, M. A.; Klapötke, T. M. Z. Naturforsch. B 2004, 59, 148. (20) Kosnik, S. C.; Nascimento, M. C.; Rawson, J. M.; Macdonald, C. L. B. Dalton Trans. 2017, 46, 9769. (21) Leitch, A. A.; McKenzie, C. E.; Oakley, R. T.; Reed, R. W.; Richardson, J. F.; Sawyer, L. D. Chem. Commun. 2006, 1088.

AUTHOR INFORMATION

Corresponding Authors

*[email protected] *[email protected] ORCID

Richard T. Oakley: 0000-0002-7185-2580 Notes

The authors declare no competing financial interest.

■ ■

ACKNOWLEDGMENTS We thank the NSERC of Canada for financial support and the Diamond Light Source for access to beamline I19. REFERENCES

(1) (a) Sato, O. Nat. Chem. 2016, 8, 644. (b) Raman, K. V.; Kamerbeek, A. M.; Mukherjee, A.; Atodiresei, N.; Sen, T. K.; Lazic, P.; Caciuc, V.; Michel, R.; Stalke, D.; Mandal, S. K.; Blügel, S.; Mü n zenberg, M.; Moodera, J. S. Nature 2013, 493, 509. (c) Bousseksou, A.; Molnár, G.; Salmon, L.; Nicolazzi, W. Chem. Soc. Rev. 2011, 40, 3313. (d) Ratera, I.; Veciana, J. Chem. Soc. Rev. 2012, 41, 303. (2) (a) Decurtins, S.; Gütlich, P.; Hasselbach, K. M.; Hauser, A.; Spiering, H. Inorg. Chem. 1985, 24, 2174. (b) Sato, O.; Iyoda, T.; Fujishima, A.; Hashimoto, K. Science 1996, 272, 704. (c) Hauser, A. Top. Curr. Chem. 2004, 234, 155. (d) Sato, O.; Tao, J.; Zhang, Y.-Z. Angew. Chem., Int. Ed. 2007, 46, 2152. (e) Spin Crossover Materials: Properties and Applications; Halcrow, M. A., Ed.; J. Wiley & Sons: Chichester, U.K., 2013. (f) Guionneau, P. Dalton Trans. 2014, 43, 382. (f1) Hayami, S.; Holmes, S. M.; Halcrow, M. A. J. Mater. Chem. C 2015, 3, 7775. (g) Brooker, S. Chem. Soc. Rev. 2015, 44, 2880. (h) Rosario-Amorin, D.; Dechambenoit, P.; Bentaleb, A.; Rouzières, M.; Mathonière, C.; Clérac, R. J. Am. Chem. Soc. 2018, 140, 98. (3) (a) Hicks, R. G. Nat. Chem. 2011, 3, 189. (b) Itkis, M. E.; Chi, X.; Cordes, A. W.; Haddon, R. C. Science 2002, 296, 1443. (c) Pal, S. K.; Bag, P.; Sarkar, A.; Chi, X.; Itkis, M. E.; Tham, F. S.; Donnadieu, B.; Haddon, R. C. J. Am. Chem. Soc. 2010, 132, 17258. (d) Fumanal, M.; Novoa, J. J.; Ribas-Arino, J. Chem. - Eur. J. 2017, 23, 7772. (e) Kumar, S.; Kumar, Y.; Keshri, S. K.; Mukhopadhyay, P. Magnetochemistry 2016, 2, 42. (f) Rawson, J. M.; Hayward, J. J. In Spin-Crossover Materials: Properties and Applications; Halcrow, M. A., Ed.; J. Wiley & Sons: Chichester, U.K., 2013; p 235. (g) Phan, H.; Lekin, K.; Winter, S. M.; Oakley, R. T.; Shatruk, M. J. Am. Chem. Soc. 2013, 135, 15674. (h) Lekin, K.; Phan, H.; Winter, S. M.; Wong, J. W. L.; Leitch, A. A.; Laniel, D.; Yong, W.; Secco, R. A.; Tse, J. S.; Desgreniers, S.; Dube, P. A.; Shatruk, M.; Oakley, R. T. J. Am. Chem. Soc. 2014, 136, 8050. (i) Li, T.; Tan, G.; Shao, D.; Li, J.; Zhang, Z.; Song, Y.; Sui, Y.; Chen, S.; Fang, Y.; Wang, X. J. Am. Chem. Soc. 2016, 138, 10092. (j) Fatila, E. M.; Mayo, R. A.; Rouzières, M.; Jennings, M. C.; Dechambenoit, P.; Soldatov, D. V.; Mathonière, C.; Clérac, R.; Coulon, C.; Preuss, K. E. Chem. Mater. 2015, 27, 4023. (4) Rawson, J. M.; McManus, J. D. Coord. Chem. Rev. 1999, 189, 135. (5) (a) Barclay, T. M.; Cordes, A. W.; George, N. A.; Haddon, R. C.; Oakley, R. T.; Patenaude, G. W.; Reed, R. W.; Zhang, H.; Palstra, T. T. M.; Richardson, J. F. Chem. Commun. 1997, 873. (b) Alberola, A.; Eisler, D.; Less, R. J.; Navarro-Moratalla, E.; Rawson, J. M. Chem. Commun. 2010, 46, 6114. (c) Alberola, A.; Eisler, D. J.; Harvey, L.; Rawson, J. M. CrystEngComm 2011, 13, 1794. (d) Alberola, A.; Clements, O. P.; Collis, R. J.; Cubbitt, L.; Grant, C. M.; Less, R. J.; Oakley, R. T.; Rawson, J. M.; Reed, R. W.; Robertson, C. M. Cryst. Growth Des. 2008, 8, 155. (e) Vela, S.; Reardon, M. B.; Jakobsche, C. E.; Turnbull, M. M.; Ribas-Ariño, J.; Novoa, J. J. Chem. - Eur. J. 2017, 23, 3479. (f) Alberola, A.; Collis, R. J.; Humphrey, S. M.; Less, R. J.; Rawson, J. M. Inorg. Chem. 2006, 45, 1903. (6) (a) Barclay, T. M.; Cordes, A. W.; George, N. A.; Haddon, R. C.; Itkis, M. E.; Mashuta, M. S.; Oakley, R. T.; Patenaude, G. W.; Reed, R. W.; Richardson, J. F.; Zhang, H. J. Am. Chem. Soc. 1998, 120, 352. (b) Fujita, W.; Awaga, K. Science 1999, 286, 261. (c) McManus, G. D.; D

DOI: 10.1021/jacs.7b13699 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX