Nuclear magnetic resonance and mass spectra of organomercury hydridesand deuteridesPeter J. Craig,* Michael I. Needham, Naaman Ostah, Grace H. Stojak, Martyn Symons andPaul Teesdale-SpittleDepartment of Chemistry, De Montfort Univerdty, The Gateway, Leicester LEI 9BH, UKA series of new organomercury hydrides, including aromatic and fluoroaromatic compounds, have beencharacterised by mass and NMR spectroscopy. Unusual features of these compounds included an unusuallydownfield chemical shift for the Hg-H proton, confirmed by deuterium substitution, and a high value for the199Hg-H coupling constant. The half-life for dilute solutions of methylmercury hydride was in the region of1-3 h.We recently reported the existence and preliminary characterisa-tion of four new organomercury hydrides [HgH(Me), HgH(Et),HgH(Ph) and HgH(C,F,)].'. This series of compounds hadbeen thought to be unstable or transient in nature, possiblyoccurring as intermediates in the reduction of organomercurychlorides or acetates by metal hydride~.,-~ The reaction withNaBH, is the second stage of the solvomercuration-demercur-ation route to Markovnikov alcohols and ethers, etc7 Allworkers prior to 1992 had assumed that the organomercuryhydride species decomposed rapidly in solution according toequation (l).' Furthermore, this reaction had been generally4HgCI(R) + NaBH, + 40H- --+4RH + 4Hg + 4C1- + Na' + H2B03- + H,O (1)discussed in the context of a reduction in basic media.The onlyspeculation as to the possibility of a greater stability forHgH(Me) was that made by Devaud.'We have investigated the reduction of organometallic speciesby metal hydrides and similar reagents over many years as partof our work on the volatilisation and derivatisation oforganometallics in the environment prior to analysis byinterfaced gas chromatography (GC) mass spectroscopy (MS)and similar methods.' These reductions (e.g.by NaBH,) arenormally carried out in acid media, following an acid extractioncarried out to remove the organometallic species from asediment or biological matrix in the natural environment. In thedevelopment of a process for the analysis of HgCl(Me) andinorganic Hg" in the environment, we noticed that solutionsof reduced methylmercury species were stable enough forcharacterisation. The precursor mercury compounds [asHgCl(Me) or environmental methylmercury] were reduced atpH 4 in aqueous solution with aqueous NaBH, and diffusedupwards to an interface of a benzene layer on top of the water,where they dissolved and were identified by NMR and MS.233Alternatively, the hydrides may be purged from solution byhelium and condensed in a cold trap prior to direct transfer in aconnecting line to a quartz-furnace atomic absorption systemwhere the mercury atoms are detected.2 The hydride HgH(Me)has also been characterised independently by Baldi and co-workers.oIn this paper we report a series of 24 new organomercuryhydride and deuteride derivatives and discuss the unusualvalues observed for the Hg-H chemical shifts and the '"Hg-Hcoupling constants.These are correlated with the NMRparameters of some alkylmercury hydrides recently reported byKwetkat and Kitching. l 1Ex per imen t a1PreparationsThe hydrides and deuterides were prepared by the reduction ofthe corresponding organomercury chloride (0.8-1.2 mmol, usedas received) with NaBH, in acid solution as reported forHgH(Me).'q3 The general procedure was as follows. Thechloride HgCI( C6F4Me) (2.0 mg, 0.005 mmol) was dissolvedin a pH 3.5 acetate buffer (sodium acetate, BDH; acetic acid,Fisons; Millipore Q water) filled to the neck of a 50 cm3volumetric flask. To this C6D6 (Aldrich, 0.5 cm3) was added.Sodium tetrahydroborate (Aldrich, 4% in water, 1 cm3) wasadded by syringe and was allowed to stand for 15 min, theC,$6 layer was then removed, dried over magnesium sulfateand then submitted for NMR and GC-MS analysis.For the corresponding deuteride, sodium tetradeuterioborate(Aldrich) was used instead of sodium tetrahydroborate.GC-MS AnalysisA HP5890 gas chromatograph was coupled to a VG-Trio-3mass spectrometer. Separation was achieved using a CpSil-8CB capillary column, 10 m in length and 0.23 mm internaldiameter with a film thickness of 0.12 mm.The GC operatingconditions were as follows: initial column temperature 50 OC,rate of increase 20°C min--', final column temperature 220and injector temperature 100 "C. The mass spectrometer wasoperated in the electron impact (EI) mode at 70 eV (1.12 xlo-' J) and mass range of 35-650.NMR spectroscopyThe NMR spectra were obtained on a Bruker AC250 at 250.I3MHz using C6D, as the solvent at 303.3 K. Chemical shiftswere measured relative to residual protons in C6D6 at 6 7.16and are stated in ppm with respect to SiMe,.Results and DiscussionMass spectroscopyWe have established the identity and existence of the newhydrides or deuterides by MS methods and have studied severalin some detail. These identifications depend on thecharacteristic isotopic distribution within the mercury atom(Table 1). Mercury isotopes range in mass from 196 to 204, butwithin this distribution there are no isotopes at 197 or 203.Hence, the existence of peaks in the mass spectra at m/z 197 orJ. Chem. SOC., Dalton Trans., 1996, Pages 153-156 15203 suggests the presence of 196Hg-H or 202Hg-H.Similarly apeak at m/z 205 suggests the occurrence of '04Hg-H and one atm/z 206 '04Hg-D. The use of characteristic isotope patterns formercury, and also of characteristic isotope absences, is integralto the identification of the hydrides and deuterides. Thesearguments are developed further in the following inter-pretations.For identification of the presence of HgH(R) and HgD(R)(R = aryl) we chiefly utilised the presence of the highestmercury isotope ""Hg. For each postulated species we observethe HgH(R)+, HgD(R)+, HgH' and HgD' as ,04HgR + oneor + two mass units (Table 2). The compounds from which theHgH(R) or HgD(R) series were made [ i e . HgCI(R)] do not90 -80 -70 -60 -Table I Mercury isotopes *3330Mass Abundance (%)196 0.15198 10.10199 17.00200 23.1020 1 13.20202 29.65204 6.80* 1 9 7 ~ g and '03Hg do not occur.See text.50 -40 -30 -20 -10-0-,329. I , . ,60 -50 -40 -30-20-10-0,70 8ol I331 'OIHgD(R)+334'=HgD(R)+ 204Hg D (R)+/198HgR+ 330 I 337329.3323500 -N> z3000 -202HgH(R)+334t"+17'2 (+ 17)333I201HgR+332I204HgH (R)'336335 I335326 328 330 332 334 336 338m/zFig. 1 Mass spectrum of (0) HgH(C6F3H2-3,5) and (b) HgD(C6F3H,-3-5)show peaks at these m/z values. Decomposition of the singlespecies HgH(R) [or similarly HgD(R)] produces Hg'R ions aswell as the parent HgH(R)+ [or HgD(R)+]. Hence the mass-spectral cluster peaks in these areas are mixed although derivedfrom a single compound.Clearly the existence of HgH(R)+ orHgD(R)+ peaks (Fig. 1) demonstrates the existence of thehydride or deuteride respectively, as does the absence of spacesin the peaks owing to the intervals that would occur if the peakswere caused only by HgR+ or Hg+ (e.g. at m/z 197 or 203showing HgH' as 19'Hg and '03Hg do not exist). Peaks atm/z 205 ('04HgH+) and 203 (202HgH+), and 206 ('04HgDf)and 203 ('OlHgD'), are especially significant as there wouldotherwise have been spaces here. The hydrides HgH(C,H,Me-p ) and HgH(C,H,NH,-p) immediately dimerised to the diary1species in the GC-MS analysis and were observed as HgR,derivatives only. The aromatic fragments of the HgH(R) orHgD(R) species appear to fragment conventionally and are notdiscussed in this context.NMR spectroscopyFor identification purposes, the 'H NMR spectra are goodfingerprints, and confirm the structures of these derivatives(Table 3 and Fig.2). The downfield chemical shifts are large,with a clear gap between the alkyl and aryl derivatives. The1H-199Hg coupling constants are also large, increasing as theshifts decrease, with gaps between alkyl and aryl, and betweencyclohexyl and perfluoroalkyl ligands. We suggest the shiftsdepend on direct shielding and a local paramagnetic field fromelectrons on the mercury. Since the hydrogen ligand is expectedto carry a small negative charge rather than a positive charge,"direct shielding would be expected to induce an upfield shift.We therefore assign these shifts to the paramagnetic term, withaverage-induced fields that add to the applied field.These arelarge because of the large spin-orbit coupling constant formercury and the very large axial distortion caused by covalento bonding." The reduction on going to aryl substituentsprobably reflects the tendency for these ligands to participate inIT bonding, thereby increasing the effective energy gap betweenthe magnetically coupled orbitals.The spin-spin coupling in this case appears to be induced, instraightforward terms, via 'H-(electr~n),-(electron)~-~~~Hgnuclear coupling within the H-Hg o bond. The isotropic termstems largely from the local s-orbital characters of electrons 1and 2. In the limit of H- as a ligand this coupling is minimised.As the local o-electron density on the HgR unit increases so thecoupling increases.Hence, the more the group R removes o-electron density from mercury, the larger the coupling shouldbecome, as indeed is observed. Thus the large coupling isindicative of a strongly covalent HgR bond, whilst the trendtcb high values shown in Table 3 and Fig. 2 is a result ofthe increasing electronegativity of the ligands. This is even+l .O2.3 ++4 e 40001 +io154 J. Chem. Soc., Dalton Trans., 1996, Pages 153-1.5Table 2 Electron-impact mass spectra of organomercury hydrides and deuterides'H ydride DeuterideHgH(R)+/HgR + HgD(R)+/HgR+HgH(C6F4Me-p) Assign- Assign- HgD(C,F,OMe-p) Assign- Assign-(m/: 360-368 expected) ment HgH+/Hg+ ment (m/z 361-369 expected)' ment HgD +/Hg + ment362 (47) 198 (33)363 (80) 199 (62)364 (92) 200 (85)365 (75) 201 (55)366 (100) "'HgH( R) + 202 (100)368 (23) "'HgH( R) + 203 ( 1 6)204 (23)205 (4)H g H ( c, F4 B r-p 1(mi- 424434 expected)426 (2)427 (39)428 (67)429 (68)430 ( 100)432 (67)434 ( I 1)HgH( C6F4H-m)(m/z 346-354 expected)346 (2)347 ( 13)348 (55)349 (84)350 (96)351 (75)352 (100)354 (23)197 (7)198 (37)199 (65)200 (84)20 1 (50) 202Hg79Br( H)( R) + and 200H g 81 Br(H!(R)+202Hg81Br(H)(R) 202 (100)and 204Hg79Br(H;(R)+204Hg81Br(H)(R) 203 (6)204 (22)205 (2)197HgH(R)+ I98 (29)199 (66)200 (89)201 (67)'04HgH( R) + 203 (31)204 (20)205 (7)"'HgH(R)+ 202 (100)HgH(C6F3H2-3,5)(m/z 328-336 expected)328 (2)329 ( 1 3)330 (55)331 (88)332 (99)333 (90)334 ( 100) "'HgH(R)+335 (1 8)336 (23) 204HgH(R) +HgH(C,F,OMe-p)(m/z 376384 expected)377 (8)378 (36)379 (60)380 (80)381 (70)382 (100)384 (25)198 (28)199 (63)200 (88)201 (65)202 (100)203 (39)204 ( 17)205 (7)361 (5)362 (12)363 (47)364 (73)365 (98)2"2HgH' 366 (63)367 (100)204HgH+ 369 (24)lg6HgH ' 425 (4)427 (22)428 (37)429 (67)430 (58)431 (100)433 (65)"'HgH'""HgH'435 (1 2)HgD(C,F,H-m)(m/z 347-355 expected)347 (10)348 ( 15)349 (50)350 (69)351 (95)352 (49)353 (100)"'HgH' 355 (24)""HgH'96HgD(R)+ 198 (28)199 (48)200 (75)201 (52)202 (100)204 (45)201 HgD( R) + 203 (10) '"HgD+"'HgD( R) + 206 (5) '04HgD+198 (38)199 (54)200 (83)201 (50)202 (100)203 (5) "'HgD+202Hg* 'Br(D)(R)+ 204 (3 1)and204Hg79Br( D)(R)+204Hg81Br(D)(R)+ 206 (3) '04HgD+208 (-) Background198 (24)199 (37)200 (66)201 (52)202 (100)204 (52)"'HgD( R) + 203 (17) "lHgD+'04HgD( R) + 206 (8) '04HgD+HgD(C,F,H2-3,5)(m/z 329-337 expected)329 (1 2)330 (19)331 (55)332 (65)333 (100)"'HgH+ 334 (43)335 (87)'04HgH+ 337 (20)' 96HgD(R) +and 19*HgR+2o HgD( R)+'04HgD( R) +HgD(C,F,OMe-p)(m/z 377-385 expected)196 (20) 372 (-)198 (51) 378 (10)199 (90) 379 (72)200 (92) 380 (93)201 (61) 381 (91)"'HgH(R)+ 202 (1 00) 382 (100)204HgH(R)+ 203 (14) '02HgH+ 383 (95)204 (31)205 (2) '04HgH+ 385 (18)198 (24)199 (37)200 (68)201 (58)202 (1 00)203 (20) 'OIHgD+204 (69)206 (9) '04HgD+Background 196 ( 1 )198 (25)199 (39)200 (71)201 (52)202 (1 00)203 (15) "lHgD+204 (49)206 (8) '04HgD+2o HgD( R) +204HgD(R)+ 205 (1)a Percentage abundance is given in parentheses.m/: 196 and 360 are not detected; m/z 163, C,F4Me+. ' m/z 164, C6F4CH2D+.transmitted by the para substituents: for example, on goingfrom p-NH, to p-NO, there is a clear increase in the couplingconstant.It can be seen that coupling of aromatic-ring fluorine atomsto the proton bound to mercury should occur. It is alsoapparent that for polyfluoro species [e.g. HgH(C,F,)] whereas19F-'H coupling should (and does) occur, the situation iscomplex. The aromatic fluorine peaks in HgH(C,F,) areJ.Chem. SOC., Dalton Trans., 1996, pages 153-156 15complexity and overlapping of the predicted couplings makesdeconvolutions of the spectroscopic data somewhat intractable,although chemical shifts are as expected for the appropriateorganofluorine or aromatic grouping. The 19F-'H couplingconstants are of the order of 1.1 Hz. This coupling, through atleast two intervening atoms compares to much larger values forF-X-H of 50 Hz for some metalloidal elements.' We note thatthe presence of complex coupling between the organic group andthe proton bound to mercury demonstrates the interactionsbetween these species within the molecule, and among otherthings, the existence of the HgH(R) molecule as postulated. . The electronegativity of the organic ligands of HgH(R) canbe deduced from the partial charge that they induce on asubstituent. The partial charge on the hydrogen atom of R-Hwas used as a model.This was determined through molecular-orbital calculations using the AM 1 semi-empirical Hamilton-ian.I4 The partial charges are reported in Table 3 and arefound to correlate well with both H-Hg coupling constants andthe hydride chemical shift (Fig. 3).Continued NM,R spectroscopic observation of the hydridepeak at 20 "C in C6D, gave a half-life of 120 min for HgH(Me)in good agreement with ref. 10.Table 3 Proton NMR data for organomercury hydridesPartialHgH(R) ~ C H g f m I J(,,H)IHZ chargeHgH(C6F5) 11.9 4056 0.1822 HgH(C,F,Me-p) 12.3 3930 0.1763 HgH(C,F,OMe-p) 12.3 3930 0.1704 HgH(C,F,H-m) 12.4 3887 0.1765 HgH(C,H4Me-p) a a 0.120 .6 HgH(C6F4NH2-p) 12.6 3855 0.1767 HgH(C,F,OH-p) 12.3 b 0.1798 HgH(C6F4N02-p) 11.5 41 87 0.189IO HgH(C,F,HZ-2,5) 12.85 3743 0.15911 HgH(C6H4C0,Me-o) 13.0 3183 0.15712 HgH(C,H,CO,Me-p) 13.64 2986 0.13613 HgH(Me) 17.2 2650 0.06714 HgH(cH2C6Hll)' 17.3 2441 0.07215 HgH(CH2C6H11)' 17.6 2409 -16 HgH(C7H13)' 17.1 2308 0.07417 HgH(Ph) 13.3d 2936 0.13018 HgH(C,H,oOMe-o)' 17.5 2314 0.09019 HgH(C6HloD-I)' He, 16.1 2302 -9 HgH(C,F,Br-p) 11.9 4054 0.18120 Ha, 17.3 2307 -Not observed due to rapid decomposition to the HgR, derivative.* Not observed. ' Ref. 11. 6 14.1 in ref. 3.10810.16iIo.081a12"171810 I 3 $0'21;1612;13 II ' ~ 3---,2000 3000 4000 5000 11.5 13.5 15.5 17.5J (Hg-H)/Hz WgH(R)IFig.3 Correlation of the electron-withdrawing nature of the R groupof HgH(R) [measured by partial charge (relative to e = 1 x 1.602 xC) on H of R-HI with (a) J(Hg-H) and (b) 6[HgH(R)]compressed, with couplings existing for a given fluorine from twofurther adjacent fluorine nuclei, the proton bound to mercuryand also '"Hg. With regard to the 'H NMR spectra, in all casesthe proton occurs as a complex multiplet (demonstrating coup-ling with the organic group), with coupling occurring within amaximum band width of 20 Hz. For HgH(C,F,) this multipletappears to be an overlapping triplet of triplet of doublets system.Further investigations on the ' 3C and Ig9Hg NMR spectra are inprogress and will be reported at a later date.At present the smalldifferences in the chemical' shifts of the aromatic nuclei and theAcknowledgementsReceipt of a stipend from De Montfort University is grate-fully acknowledged (G. H. s.) as is support from the BritishCouncil ARC and Alliance Schemes (Germany and France)(P. J. C.). We particularly acknowledge the gift of precursorfluorinated aromatic organomercury chlorides from ProfessorG. B. Deacon, Monash University, Clayton, Victoria,Australia.References1 P. J. Craig, D. Mennie, N. Ostah, 0. F. X. Donard and F. Martin,Analyst, 1992,117,823.2 P. J. Craig, D. Mennie, M. Needham, N. Ostah, 0. F. X. Donardand F. Martin, J. Organomet. Chem., 1993,447, 5.3 P. J. Craig, H. Garraud, S. H. Laurie, D. Mennie and G. H. Stojak,J. Organomet. Chem., 1994,468,7.4 J. L. Wardell, in Comprehensive Organometallic Chemistry, eds.G. Wilkinson, F. G. A. Stone and E. W. Abel, Pergamon, Oxford,1982, vol. 2, p. 966.5 W. Kitching, A. R. Atkins, G. Wiekham and V. Alberts, J. Org.Chem., 1981,46,563.6 C. L. Hill and G. M. Whitsides, J. Am. Chem. SOC., 1974,%, 870.7 R. L. Larock, in Solvomercuration-Demercuration Reactions in8 M. Devaud, J. Organomet. Chem., 1981,22O,'C27.9 J. R. Ashby and P. J. Craig, Sci. Total Environ., 1989,78, 219.Organic Synthesis, Springer, Berlin, 1986.10 M. B. Filipelli, F. Baldi, F. E. Brinckman and G. J. Olson, Environ.Sci. Technol., 1992,26, 1457.11 K. Kwetkat and W. Kitching, J. Chem. SOC., Chem. Commun., 1994,345.12 M. J. Almond, D. A. Rice, L. A. Sheridan, P. J. Craig, G. Stojak,M. C. R. Symons and U. S. Rai, J, Chem. Soc., Faraday Trans.,1994,3153.13 C. J. Jameson, in Multinuclear NMR, ed. J. Mason, Plenum,New York, 1987, p. 437.14 M. J. S. Dewar, E. G. Zoebisch, E. F. Healey and J. J. P. Stewart,J. Am. Chem. SOC., 1985,107,3902.Received 10th July 1995; Paper 5/04489H156 J. Chem. SOC., Dalton Trans., 1996, Pages 153-15