Evaluation of electronics, electrostatics and hydrogen bond cooperative in the binding of cyanide and fluoride by lewis acidic ferrocenylboranes

Article abstract of DOI:10.1021/ic901673u

Synthetic approaches based on the direct borylation of ferrocene by BBr₃, followed by boryl substituent modification, or on the lithiation of ferrocene derivatives and subsequent quenching with the electrophile FBMeS₂, have given access to a range of ferrocene derivatized Lewis acids with which to conduct a systematic study of fluoride and cyanide binding. In particular, the effects of borane electrophilicity, net charge, and ancillary ligand electronics/cooperativity on the binding affinities for these anions have been probed by a combination of NMR, IR, mass spectrometric, electrochemical, crystallographic, and UV-vis titration measurements. In this respect, modifications made at the para position of the boron-bound aromatic substituents exert a relatively minor influence on the binding constants for both fluoride and cyanide, as do the electronic properties of peripheral substituents at the 1 ′- position (even for cationic groups). By contrast, the influence of a CH₂NMe₃ ⁺ substituent in the 2- position is found to be much more pronounced (by >3 orders of magnitude), reflecting, at least in part, the possibility in solution for an additional binding component utilizing the hydrogen bond donor capabilities of the methylene CH₂ group. While none of the systems examined in the current study display any great differentiation between the binding of F^- and CN^- (and indeed some, such as FcBMeS₂, bind both anions with equal affinity within experimental error), much weaker boronic ester Lewis acids will bind fluoride (but give a negative response for cyanide). Thus, by the incorporation of an irreversible redox-matched organic dye, a two-component [BMes₂/B(OR)₂] dosimeter system can be developed capable of colorimetrically signaling the presence of fluoride and cyanide in organic solution by Boolean AND/NOT logic.

Full text of DOI:10.1021/ic901673u

Inorg. Chem. 2010, 49, 157–173 157
DOI: 10.1021/ic901673u
Evaluation of Electronics, Electrostatics and Hydrogen Bond Cooperativity in the
Binding of Cyanide and Fluoride by Lewis Acidic Ferrocenylboranes
Alexander E. J. Broomsgrove,† David A. Addy,† Angela Di Paolo,† Ian R. Morgan,†,‡ Christopher Bresner,†,‡
Victoria Chislett,† Ian A. Fallis,*,‡ Amber L. Thompson,† Dragoslav Vidovic,† and Simon Aldridge*,†
†Inorganic Chemistry, University of Oxford South Parks Road, Oxford, United Kingdom, OX1 3QR and
‡School of Chemistry, Cardiff University, Main Building, Park Place, Cardiff, United Kingdom, CF10 3AT
Received August 21, 2009
Synthetic approaches based on the direct borylation of ferrocene by BBr3, followed by boryl substituent modification, or
on the lithiation of ferrocene derivatives and subsequent quenching with the electrophile FBMes2, have given access to
a range of ferrocene derivatized Lewis acids with which to conduct a systematic study of fluoride and cyanide binding.
In particular, the effects of borane electrophilicity, net charge, and ancillary ligand electronics/cooperativity on the
binding affinities for these anions have been probed by a combination of NMR, IR, mass spectrometric,
electrochemical, crystallographic, and UV-vis titration measurements. In this respect, modifications made at the
para position of the boron-bound aromatic substituents exert a relatively minor influence on the binding constants for
both fluoride and cyanide, as do the electronic properties of peripheral substituents at the 10- position (even for cationic
groups). By contrast, the influence of a CH2NMe3þ substituent in the 2- position is found to be much more pronounced
(by >3 orders of magnitude), reflecting, at least in part, the possibility in solution for an additional binding component
utilizing the hydrogen bond donor capabilities of the methylene CH2 group. While none of the systems examined in the
current study display any great differentiation between the binding of F- and CN- (and indeed some, such as
FcBMes2, bind both anions with equal affinity within experimental error), much weaker boronic ester Lewis acids will
bind fluoride (but give a negative response for cyanide). Thus, by the incorporation of an irreversible redox-matched
organic dye, a two-component [BMes2/B(OR)2] dosimeter system can be developed capable of colorimetrically
signaling the presence of fluoride and cyanide in organic solution by Boolean AND/NOT logic.
Introduction
array of Lewis acidic centers,10a,b and the affinity of cyanide
for three-coordinate boranes (even in the presence of water)
has been known for more than 45 years.12 In a similar vein, a
number of the studies have recently demonstrated the use of
Lewis acid receptors containing the -BMes2 (Mes = 2,4,6-
Me3C6H2) function to detect cyanide,7 in one case offering
remarkable, selective binding in aqueous solution.7c
The selective detection of CN- and F- (and of their
conjugate acids HCN and HF) constitute significant chemi-
cal challenges both from a fundamental supramolecular
perspective and from a more applied viewpoint, for example,
in environmental and medical monitoring.1,2 While consider-
able research effort has been expended on the development of
sensors for fluoride-containing species, encompassing a
range of host/guest strategies to bind the target analyte,3-6
cyanide detection has received less attention.7-11 Neverthe-
less, a number of cyanide receptors have been reported in the
recent literature incorporating an appropriately positioned
(3) For recent reviews of anion binding, see, for example: (a) de Silva, A.
P.; Gunaratne, H. Q. N.; Gunnlaugsson, T.; Huxley, A. J. M.; McCoy, C. P.;
Rademacher, J. T.; Rice, T. E. Chem. Rev. 1997, 97, 1515. (b) Schmidtchen, F.
P.; Berger, M. Chem. Rev. 1997, 97, 1609. (c) Gale, P. A.; Sessler, J. A.; Kral, V.
Chem. Commun. 1998, 1. (d) Beer, P. D.; Gale, P. A. Angew. Chem., Int. Ed.
2001, 40, 486. (e) Lavigne, J. J.; Anslyn, E. V. Angew. Chem., Int. Ed. 2001, 40,
ꢀ
ꢀ
3118. (f) Martínez-Manez, R.; Sancernon, F. Chem. Rev. 2003, 103, 4419.
(g) Best, M. D.; Tobey, S. L.; Anslyn, E. V. Coord. Chem. Rev. 2003, 240, 3.
(h) Sessler, J. A.; Camiolo, S.; Gale, P. A. Coord. Chem. Rev. 2003, 240, 17.
(i) Llinares, J. M.; Powell, D.; Bowman-James, K. Coord. Chem. Rev. 2003, 240,
57. (j) Loeb, S.; Bondy, C. R. Coord. Chem. Rev. 2003, 240, 77. (k) Choi, K.;
Hamilton, A. D. Coord. Chem. Rev. 2003, 240, 101. (l) Wedge, T. J.; Hawthorne,
M. F. Coord. Chem. Rev. 2003, 240, 111. (m) Beer, P. D.; Hayes, E. J. Coord.
Chem. Rev. 2003, 240, 167. (n) O'Neill, E. J.; Smith, B. D. Coord. Chem. Rev.
2006, 250, 3068. (o) Filby, M. H.; Steed, J. W. Coord. Chem. Rev. 2006, 250,
3200. (p) Gale, P. A.; Quesada, R. Coord. Chem. Rev. 2006, 250, 3219.
*To whom correspondence should be addressed. E-mail: Simon.
Aldridge@chem.ox.ac.uk.
(1) (a) Yang, Y.-C.; Baker, J. A.; Ward, J. R. Chem. Rev. 1992, 92, 1729.
(b) Baskin, S. I.; Brewer, T. G. Medical Aspects of Chemical and Biological
Warfare; Sidell, F. R., Takafuji, E. T., Franz, D. R., Eds.; TMM Publications: New
ꢀ
York, 1997; pp 271-286. (c) Sohn, H.; Letant, S.; Sailor, M. J.; Trogler, W. C. J.
Am. Chem. Soc. 2000, 122, 5399. (d) Bresner, C.; Aldridge, S.; Fallis, I. A.;
Jones, C.; Ooi, L.-L. Angew. Chem., Int. Ed. 2005, 44, 3606. (e) Colquhoun, J.
Perspect. Biol. Med. 1997, 41, 29. (f) Carton, R. J. Fluoride 2006, 39, 163.
(2) (a) Holland, M. A.; Kozlowski, L. M. Clin. Pharm. 1986, 5, 737. (b)
Baud, F. Hum. Exp. Toxicol. 2007, 26, 191.
€
(q) Hudnall, T. W.; Chiu, C. W.; Gabbaı, F. P. Acc. Chem. Res. 2009, 42, 388.
(r) Hudson, Z. M.; Wang, S. Acc. Chem. Res. 2009 [Online].
r
2009 American Chemical Society
Published on Web 12/03/2009
pubs.acs.org/IC

158 Inorganic Chemistry, Vol. 49, No. 1, 2010
Broomsgrove et al.
An alternative approach utilizing aryl boronic acids and the
ready cleavage of the B-OH function in the presence of
aqueous HCN has also been employed for cyanide sensing.9
Given the facts that (i) systems of the type ArBMes2 are
known to be air- and moisture-stable, (ii) under appropriate
conditions such compounds are known to bind cyanide,7 and
(iii) ferrocene functionalized boranes are known to undergo
large electrochemical shifts on anion binding,4a,p we have
recently begun a program to investigate the electrochemical
detection of cyanide using ferrocenyl boranes. Moreover,
given that an electrochemical response, coupled with an
(6) For bidentate boron-centered Lewis acids, see, for example: (a) Katz,
H. E. J. Am. Chem. Soc. 1985, 107, 1420. (b) Katz, H. E. J. Org. Chem. 1985,
50, 5027. (c) Jia, L.; Yang, X.; Stern, C.; Marks, T. J. Organometallics 1994, 13,
(4) For previous examples of fluoride binding by boronic acid/esters, see,
for example: (a) Dusemund, C.; Sandanayake, K. R. A. S.; Shinkai, S. J.
Chem. Soc., Chem. Commun. 1995, 333. (b) Yamamoto, H.; Ori, A.; Ueda, K.;
Dusemund, C.; Shinkai, S. Chem. Commun. 1996, 407. (c) Cooper, C. R.;
Spencer, N.; James, T. D. Chem. Commun. 1998, 1365. (d) Shiratori, H.; Ohno,
T.; Nozaki, K.; Osuka, A. Chem. Commun. 1999, 2181. (e) Yuchi, A.; Tatebe, A.;
Kani, S.; James, T. D. Bull. Chem. Soc. Jpn. 2001, 74, 509. (f) Ward, C. J.; Patel,
P.; James, T. D. Chem. Lett. 2001, 406. (g) Nicolas, M.; Fabre, B.; Simonet, J.
Electrochim. Acta 2001, 46, 1179. (h) Fabre, B.; Lehmann, U.; Schl€uter, A. D.
Electrochim. Acta 2001, 46, 2855. (i) DiCesare, N.; Lakowicz, J. R. Anal.
Biochem. 2002, 301, 111. (j) Arimori, S.; Davidson, M. G.; Fyles, T. M.; Hibbert,
T. G.; James, T. D.; Kociok-Kohn, G. I. Chem. Commun. 2004, 1640. (k) Badugu,
R.; Lakowicz, J. R.; Geddes, C. D. J. Fluoresc. 2004, 14, 693. (l) Kubo, Y.;
Kobayashi, A.; Ishida, T.; Misawa, Y.; James, T. D. Chem. Commun. 2005, 2846.
(m) Kubo, Y.; Ishida, T.; Kobayashi, A.; James, T. D. J. Mater. Chem. 2005, 15,
2889. (n) Badugu, R.; Lakowicz, J.; Geddes, C. D. Curr. Anal. Chem. 2005, 1,
157. (o) Badugu, R.; Lakowicz, J. R.; Geddes, C. D. Sens. Actuators, B 2005,
104, 103. (p) Swamy, K. M. K.; Lee, Y. J.; Lee, H. N.; Chun, J.; Kim, Y.; Kim,
S.-J.; Yoon, J. J. Org. Chem. 2006, 71, 8626. (q) Kubo, Y.; Ishida, T.; Minami, T.;
James, T. D. Chem. Lett. 2006, 35, 996. (r) Neumann, T.; Dienes, Y.;
Baumgartner, T. Org. Lett. 2006, 8, 495. (s) Tan, W.; Zhang, D.; Wang, Z.;
Liu, C.; Zhu, D. J. Mater. Chem. 2007, 17, 1964. (t) Oehlke, A.; Auer, A. A.;
Jahre, I.; Walfort, B.; Rueffer, T.; Zoufala, P.; Lang, H.; Spange, S. J. Org. Chem.
2007, 72, 4328. (u) Kataoka, K.; James, T. D.; Kubo, Y. J. Am. Chem. Soc. 2007,
€
3755. (d) Kohler, K.; Piers, W. E.; Sin, X.; Feng, Y.; Bravakis, A. M.; Jarvis, A. P.;
Collins, S.; Clegg, W.; Yap, G. P. A.; Marder, T. B. Organometallics 1998, 17,
3557. (e) Williams, V. C.; Piers, W. E.; Clegg, W.; Elsegood, M. R. J.; Collins, S.;
Marder, T. B. J. Am. Chem. Soc. 1999, 121, 3244. (f) Williams, V. C.; Dai, C.; Li,
Z.; Collins, S.; Piers, W. E.; Clegg, W.; Elsegood, M. R. J.; Marder, T. B. Angew.
Chem., Int. Ed. 1999, 38, 3695. (g) Metz, M. V.; Schwartz, D. J.; Stern, C. L.;
Nickias, P. N.; Marks, T. J. Angew. Chem., Int. Ed. 2000, 39, 1312. (h) Williams,
V. C.; Irvine, G. J.; Piers, W. E.; Li, Z. M.; Collins, S.; Clegg, W.; Elsegood, M. R.
J.; Marder, T. B. Organometallics 2000, 19, 1619. (i) Piers, W. E.; Irvine, G.;
Williams, V. C. Eur. J. Inorg. Chem. 2000, 2131. (j) Hoefelmeyer, J. D.; Schulte,
€
M.; Tschinkl, M.; Gabbaı, F. P. Coord. Chem. Rev. 2002, 235, 93. (k) Henderson,
L. D.; Piers, W. E.; Irvine, G. J.; McDonald, R. Organometallics 2002, 21, 340. (l)
Metz, M. V.; Schwartz, D. J.; Stern, C. L.; Marks, T. J.; Nickias, P. N.
Organometallics 2002, 21, 4159. (m) Emslie, D. J. M.; Piers, W. E.; Parvez,
€
M. Angew. Chem., Int. Ed. 2003, 42, 1252. (n) Gabbaı, F. P. Angew. Chem., Int
Ed. 2003, 42, 2218. (o) Lewis, S. P.; Taylor, N. J.; Piers, W. E.; Collins, S. J. Am.
Chem. Soc. 2003, 125, 14686. (p) Ghesner, I.; Piers, W. E.; Parvez, M.;
McDonald, R. Organometallics 2004, 23, 3085. (q) Morrison, D. J.; Piers, W.
ꢀ
€
E.; Parvez, M. Synlett 2004, 13, 2429. (r) Sole, S.; Gabbaı, F. P. Chem. Commun.
2004, 1284. (s) Piers, W. E. Adv. Organomet. Chem. 2005, 52, 1. (t) Melaimi, M.;
€
Gabbaı, F. P. Adv. Organomet. Chem. 2005, 52, 61. (u) Scheibitz, M.; Heilmann,
J. B.; Winter, R. F.; Bolte, M.; Bats, J. W.; Wagner, M. Dalton Trans. 2005, 159.
(v) Scheibitz, M.; Bolte, M.; Bats, J. W.; Lerner, H.-W.; Nowik, I.; Herber, R. H.;
Krapp, A.; Lein, M.; Holthausen, M.; Wagner, M. Chem.;Eur. J. 2005, 11, 584.
€
129, 15126. (v) Agou, T.; Kobayashi, J.; Kim, Y.; Gabbaı, F. P.; Kawashima, T.
Chem. Lett. 2007, 36, 976. (w) Day, J. K.; Bresner, C.; Coombs, N. D.; Fallis, I.
A.; Ooi, L.-L.; Aldridge, S. Inorg. Chem. 2008, 47, 793. (x) Galbraith, E.; Fyles,
T. M.; Marken, F.; Davidson, M. G.; James, T. D. Inorg. Chem. 2008, 47, 6236.
(y) Ting, R.; Harwig, C. W.; Lo, J.; Li, Y.; Adam, M. J.; Ruth, T. J.; Perrin, D. M. J.
Org. Chem. 2008, 73, 4662. (z) Ting, R.; Harwig, C.; auf dem Keller, U.;
McCormick, S.; Austin, P.; Overall, C. M.; Adam, M. J.; Ruth, T. J.; Perrin, D. M.
J. Am. Chem. Soc. 2008, 130, 12045. (aa) Xu, Z.; Kim, S. K.; Han, S. J.; Lee, C.;
Kociok-Kohn, G.; James, T. D.; Yoon, J. Eur. J. Org. Chem. 2009, 3058.
(5) For fluoride binding by other three-coordinate boron-containing
systems, see, for example: (a) Yamaguchi, S.; Akiyama, S.; Tamao, K. J.
Am. Chem. Soc. 2001, 123, 11372. (b) Yamaguchi, S.; Shirasaka, T.; Akiyama,
S.; Tamao, K. J. Am. Chem. Soc. 2002, 124, 8816. (c) Miyata, M.; Chujo, Y.
Polym. J. 2002, 34, 967. (d) Kubo, Y.; Yamamoto, M.; Ikeda, M.; Takeuchi, M.;
Shinkai, S.; Tamao, K. Angew. Chem., Int. Ed. 2003, 42, 2036. (e) Herberich, G.
E.; Englert, U.; Fischer, A.; Wiebelhaus, D. Eur. J. Inorg. Chem. 2004, 4011. (f)
Liu, Z. Q.; Shi, M.; Li, F. Y.; Fang, Q.; Chen, Z. H.; Yi, T.; Chuang, C. H. Org.
Lett. 2005, 7, 5481. (g) Badr, I. H. A.; Meyerhoff, M. E. J. Am. Chem. Soc. 2005,
127, 5318. (h) Agou, T.; Kobayashi, J.; Kawashima, T. Org. Lett. 2005, 7, 4373.
(i) Xu, S.; Chen, K.; Tian, H. J. Mater. Chem. 2005, 15, 2676. (j) Parab, K.;
€
€
ꢀ
(w) Gabbaı, F. P. ACS Symp. Ser. 2006, 917, 208. (x) Melaımi, M.; Sole, S.;
€
Chiu, C.-W.; Wang, H.; Gabbaı, F. P. Inorg. Chem. 2006, 45, 8136. (y) Chase, P.
A.; Henderson, L. D.; Piers, W. E.; Parvez, M.; Clegg, W.; Elsegood, M. R. J.
€
Organometallics 2006, 25, 349. (z) Lee, M. H.; Gabbaı, F. P. Inorg. Chem. 2007,
46, 8132.
(7) For examples of cyanide binding by triarylboranes, see, for example,
ref 5j and: (a) Zhou, J.; Lancaster, S. J.; Walker, D. A.; Beck, S.; Thornton-
Pett, M.; Bochmann, M. J. Am. Chem. Soc. 2001, 123, 223. (b) Vei, I. C.;
Pascu, S. I.; Green, M. L. H.; Green, J. C.; Schilling, R. E.; Anderson, G. D. W.;
€
Rees, L. H. Dalton Trans. 2003, 2550. (c) Hudnall, T. W.; Gabbaı, F. P. J. Am.
€
Chem. Soc. 2007, 127, 11978. (d) Chiu, C.-W.; Gabbaı, F. P. Dalton Trans. 2008,
814. (e) Huh, J.-O.; Do, Y.; Lee, M. H. Organometallics 2008, 27, 1022. (f) Chiu,
€
C. W.; Kim, Y.; Gabbaı, F. P. J. Am. Chem. Soc. 2009, 131, 60. (g) Kim, Y.; Zhao,
€
H.; Gabbaı, F. P. Angew. Chem., Int. Ed. 2009, 48, 4957.
(8) For examples of boron subphthalocyanines in cyanide sensing, see, for
ꢀ~
´
example: (a) Ros-Lis, J. V.; Martınez-Manez, R.; Soto, J. Chem. Commun.
2005, 5260. (b) Palomares, E.; Martínez-Díaz, M. V.; Torres, T; Coronado, E.
Adv. Funct. Mater. 2006, 16, 1166.
(9) For examples of cyanide sensing using boronic acids in aqueous
€
Venkatasubbaiah, K.; Jakle, F. J. Am. Chem. Soc. 2006, 128, 12879. (k) Chiu,
€
C.-W.; Gabbaı, F. P. J. Am. Chem. Soc. 2006, 128, 14248. (l) Agou, T.;
solution making use of the equilibrium ArB(OH)2
þ
3HCN f
[ArB(CN)3]- þ H2O þ [H3O]þ, see, for example: (a) Badugu, R.; Lakowicz,
J. R.; Geddes, C. D. Anal. Biochem. 2004, 327, 82. (b) Badugu, R.; Lakowicz, J.
R.; Geddes, C. D. Anal. Chim. Acta 2004, 522, 9. (c) Badugu, R.; Lakowicz, J.
R.; Geddes, C. D. J. Am. Chem. Soc. 2005, 127, 3635. (d) Badugu, R.; Lakowicz,
J. R.; Geddes, C. D. Dyes Pigm. 2005, 64, 49.
Kobayashi, J.; Kawashima, T. Inorg. Chem. 2006, 45, 9137. (m) Sakuda, E.;
Funahashi, A.; Kitamura, N. Inorg. Chem. 2006, 45, 10670. (n) Liu, X. Y.; Bai, D.
R.; Wang, S. Angew. Chem., Int. Ed. 2006, 45, 5475. (o) Hudnall, T. W.;
€
€
Melaımi, M.; Gabbaı, F. P. Org. Lett. 2006, 8, 2747. (p) Bresner, C.; Day, J. K.;
Coombs, N. D.; Fallis, I. A.; Aldridge, S.; Coles, S. J.; Hursthouse, M. B. Dalton
Trans. 2006, 3660. (q) Lee, M. H.; Agou, T.; Kobayashi, J.; Kawashima, T.;
(10) For examples of cyanide binding/sensing by transition metal systems,
€
Gabbaı, F. P. Chem. Commun. 2007, 1133. (r) Sun, Y.; Ross, N.; Zhao, S.-B.;
€
see, for example: (a) Warzeska, S.; Kramer, R. Chem. Commun. 1996, 499. (b)
Huszarik, K.; Jia, W.-L.; Wang, R.-Y.; Macartney, D.; Wang, S. J. Am. Chem.
Soc. 2007, 129, 7510. (s) Cui, Y.; Lu, Z.-H.; Wang, S. Dalton Trans. 2007, 2634.
(t) Boshra, R.; Venkatasubbaiah, K.; Doshi, A.; Lalancette, R. A.; Kakalis, L.;
€
Ahlers, B.; Cammann, K.; Warzeska, S.; Kramer, R. Angew. Chem., Int. Ed.
Engl. 1996, 35, 2141. (c) Kim, Y.-H.; Hong, J.-I. Chem. Commun. 2002, 512. (d)
Anzenbacher, P., Jr.; Tyson, D. S.; Jursíkova, K.; Castellano, F. N. J. Am. Chem.
ꢀ
€
Jakle, F. Inorg. Chem. 2007, 46, 10174. (u) Dorsey, C. L.; Jewula, P.; Hudnall, T.
Soc. 2002, 124, 6232. (e) Chow, C.-F.; Lam, M. H. W.; Wong, W.-Y. Inorg.
Chem. 2004, 43, 8387. (f) Liu, H.; Shao, X.-B.; Jia, M.-X.; Jiang, X.-K.; Li, Z.-T.;
Chen, G.-J. Tetrahedron 2005, 61, 8095. (g) Poland, K.; Topoglidis, E.; Durrant,
J. R.; Palomares, E. Inorg. Chem. Commun. 2006, 9, 1239.
W.; Hoefelmeyer, J. D.; Taylor, T. J.; Honesty, N. R.; Chiu, C.-W.; Schulte, M.;
€
Gabbaı, F. P. Dalton Trans. 2008, 4442. (v) Hudnall, T. W.; Kim, Y.-K.;
€
Bebington, M. W. P.; Bourissou, D.; Gabbaı, F. P. J. Am. Chem. Soc. 2008,
130, 10890. (w) Kawachi, A.; Tani, A.; Shimada, J.; Yamamoto, Y. J. Am. Chem.
(11) For organic cyanide sensors, see, for example: (a) Ros-Lis, J. V.;
€
´
ꢀ~
Soc. 2008, 130, 4222. (x) Boshra, R.; Doshi, A.; Jakle, F. Organometallics 2008,
Martınez-Manez, R.; Soto, J. Chem. Commun. 2002, 2248. (b) Tomasulo, M.;
27, 1534. (y) Zhao, Q.; Li, F.; Liu, S.; Yu, M.; Liu, Z.; Yi, T.; Huang, C. Inorg.
Chem. 2008, 47, 9256. (z) Zhou, G.; Baumgarten, M.; M€ullen, K. J. Am. Chem.
Raymo, F. M. Org. Lett. 2005, 7, 4633. (c) García, F.; García, J. M.; García-
ꢀ~
ꢀ
Acosta, B.; Martínez-Manez, R.; Sancenon, F.; Soto, J. Chem. Commun. 2005,
2790. (d) Chung, Y.; Lee, H.; Ahn, K. H. J. Org. Chem. 2006, 71, 9470. (e)
Tomasulo, M.; Sortino, S.; White, A. J. P.; Raymo, F. M. J. Org. Chem. 2006, 71,
744. (f) Yang, Y.-K.; Tae, J. Org. Lett. 2006, 8, 5721. (g) Chung, Y. M.; Raman,
B.; Kim, D.-S.; Ahn, K. H. Chem. Commun. 2006, 186. (h) Chen, C.-L.; Chen,
Y.-H.; Chen, C.-Y.; Sun, S.-S. Org. Lett. 2006, 8, 5053.
€
Soc. 2008, 130, 12477. (aa) Kim, Y.; Gabbaı, F. P. J. Am. Chem. Soc. 2009, 131,
3363. (bb) Sun, Y.; Wang, S. Inorg. Chem. 2009, 48, 3755. (cc) Rao, Y.-L.; Wang,
S. Inorg. Chem. 2009, 48, 7689. (dd) Hudson, Z. M.; Zhao, S.-B.; Wang, R.-Y.;
Wang, S. Chem.;Eur. J. 2009, 15, 6131. (ee) Agou, T.; Sekine, M.; Kobayashi,
J.; Kawashima, T. Chem. Commun. 2009, 1894.

Article
Inorganic Chemistry, Vol. 49, No. 1, 2010 159
appropriate redox-active organic dye, can be used to produce
an amplified colorimetric output,4b we have also targeted
systems capable of colorimetric sensing. In this respect, a
major sensing challenge stems from the potential for compe-
titive binding of fluoride at strongly Lewis acidic -BMes2-
based receptors. Previous work by the group of Gabba€ı, for
example, has demonstrated the importance not only of
electronics (i.e., of borane Lewis acidity) but also of steric
factors in determining the relative binding affinities of borane
receptors for cyanide and fluoride.7c Moreover, the difficul-
ties in using the relative proton affinities (or even pKa’s) of
cyanide/fluoride in a predictive sense are amply illustrated
by the very strong solvent dependence of the relative basicities
of F- and CN- [HF: pKa 3 (in H2O), 15 (in DMSO); HCN:
pKa 9 (in H2O), 13 (in DMSO)].13 Thus, in order to put the
binding of these two important analytes by ferrocene-
derivatized Lewis acids within a systematic framework, we
have undertaken a synthetic study designed to access a range
of such systems, offering systematic variation in the nature of
the borane substituent, overall charge, and ancillary ligand
sterics/electronics. The potential for anion binding by hybrid
Lewis acid/hydrogen-bondreceptors hasalso beenexamined.
While these studies targeted a rational, modular approach to
understand factors important in sensor design, a more empi-
rical approach is also reported, namely, a two-component
sensor system which offers a logical (colorimetric) solution
for discriminating between fluoride and cyanide.14
CFCl3, respectively. The 13C signals due to boron-bound C5H4
or CN carbon atoms were typically broad or not observed.
Infrared spectra were measured for each compound pressed
into a disk with an excess of dried KBr or as a solution in an
appropriate solvent on a Nicolet 500 FT-IR spectrometer.
Mass spectra were measured by the EPSRC National Mass
Spectrometry Service Centre, Swansea University, or by the
departmental service. Perfluorotributylamine was used as a
standard for high-resolution measurements. Elemental micro-
analyses were carried out at London Metropolitan University.
Abbreviations: b = broad, s = singlet, d = doublet, t = triplet,
q = quintet, sept = septet, m = multiplet; Fc = ferrocenyl, (η5-
C5H5)Fe(η5-C5H4); Fc* = 10,20,30,40,50-pentamethyl-ferrocenyl,
(η5-C5Me5)Fe(η5-C5H4); 1,10-fc = 1,10-ferrocenediyl, (η5-C5-
H4)2Fe; 1,2-fc = 1,2-ferrocenediyl, (η5-C5H5)Fe(η5-C5H3-1,2);
Mes = mesityl, 2,4,6-Me3C6H2; Xyl = 2,6-Me2C6H3; XylF
4-F-2,6-Me2C6H2; XylOMe = 4-MeO-2,6-Me2C6H2.
=
ii. Syntheses. FcBMes2 (1) and Fc*BMes2 (2). Compounds
1 and 2 were prepared by a common method, exemplified for 1.
To a solution of FcBBr2 (2.00 g, 5.62 mmol) in diethyl ether
(50 mL) was added dropwise mesityllithium (2.6 equiv) also in
diethyl ether (ca. 50 mL) and the reaction mixture stirred for
18 h. At this point, 11B NMR indicated complete conversion to a
single product (δB 76). After the removal of volatiles in vacuo,
extraction into hexanes (ca. 50 mL), and cooling to -30 ꢀC,
1 was obtained as a red powder (yield: 1.52 g, 62%). Single
crystals suitable for X-ray diffraction were obtained by the slow
evaporation of pentane from a concentrated solution. 1H NMR
(300 MHz, [D6]benzene, 20 ꢀC): δ 1.91 (s, 6H, para-CH3 of Mes),
2.21 (s, 12H, ortho-CH3 of Mes), 3.65 (s, 5H, Cp), 4.09 (m, 2H,
C5H4), 4.26 (m, 2H, C5H4), 6.54 (s, 4H, aromatic CH of Mes).
13C NMR (126 MHz, [D6]benzene, 20 ꢀC): δ 21.0 (para-CH3 of
Mes), 24.7 (ortho-CH3 of Mes), 69.6 (Cp), 73.8, 79.6 (C5H4),
128.7 (aromatic CH of Mes), 137.9 (para-quaternary of Mes),
139.2 (ortho-quaternary of Mes), boron-bound quaternary
carbons not observed. 11B (96 MHz, [D6]benzene, 20 ꢀC): δ
76. MS(EI): 434 (100%) Mþ; exact mass (calcd for Mþ, 10B, 56Fe
isotopomer) = 433.1899, (measd) 433.1899. UV/vis (aceto-
nitrile): λmax = 510 nm, ε = 1310 mol-1 cm-1 dm3. E1/2 versus
FcH/FcHþ (peak-to-peak separation) = þ150 (98) mV in
dichloromethane; þ181 (80) mV in acetonitrile. Elem micro-
analysis calcd (for 1, C28H31BFe): C, 77.40; H, 7.20. Found: C,
77.37; H, 7.08. Compound 2 was obtained using a similar
procedure from Fc*BBr2 as a purple powder (yield: 0.23 g,
44%). Data for 2 is as follows. 1H NMR (300 MHz,
[D1]chloroform, 20 ꢀC): δ 1.66 (s, 15H, C5Me5), 2.19 (s, 6H,
para-CH3 of Mes), 2.30 (s, 12H, ortho-CH3 of Mes), 4.05 (m,
2H, C5H4), 4.15 (m, 2H, C5H4), 6.70 (s, 4H, CH of Mes). 13C
NMR (126 MHz, [D6]benzene, 20 ꢀC): δ 11.6 (CH3 of C5Me5),
21.3 (para-CH3 of Mes), 25.0 (ortho-CH3 of Mes), 79.9
(quaternary of C5Me5), 81.0, 82.1 (C5H4), 128.2 (aromatic CH
of Mes), 136.9 (para-quaternary of Mes), 139.1 (ortho-quaternary
of Mes), boron-bound quaternary carbons not observed. 11B
(96 MHz, [D6]benzene, 20 ꢀC): δ 77. MS(EI): 504 (100%) Mþ;
exact mass (calcd for Mþ, 10B, 56Fe isotopomer), 503.2682,
Founds: 503.2677. UV/vis (acetonitrile): λmax = 542 nm, ε =
1420 mol-1 cm-1 dm3. E1/2 versus FcH/FcHþ (peak-to-peak
separation) = -194 (72) mV in dichloromethane; -176 (75) mV
in acetonitrile.
Experimental Section
i. General Considerations. Manipulations of air-sensitive
reagents were carried out under a nitrogen or argon atmosphere
using standard Schlenk line or drybox techniques. Nondeu-
teriated solvents were dried using a commercially available
Braun Solvent Purification System. [D6]Benzene, [D]chloro-
form, and [D2]dichloromethane (Goss) were degassed and dried
over potassium ([D6]benzene) or molecular sieves ([D]chloro-
form, [D2]dichloromethane) prior to use. Triethylamine (Alfa)
was dried over sodium wire before use; the tetra-n-butylammo-
nium salts of fluoride and cyanide were dried to a constant
weight in vacuo, analyzed for composition (i.e., for state of
hydration),15 and stored under an atmosphere of dry argon until
use. The known compounds FcBBr2, Fc*BBr2, MesLi, 1-iodo-
2,6-dimethyl-4-fluorobenzene, 1-bromo-2,6-dimethyl-4-meth-
oxybenzene, 1,10-fcBr2, FcBO2C2H2Ph2 (16), and Fc*BO2C2-
H2Ph2 (17) were prepared by literature procedures.4p,16 All
other reagents were used as received from commercial sources.
NMR spectra were measured on a Varian Mercury VX-300 or
Bruker AVII 500 FT-NMR spectrometer. Residual signals of
solvent were used as a reference for 1H and 13C NMR; 11B and
19F NMR spectra were referenced with respect to Et2O BF3 and
3
(12) Havir, J. Collect. Czech. Chem. Commun. 1961, 26, 1775.
(13) See, for example: http://www2.lsdiv.harvard.edu/labs/evans/pdf/
evans_pKa_table.pdf (accessed Nov 2009).
(14) A preliminary communication of part of this work has previously
been published: Broomsgrove, A. E. J.; Addy, D. A.; Bresner, C.; Fallis, I.
A.; Thompson, A. L.; Aldridge, S. Chem.;Eur. J. 2008, 14, 7525.
(15) The compositions of the tetrabutylammonium fluoride and cyanide
hydrates used in anion binding studies (and prepared by prolonged drying in
vacuo) were determined to be [nBu4N]F 4H2O and [nBu4N]CN 2H2O by
FcB(Xyl)2 (3), FcB(XylF)2 (4), and FcB(XylOMe
) (5). The
2
three compounds were prepared by a common method, exem-
plified for 4. To a solution of 1-iodo-2,6-dimethyl-4-fluoroben-
zene (0.50 g, 2.00 mmol) in diethyl ether (20 mL) was added
dropwise n-butyllithium (1.0 equiv) at -35 ꢀC and the reaction
mixture warmed to room temperature. After stirring for 4 h, a
solution of FcBBr2 (0.41 equiv) also in diethyl ether (ca. 30 mL)
was added to the reaction mixture, which was then stirred for a
further 18 h. At this point, monitoring by 11B NMR spectro-
scopy indicated conversion to two products (giving rise to
3
3
elemental microanalysis.
(16) (a) FcBBr2:Renk, T.; Ruff, W.; Siebert, W. J. Organomet. Chem.
1976, 120, 1. (b) MesLi:Rybinskaya, M. I.; Kreindlin, A. Z.; Fadeeva, S. S.;
Petrovskii, P. V. J. Organomet. Chem. 1988, 345, 341. (c) 1-I-2,6-Me2-4-F-C6H2:
Dewar, M. S.; Takeuchi, Y. J. Am. Chem. Soc. 1967, 89, 390. (d) 1-Br-2,6-Me2-
4-OMe-C6H2: Kang, H.; Facchetti, A.; Stern, C. L.; Rheingold, A. L.; Kassel, W.
S.; Marks, T. J. Org. Lett. 2005, 7, 3721. (e) 1,10-fcBr2: Shafir, A.; Power, M. P.;
Whitener, G. D.; Arnold, J. Organometallics 2000, 19, 3978.

160 Inorganic Chemistry, Vol. 49, No. 1, 2010
Broomsgrove et al.
signals at δB 76 and 51). After removal of the volatiles in vacuo,
purification by column chromatography (5% Et2O in hexanes)
yielded FcB(XylF)2 (4) as a red powder. Yield: 0.25 g, 70%.
Single crystals suitable for X-ray diffraction were obtained by
the slow evaporation of solvent from a solution in diethyl ether.
1H NMR (300 MHz, [D]chloroform, 20 ꢀC): δ 2.39 (s, 12H,
ortho-CH3 of XylF), 4.15 (s, 5H, Cp), 4.46 (m, 2H, C5H4), 4.74
solid. Purification of the crude product by column chromato-
graphy (hexanes to 5% Et2O/hexanes) yielded 1,10-fc(Br)BMes2
(6) as a blood-red microcrystalline solid. Yield: 0.85 g, 90%.
Single crystals suitable for X-ray diffraction were obtained by
slow evaporation of a concentrated solution in diethyl ether. 1H
NMR (300 MHz, [D]chloroform, 20 ꢀC): δ 2.27 (s, 6H, para CH3
of Mes), 2.37 (s, 12H, ortho-CH3 of Mes), 4.16 (AB multiplet,
2H, C5H4), 4.35 (AB multiplet, 2H, C5H4), 4.57 (AB multiplet,
2H, C5H4), 4.87 (AB multiplet, 2H, C5H4), 6.80 (s, 4H, CH of
Mes). 13C NMR (126 MHz, [D]chloroform, 20 ꢀC): δ 21.0 (para-
CH3 of Mes), 24.5 (ortho-CH3 of Mes), 69.2, 71.0, 78.1, 80.9
(C5H4), 128.2 (aromatic CH of Mes), 137.1 (para-quaternary of
Mes), 138.9 (ortho-quaternary of Mes), boron-bound quatern-
ary carbons not observed. 11B (96 MHz, [D]chloroform, 20 ꢀC):
(m, 2H, C5H4), 6.70 (d, J = 9 Hz, 4H, aromatic CH of XylF). 13
C
NMR (126 MHz, [D]chloroform, 20 ꢀC): δ 24.6 (d, 4JCF = 2 Hz,
ortho-CH3 of XylF), 69.5 (Cp), 74.0, 79.2 (C5H4), 113.9 (d,
2JCF = 19 Hz, aromatic CH of XylF), 141.6 (d, JCF = 8 Hz,
1
ortho-quaternary of MesF), 160.7 (d, JCF = 245 Hz, CF of
1
XylF). 19F NMR (282 MHz, [D]chloroform, 20 ꢀC): δ -116.8 (t,
J = 9 Hz, para-F of XylF). 11B (96 MHz, [D]chloroform, 20 ꢀC):
77. MS(EI): 442 (100%) Mþ; exact mass (calcd for Mþ, 10B, 56Fe
isotopomer), 441.1409; found, 441.1413. UV/vis (acetonitrile):
λmax = 509 nm, ε = 1406 mol-1 cm-1 dm3. E1/2 versus FcH/
FcHþ (peak-to-peak separation) = þ184 (60) mV in dichloro-
methane. Elem microanalysis calcd (for 4, C26H25BF2Fe): C,
70.58; H, 5.70. Found: C, 70.68; H, 5.66. FcB(Xyl)2 (3) was
prepared from 1-lithio-2,6-dimethylbenzene in an analogous
manner. Yield: 0.25 g, 56%. 1H NMR (300 MHz,
[D]chloroform, 20 ꢀC): δ 2.42 (s, 12H, ortho-CH3 of Xyl), 4.16
(s, 5H, Cp), 4.51 (m, 2H, C5H4), 4.71 (m, 2H, C5H4), 6.98 (AB
mult, 4H, meta-CH of Xyl), 7.12 (AB mult, 2H, para-CH of
Xyl). 13C NMR (126 MHz, [D]chloroform, 20 ꢀC): δ 24.7
(ortho-CH3 of Xyl), 69.5 (Cp), 73.7, 79.4 (C5H4), 127.2 (meta-
CH of Xyl), 127.5 (para-CH of Xyl), 139.0 (ortho-quaternary of
Xyl), 145.7 (boron-bound quaternary of Xyl). 11B (96 MHz,
[D]chloroform, 20 ꢀC): 74.5. MS(EI): 406.2 (100%) Mþ; exact
mass (calcd for Mþ, 10B, 56Fe isotopomer), 406.1550; found,
406.1551. UV/vis (acetonitrile): λmax = 509 nm, ε = 1280 mol-1
cm-1 dm3. E1/2 versus FcH/FcHþ (peak-to-peak separation) =
þ153 (98) mV in dichloromethane. Elem microanalysis calcd
(for 3, C26H27BFe): C, 76.84; H, 6.70. Found: C, 76.84; H, 6.74.
δ 78. MS(EIþ): 512 (100%) Mþ; exact mass (calcd for Mþ, 10
B
isotopomer), 511.1004; found, 511.1004. UV/vis (acetonitrile):
λmax = 510 nm, ε = 1167 mol-1 cm-1 dm3. E1/2 versus FcH/
FcHþ (peak-to-peak separation) = þ265 (111) mV in dichloro-
methane. Elem microanalysis calcd (for 6, C28H30BBrFe): C,
65.50; H, 5.89. Found: C, 65.62; H, 5.85.
1,10-fc(CH2NMe2)BMes2 (7). To a solution of 6 (0.40 g, 0.78
mmol) in tetrahydrofuran (20 mL) at -78 ꢀC was added
dropwise t-butyllithium (0.92 mL of a 1.7 M solution in pentane,
1.56 mmol) and the reaction mixture stirred for 30 min at
that temperature. Solid N,N-dimethylmethylideneammonium
iodide (0.29 g, 1.56 mmol) was then added and the reaction
mixture allowed to warm to room temperature overnight. The
resulting blood-red solution was diluted with ethyl acetate
(ca. 50 mL) and washed with water (25 mL) and brine
(25 mL). The organic fractions were then dried (MgSO4) and
volatiles removed in vacuo to yield crude 7, which was purified
by column chromatography (hexanes to 5% ethyl acetate/
hexanes, then hexanes to 4% ethyl acetate and 1% NEt3, and
finally hexanes to 49% ethyl acetate and 1% NEt3) to give 7 as a
blood-red solid. Yield: 0.26 g, 69%. 1H NMR (300 MHz,
[D]chloroform, 20 ꢀC): δ 2.01 (s, 6H, NMe2), 2.20 (s, 6H,
para-CH3 of Mes), 2.34 (s, 12H, ortho-CH3 of Mes), 2.79
(s, 2H, CH2 of CH2NMe2), 4.06 (AB multiplet, 2H, C5H4),
4.09 (AB multiplet, 2H, C5H4), 4.32 (AB multiplet, 2H, C5H4),
4.55 (AB multiplet, 2H, C5H4), 6.74 (s, 4H, CH of Mes). 13C
NMR (126 MHz, [D]chloroform, 20 ꢀC): δ 20.0 (para-CH3 of
Mes), 23.5 (ortho-CH3 of Mes), 43.6 (CH3 of NMe2), 57.3 (CH2
of CH2NMe2), 67.4, 69.0, 70.4, 73.4, 78.9, 82.5 (C5H4), 127.2
(aromatic CH of Mes), 135.9 (para-quaternary of Mes), 137.9
(ortho-quaternary of Mes), 141.7 (boron-bound quaternary of
Mes). 11B (96 MHz, [D]chloroform, 20 ꢀC): δ 78. MS(ESþ): 492
(100%); M þ H); exact mass (calcd for Mþ, 10B, 54Fe iso-
topomer), 489.2603; found, 489.2603. UV/vis (acetonitrile):
λmax = 499 nm, ε = 1104 mol-1 cm-1 dm3. E1/2 versus FcH/
FcHþ (peak-to-peak separation) = þ297 (62) mV in dichlor-
omethane.
[1,10-fc(CH2NMe3)BMes2]þI- (8). To a solution of 7 (0.049 g,
0.10 mmol) in hexanes (5 mL) was added dropwise methyl iodide
(0.5 mL, 8.03 mmol) and the reaction mixture stirred for 1 h at
room temperature. Volatiles were removed in vacuo, yielding
crude 8 as a pale pink powder, which was recrystallized from
dichloromethane/hexanes as single crystals suitable for X-ray
diffraction. Yield: 0.063 g, 99%). 1H NMR (300 MHz,
[D]chloroform, 20 ꢀC): δ 2.22 (s, 6H, para-CH3 of Mes), 2.38
(s, 12H, ortho-CH3 of Mes), 2.99 (s, 6H, CH3 of NMe3þ), 3.85 (s,
2H, CH2 of CH2NMe3þ), 4.25 (AB multiplet, 2H, C5H4), 4.39
(AB multiplet, 2H, C5H4), 4.56 (AB multiplet, 2H, C5H4), 4.85
(AB multiplet, 2H, C5H4), 6.79 (s, 4H, CH of Mes). 13C NMR
(126 MHz, [D]chloroform, 20 ꢀC): δ 20.0 (para-CH3 of Mes),
23.6 (ortho-CH3 of Mes), 51.4 (CH3 of NMe3þ), 65.4 (CH2 of
CH2NMe3þ), 70.9, 71.5, 72.4, 74.6 (C5H4), 79.8 (quaternary C
of C5H4), 127.5 (aromatic CH of Mes), 136.3 (para-quaternary
of Mes), 137.9 (ortho-quaternary of Mes), 141.4 (boron-bound
quaternary of Mes); boron-bound quaternary C of C5H4 not
FcB(XylOMe
) (5) was prepared from 1-lithio-2,6-dimethyl-4-
2
methoxybenzene in an analogous manner. Yield: 0.35 g, 67%.
Samples prepared using this method are invariably contami-
nated by ca. 3% of a second ferrocene-containing species, which
has been shown by mass spectrometry and NMR studies to be
FcB(XylOMe)(Ar) (where Ar = 1-methoxy-3-methyl-4-bromo-
5-methylbenzene). The aryl substituent is derived from a com-
peting deprotonation reaction ortho to the OMe substituent
inherent in the reaction of 1-bromo-2,6-dimethyl-4-methoxy-
1
benzene with n-butyllithium. H NMR (300 MHz, [D]chloro-
form, 20 ꢀC): δ 2.39 (s, 12H, ortho-CH3 of XylOMe), 3.78 (s, 6H,
OCH3 of XylOMe), 4.14 (s, 5H, Cp), 4.48 (m, 2H, C5H4), 4.68 (m,
2H, C5H4), 6.53 (s, 4H, aromatic CH of XylOMe). 13C NMR (126
MHz, [D]chloroform, 20 ꢀC): δ 24.8 (ortho-CH3 of XylOMe),
54.8 (OCH3 of XylOMe), 69.4 (Cp), 73.4, 79.3 (C5H4), 112.7
(aromatic CH of XylOMe), 141.0 (ortho-quaternary of XylOMe),
158.9 (para-quaternary of XylOMe). 11B (96 MHz,
[D]chloroform, 20 ꢀC): 75.0. MS(EI): 466 (100%) Mþ; exact
mass (calcd for Mþ, 10B, 56Fe isotopomer), 465.1801; found,
465.1797. UV/vis (acetonitrile): λmax = 502 nm, ε = 1075 mol-1
cm-1 dm3. E1/2 versus FcH/FcHþ (peak-to-peak separation) =
þ95 (87) mV in dichloromethane.
1,10-fc(Br)BMes2 (6). To a solution of 1,10-fcBr2 (0.63 g, 1.84
mmol) in tetrahydrofuran (20 mL) at -78 ꢀC was added,
dropwise, n-butyllithium (1.15 mL of a 1.6 M solution in
hexanes, 1.84 mmol), and the reaction mixture stirred for 30
min at this temperature. To the precipitate so formed was added
dimesitylboron fluoride (0.58 g, 1.93 mmol, 90% purity as
received from Aldrich) as a solution in tetrahydrofuran (5 mL)
and the reaction mixture allowed to warm to room temperature
over a period of ca. 4 h. The resulting blood-red solution
was diluted with diethyl ether (75 mL) and washed with water
(50 mL) and brine (50 mL). The organic fraction was then dried
(MgSO4) and volatiles removed in vacuo to yield a blood-red

Article
Inorganic Chemistry, Vol. 49, No. 1, 2010 161
observed. 11B (96 MHz, [D]chloroform, 20 ꢀC): no signal
observed. MS(ESþ): 506 (94%) Mþ, 447 (100%) (M - NMe3)þ;
exact mass (calcd for Mþ, 10B, 54Fe isotopomer), 503.2759;
found, 503.2768. UV/vis (acetonitrile): λmax = 487 nm, ε =
898 mol-1 cm-1 dm3. E1/2 versus FcH/FcHþ (peak-to-peak
separation) = þ314 (62) mV in dichloromethane. Elem anal.
dm3. E1/2 versus FcH/FcHþ (peak-to-peak separation) =
þ367 (91) mV in dichloromethane. Elem anal. (%) calcd (for
10 1/2CH2Cl2): C, 57.73; H, 6.27; N, 2.07. Found: C, 57.55; H,
3
6.01; N, 2.14.
[1,2-fc(CH2NMe2H)BMes2]þCl- (11). To a solution of 9
(0.20 g, 0.41 mmol) in diethyl ether (ca. 30 mL) at -30 ꢀC was
added a solution of hydrochloric acid in diethyl ether (0.8 mL of
a 1 M solution, 0.81 mmol) and the reaction mixture stirred for
30 min at -30 ꢀC. Volatiles were removed in vacuo, and the
resulting pink solid recrystallized from dichloromethane/hex-
anes to yield 11 as a pale red crystalline solid. Yield: 0.12 g,
55%. 1H NMR (300 MHz, [D]chloroform, 20 ꢀC): δ 1.95 (s, 6H,
para-CH3 of Mes), 2.22 (s, 18H, overlapping signals: CH3 of
NMe2Hþ and ortho-CH3 of Mes), 3.59 (s, 1H, CH of CH2-
NMe3þ), 4.13 (s, 1H, CH of C5H3), 4.20 (s, 5H, Cp), 4.34 (s, 1H,
CH of CH2NMe3þ), 4.59 (s, 1H, CH of C5H3), 4.74 (s, 1H, CH
of C5H3), 5.67 (s, 1H, NH), 6.76 (s, 4H, aromatic CH of Mes).
13C NMR (126 MHz, [D]chloroform, 20 ꢀC): δ 20.0 (para-CH3
of Mes), 24.8 (ortho-CH3 of Mes), 44.3 (CH3 of NMe2Hþ), 56.0
(CH2 of CH2NMe3þ), 70.3 (Cp), 70.5, 71.8, 72.7, 80.3, 82.1
(C5H3), 127.5 (aromatic CH of Mes), 137.0 (para-quaternary of
Mes), 138.2 (ortho-quaternary of Mes), boron-bound quatern-
ary of Mes not observed. 11B NMR (96 MHz, [D]chloro-
form, 20 ꢀC): δ 78 (br). MS(EI): 491 (100%; M - H)þ; exact
mass (calcd for (M - H)þ, 10B isotopomer), 490.2478; found,
490.2479. UV/vis (dichloromethane): λmax = 502 nm, ε = 1120
mol-1 cm-1 dm3.
(%) calcd (for 8 2H2O): C, 57.39; H, 6.78; N, 2.09. Found: C,
3
57.19; H, 7.00; N, 2.07.
1,2-fc(CH2NMe2)BMes2 (9). To a solution of (N,N-dimethyl-
aminomethyl)ferrocene (1.50 g, 6.17 mmol) in diethyl ether
(ca. 50 mL) at 0 ꢀC was added dropwise n-butyllithium (1.0
equiv) and the reaction mixture stirred at 0 ꢀC for 1 h. After
warming to room temperature and stirring for a further 12 h,
dimesitylboron fluoride (1.66 g, 6.17 mmol) in diethyl ether
(ca. 50 mL) was then added slowly to the reaction mixture
at -78 ꢀC. After stirring for a further 1 h, and subsequent
warming to room temperature, volatiles were removed in vacuo
and the solid extracted into hexanes (ca. 10 mL). Cooling to
-30 ꢀC led to the formation of 9 as a purple powder, which was
recrystallized from a concentrated pentane solution at -80 ꢀC.
Yield: 1.66 g, 55%. 1H NMR (300 MHz, [D]chloroform, 20 ꢀC):
δ 1.71 (s, 6H, para-CH3 of Mes), 2.17 (s, 6H, NMe2), 2.25 (s,
12H, ortho-CH3 of Mes), 3.04 (AB m, 2H, CH2 of CH2NMe2),
4.08 (s, 5H, Cp), 4.25 (m, 1H, CH of C5H3), 4.46 (m, 1H, CH of
C5H3), 4.83 (m, 1H, CH of C5H3), 6.65 (s, 4H, aromatic CH of
Mes). 13C NMR (126 MHz, [D]chloroform, 20 ꢀC): δ 21.1 (para-
CH3 of Mes), 24.7 (ortho-CH3 of Mes), 44.9 (CH3 of NMe2),
57.0 (CH2 of CH2NMe2), 70.2 (Cp), 71.8, 76.8, 77.4, 79.4, 94.3
(C5H3), 128.7 (aromatic CH of Mes), 137.2 (para-quaternary of
Mes), 139.6 (ortho-quaternary of Mes), 141.1 (boron-bound
quaternary of Mes). 11B NMR (96 MHz, [D]chloroform, 20 ꢀC):
δ 76 (br). MS(EI): 491 (weak) Mþ, 243 (54%) (M - BMes2)þ,
199 (45%) (M - BMes2NMe2)þ, 186 (7%) (M - BMes2CH2-
NMe2)þ; exact mass (calcd for Mþ, 10B isotopomer), 490.2478;
found, 490.2475. UV/vis (dichloromethane): λmax = 510 nm,
ε = 1500 mol-1 cm-1 dm3. Elem anal. (%) calcd (for 9): C,
75.74; H, 7.80; N, 2.85. Found: C, 75.77; H, 7.81; N, 2.85.
[1,2-fc(CH2NMe3)BMes2]þI- (10). To a solution of 9 (0.20 g,
0.41 mmol) in hexanes (ca. 30 mL) was added methyl iodide
(0.05 mL, 0.81 mmol) and the reaction mixture stirred at room
temperature for 12 h. The pink precipitate of 10 so formed was
collected by filtration and dried in vacuo. Single crystals suitable
for X-ray diffraction were obtained by recrystallization from
[1,2-fc(CH2NMe2H)B(Mes)OH]þ[BF4]- (13). To a solution
of 9 (0.20 g, 0.41 mmol) in diethyl ether (ca. 30 mL) was added
tetrafluoroboric acid diethyl ether complex (2.0 equiv), and the
reaction mixture stirred at room temperature for 12 h. The
resulting orange precipitate was collected by filtration and
recrystallized from dichloromethane/hexanes as orange crystals
of 13 suitable for X-ray diffraction. Yield: 0.01 g, 60%. 1H NMR
(300 MHz, [D]chloroform, 20 ꢀC): δ 2.22 (s, 6H, ortho-CH3 of
Mes), 2.68 (m, 6H, NMe2H), 2.87 (s, 3H, para-CH3 of Mes), 4.15
(s, 5H, Cp), 4.22 (s, 1H, CH of C5H3), 4.32 (br AB m, 2H, CH2 of
CH2NMe2Hþ), 4.42 (s, 1H, CH of C5H3), 4.65 (s, 1H, CH of
C5H3), 6.55 (s, 1H, OH), 6.75 (s, 2H, aromatic CH of Mes), 7.57
(br s, 1H, NH). 13C NMR (126 MHz, [D]chloroform, 20 ꢀC): δ
21.2 (para-CH3 of Mes), 22.5 (ortho-CH3 of Mes), 40.6 (CH3 of
NMe2Hþ), 42.2 (CH3 of NMe2Hþ), 58.6 (CH2 CH2NMe2Hþ),
69.3 (Cp), 69.8, 69.9, 70.6, 71.0, 72.2 (C5H3), 126.9 (aromatic
CH of Mes), 127.6 (para-quaternary of Mes), 137.7 (ortho-
quaternary of Mes), 138.4 (boron-bound quaternary of Mes).
11B NMR (96 MHz, [D]chloroform, 20 ꢀC): δ 51. MS(ESþ): 390
(22%) Mþ; exact mass (calcd for Mþ, 11B isotopomer),
390.0715; found, 390.0707.
1
dichloromethane/diethyl ether. Yield: 0.19 g, 70%. H NMR
(300 MHz, [D]chloroform, 20 ꢀC): δ 2.00 (br s, 6H, ortho-CH3 of
Mes), 2.20 (s, 6H, para-CH3), 2.78 (s, 9H, CH3 of [NMe3]þ), 4.08
(br s, 1H, CH of CH2NMe2), 4.32 (s, 5H, Cp), 4.60 (s, 1H, CH of
C5H3), 4.81 (s, 1H, CH of C5H3), 5.44(s, 1H, CH of C5H3), 5.50
(s, 1H, CH of CH2NMe2), 6.76 (br s, 4H, aromatic CH of Mes).
1H NMR (300 MHz, [D2]dichloromethane, -30 ꢀC): δ 1.77
(s, 3H, ortho-CH3 of Mes), 2.05 (s, 3H, ortho-CH3 of Mes), 2.22
(s, 9H, overlapping signals: both para- and one ortho-CH3 of
Mes), 2.58 (s, 9H, CH3 of NMe3þ), 3.01 (s, 3H, ortho-CH3 of
Mes), 4.01 (m, 1H, CH of CH2NMe3þ), 4.35 (s, 5H, CH of
C5H3), 4.57 (s, 1H, CH of C5H3), 4.88 (s, 1H, CH of C5H3), 5.08
(m, 1H, CH of CH2NMe3þ), 5.33 (s, 1H, CH of C5H3), 6.67 (s,
1H, aromatic CH of Mes), 6.75 (s, 1H, aromatic CH of Mes),
6.83 (s, 1H, aromatic CH of Mes), 7.00 (s, 1H, aromatic CH of
Mes). 13C NMR (126 MHz, [D]chloroform, 20 ꢀC): δ 20.0 (para-
CH3 of Mes), 23.6 (br, ortho-CH3 of Mes), 50.9 (CH3 of
NMe3þ), 64.5 (CH2 of CH2NMe3þ), 70.9 (Cp), 73.6, 79.1,
79.5, 81.8, 82.3 (C5H3), 127.7 (br, aromatic CH of Mes), 137.3
(br, para-quaternary of Mes), 138.6 (br, ortho-quaternary of
Mes), 140.0 (boron-bound quaternary of Mes). 11B NMR (96
MHz, [D]chloroform, 20 ꢀC): δ 80 (br). MS(ESþ): 506 (39%)
Mþ, 447 (100%; M - NMe3)þ, 258 (5%; M - BMes2)þ, 199
(17%; M - BMes2 - NMe3)þ; exact mass (calcd for Mþ, 10B
and 54Fe isotopomer), 503.2759; found, 503.2754. UV/vis
(dichloromethane): λmax = 502 nm, ε = 1300 mol-1 cm-1
FcB(MesF)F (14). To a solution of 1,3,5-tris(trifluoro-
methyl)benzene (0.7 mL, 3.76 mmol) in diethyl ether (ca. 10
mL) was added dropwise n-butyllithium (1.0 equiv) at -78 ꢀC,
and the reaction mixture warmed to room temperature. After
stirring for 18 h, a solution of FcBBr2 (0.47 equiv) also in diethyl
ether (ca. 50 mL) was added to the reaction mixture, which was
then stirred for a further 18 h. At this point, monitoring by 11
B
NMR spectroscopy indicated complete conversion to a single
product (δB 51). Removal of volatiles in vacuo, extraction into
hexanes (ca. 70 mL), and cooling to -30 ꢀC yielded 14 as a red
powder. Yield: 0.48 g, 54%. Single crystals suitable for X-ray
diffraction were obtained by cooling a concentrated solution in
hexanes to -30 ꢀC. 1H NMR (300 MHz, [D6]benzene, 20 ꢀC): δ
3.52 (s, 5H, Cp), 3.57 (m, 2H, C5H4), 3.68 (m, 2H, C5H4), 6.60 (s,
2H, aromatic CH of MesF). 13C NMR (126 MHz,
[D]chloroform, 20 ꢀC): 69.0 (Cp), 73.1 (C5H4), 74.8 (d, 3JCF
=
4 Hz, C5H4), 122.5 (q, 1JCF = 273 Hz, para-CF3 of MesF), 123.3
(q, 1JCF = 276 Hz, ortho-CF3 of MesF), 125.9 (sept, 3JCF = 4
2
Hz, aromatic CH of MesF), 132.5 (q, JCF = 35 Hz, para-
quaternary of MesF), 134.4 (q, 2JCF = 34 Hz, ortho-quaternary
of MesF), 138.6 (br, boron-bound quaternary of MesF). 11B (96

162 Inorganic Chemistry, Vol. 49, No. 1, 2010
Broomsgrove et al.
MHz, [D6]benzene, 20 ꢀC): 51. 19F NMR (282 MHz,
[D6]benzene, 20 ꢀC): δ -49.1 (br s, BF), -57.2 (s, ortho-CF3
of MesF), -63.1 (s, para-CF3 of MesF). MS(EI): 496 (100%)
Mþ; exact mass (calcd for Mþ, 10B isotopomer), 495.0174;
found, 495.0173.
FcB(MesF)Me (15). To a solution of 14 (0.20 g, 0.40 mmol)
in diethyl ether (10 mL) was added dropwise methyllithium
(1.0 equiv) at -78 ꢀC, and the reaction mixture warmed to room
temperature. After stirring for 2 h, monitoring by 11B NMR
spectroscopy indicated complete conversion to a single product
(δB 72). After the removal of volatiles in vacuo, extraction into
pentane (ca. 20 mL), and cooling to -30 ꢀC, 15 was obtained
as a red powder. Yield: 0.12 g, 58%. 1H NMR (300 MHz,
[D6]benzene, 20 ꢀC): δ 0.57 (sept, 6JHF = 2 Hz, 3H, Me), 3.39 (s,
5H, Cp), 3.43 (m, 2H, C5H4), 3.61 (m, 2H, C5H4), 6.52 (s, 2H,
aromatic CH of MesF). 13C NMR (126 MHz, [D]chloroform,
Mes), 30.0 (br, ortho-CH3 of Mes), 61.6 (NCH2 of [nBu4N]þ),
76.0, 81.2 (C5H4), 82.0 (quaternary of C5Me5), 133.2 (aromatic
CH of Mes), 134.6 (para-quaternary of Mes), 146.0 (ortho-
quaternary of Mes), 175.6 (CN-), boron-bound quaternary
carbons not observed. 11B (96 MHz, [D]chloroform, 20 ꢀC):
δ -17. MS(ES-): 530 (100%) [2 CN]-; exact mass (calcd for
3
[2 CN]-, 56Fe, 10B isotopomer), 529.2723; found, 529.2730.
3
UV/vis (acetonitrile): λmax = 481 nm, ε = 295 mol-1 cm-1
dm3. IR (CHCl3): 2162 cm-1 st, ν(CN). E1/2 versus FcH/FcHþ
(peak-to-peak separation) = -691 (95) mV in acetonitrile.
[K(18-crown-6)]þ[FcB(Xyl)2 CN]- ([K(18-crown-6)]þ[3 CN]-).
3
3
The reactions of Lewis acids 3, 4, 5, and 15 with KCN and 18-
crown-6 were carried out in a similar manner, illustrated here for
3. To a solution of 3 (0.045 g, 0.11 mmol) in chloroform-d (4 mL)
was added KCN and 18-crown-6 (1 equiv of each). The reaction
mixture was stirred for 1 h at 20 ꢀC, after which it was judged to be
complete, by the existence of a single resonance in the 11B NMR
spectrum at δB ca. -15 ppm. Single crystals suitable for X-ray
diffraction were obtained by layering the reaction solution with
pentane. Yield: 0.075 g, 81%. 1H NMR (300 MHz, [D]chloroform,
20 ꢀC): δ 2.20 (br s, 12H, ortho-CH3 of Xyl), 3.53 (s, 24H, 18-
crown-6), 3.98 (s, 5H, Cp), 4.12 (br s, 2H, C5H4), 4.25 (br s, 2H,
C5H4), 6.71-6.82 (overlapping m, 6H, ortho- and para-CH of Xyl).
13C NMR (126 MHz, [D]chloroform, 20 ꢀC): δ 25.3 (ortho-CH3 of
Xyl), 67.5 (C5H4), 68.2 (Cp), 70.1 (18-crown-6), 75.7 (C5H4), 123.0
(meta-CH of Xyl), 127.8 (para-CH of Xyl), 141.8 (ortho-quatern-
ary of Xyl), 153.0 (boron-bound quaternary of Xyl). 11B (96 MHz,
20 ꢀC): 10.9 (CH3), 69.0 (Cp), 73.2, 75.4 (C5H4), 122.9 (q, 1JCF
=
272 Hz, para-CF3 of MesF), 123.7 (q, J = 275 Hz, ortho-CF3 of
MesF), 125.8 (sept, 3JCF = 4 Hz, aromatic CH of MesF), 130.6
(q, 2JCF = 34 Hz, para-quaternary of MesF), 131.7 (q, 2JCF = 33
Hz, ortho-quaternary of MesF), 149.0 (br, boron-bound qua-
ternary of MesF). 11B (96 MHz, [D6]benzene, 20 ꢀC): δ 72. 19
F
NMR (282 MHz, [D6]benzene, 20 ꢀC): δ -56.3 (s, ortho-CF3 of
MesF), -62.9 (s, para-CF3 of MesF). MS(EI): 492 (100%) Mþ;
exact mass (calcd for Mþ, 10B isotopomer), 491.0425; found,
491.0423.
[nBu4N]þ[FcBMes2 CN]- ([nBu4N]þ[1 CN]-). A mixture of 1
3
3
[D]chloroform, 20 ꢀC): -16. MS(ES-): 432.2 (100%) [3 CN]-;
(0.05 g, 0.12 mmol) and tetra-n-butylammonium cyanide dihy-
drate (1.05 equiv) in [D]chloroform (5 mL) was stirred for 1 h, at
which point the reaction was judged to be complete by 11B NMR
spectroscopy (quantitative conversion to a single resonance at
δB -16). Layering of the reaction mixture with diethyl ether led
3
exact mass (calcd for [3 CN]-, 56Fe, 10B isotopomer), 429.1675;
3
found, 429.1670. UV/vis (dichloromethane): λmax = 457 nm, ε =
142 mol-1 cm-1 dm3. IR (CHCl3): 2164 cm-1 st, ν(CN). E1/2 versus
FcH/FcHþ (peak-to-peak separation) = -327 (90) mV in acet-
to the formation of [nBu4N]þ[1 CN]- as orange crystals suitable
onitrile. Elem microanalysis calcd (for [K(18-crown-6)]þ[3 CN]-,
3
3
for X-ray diffraction. Yield: 0.089 g, 90%. 1H NMR (300 MHz,
[D]chloroform, 20 ꢀC): δ 0.95 (m, 12H, CH3 of [nBu4N]þ), 1.36
(m, 8H, CH2 of [nBu4N]þ), 1.46 (m, 8H, CH2 of [nBu4N]þ), 2.09
(s, 18H, ortho- and para-CH3 of Mes), 2.92 (m, 8H, NCH2 of
[nBu4N]þ), 3.90 (s, 5H, Cp), 4.02 (m, 2H, C5H4), 4.1 (br m, 2H,
C5H4), 6.48 (s, 4H, aromatic CH of Mes). 13C NMR (126 MHz,
[D]chloroform, 20 ꢀC): δ 13.8 (CH3 of [nBu4N]þ), 19.8, 24.1
(CH2 of [nBu4N]þ), 20.9 (para-CH3 of Mes), 25.5 (br, ortho-CH3
of Mes), 58.6 (NCH2 of [nBu4N]þ), 68.1 (Cp), 67.4, 75.5 (C5H4),
128.9 (aromatic CH of Mes), 131.3 (para-quaternary of Mes),
141.7 (ortho-quaternary of Mes), 176.0 (CN-), boron-bound
quaternary carbons not observed. 11B (96 MHz, [D]chloroform,
C39H51BFeKNO6): C, 63.65; H, 6.99; N, 1.90. Found: C, 63.64; H,
7.19; N, 1.80. Data for [K(18-crown-6)]þ[FcB(XylF)2 CN]-, that
3
is, [K(18-crown-6)]þ[4 CN]-. Yield: 0.067 g, 71%. 1H NMR (300
3
MHz, [D]chloroform, 20 ꢀC): δ 2.17 (br, s, 12H, ortho-CH3 of
XylF), 3.53 (s, 24H, 18-crown-6), 3.96 (s, 5H, Cp), 4.08 (m, 2H,
C5H4), 4.14 (m, 2H, C5H4), 6.45 (d, 3JHF = 9 Hz, 4H, aromatic CH
of XylF). 13C NMR (126 MHz, [D]chloroform, 20 ꢀC): 25.4 (s, CH3
of XylF), 67.5 (C5H4), 67.9 (Cp), 70.0 (18-crown-6), 75.4 (C5H4),
113.6 (d, 2JCF =16 Hz, aromatic CH of XylF), 143.7 (d, 3JCF = 6
Hz, ortho-quarternary of XylF), 159.9 (d, 1JCF = 238 Hz, para-CF
of Xylf). 11B (96 MHz, [D]chloroform, 20 ꢀC): -14. 19F NMR (282
MHz, [D]chloroform, 20 ꢀC): -124.5 (t, 3JFH = 9 Hz, para-F of
XylF). MS(ES-): 468 (100%) [4 CN]-; exact mass (calcd for
20 ꢀC): -16. MS(ES-): 460 (100%) [1 CN]-; exact mass (calcd
3
3
[4 CN]-, 56Fe, 10B isotopomer), 467.1439; found, 467.1437. UV/
for [1 CN]-, 10B, 54Fe isotopomer), 457.1995; found, 457.1988.
3
3
vis (dichloromethane): λmax = 463 nm, ε = 210 mol-1 cm-1 dm3.
IR (CHCl3): 2171 cm-1 st, ν(CN). E1/2 versus FcH/FcHþ (peak-
to-peak separation) = -330 (85) mV in acetonitrile. Elem micro-
UV/vis (chloroform): λmax = 460 nm, ε = 170 mol-1 cm-1 dm3.
IR (CH2Cl2): 2162 cm-1 st, ν(CN). E1/2 versus FcH/FcHþ
(peak-to-peak separation) = -383 (100) mV in acetonitrile.
Elem microanalysis calcd (for [nBu4N]þ[1 CN]- CHCl3,
C46H68BCl3FeN2): C, 67.21; H, 8.34; N, 3.41. Found: C,
66.83; H, 8.28; N, 3.23.
[nBu4N]þ[Fc*BMes2 CN]- ([nBu4N]þ[2 CN]-). A mixture of
2 (0.05 g, 0.10 mmol) and tetra-n-butylammonium cyanide
analysis calcd (for [K(18-crown-6)]þ[4 CN]-, C39H49BF2-
3
3
3
FeKNO6): C, 60.68; H, 6.40; N, 1.82. Found: C, 60.63; H, 6.32;
CN]-, that is,
N, 1.75. Data for [K(18-crown-6)]þ[FcB(XylOMe
)
2
3
[K(18-crown-6)]þ[5 CN]-. Yield: 0.105 g, 92%. H NMR (300
1
3
3
3
MHz, [D]chloroform, 20 ꢀC): δ 2.18 (br s, 12H, ortho-CH3 of
XylOMe), 3.53 (s, 24H, 18-crown-6), 3.69 (s, 6H, OCH3 of XylOMe),
3.98 (s, 5H, Cp), 4.07 (m, 2H, C5H4), 4.16 (br s, C5H4) 6.33 (s, 4H,
aromatic CH of XylOMe). 13C NMR (126 MHz, [D]chloroform,
20 ꢀC): δ25.6 (ortho-CH3 of XylOMe), 54.6(OCH3 of XylOMe), 67.9
(Cp), 70.0 (18-crown-6), 75.6, 77.1 (C5H4), 113.1 (aromatic CH of
XylOMe), 143.0 (ortho-quaternary of XylOMe), 155.4 (para-quatern-
ary of XylOMe). 11B (96 MHz, [D]chloroform, 20 ꢀC): -15.
MS(ES-): 492.2 (100%) [5 CN]-; exact mass (calcd for [5 CN]-,
dihydrate (1.05 equiv) in [D]chloroform (5 mL) was stirred for
1 h, after which time the reaction was deemed complete by 11
B
NMR spectroscopy (quantitative conversion to a single reso-
nance at δB -17). Concentration of the reaction mixture
and layering with diethyl ether led to the formation of
[nBu4N]þ[2 CN]- as an orange microcrystalline solid. Yield:
3
1
0.081 g, 91%. H NMR (300 MHz, [D]chloroform, 20 ꢀC): δ
0.96 (m, 12H, CH3 of [nBu4N]þ), 1.34 (m, 8H, CH2 of [nBu4N]þ),
1.47 (m, 8H, CH2 of [nBu4N]þ), 1.75 (s, 15 H, C5Me5), 1.95 (s,
6H, para-CH3 of Mes), 2.06 (s, 12H, ortho-CH3 of Mes), 2.94
(m, 8H, NCH2 of [nBu4N]þ), 3.52 (s, 2H, C5H4), 3.68 (s, 2H,
C5H4), 6.43 (s, 4H, aromatic CH of Mes). 13C NMR (126 MHz,
[D6]benzene, 20 ꢀC): δ 15.5 (CH3 of C5Me5), 17.4 (CH3 of
[nBu4N]þ), 23.4, 27.5 (CH2 of [nBu4N]þ), 24.7 (para-CH3 of
3
3
56Fe, 10B isotopomer), 489.1886; found, 489.1886. UV/vis
(dichloromethane): λmax = 454 nm, ε = 162 mol-1 cm-1 dm3.
IR (CHCl3): 2164 cm-1 st, ν(CN). E1/2 versus FcH/FcHþ (peak-to-
peak separation) = -471(60) mV in acetonitrile. Data for [K(18-
crown-6)]þ[FcB(MesF)Me CN]-, that is, [K(18-crown-6)]þ-
3
[15 CN]-. Yield: 0.089 g, 90%. 1H NMR (300 MHz,
3

Article
Inorganic Chemistry, Vol. 49, No. 1, 2010 163
[D]chloroform, 20 ꢀC): δ 0.46 (s, 3H, CH3), 3.63 (s, 24H, 18-crown-
6), 4.00 (m, 1H, C5H4), 4.01 (m, 2H, C5H4), 4.12 (s, 5H, Cp), 4.31
(m, 1H, C5H4), 7.75 (s, 2H, aromatic CH of Mes). 13C NMR (126
MHz, [D]chloroform, 20 ꢀC): δ 8.3 (br, CH3), 65.8, 67.4 (C5H4),
68.4 (Cp), 70.1 (18-crown-6), 123.7 (q, 1JCF = 272 Hz, para-CF3 of
MesF), 125.1 (q, J = 275 Hz, ortho-CF3 of MesF), 125.6 (br,
aromatic CH of MesF), 125.8 (q, 2JCF = 33 Hz, para-quaternary of
MesF), 136.9 (q, 2JCF = 30 Hz, ortho-quaternary of MesF), 145.2
(br, boron-bound quaternary of MesF). 11B (96 MHz,
was obtained merely for verification of connectivity, and no
discussion of the metrical data is attempted.
In the case of [K(18-crown-6)]þ[4 CN]-, there was a small
3
amount of disorder present in one crown ether and a phenyl
group exhibited some translational disorder, which were mod-
eled with two sites and refined occupancies with isotropic
displacement parameters. Similarly, [K(18-crown-6)]þ-
[15 CN]- exhibited disorder in one of the CF3 groups, which
3
was modeled with isotropic displacement parameters. On refin-
ing [K(18-crown-6)]þ[5 CN]-, it became evident that there was
[D]chloroform, 20 ꢀC):
δ
-14. 19F NMR (300 MHz,
3
[D]chloroform, 20 ꢀC): -50.8 (s, ortho-CF3 of MesF), -62.8 (s,
a large residual electron density peak approximately 1.7 A from
C(25). This was thought to result from a small amount of
residual bromine from the previous preparation step, which
was modeled as ca. 4% occupied. It is possible that the peak is
spurious, due for example to absorption. However, the data are
otherwise good, and there is no indication of any other difficul-
ties.
para-CF3 of MesF). MS(ES-): 518 (100%) [15 CN]-; exact mass
3
(calcd for [15 CN]-, 10B, 54Fe isotopomer), 515.0514; found,
515.0525.
3
iii. Crystallographic Method. Data for 1, 4, 6, 8 CH2Cl2,
10 1/2CH2Cl2, 13, 14, [nBu4N]þ[1 CN]- CHCl3, [K(18-crown-
3
þ3
3
3
6)] [3 CN]-, [K(18-crown-6)]þ[4 CN]-, [K(18-crown-6)]þ-
3
3
[5 CN]-, and [K(18-crown-6)]þ[15 CN]- were collected on a
Selected structural details for the new compounds are in-
cluded in Table 1, and full crystallographic data for all struc-
tures have been deposited with the Cambridge Crystallographic
Data Centre, CCDC 671665, 671666, and 744798-744807.
Copies of these data can be obtained free of charge from The
Cambridge Crystallographic Data Centre via www.ccdc.cam.
ac.uk/data_request/cif.
3
3
Nonius KappaCCD diffractometer (Mo KR radiation (λ=
0.71073 A) at 150(2) K with an Oxford Cryosystems Cryostream
N2 open-flow cooling device.17a Data were processed using the
DENZO-SMN package, including interframe scaling (which
was carried out using Scalepack within DENZO-SMN).17b The
structures were solved using SIR92 (for 4, 6, 8 CH2Cl2, 13,
3
[K(18-crown-6)]þ[3 CN]-, [K(18-crown-6)]þ[4 CN]-, [K(18-
iv. Electrochemical Method. Electrochemical analyses were
carried out using the following conditions: electrolyte, 0.1 M
[nBu4N]þ[PF6]- in dichloromethane or acetonitrile; reference
electrode standard, 0.1 M [nBu4N]þ[PF6]-, 0.01 M AgNO3 in
acetonitrile. Following degassing of the electrolyte solution with
argon, background cyclic voltammetry (CV) scans were mea-
sured, and a sample (ca. 2-5 mg) of the ferrocene functionalized
Lewis acid was added to the solution. Further degassing served
to purge the solution of any additional dissolved oxygen and
agitate the solid Lewis acid to dissolve the compound, prior to
spectral acquisition. Further CV scans were measured on the
3
3
crown-6)]þ[5 CN]-, and [K(18-crown-6)]þ[15 CN]-),17c or
3
3
SHELXS (for 1, 10 1/2CH2Cl2, 14, and [nBu4N]þ[1 CN]-
3
3
3
CHCl3).17d Refinement was carried out using full-matrix least-
squares within the CRYSTALS suite,17e on either F2 (for 4, 6,
8 CH2Cl2, 13, 14, [K(18-crown-6)]þ[3 CN]-, [K(18-crown-
3 þ
3
6)] [4 CN]-, and [K(18-crown-6)]þ[5 CN]-) or F (for [K(18-
3
3
crown-6)]þ[15 CN]- only), or with SHELXTL (for 1, 10
3
3
1/2CH2Cl2, and [nBu4N]þ[1 CN]- CHCl3).17d In general, all
3
3
non-hydrogen atoms were refined with anisotropic displace-
ment parameters; however, this was not always possible where
addition of aliquots of solid [nBu4N]F 4H2O (or KF/18-crown-
there was disorder present (e.g., [K(18-crown-6)]þ[4 CN]-) or
the data were of poor quality (e.g., 14). For 4, 6, 8 CH2Cl2, 13,
3
3
6), and on the addition of ferrocene as a reference. Electro-
chemical data reported in the text are for solutions in acetonitrile
or dichloromethane, referenced with respect to the ferrocene/
ferrocenium couple; peak-to-peak separations are listed (in
parentheses) in the experimental data.
3
14, [K(18-crown-6)]þ[3 CN]-, [K(18-crown-6)]þ[4 CN]-,
3
3
3
[K(18-crown-6)]þ[5 CN]-, and [K(18-crown-6)]þ[15 CN]-,
3
the majority of hydrogen atoms were visible in the difference
map, and the general technique was to refine their positions
and isotropic displacement parameters using restraints prior
to inclusion into the model with riding constraints. For
compounds 1, 10 1/2CH2Cl2, and [nBu4N]þ[1 CN]- CHCl3,
v. Binding Constant Determinations. Binding constants were
evaluated for 1, 3, 4, 5, 6, and 8 in dichloromethane by the
6r
ꢀ
method reported by Sole and Gabba€ı; the program LabFit
3
3
3
(www.labfit.net) was used to fit the experimentally determined
data of A/Ao versus anion concentration. The binding constants
for [1,2-fc(CH2NMe3)BMes2]þI- (10) for both fluoride and
cyanide, however, were found to be too large to be reliably
determined by this method and were therefore determined by a
competition experiment between 10 and PhBMes2.5w
hydrogen atoms were added geometrically and refined using a
riding model only.
The data for 14 are very weak and twinned, which together
with the pseudosymmetry present meant the refinement was
unstable. It was therefore controlled using partial shifts and
shift-limiting restraints within CRYSTALS.17c In addition,
there was disorder in the nonfunctionalized Cp rings, and the
poor quality of the data meant that a partially isotropic refine-
ment was necessary with thermal restraints used for the CF3
fluoride atoms. Despite the difficulties with this structure, the
only real doubt concerning the connectivity is whether there is a
fluorine or an OH bound to the boron. Given the similarity in
scattering and the quality of the data, it was not possible to tell
from the crystallography; however, NMR experiments carried
out on the single crystals were conclusive, so fluorine was used in
the model. The structure of compound 14 (along with that of 13)
Results and Discussion
i. Syntheses. The synthesis of simple ferrocenylboranes
of the type (η5-C5R5)Fe(η5-C5H4BAr2) (e.g., 1-5,
Scheme 1) can readily be accomplished in yields of up to
60% by the reaction of slightly more than 2 equiv of the
respective aryllithium reagent with FcBBr2. This metho-
dology offers a versatile approach to a range of Lewis
acids which are both air-stable (provided that each aryl
substituent features two ortho-methyl groups) and sys-
tematically tunable (through the electronic properties of
the para substituent). This approach removes the need to
the isolate the corresponding (halo)diarylborane,
(Hal)BAr2, as would be a prerequisite, for example, in a
synthetic route involving the intermediacy of ferrocenyl-
lithium. Diethyl ether proves to be the optimal reaction
medium, with related chemistry carried out in thf, for
(17) (a) Cosier, J.; Glazer, A. M. J. Appl. Crystallogr. 1986, 19, 105. (b)
Denzo: Otwinowski, Z.; Minor, W. Methods Enzymol.; Carter, C. W., Sweet, R.
M., Eds.; Academic Press: New York, 1996; Vol. 276, p 307. (c) Sir-92: Altomare,
A.; Cascarano, G.; Giacovazzo, C.; Guagliardi, A.; Burla, M. C.; Polidori, G.;
Camalli, M. J. Appl. Crystallogr. 1994, 27, 435. (d) Sheldrick, G. M. Acta
Crystallogr., Sect. A 2008, 64, 112. (e) CRYSTALS: Betteridge, P. W.;
Carruthers, J. R.; Cooper, R. I.; Prout, J.; Watkin, D. J. J. Appl. Crystallogr.
2003, 36, 1487.

164 Inorganic Chemistry, Vol. 49, No. 1, 2010
Broomsgrove et al.
Table 1. Crystallographic Data for 1, 4, 6, 8 CH2Cl2, 10 1/2CH2Cl2, 13, 14, [nBu4N]þ[1 CN]- CHCl3, and [K(18-crown-6)]þ[X CN]- (X=3, 4, 5, 15)
3
3
3
3
3
1
4
6
empirical formula
CCDC deposition number
fw
C
28H31BFe
C26H25BF2Fe
744798
442.14
150(2)
0.71073
orthorhombic
P212121
7.9136(1), 10.5933(2), 25.2332(5)
90, 90, 90
2115.33(6), 4
1.388
C28H30BBrFe
744799
513.11
671665
434.19
120(2)
0.71073
monoclinic
P21/n
temp (K)
150(2)
wavelength (A)
cryst syst
space group
0.71073
monoclinic
P21/n
unit cell lengths: a, b, c (A)
R, β, γ (deg)
10.298(1), 15.728(1), 14.027(1)
10.5687(1), 15.5960(3), 14.8927(2)
90, 104.958(3), 90
2194.84(16), 4
1.314
0.699
920
90, 106.9697(8), 90
2347.87(6)
1.452
2.357
1056
volume (A3), Z
density (calcd) (Mg/m3)
abs coeff (mm-1
F(000)
)
0.740
920
cryst size (mm3)
0.08 ꢀ 0.06 ꢀ 0.01
0.04 ꢀ 0.10 ꢀ 0.24
0.08 ꢀ 0.21 ꢀ 0.33
θ range for data collection (deg) 2.99-25.03
index ranges (h, k, l)
5.153-27.489
5.129-27.488
-12 to þ12, -18 to þ18, -16 to þ16 -10 to þ10, -13 to þ13, -32 to þ32 -13 to þ13, -20 to þ17, -19 to þ19
no. of reflns collected
no. of indep reflns/Rint
completeness to θmax (%)
absorption correction
max. and min transmission
refinement method
22801
3855 (0.1177)
99.7
22029 24590
4674 (0.064)
0.989
5349 (0.030)
0.992
semiempirical from equivs
0.993 and 0.946
full matrix least sq (F2)
3855/0/277
semiempirical from equivs
0.97 and 0.92
semiempirical from equivs
0.83 and 0.67
full matrix least sq (F2)
2746/0/272
0.9785
full matrix least sq (F2)
5348/0/280
0.9771
no. of data/restraints/params
goodness-of-fit on F2
final R indices [I > 2σ(I)]
R indices (all data)
1.094
R1 = 0.0658, wR2 = 0.1190
R1 = 0.1189, wR2 = 0.1463
þ0.413 and -0.421
R1 = 0.0349, wR2 = 0.0760
R1 = 0.0479, wR2 = 0.0808
þ0.43 and -0.39
0.51(2)
R1 = 0.0357, wR2 = 0.0802
R1 = 0.0441, wR2 = 0.0863
þ0.42 and = -0.89
largest peak/hole (e A-3
)
absolute structure parameter
8 CH2Cl2
3
10 1/2CH2Cl2
3
13
empirical formula
CCDC deposition number
fw
C
33H43BCl2FeIN
744800
C32.50H42BClFeIN
744801
675.68
150(2)
C22H29B2F4FeNO
744802
476.94
150(2)
718.18
150(2)
temp (K)
wavelength (A)
cryst syst
space group
0.71073
monoclinic
P21/c
22.7312(3), 15.5761(2), 9.1112(1)
90, 101.2240(5), 90
3164.24(7)
1.507
1.644
1464
0.71073
triclinic
0.71073
monoclinic
P21/c
10.5395(1), 8.1667(1), 28.2528(3)
90, 112.0243(4), 90
2254.34(4)
1.405
0.715
992
P1
8.1684(1), 19.1107(2), 20.674(3)
78.271(1), 87.789(1), 89.770(1)
3157.5(4)
1.421
1.561
unit cell lengths: a, b, c (A)
R, β, γ (deg)
volume (A3), Z
density (calcd) (Mg/m3)
abs coeff (mm-1
F(000)
)
1380
0.02 ꢀ 0.08 ꢀ 0.44
cryst size (mm3)
0.02 ꢀ 0.12 ꢀ 0.50
0.12 ꢀ 0.17 ꢀ 0.20
θ range for data collection (deg) 5.185-27.488
index ranges (h, k, l)
5.10-27.47
5.159-27.484
-29 to þ29, -20 to þ20, -11 to þ11 -10 to þ10, -24 to þ24, -24 to þ26 -13 to þ13, -10 to þ10, -36 to þ31
no. of reflns collected
no. of indep reflns/Rint
completeness to θmax (%)
absorption correction
max. and min transmission
refinement method
46216
7202 (0.046)
40730 49654
14250 (0.0397)
0.985
5152 (0.048)
0.992
0.992
semiempirical from equivs
0.97-0.81
semiempirical from equivs
0.966-0.892
semiempirical from equivs
0.92-0.87
full matrix least sq (F2)
7202/0/352
0.9611
full matrix least sq (F2)
14250/0/694
1.204
full matrix least sq (F2)
5152/0/280
0.9482
no. of data/restraints/params
goodness-of-fit on F2
final R indices [I > 2σ(I)]
R indices (all data)
R1 = 0.0341, wR2 = 0.0676
R1 = 0.0545, wR2 = 0.0732
þ1.28 and -0.86
R1 = 0.0500, wR2 = 0.1037
R1 = 0.0683, wR2 = 0.1098
þ0.099 and -1.002
R1 = 0.0360, wR2 = 0.0770
R1 = 0.0586, wR2 = 0.0832
þ0.50 and -0.58
largest peak/hole (e A-3
)
14
[nBu4N]þ[1 CN]- CHCl3
[K(18-crown-6)]þ[3 CN]-
3
3
3
empirical formula
C19H11BF10Fe
744803
495.94
150(2)
C46H68BCl3FeN2
C39H51BFeKNO6
744804
735.59
150(2)
CCDC deposition number
fw
temp (K)
671666
822.03
150(2)
wavelength (A)
cryst syst
space group
0.71073
monoclinic
P21/c
32.6412(5), 7.5481(1), 23.8131(3)
90, 111.3350(7), 90
5464.98(13)
1.808
0.71073
monoclinic
P21/c
11.951(1), 17.294(1), 21.529(1)
90, 92.344(1), 90
4445.82(12), 4
1.228
0.71073
monoclinic
P21/n
11.8766(1), 17.6012(2), 18.2248(2)
90, 91.7759(6), 90
3807.93(7)
1.283
unit cell lengths: a, b, c (A)
R, β, γ (deg)
volume (A3), Z
density (calcd) (Mg/m3)
abs coeff (mm-1
)
0.930
0.553
0.551

Article
Inorganic Chemistry, Vol. 49, No. 1, 2010 165
Table 1. Continued
14
[nBu4N]þ[1 CN]- CHCl3
[K(18-crown-6)]þ[3 CN]-
3
3
3
F(000)
cryst size (mm3)
θ range for data collection (deg) 5.119-27.479
index ranges (h, k, l)
no. of reflns collected
no. of indep reflns/Rint
completeness to θmax (%)
absorption correction
max. and min transmission
refinement method
2952
0.14 ꢀ 0.18 ꢀ 0.18
1760
0.60 ꢀ 0.20 ꢀ 0.06
1560
0.20 ꢀ 0.24 ꢀ 0.26
5.11-27.46
5.146-27.546
-42 to þ42, -8 to þ9, -30 to þ30 -15 to þ15, -22 to þ20, -25 to þ27 -15 to þ15, -22 to þ22, -23 to þ23
82235
12437 (0.090)
38935 47594
9981 (0.0778)
98.9
integration
0.935 and 0.680
full matrix least sq (F2)
9981/45/474
8689 (0.061)
0.987
0.991
semiempirical from equivs
0.88 and 0.79
semiempirical from equivs
0.90 and 0.83
full matrix least sq (F2)
7433/1294/524
1.2015
full matrix least sq (F2)
8687/0/442
0.9467
no. of data/restraints/params
goodness-of-fit on F2
final R indices [I > 2σ(I)]
R indices (all data)
1.023
R1 = 0.0540, wR2 = 0.1824
R1 = 0.1267, wR2 = 0.3820
þ0.71 and -0.59
R1 = 0.0528, wR2 = 0.1115
R1 = 0.1075, wR2 = 0.1320
þ0.511 and -0.457
R1 = 0.0433, wR2 = 0.0988
R1 = 0.0704, wR2 = 0.1095
þ0.52 and -0.60
Largest peak/hole (e A-3
)
[K(18-crown-6)]þ[4 CN]-
[K(18-crown-6)]þ[5 CN]-
[K(18-crown-6)]þ[15 CN]-
3
3
3
empirical formula
C39H49BF2FeKNO6
744805
771.57
C41H54.96BBr0.04FeKNO8
744806
798.76
150(2)
C33H38BF9FeKNO6
744807
821.41
150(2)
CCDC deposition number
fw
temp (K)
150(2)
wavelength (A)
cryst syst
space group
0.71073
triclinic
P1
0.71073
monoclinic
P21/n
12.0544(1), 27.3288(3), 12.4741(2)
90, 97.5901(7), 90
4073.37(9)
0.71073
triclinic
P1
11.7599(2), 11.9146(2), 14.1449(3)
102.9276(8), 101.1335(9), 100.6566(9)
1840.60(6)
1.482
0.612
unit cell dimensions: a, b, c (A) 12.6363(1), 14.9996(2), 20.2746(2)
R, β, γ (deg)
91.3387(5), 90.2094(5), 97.8732(5)
volume (A3), Z
3805.48(7)
1.347
0.562
1624
0.07 ꢀ 0.35 ꢀ 0.40
density (calcd) (Mg/m3)
1.302
0.562
abs coeff (mm-1
F(000)
)
1693.4
0.05 ꢀ 0.21 ꢀ 0.45
844
0.04 ꢀ 0.22 ꢀ 0.26
crystal size (mm3)
θ range for data collection (deg) 5.133-27.526
index ranges (h, k, l)
5.104-27.481
5.102-27.464
-16 to þ16, -19 to þ19, -19 to þ26 -15 to þ15, -32 to þ35, -16 to þ16 -15 to þ14, -15 to þ15, 0 to þ18
no. of reflns collected
no. of indep reflns/Rint
completeness to θmax (%)
absorption correction
max. and min transmission
refinement method
55708
17258 (0.036)
46021 36759
9208 (0.047)
0.986
8362 (0.046)
0.992
semiempirical from equivs
0.98 and 0.86
full matrix least sq (F)
5982/12/467
1.0838
0.984
semiempirical from equivs
0.96 and 0.83
semiempirical from equivs
0.97 and 0.71
full matrix least sq (F2)
17258/0/884
1.0042
full matrix least sq (F2)
9208/6/483
0.9341
no. of data/restraints/params
goodness-of-fit on F2
final R indices [I > 2σ(I)]
R indices (all data)
R1 = 0.0688, wR2 = 0.1260
R1 = 0.0786, wR2 = 0.1349
þ1.18 and -0.92
R1 = 0.0592, wR2 = 0.1208
R1 = 0.0831, wR2 = 0.1315
þ0.92 and -0.72
R1 = 0.0517, wR2 = 0.0647
R1 = 0.0799, wR2 = 0.0849
þ1.22 and -0.83
largest diff. peak/ hole (e A-3
)
example, leading to side reactions derived from Lewis
acid promoted ring-opening chemistry. Thus, for exam-
ple, FcB(Mes)O(CH2)4Br is the predominant organome-
tallic product obtained from the reaction of FcBBr2 with
MesMgBr in thf (see the Supporting Information).
Scheme 1. Syntheses of Diarylboryl Complexes 1-5a
Lewis acids 1-5 have been characterized by standard
spectroscopic and analytical techniques and, in the cases
of 1 and 4, by single-crystal X-ray diffraction. Particularly
diagnostic are the low-field 11B NMR resonances (δB ca.
75 ppm) characteristic of triarylboranes, and redox po-
tentials for the Fe(II)/Fe(III) couple which are shifted
anodically (by þ95 to þ184 mV) with respect to the
parent ferrocene/ferrocenium system, consistent with
the π-electron-withdrawing capabilities of the diarylboryl
(-BAr2) substituents.4w,18 For the series of compounds 1,
3, 4, and 5, which differ only in the nature of the para-aryl
substituent, the trend in the measured values of E1/2 is
a Typical reagents and conditions: (i) ArLi (ca. 2.5 equiv), diethyl
ether, 18 h at 20 ꢀC, 40-70%.
consistent with that expected from the respective Ham-
mett parameters (σp = þ0.06, 0, -0.17, and -0.27 for F,
H, Me, and OMe).19 For the Cp* derivative (2), the
electron-donating capabilities of the five methyl groups
(18) See, for example: (a) Carpenter, B. E.; Piers, W. E.; McDonald, R.
Can. J. Chem. 2001, 79, 291. (b) Carpenter, B. E.; Piers, W. E.; Parvez, M. E.;
Yap, G. P. A.; Rettig, S. J. Can. J. Chem. 2001, 79, 857.
(19) Hansch, C.; Leo, A.; Taft, R. W. Chem. Rev. 1991, 91, 165.

166 Inorganic Chemistry, Vol. 49, No. 1, 2010
Broomsgrove et al.
Figure 1. Molecular structures of FcBMes2 (1), FcBXylF (4), 1,10-fc(Br)BMes2 (6), and the cationic components of [1,10-fc(CH2NMe3)-
2
BMes2]þI- CH2Cl2 (8 CH2Cl2) and [1,2-fc(CH2NMe3)BMes2]þI- 1/2CH2Cl2 (10 1/2CH2Cl2). Hydrogen atoms, solvent molecules, and counterions
3
3
3
3
omitted for clarity; thermal ellipsoids set at the 40% probability level. Key bond lengths (A) and angles (deg), for 1: B(1)-C(1) 1.546(7), B(1)-C(11)
1.581(7), B(1)-C(20) 1.597(7), C(1)-B(1)-C(11) 119.7(4), C(1)-B(1)-C(20)122.0(4), C(11)-B(1)-C(20) 118.1(4). For4: B(7)-C(3) 1.547(4), B(7)-C(8)
1.609(4), B(7)-C(17) 1.588(4), C(3)-B(7)-C(8) 119.8(2), C(3)-B(7)-C(17) 118.0(2), C(8)-B(7)-C(17) 121.9(2). For 6: B(13)-C(9) 1.552(3), B(13)-C-
(14) 1.589(3), B(13)-C(23) 1.595(3), C(2)-Br(1) 1.879(3), C(9)-B(13)-C(14) 117.6(2), C(9)-B(13)-C(23) 125.1(2), C(14)-B(13)-C(23) 117.3(2). For
8 CH2Cl2: B(8)-C(4) 1.552(4), B(8)-C(9) 1.596(4), B(8)-C(18) 1.598(4), C(4)-B(8)-C(9) 125.2(2), C(4)-B(8)-C(18) 115.7(2), C(9)-B(8)-C(18)
3
118.8(2), C(5)-C(4)-B(9) 124.5(2). For 10 1/2CH2Cl2: B(1)-C(101) 1.557(6), B(1)-C(111) 1.595(6), B(1)-C(120) 1.592(6), C(101)-B(1)-C(111)
3
124.7(4), C(101)-B(1)-C(120) 114.6(4), C(111)-B(1)-C(120) 120.5(4), B(1)-C(101)-C(102) 134.4(4).
give rise to a ca. 350 mV cathodic shift in E1/2 compared to
Cp-substituted 1.20
electron-withdrawing 2,4,6-(CF3)3C6H2 (MesF) substituent
met with only partial success (Scheme 2). Thus, the
reaction of in situ generated MesFLi with FcBBr2 under
conditions analogous to those used to synthesize com-
plexes 1-5 does not lead to the isolation of the corre-
The molecular structures of 1 and 4 in the solid state are
qualitatively very similar (Figure 1 and Table 1), each
displaying a propeller-like alignment of the two aryl
substituents presumably enforced on steric grounds.
The angles between the least-squares BC3 and aryl ring
least-squares planes are 62.1 and 62.2ꢀ (for 1) and 59.8
and 63.9ꢀ (for 4), and related B-C distances for the two
compounds are statistically indistinguishable. Bending of
the -BAr2 unit toward the iron center;such that the
boron atom lies out of the least-squares plane defined by
the Cp substituent;is typically found for strongly Lewis
acidic boryl substituents (e.g., FcBBr2: —Cp cen-
troid-Cipso-B = 162ꢀ)21 but is much less marked in
both 1 and 4. Thus, Cp centroid-Cipso-B angles of 172.7
and 177.1ꢀ are found for 1 and 4, respectively. This near
linearity almost certainly reflects the relatively bulky
nature of the BAr2 (Ar = Mes, XylF) substituents in each
case, which constitutes a steric impediment to bending.21b
Attempts to extend this synthetic methodology
to systems featuring the bulkier and more strongly
sponding diarylboryl derivative FcBMesF but, rather,
2
the fluoro(aryl)borane FcB(MesF)F (14) in ca. 54%
yield. Similar chemistry resulting in incomplete substitu-
tion of boron-bound halides by MesFLi and in B-F bond
formation via activation of a CF3 substituent has been
reported by Dillon and co-workers.22 Thus, the reaction
of BCl3 with ca. 0.5 equiv of MesFLi in a mixed diethyl
ether/hexanes/heptane solvent system has been shown to
yield MesF2BF together with MesFBCl2. While 14 can be
characterized by standard spectroscopic, analytical, and
crystallographic methods (see Supporting Information),
its use in fluoride and cyanide binding experiments ap-
pears to be limited by the lability of the boron-bound
substituents (vide infra). Attempts to introduce a second
(different) aryl group by fluoride substitution proved to
be unsuccessful; reaction with MesLi under the relatively
forcing conditions required to bring about any conver-
sion of 14 leads to the formation of FcBMes2 (1), resulting
from displacement of both F and MesF substituents.
Simple displacement of fluoride can be accomplished by
the use of the much less bulky methyllithium, to give
(20) (a) Connelly, N. G.; Geiger, W. E. Chem. Rev. 1996, 96, 877. (b)
Noviandri, I.; Brown, K. N.; Fleming, D. S.; Gulyas, P. T.; Lay, P. A.; Masters, A.
F.; Phillips, L. J. Phys. Chem. B 1999, 103, 6713.
€
(21) (a) Appel, A.; Jakle, F.; Priermeier, T.; Schmid, R.; Wagner, M.
Organometallics 1996, 15, 1188. For a related recent study of a less sterically
encumbered diarylborylferrocene with a large tilt angle, see: (b) Kaufmann, L.;
Vitze, H.; Bolte, M.; Lerner, H.-W.; Wagner, M. Organometallics 2008, 27, 6215.
(22) Cornet, S.; Dillon, K.; Entwhistle, C. D.; Fox, M. A.; Goeta, A. E.;
Goodwin; Marder, T. B.; Thompson, A. L. Dalton Trans. 2003, 4395.

Article
Inorganic Chemistry, Vol. 49, No. 1, 2010 167
Scheme 2. Syntheses of MesF-Substituted Lewis Acids 14 and 15a
a Reagents and conditions: (i) MesFLi (2.1 equiv), diethyl ether, 18 h at 20 ꢀC, 54%; (ii) MeLi (1.0 equiv), diethyl ether, -78 to þ20 ꢀC then 2 h at
20 ꢀC, 58%.
Scheme 3. Syntheses of 10-Functionalized Ferrocenyl Dimesitylboranesa
a Reagents and conditions: (i) nBuLi (1.0 equiv), thf, -78 ꢀC, 30 min, then Mes2BF (0.94 equiv), thf, -78 to þ20 ꢀC, chromatographic purification,
t
90%; (ii) BuLi (2.0 equiv), thf/pentane, -78 ꢀC, 30 min, then [Me2NCH2]þI- (2.0 equiv), -78 to þ20 ꢀC, then 12 h at 20 ꢀC, chromatographic
purification, 69%; (iii) MeI (ca. 8 equiv), hexanes, 20 ꢀC, 1 h, 99%.
FcB(MesF)Me (15), which is competent for the binding of
anions such as fluoride and cyanide, but which;by virtue
of the presence of only one ortho-disubstituted aryl
substituent;proves to be air- and moisture-sensitive
and thus of little practical use in sensing/dosimetry ap-
plications.
complex 10 (closest C-H I contact: 3.01 A, vide
3 3 3
infra).23
Related systems featuring a 1,2 disposition of the
-BMes2 and pendant amine/ammonium functionalities
can be accessed from (N,N-dimethylaminomethyl)-
ferrocene utilizing simple ortho-directed lithiation chem-
istry.24 Thus, charge-neutral 1,2-fc(CH2NMe2)BMes2 (9)
and cationic [1,2-fc(CH2NMe3)BMes2]þ (10, as the iodide
salt) can be accessed in 55 and 39% overall yields from
FcCH2NMe2, as outlined in Scheme 4. Spectroscopic and
analytical data for both compounds are consistent with
the proposed formulations, with the structure of 10 (as the
dichloromethane hemisolvate) in the solid state being
confirmed crystallographically (Figure 1). In contrast to
the 1,10-disubstituted system 8, the increase in steric
crowding at the BMes2 unit implicit in functionalization
at the 2- position is presumably responsible for a widening
of the B(1)-C(101)-C(102) angle [134.4(4) cf. 124.5(2)ꢀ
for 8]. These steric effects are also evident in the 1H NMR
spectrum of 10 in [D]chloroform. Thus, a single broad
resonance is observed for the four mesityl ortho-methyl
groups at 20 ꢀC, which sharpens on warming, and which
can be resolved into four separate signals on cooling
to -30 ꢀC, consistent with restricted rotation about the
C(Cp)-B bond. As with the isomeric complex 8, the
Fe(II)/Fe(III) redox potential for compound 10 experi-
ences a significant anodic shift (þ217 mV with respect to
1/1þ; þ367 mV with respect to ferrocene/ferrocenium).
An alternative to methylation of the tertiary amine to
introduce a positively charged ammonium function is
simple protonation. In terms of anion binding motifs,
Syntheses of diarylborylferrocenes bearing additional
functionality can readily be accomplished via alternative
approaches which make use of lithioferrocene reagents.
Thus, systems containing pendant -CH2NMe2 and re-
lated substituents at either the 10- or 2- positions can be
synthesized by making use of either 1,10-dibromoferro-
cene or commercially available (N,N-dimethylamino-
methyl)ferrocene as starting materials (Schemes 3 and
4). Sequential lithiation of 1,10-fcBr2 and quenching with
Mes2BF and [Me2NdCH2]þI- leads to the stepwise for-
mation of 1,10-fc(Br)BMes2 (6) and 1,10-fc(CH2NMe2)-
BMes2 (7); 7 can readily be methylated at nitrogen by
methyl iodide to generate the cationic ammonium-func-
tionalized borane [1,10-fc(CH2NMe3)BMes2]þ (8) as the
iodide salt. Spectroscopic data are in line with the pro-
posed formulations for compounds 6-8, with crystal-
lographic confirmation being possible in the cases of 6
and 8 (the latter as the dichloromethane solvate;
Figure 1). Of note is the considerable anodic shift in the
redox potential for 8 (þ164 mV with respect to 1/1þ; þ314
mV with respect to ferrocene/ferrocenium), consistent
with the presence of an additional pendant cationic
fragment. Structurally, the borane fragments within Le-
wis acids 6 and 8 bear close resemblance to that found in
1. Moreover, there appears to be no contact in the solid
state between the tricoordinate boron center and the
iodide counterion in 8; the closest contact of I- to the
(23) These distances fall comfortably within the sum of the Van der Waals
radii for hydrogen and iodine (3.35 A): Emsley, J. The Elements; OUP:
Oxford, 1995.
(24) See, for example: Marr, G.; Moore, R. E.; Rockett, B. W. J. Chem.
Soc. C 1968, 24.
cationic component of 8 occurs via a C-H I interac-
3 3 3
tion (2.92 A) involving one of the methyl groups of the
NMe3þ unit. A similar pattern of intermolecular contacts
is observed in the solid-state structure of the isomeric

168 Inorganic Chemistry, Vol. 49, No. 1, 2010
Broomsgrove et al.
Scheme 4. Syntheses of 2-Functionalized Ferrocenyl Dimesitylboranesa
a Reagents and conditions: (i) nBuLi (1.0 equiv), diethyl ether, 0 to 20 ꢀC, then 12 h at 20 ꢀC, then Mes2BF (1.0 equiv), diethyl ether, -78 to þ20 ꢀC,
55%; (ii) MeI (ca. 2 equiv), hexanes, 20 ꢀC, 12 h, 70%; (iii) HCl (as a solution in Et2O, ca. 2 equiv), diethyl ether, -30 ꢀC, 30 min, 55%; (iv) bench (i.e., wet)
chloroform, 144 h, quant. by NMR; (v) tetrafluoroboric acid (2.0 equiv), diethyl ether, 20 ꢀC, 12, 60%.
such reactivity necessarily generates a tertiary ammonium
function and hence the potential for stronger host/guest
interactions via hydrogen-bond/Lewis acid cooperativi-
ty.5o In practice, however, the applicability of this
mode of anion binding appears to be limited by ready
B-C bond cleavage in the presence of a pendant
-CH2NMe2Hþ functionality. Thus, although the reac-
tion of 9 with dry HCl in diethyl ether does lead to the
isolation of [1,2-fc(CH2NMe2H)BMes2]þCl- (11) in ca.
55% yield (after recrystallization from dichloromethane/
hexanes), the corresponding reaction with (wet) tetra-
fluoroboric acid leads instead to the formation of the
hydrolysis product [1,2-fc(CH2NMe2H)B(Mes)OH]þ-
[BF4]- (13), which has been characterized by standard
spectroscopic/analytical techniques and X-ray crystallo-
graphy (see the Supporting Information). The presence of
a proximal -CH2NMe2Hþ group presumably activates
the (otherwise robust) B-Mes linkages to substitution via
nucleophilic attack at boron by water. Protonation by the
tertiary ammonium salt facilitates a loss of the mesityl group
as mesitylene (as demonstrated by 1H NMR moni-
toring of the reaction mixture), with the much more rapid
kinetics of hydrolysis observed for 11 (versus the corres-
ponding 1,10 isomer) being consistent with such a mechan-
ism. Complete hydrolytic decomposition of 11 is observed in
bench (i.e., wet) [D]chloroform over a period of 72 h, while
the analogous 1,10 disubstituted compound has a half-life
of ca. 10 days under identical conditions. Moreover, moni-
toring of the reaction of 11 with water over a longer time
frame by electrospray mass spectrometry is consistent with
the loss of both boron-bound mesityl substituents to give
[1,2-fc(CH2NMe2H)B(OH)2]þ. As such, the ready frag-
mentation of -NMe2Hþ functionalized receptors under
conditions other than those which are strictly nonaqueous
led to their abandonment as potential fluoride/cyanide
sensors. As it happens, the C-H bonds of proximal
CH2NMe3þ groups prove to be a more convenient source
of hydrogen-bond donors and exert a marked effect on
anion binding affinities, with little detriment to the hydro-
lytic stability of key B-C linkages (vide infra).
ii. Anion Binding. The propensity of the above range of
ferrocene-derivatized Lewis acids to interact with fluoride or
cyanide in nonaqueous media has been probed by NMR,
IR, mass spectrometric, electrochemical, crystallographic,
and UV-vis titration approaches with a view to (i) system-
atically probing the effects of borane electrophilicity, ancil-
lary functional groups, and net charge on the binding
of these anions and (ii) exploiting such systems in the
colorimetric sensing of fluoride and cyanide.
The binding of cyanide (as either KCN/18-crown-6 or
[nBu4N]þ[CN]- 2H2O)15 by the parent system FcBMes2
3
(1) in a range of solvents (chloroform, dichloromethane,
acetonitrile) can readily be demonstrated by a combina-
tion of spectroscopic techniques. Thus, the changes in 11B
NMR chemical shift (δB 76 to -16) and IR-detected
cyanide stretching frequency (2080 to 2162 cm-1) are in
line with previous reports of cyanide complexation to
boron-based Lewis acids.7 In addition, negative ion ESI-
MS sampling of the reaction mixture reveals a “flag-pole”
mass spectrum with an isotopic profile and measured
exact mass consistent with the formulation [1 CN]-.14
3
The thermodynamics of CN- binding by 1 can readily be
assessed by monitoring the intensity of the UV/vis band at
510 nm as a function of cyanide concentration. A binding
constant of 8.3(2.0) ꢀ 104 mol-1 dm3 can be determined
by fitting the resulting curve of absorbance versus
cyanide concentration in dichloromethane solution (see
the Supporting Information). This figure is significantly
less than that recently reported for a BMes2-derivatized
BODIPY system (5 ꢀ 107 mol-1 dm3),7e although the

Article
Inorganic Chemistry, Vol. 49, No. 1, 2010 169
Table 2. Binding Constants of Receptors 1, 3, 4, 5, 6, 8, and 10 for Fluoride and Cyanidea
receptor
KF/mol-1 dm3
KCN/mol-1 dm3
E1/2/mVb
FcBMes2 (1)
FcB(Xyl)2 (3)
7.8(1.2) ꢀ 104
4.4(0.5) ꢀ 105
4.3(0.7) ꢀ 105
6.6 (0.4) ꢀ 104
2.8(0.7) ꢀ 105
9.4(3.6) ꢀ 105
5.6(2.3) ꢀ 109
8.3(2.0) ꢀ 104
1.4(0.2) ꢀ 105
7.7(1.8) ꢀ 104
1.5(0.2) ꢀ 105
6.5(0.8) ꢀ 104
5.8(1.7) ꢀ 105
5.6 (2.4) ꢀ 109
þ131
þ153
þ184
þ95
FcB(XylF)2 (4)
FcB(XylOMe)2 (5)
1,10-fc(Br)BMes2 (6)
þ169
þ314
þ367
[1,10-fc(CH2NMe3)BMes2]þI- (8)
[1,2-fc(CH2NMe3)BMes2]þI- (10)c
a Conditions: dichloromethane, [receptor] = ca. 4 ꢀ 10-4 mol dm-3
.
b Measured in dichloromethane and referenced with respect to ferrocene/
ferrocenium. c Determined from competition experiments with PhBMes2.
more competitive nature of the dichloromethane reaction
medium (vs thf) together with the differing electronic
properties of the BODIPY/Fc substituents are presum-
ably contributory factors.25 Under conditions analogous
to those used in the cyanide binding experiments, 1 can
also be shown to bind fluoride, a competing affinity which
can readily be understood in terms of the known (high)
B-F bond strength26 and the comparable basicity re-
ported for F- (cf. CN-) in nonaqueous media (pKa’s in
DMSO: HF 15, HCN 13).13 The binding constant of 1 for
fluoride in dichloromethane solution determined from
UV/vis titration data [7.8(1.2) ꢀ 104 mol-1 dm3; see the
Supporting Information] is similar to those measured
previously for the related Lewis acids BMes3 and tris(9-
anthryl)borane [3.3(0.4) ꢀ 105 and 2.8(0.3) ꢀ 105 mol-1
dm3, respectively].5a,6r Moreover, although the binding
constants for 1 with fluoride and cyanide cannot be
separated within experimental error, stronger binding of
cyanide is implied by 11B NMR-monitored competition
experiments. Thus, cyanide will displace fluoride from
[FcBMes2 F]-, while [FcBMes2 CN]- is stable in the
With respect to the binding of fluoride, it is evident that
changes made at the para-substituent of the boryl sub-
stituent exert a relatively minor influence on KF. Thus
receptors 1, 3, 4, and 5 featuring para-Me, H, F, and OMe
substituents, respectively, give rise to binding constants of
7.8(1.2), 44(5), 43(7), and 6.6(0.4) ꢀ 104 mol-1 dm3. This
trend is broadly consistent with the electron withdrawing/
donating properties of the para substituents, as reflected
not only by their conventional Hammett parameters
(σp = -0.17, 0, 0.06, and -0.27, respectively)19 but also
in the measured Fe(II)/Fe(III) redox potentials, E1/2
(Table 2).28 Thus, 4 is oxidized at a potential of þ53
mV (with respect to 1), while the corresponding shift for 5
is -36 mV. Interestingly, although the binding constant
of 4 for F- is a factor of ∼5 greater than that of 1, the
affinity of the two receptors for cyanide is identical within
experimental error; indeed, the cyanide binding affinities
of receptors 1, 3, 4, and 5 do not appear to differ by more
than a factor of 2 across all four systems.
Relatively minor changes in the binding capabilities for
fluoride and cyanide are also brought about by the
introduction of additional functional groups at the 10
position of the ferrocene core. Thus, the values of KF and
KCN determined for 1,10-fc(Br)BMes2 (6), and even [1,10-
fc(CH2NMe3)BMes2]þ (8, as the iodide salt), are within
ca. 1 order of magnitude of those determined for the
parent receptor 1. The anion binding capabilities of the
isomeric system [1,2-fc(CH2NMe3)BMes2]þ (10, as the
iodide salt), however, show a marked enhancement, with
values of KF and KCN being determined [5.6(2.3) and
5.6(2.4) ꢀ 109 mol-1 dm3], which are (i) >3 orders of
magnitude greater than that for the 1,10 isomer 8,
(ii) comparable to the cyanide binding constant measured
for [4-Me3NC6H4BMes2]þ (KCN = 4 ꢀ 108 mol-1 dm3 in
water/DMSO),7c and (iii) compatible with the observed
capability of chloroform solutions of 10 to sequester
fluoride from aqueous solution.
3
3
presence of excess fluoride.27
The use of positively charged peripheral functional
groups has been reported to lead to marked enhancement
in the anion affinity of borane Lewis acids and to the
possibility of binding in protic media.7c With this in mind,
we have sought to determine the effect of this and other
factors on the binding of fluoride and cyanide by ferro-
cene-derivatized Lewis acids. The binding constants (for
both F- and CN-) of 1; the 4-functionalized 2,6-xylyl
boranes FcB(Xyl)2 (3), FcB(XylF)2 (4), and FcB-
(XylOMe
)
(5); the 1,10-difunctionalized ferrocþen-es fc-
2
(Br)BMes2 (6) and [1,10-fc(CH2NMe3)BMes2] I (8);
and the isomeric 1,2-disubstituted cation [1,2-fc-
(CH2NMe3)BMes2]þI- (10) have been determined by
UV-vis titration techniques, and the results are outlined
in Table 2 (see the Supporting Information for experi-
mental details). These allow some systematic conclusions
to be drawn concerning the effect of net charge, borane
substituent, and ancillary ligand electronics/substitution
pattern on the strength of F-/CN- binding.
While the very similar E1/2 potentials for 8 and 10 (þ314
and þ367 mV, respectively, with respect to ferrocene/
ferrocenium) reflect the additive electron-withdrawing
þ
effects of the CH2NMe3 and BMes2 groups on the
electron density at the ferrocenediyl core, the marked
differences in both fluoride and cyanide binding between
8 and 10 presumably reflect, at least in part, the differing
(25) An indication that binding constants for fluoride are typically higher
in thf than in chlorocarbon solvents comes from the fact that a value of KF
=
3.3(0.4) ꢀ 105 mol-1 dm3 has been determined for BMes3 in thf,5e while a null
response was observed for the same receptor/analyte combination in
chloroform5e or dichloromethane.
€
(28) Related studies have recently been reported by Gabba€ı and by Jakle
(26) A gas-phase fluoride affinity of 385 kJ mol-1 has been determined for
and Norton, detailing the effects on the reduction potentials of triarylbor-
anes of systematic replacement of mesityl groups with either C6F5 or 2,6-
Me2C6H2NMe3þ: (a) Cummings, S. A.; Iimura, M.; Harlan, C. J.; Kwaan, R.
€
BF3:Mallouk, T. E.; Rosenthal, G. L.; Muller, G.; Brusasco, R.; Bartlett, N.
Inorg. Chem. 1984, 23, 3167.
€
J.; Trieu, I. V.; Norton, J. R.; Bridgewater, B. M.; Jakle, F.; Sundararaman, A.;
(27) The reversibility of fluoride and cyanide binding by ferrocene
derivatized boranes could readily be established by the use of AlCl3 (in the
case of fluoride adducts) or HCl (cyanide adducts).
€
Tilset, M J. Am. Chem. Soc. 2006, 25, 1565. (b) Chiu, C.-W.; Kim, Y.; Gabbaı, F. P.
J. Am. Chem. Soc. 2008, 131, 60.

170 Inorganic Chemistry, Vol. 49, No. 1, 2010
Broomsgrove et al.
Figure 2. Proposed C-H F interaction in 1,2-fc(CH2NMe3)BMes2F.
3 3 3
þ
proximities of the cationic -NMe3 function to the
fluoride by a mixed borane/amide receptor.29 The addi-
-BMes2 binding site [in the solid state at least, the respec-
tive B N separations are 4.26 and 5.64 A]. A similar
tional C-H F-B interaction in 1,2-fc(CH2NMe3)-
3 3 3
BMes2F is thus presumably a factor in its enhanced anion
binding compared to its 1,10 isomer.
3 3 3
enhancement in Lewis acidity as a function of reduced
separation (and increasedelectrostatic interaction) between
a borane receptor site and a pendant cationic group has
been used to rationalize the increased fluoride binding
affinity of [2-Me3NC6H4BMes2]þ, over its 1,4-disubstitut-
ed isomer.7c An additional factor in the stronger anion
binding by 10 (over 8), which has recent literature
precedent, also stems from the close proximity of
Intriguingly, the cyanide binding constant measured
for 10 also appears to be very much enhanced compared
to that for 8; the binding constants of 10 for fluoride and
cyanide determined by titration methods are essentially
identical (Table 2). Moreover, this finding is also consis-
tent with the results of direct competition experiments.
Thus, if a dichloromethane solution containing 1 equiv of
fluoride and 1 equiv of cyanide (both as the [nBu4N]þ
salts) is added to a solution of 10, 11B NMR monitoring
reveals the presence of a mixture of the cyanide (δB -17
ppm) and fluoride adducts (δB 6 ppm). Moreover, the fact
that addition of excess cyanide to a solution of 1,2-
fc(CH2NMe3)BMes2F leads to the formation of 1,2-fc-
(CH2NMe3)BMes2CN (but the reverse reaction is not
spontaneous) implies that, if anything, cyanide binding is
a slightly more thermodynamically favorable process.
This situation contrasts with that observed for
[1,8-(Mes2B)(Me3NCH2)C10H6]þ, which binds fluoride
(in thf solution) ca. 3 orders of magnitude more strongly
than cyanide;a finding attributed to the greater steric
bulk of the cyanide anion and the sterically congested
binding cavity.7d That no similar steric effect on cyanide
binding is apparently observed with 10 presumably re-
flects a smaller degree of steric crowding at the binding
site, which in turn can be related to the geometric con-
sequences of ortho-substitution of the five-membered
cyclopentadienyl ring.
Anion binding studies with the closely related tertiary
ammonium system [1,2-fc(CH2NMe2H)BMes2]þCl- (11)
are not straightforward, being complicated by hydrolysis
of the mesityl B-C linkages in the presence of water (such
as that provided by hydrated sources of fluoride or
cyanide). As such, anion binding by these systems was
not pursued any further. Lability of the boron-bound
substituent(s) under anion complexation conditions also
proves to be a feature of the chemistry of FcB(MesF)F
(14). Thus, NMR monitoring of the reactions of 14
toward fluoride and cyanide sources are consistent
þ
the -NMe3 and -BMes2 groups.5k Chiu and Gabba€ı
have reported a naphthalenediyl system featuring -
þ
BMes2 and -CH2NMe3 substituents in the 1 and 8
positions which binds fluoride via both a conventional
Lewis acid/base interaction (at boron) and a C-H F-B
3 3 3
hydrogen bond utilizing one of the methylene hydrogens of
the CH2NMe3þ group. The presence of such a cooperative
binding motif in the solid state is manifested by a C
F
3 3 3
distance of 2.826(4) Afor the fluoride adduct, and by a 1JHF
coupling constant of 9.2 Hz in chloroform solution. In the
case of the fluoride adduct formed on the addition of excess
[(Me2N)3S]þ[Me3SiF2]- to a solution of 10 in [D]chloro-
form, the existence of an analogous (borane/hydrogen-
bond) cooperative binding mode (Figure 2) is signaled by
(i) an upfield shift in the 11B NMR signal from δB 80 to 5.7
1
ppm, (ii) a downfield shift in the H NMR resonance for
one of the diastereotopic methylene protons from δH 5.08
to 5.54 ppm, and (iii) the collapse, on broad-band 19F
decoupling, of the doublet of doublets coupling pattern
observed for the latter signal (1JHF = 6.6 Hz, 2JHH = 12.2
Hz; Figure 2) to a simple doublet. The magnitude of both
1
the coordination-induced shift in the methylene H reso-
nance and the 1JHF coupling constant itself are indicative of
a weaker C-H F interaction in the case of 1,2-
3 3 3
fc(CH2NMe3)BMes2F compared to 1,8-(Mes2FB)-
(Me3NCH2)C10H6;5k a more similar JHF coupling constant
(7.2 Hz) has, however, been determined for
N-H F-B interaction implicated in the binding of
a
3 3 3
(29) Hudnall, T. W.; Bondi, J. F.; Gabba€ı, F. P. Main Group Chem. 2006,
5, 319.

Article
Inorganic Chemistry, Vol. 49, No. 1, 2010 171
Figure 3. Molecular structures of the anionic component of [nBu4N]þ[1 CN]- CHCl3 and of [K(18-crown-6)]þ[X CN]- (X = 3, 4, 5, 15). Hydrogen
3
3
3
atoms (except that attached to C(1S)) and [nBu4N]þ counterion omitted for clarity; thermal ellipsoids set at the 40% probability level. Key bond lengths
and angles, for [1 CN]- CHCl3: B(1)-C(1) 1.639(4), B(1)-C(20) 1.621(3), N(1)-C(20) 1.150(3), N(1)-C(1S) 3.085, B(1)-C(20)-N(1) 169.8(3).
3
3
For [K(18-crown-6)]þ[3 CN]-: B(7)-C(2) 1.637(3), B(7)-C(16) 1.624(3), N(17)-C(16) 1.147(3), N(17)-K(31) 2.765, B(7)-C(16)-N(17) 170.2(2). For
3
[K(18-crown-6)]þ[4 CN]-: B(39)-C(34) 1.632(5), B(39)-C(40) 1.620(5), N(41)-C(40) 1.148(4), N(41)-K(101) 2.762, B(39)-C(40)-N(41) 171.4(3).
3 þ
For [K(18-crown-6)] [5 CN]-: B(12)-C(10) 1.633(4), B(12)-C(33) 1.621(4), N(34)-C(33) 1.148(3), N(34)-K(35) 2.755, B(12)-C(33)-N(34) 171.1(3).
3
For [K(18-crown-6)]þ[15 CN]-: B(7)-C(2) 1.635(4), B(7)-C(26) 1.624(5), N(27)-C(26) 1.147(4), N(27)-K(34) 2.789, B(7)-C(26)-N(27) 177.8(3).
3
with the formation of [FcBF3]- and [FcB(CN)3]-, respec-
tively.1d,30 Substitution of a methyl group for the boron-
bound fluoride to give FcB(MesF)Me (15) appears to
prevent such exchange processes;to the extent that
response is consistent with the conversion of a pendant
three-coordinate boryl Lewis acid to an anionic four-
coordinate borate and is mirrored by changes in electro-
chemical behavior. Thus, a cathodic shift of ca. -560 mV
is measured for 1 on the addition of cyanide {E1/2 = -383
[K(18-crown-6)]þ[15 CN]- can be isolated from the re-
3
action of 15 with KCN/18-crown-6;but the moisture
sensitivity of 15 precludes reliable measurements of bind-
ing constants using soluble fluoride/cyanide sources.
Structural authentication of the mode of anion binding
in the cases of [nBu4N]þ[1 CN]- CHCl3 and [K(18-
and þ181 mV for [1 CN]- and 1, respectively, in acet-
3
onitrile}, which mirrors the behavior of related ferrocene-
derivatized Lewis acids on coordination of bases such as
fluoride or trimethylphosphine.4a,w,18 For each of the
cyanide adducts, further supramolecular interactions
can be identified resulting from the residual Lewis basi-
city of the nitrogen atom of the coordinated cyanide
molecule. In the case of [nBu4N]þ[1 CN]- CHCl3, this
3
3
crown-6)]þ[X CN]- (X = 3, 4, 5, 15) has been obtained
3
by X-ray crystallography (Figure 3), and a consistent
pattern of geometric structure arises from analyses of
each of these adducts. Thus, an essentially linear C-bound
cyanide adduct is found in each case, with metrical
parameters for the BCN unit in agreement with previous
3
3
takes the form of an intermolecular hydrogen bond
between N(1) and the hydrogen atom of the chloroform
solvate molecule [d(N H) = 2.143 A; d(N C) =
3 3 3
3 3 3
reports of cyanide/borane complexes [d(B-C)
=
3.085(3) A; —C-N C = 151.5(2)ꢀ]. In the cases of
þ3 3 3
1.621(3), 1.624(3), 1.620(5), 1.621(4), 1.624(5) A,
—B-C-N = 169.8(3), 170.2(2), 171.4(3), 171.1(3)ꢀ,
177.8(3) for [1 CN]-, [3 CN]-, [4 CN]-, [5 CN]-, and
the [K(18-crown-6)] salts, an additional CN- to Lewis
acid interaction is observed between the cyanide nitro-
gen and the potassium counterion {e.g. d(K N) =
3
3
3
3
3 3 3
[15 CN]-, respectively].7c,e,31 No statistically significant
2.762, 2.789 A for [4 CN]- and [15 CN]-, respectively}.
The coordination geometry at the potassium center in
3
3
3
trend in either the B-C or C-N distances can be seen for
the series of para-substituted systems [1 CN]-, [3 CN]-,
[K(18-crown-6)]þ[4 CN]- is then completed by a second-
ary K O interaction [d(K O) = 2.812 A] invol-
ving one of the oxygen atoms from a neighboring
3
3
3
[4 CN]-, and [5 CN]-.
3
3
3 3 3
3 3 3
Significant elongation of the ferrocenyl B-Cipso bond
is observed on cyanide binding [d(B-Cipso) = 1.639(4),
1.632(5) A for [1 CN]- and [4 CN]-, cf., 1.546(7),
[K(18-crown-6)]þ moiety; a reciprocal K O interaction
3 3 3
involving O(65) then links the two adjacent units to give a
centro-symmetric supramolecular dimer (Figure 3). In
3
3
1.546(5) A for the free receptors]. The latter structural
the case of [K(18-crown-6)]þ[15 CN]-, by contrast,
the coordination geometry at each potassium center is
3
(30) Yao, H.; Kuhlman, M. L.; Rauchfuss, T. B.; Wilson, S. R. Inorg.
Chem. 2005, 44, 6256.
(31) Kuz’mina, L. G.; Struchkov, Y. T.; Lemenoksky, D. A.; Urazowsky,
I. F. J. Organomet. Chem. 1984, 277, 147.
completed by weak C-F K interactions [d(K F) =
3 3 3
3 3 3
3.033, 3.434 A] involving two of the fluorine atoms of the
para-CF3 group of an adjacent B(MesF) unit. These

172 Inorganic Chemistry, Vol. 49, No. 1, 2010
Broomsgrove et al.
Figure 4. UV/vis spectra of acetonitrile/methanol (>100:1) solutions containing Lewis acid receptors 2 or 17 (0.5 mM) and tetrazolium violet (1.0 mM) in
the absence (gray trace) and presence (black trace) of added anion: (a) 2 with F-, (b) 2 with CN-, (c) 17 with F-, and (d) 17 with CN-.
interactions fall comfortably within the sum of the van
der Waals’ radii of potassium and fluorine [4.22 A],23 and
thereby generating a colorimetric reporter response.4b
Potential candidates as oxidants include the tetrazolium
dyes, which in addition to offering a range of compatible
redox potentials give rise (on reduction to the correspond-
ing formazan) to very large changes in extinction coeffi-
cient at suitable wavelengths in the visible region of the
spectrum.32 Moreover, in contrast to previously reported
redox indicators,4b the conversion of a tetrazolium to a
formazan is a two-electron plus proton reduction (i.e., is
effectively hydridic) and under normal conditions is
irreversible. Hence, the color changes that characterize
fluoride or cyanide binding are rendered irreversible,
thereby converting the system from a sensor in the
classical sense to an effective dosimeter. By tuning
the electrochemical window determined by the receptor
and receptor/analyte complex, it has proved possible
to select redox-matched tetrazolium dyes which give
rise to an appropriate color change. Thus, 2 and
link adjacent [K(18-crown-6)]þ[15 CN]-, units into a
3
loosely bound one-dimensional coordination polymer in
the solid state.
iii. Colorimetric Anion Sensing/Dosimetry. Each of the
air-stable ferrocene-functionalized receptors 1, 3-6, 8,
and 10 has an approximately equal affinity in dichloro-
methane solution for fluoride to that for for cyanide
(Table 2). In the case of the parent system 1, a larger
binding constant for cyanide is implied by 11B NMR
monitored competition experiments. However, in terms
of developing a workable sensor system, the BMes2
complexes we have examined do not, in isolation, possess
the ability to differentiate between exposure to fluoride
and cyanide, although their interaction with other com-
peting anionic analytes is negligible. Discrimination be-
tween these two analytes can be achieved by the use of a
weaker Lewis acid receptor. Thus, boronic ester systems
related to 1 and 2 [i.e., FcB(OR)2 and Fc*B(OR)2, where,
for example (OR)2 = OC(H)PhC(H)PhO; 16 and 17]
have previously been shown to be selective for the binding
of fluoride.4w Hence, a two-component system can be
envisaged featuring, for example, 1 and 16, which uses
AND/NOT Boolean logic to distinguish between fluoride
and cyanide in nonaqueous solution.
[2 CN]- are oxidized at potentials (-176 and -691
3
mV, respectively) which are shifted ca. -300 mV cath-
odically with respect to 1/[1 CN]- and are thus compa-
3
tible with the use of tetrazolium violet as the redox
indicator (Chart 1).
The results of monitoring the exposure of 2 to either
cyanide or fluoride, in the presence of tetrazolium
violet, both by UV/vis spectroscopy and colorimetrically,
are shown in Figures 4 and 5. Thus, 2 is shown to
give a colorimetric response on exposure to both
fluoride and cyanide. Such a receptor/dye combina-
tion proves to be competent for visual detection
While the anion binding event can readily be moni-
tored by electrochemical measurements (e.g., by a shift
from þ181 to -383 mV on going from 1 to [1 CN]- in
3
acetonitrile), an attractive alternative involves the use of a
redox-matched dye which will oxidize an electron-rich
fluoride/cyanide adduct but not the “free” receptor,
(32) Nineham, A. W. Chem. Rev. 1955, 55, 355.

Article
Inorganic Chemistry, Vol. 49, No. 1, 2010 173
Conclusions
Synthetic approaches based on the direct borylation of
ferrocene by BBr3, followed by boryl substituent modification,
or on the lithiation of ferrocene derivatives and subsequent
quenching with the electrophile FBMes2, have given access to
a range of Lewis acids with which to conduct a systematic
study of fluoride and cyanide binding. In particular, the effects
of borane electrophilicity, net charge, and ancillary ligand
electronics/cooperativity have been examined. In this respect,
modifications made at the para position of the boron-bound
aromatic substituents exert a relatively minor influence on the
binding constants for both fluoride and cyanide, as do the
electronic properties of peripheral substituents at the ferroce-
nyl 10- position (even for cationic substituents). By contrast,
the influence of a CH2NMe3þ substituent in the 2- position is
found to be much more pronounced, reflecting, at least in part,
the existence in solution of an additional binding component
utilizing the hydrogen-bond donor capabilities of the methy-
lene CH2 group. While none of the systems examined in the
current study display any great differentiation between the
binding of fluoride and cyanide (and indeed some, such as 1,
bind both anions with equal affinity, within experimental
error), much weaker boronic ester Lewis acids will bind
fluoride (but give a negative response for cyanide). Thus, by
the incorporation of a suitable redox-matched organic dye, a
two-component dosimeter system can be developed capable of
colorimetrically signaling the presence of fluoride and cyanide
in organic solution by Boolean AND/NOT logic.
Figure 5. Colorimetric responses of receptor molecules 2 and 17 to the
additionofCN- (left-hand pair), fluoride(middle pair), andchloride(null
response control, right-hand pair) in acetonitrile/methanol (>100:1) in
the presence of tetrazolium violet.
Chart 1. Tetrazolium Violet
down to 25-40 nmol of analyte. In a similar fashion, the
weaker Lewis acid 17 can be shown to undergo a
ca. -580 mV electrochemical shift on fluoride binding
[e.g., from -169 to -749 mV in acetonitrile].4w In
combination with the same tetrazolium violet redox
dye, a colorimetric response is therefore generated on
exposure of 17 to fluoride. By contrast, a null response is
observed when excess cyanide is added to solutions of 17
in acetonitrile/methanol under identical conditions
(Figures 4 and 5).33 Thus, while the stronger Lewis acid
2 gives positive colorimetric responses to both cyanide
AND fluoride, 17 senses fluoride but NOT cyanide under
the same conditions.
Acknowledgment. The financial support of the EPSRC
and the Joint Grants Scheme is gratefully acknowledged, as
are the efforts of the EPSRC National Mass Spectrometry
and Crystallography Services. Drs. Nick Rees and Barbara
Odell are thanked for assistance with NMR measurements.
Supporting Information Available: Complete details of all
crystal structures (CIFs); details of the determination of all
binding constants listed in Table 2; spectroscopic data for
FcB(Mes)O(CH2)4Br. This material is available free of charge
via the Internet at http://pubs.acs.org.
(33) Although boronic acids are known to bind cyanide in aqueous
solution (see ref 9), electrochemical studies of 16 in an acetonitrile solution
suggest a significant cyanide binding event only for the oxidized ferrocenium
species [16]þ.