Phenyl(trichloromethyl)mercury1

PhHgCCl3

[3294-57-3]  · C7H5Cl3Hg  · Phenyl(trichloromethyl)mercury  · (MW 396.06)

(transfers :CCl2 to many unsaturated systems, to heteroatoms, and to several s-bonds)

Physical Data: mp 117-118 °C.

Solubility: sol ether, benzene, chloroform.

Form Supplied in: colorless solid; commercially available.

Preparative Methods: prepared from PhHgBr either by reaction with Potassium t-Butoxide-Chloroform in benzene at 10 °C in 66% yield,2 or by treatment with Sodium Methoxide-Ethyl Trichloroacetate in 62% yield.3 Phenylmercury(II) Chloride furnishes PhHgCCl3 by reaction with Sodium Trichloroacetate in refluxing DME (65%).4 A two-phase approach from PhHgCl and CHCl3 using aqueous Sodium Hydroxide-Potassium Fluoride and small amounts of Benzyltriethylammonium Chloride furnishes PhHgCCl3 in 72% yield.5

Purification: recrystallization from chloroform.

Handling, Storage, and Precautions: flame-dried glassware and an atmosphere of dry nitrogen are used for the reactions. Organomercury(II) compounds should be carefully handled and skin contact and inhalation should be particularly avoided. All reactions must be conducted in a well-ventilated fume hood. The reagent is light sensitive and should be stored under nitrogen at 4 °C.

Cyclopropanation of Alkenes.

A wide and representative set of alkenes react with PhHgCCl3 to afford gem-dihalocyclopropanes in good yield (eq 1).6 Reactions are carried out under nitrogen, using dry benzene as solvent, and can be easily monitored by TLC analysis. When all the starting mercurial has been consumed, PhHgCl is filtered off, the solvent and the excess alkene are removed in vacuo, and the resulting cyclopropanes are generally purified by distillation. The simplicity of product isolation represents an important synthetic advantage of this procedure.

A distinctive feature of PhHgCCl3 as a cyclopropanating reagent is that it reacts with alkenes that are electron-poor (eq 2).6 In contrast, attempts to cyclopropanate tetrachloroethylene using t-BuOK-CHCl3 or thermal decomposition of Cl3CCO2Na give only modest (less than 10%) yields of hexachlorocyclopropane.7

PhHgCCl3 competes well in several aspects with PhHgCCl2Br as a :CCl2 transfer reagent for gem-dichlorocyclopropanation of alkenes; in particular, the former is a cheaper reagent and is more easily prepared. Conversely, the latter behaves as a milder source of :CCl2, requiring shorter reaction times that enhance its synthetic potential.6

A simple modification of the experimental protocol, involving addition of Sodium Iodide, has been developed. It decreases either the reaction time or the reaction temperature of the cyclopropanation with respect to that required when using a-haloorganomercurials.8 Following this procedure, trans-stilbene smoothly undergoes cyclopropanation in high yield (eq 3).9

Using the above-mentioned conditions, cyclohexene is cyclopropanated (3 h, 85 °C) in 91% yield, while the cyclopropane derivative is formed in only 15% yield in the absence of NaI. Interestingly, the NaI method allows cyclopropanation by PhHgCCl3 to take place at rt (eq 4).9

The activating effect of added NaI has been interpreted in terms of iodide attack at mercury and concomitant displacement of -CCl3, which generates in turn the dihalocarbene species (-CX3 -> :CX2 + X-). Consistent with that assumption, the reaction of NaI with equimolecular amounts of PhHgCCl3 in dry acetone at 25 °C yields products that arise from ketone trapping by CCl3-.9 The NaI-based procedure has some limitations. It reacts sluggishly with alkenes like tetrachloroethylene, reputed to be quite unreactive towards :CCl2. Also, in good agreement with the electrophilic nature of the intermediate, the observed trend of alkene reactivity in this process (R2C=CR2 > R2C=CHR > R2C=CH2 > RCH=CHR > RCH=CH2)9 matches well with that observed for the reactivity of carbenes generated by thermal treatment of Cl3CCO2Na in DME.

A major advantage of PhHgCCl3 in gem-dichlorocyclopropanation reactions is realized when dealing with base-sensitive alkenes. In these cases, conventional carbene generation by treatment of chloroform with strong bases does not afford significant cyclopropane formation.6 Many examples are known, and several synthetic applications have been reported. For example, 2-vinyl-4,4,6,6-tetramethyl-1,3,2-dioxaborinane is cleanly cyclopropanated with PhHgCCl3 (eq 5);10 cyclopropanation of this compound through thermal decomposition of Cl3CCO2Na as a source of :CCl2 gives the desired cyclopropane in only 4% yield.10

Diketene is also smoothly gem-dichlorocyclopropanated by PhHgCCl3 (eq 6).11 Interestingly, further reactions on the resulting dichlorocyclopropanespirolactone do not induce opening of the three-membered ring, in contrast with the behavior of the cyclopropanation product obtained by reaction of diketene with a-diazo ketones.11

Vinyl Acetate, a base-sensitive compound that cannot be cyclopropanated by traditional methods, reacts readily with trihalomercurial reagents,6 forming gem-dichlorocyclopropanes which are useful synthons leading to pyrazole rings (eq 7).12

In this respect, if the size of the cycloalkenyl acetate is reduced then the yield of metacyclophane goes down, and the 1,2-function becomes significant. Treatment of cyclohexenyl acetate with PhHgCCl3 and further reaction of the resulting gem-dichlorocyclopropane with Hydrazine yields exclusively 4,5,6,7-tetrahydroindazole (1,2-ring fusion) in 40% yield.13 2-Aminopyrimidines are also accessible by treating the a,a-dichlorocyclopropyl acetates with Guanidine (eq 8).14

Similarly, the enol acetate function of steroids reacts with PhHgCCl3 to give the corresponding cyclopropane derivatives; these are useful compounds for furnishing new skeletal rearrangements through ring-expansion reactions in basic media (eq 9).15

Depending on the strain of the system, ring-opening processes take place spontaneously under the reaction conditions, as illustrated in eq 10.16

Carbene transfer from PhHgCCl3 to 5-enyl acetate steroids occurs from the less-hindered side, the so-called a-face.17 Thus the stereochemical outcome is generally opposite to that observed when using the Simmons-Smith procedure, suggesting that precoordination of the reagent does not occur.17

The cyclopropane derived from norbornene and PhHgCCl3 also undergoes spontaneous ring expansion under the reaction conditions, affording 3,4-dichlorobicyclo[3.2.1]oct-2-ene.18 This sequence, cyclopropanation/thermal ring expansion, has been successfully applied to cyclophane synthesis by starting from substituted indenes19 and indoles.20 Recently, cyclopropanation of alkenes has been used for the preparation of 1H-cyclopropa[a]naphthalene in a multistep sequence (eq 11).21

In short, the merit of PhHgCCl3 as a :CCl2 transfer reagent to cyclopropanate alkenes is derived from experimental facts; namely, the carbene is generated under neutral conditions at moderate reaction temperatures. The absence of side reactions clearly underlines the aforementioned comments. The accepted mechanism for the thermally induced cyclopropanation of alkenes with PhHgCCl3 involves carbene generation from the a-halo organomercurial in the rate limiting step, followed by fast trapping of :CCl2 by the alkene. The body of knowledge derived from the above described reactions supports this assumption. Of special relevance in this context is the easy cyclopropanation of vinyl acetates that clearly rules out participation of the trichloromethyl anion; in fact, all attempts to cyclopropanate this base-sensitive alkene using t-BuOK-CHCl3 as carbene source were unsuccessful. Indirect evidence is provided by kinetic studies of related processes using PhHgCCl2Br as the carbene precursor.22 Further support for the proposed mechanism is obtained from gas-phase pyrolysis studies on PhHgCCl3,23 and subsequent characterization of the thermally generated carbene by matrix isolation techniques.24

Reaction with Methoxynaphthalene.

PhHgCCl3 adds to the C-3-C-4 double bond of 1-methoxynaphthalene to give, after ring expansion under the reaction conditions, 5-chloro-2,3-benzotropone (eq 12),25 a structure which was initially misassigned.25a

Reaction with Diiodoacetylene.

The addition of gem-dichlorocarbene (generated from PhHgCCl2Br) to different alkynes has been described.26 On this ground, some of the products of the reaction mixture obtained from PhHgCCl3 and IC&tbond;CI were mechanistically interpreted to arise from initial addition of dichlorocarbene to the diiodoacetylene. Using a 2:1 ratio of PhHgCCl3/IC&tbond;CI, 1-iodo-3,3,3-trichloropropyne and tetrachlorocyclopropene were formed as the major components of the reaction mixture (eq 13).27

Reaction with 1,3-Dienes.

1,1,2,2,3,3-Hexamethyl-4,5-bis(methylene)cyclopentane is an interesting 1,3-diene to check whether 1,4-addition is a feasible reaction path for a given transformation. Its reaction with PhHgCCl3 shows clearly that this possibility is a true alternative for a singlet carbene (eq 14).28

Reaction with Nitrogen- and Other Heteroatom-Substituted Compounds.

Tertiary amines react with PhHgCCl3 in boiling benzene; for instance, Et3N gives EtCl, PhHgCl, and N,N-diethyltrichlorovinylamine (Et2NCCl=CCl2) in low yield.29 Short reaction times have been noticed, suggesting an initial nucleophilic substitution reaction on mercury by the amine and -CCl3 displacement. Subsequent decomposition reactions would eventually lead to the nitrogen ylide (Et3N+-CCl2--CCl2) from which reaction products will be formed.29 These findings are consistent with the interesting behavior shown by several allylamines towards PhHgCCl3 (eq 15).30 The yield of cyclopropane increases as the nucleophilic character of the nitrogen decreases. For more nucleophilic nitrogen atoms, competitive enamine formation could be the major pathway.

The reaction of N,N-dimethylaniline with PhHgCCl3 has also been described.31 In this case, a novel class of compound is formed via electrophilic substitution on the starting amine, along with the expected products arising from an attack on nitrogen. Thermal reaction of PhHgCCl3 with the C=N double bond should offer a simple entry to aziridines. However, its synthetic utility is severely limited due to the presence of PhHgCl, formed in the reaction.32 To overcome this, cyclopropanation was attempted in the presence of NaI.9 Although remarkable examples have been recorded (eq 16), the procedure lacks generality since the PhHgI formed catalyzes aziridine ring opening at rt.32

The C=N function of azirines reacts with PhHgCCl3, giving ring-opened products (eq 17).33 A minor byproduct, identified as the corresponding N-vinylaziridine, results from further attack of :CCl2 on the C=N rather than the C=C bond of the azadiene.33

The reaction of oxygen-containing compounds with :CCl2 generated from a-halo organomercurials (PhHgXCCl2) is known, but most of the work deals with PhHgCCl2Br (reactions with ROH,34 RCO2H,35 RCHO,36 and RCOR37 have been studied). PhHgCCl3 attack at oxygen of the carbonyl group of several unsaturated steroidal ketones triggers skeletal rearrangements, and subsequent aromatization of the A ring.38 PhHgCCl3 is reported to deoxygenate pyridine N-oxides by carbene attack onto the oxygen atom to afford deoxygenated products.39

With respect to sulfur-containing substrates, sulfoximines react with PhHgCCl3 to furnish sulfur ylides in modest yields via nitrogen displacement.40 Of synthetic interest is the reaction of penams with PhHgCCl3, offering a nonbasic approach to the corresponding azetidinones via thiazolidine ring opening (eq 18).41 For these substrates, addition of NaI not only permits the use of milder reaction conditions, but results in a different type of product formation,41b probably due to -CCl3 acting as a base.

Insertion Reactions.

PhHgCCl3 easily inserts :CCl2 into the Si-H bond.42 In this respect, good yields of insertion product are obtained with silanes having bulky groups (eq 19).42a In contrast, carbenes generated following the t-BuOK-CHCl3 protocol fail to react with these silicon compounds. Optically active silanes have been used to study the mechanism.43 Since retention of configuration is mainly observed,43a the insertion process is seen to take place through a three-centered transition state.43b

Insertion of :CCl2, generated by thermal treatment of PhHgCCl3, into C-H bonds usually does not compete with the addition to C=C bonds in typical alkenes, although some examples have been described in steroidal systems.17 Several examples of insertion into Si-Br and Ge-Cl bonds are also known for PhHgCCl3.44


1. (a) Seyferth, D. PAC 1970, 23, 391. (b) Seyferth, D. ACR 1972, 5, 65. (c) Zeller, K.-P.; Straub, H. MOC 1974, 13/2b, 351. (d) Larock, R. C. Organomercury Compounds in Organic Synthesis; Springer: Berlin, 1985; Chapter X, pp 327-413.
2. (a) Reutov, O. A.; Lovtsova, A. N. DOK 1961, 139, 622. (b) Seyferth, D.; Burlitch, J. M. JACS 1962, 84, 1757. (c) Seyferth, D.; Burlitch, J. M. JOM 1965, 4, 127.
3. Schweizer, E. E.; O'Neill, G. J. JOC 1963, 28, 851.
4. Logan, T. J. OSC 1973, 5, 969.
5. Fedorynski, D.; Makosza, M. JOM 1973, 51, 89.
6. Seyferth, D.; Burlitch, J. M.; Minasz, R. J.; Mui, J. Y.-P.; Simmons, H. D., Jr.; Treiber, A. J. H.; Dowd, S. R. JACS 1965, 87, 4259.
7. Moore, W. R.; Krikorian, S. E.; La Prade, J. E. JOC 1963, 28, 1404.
8. Seyferth, D.; Mui, J. Y.-P.; Gordon, M. E.; Burlitch, J. M. JACS 1965, 87, 681.
9. Seyferth, D.; Gordon, M. E.; Mui, J. Y.-P.; Burlitch, J. M. JACS 1967, 89, 959.
10. Woods, W. G.; Bengelsdorf, I. S. JOC 1966, 31, 2769.
11. Kato, T.; Chiba, T.; Sato, R.; Yashima, T. JOC 1980, 45, 2020.
12. Parham, W. E.; Dooley, J. F. JACS 1967, 89, 985.
13. Parham, W. E.; Dooley, J. F. JOC 1968, 33, 1476.
14. Parham, W. E.; Dooley, J. F.; Meilahn, M. K.; Greidanus, J. W. JOC 1969, 34, 1474.
15. Crabbé, P.; Luche, J.-L.; Damiano, J.-C.; Luche, M.-J.; Cruz, A. JOC 1979, 44, 2929.
16. Rosen, P.; Karasiewicz, R. JOC 1973, 38, 289.
17. Bond, F. T.; Cornelia, R. H. CC 1968, 1189.
18. Jefford, C. W.; Hill, D. T.; Gore, J.; Waegell, B. HCA 1972, 55, 790.
19. Parham, W. E.; Egberg, D. C.; Montgomery, W. C. JOC 1973, 38, 1207.
20. (a) Parham, W. E.; Davenport, R. W.; Biasotti, J. B. TL 1969, 557. (b) Parham, W. E.; Davenport, R. W.; Biasotti, J. B. JOC 1970, 35, 3775.
21. (a) Müller, P.; Nguyen-Thi, H.-C. HCA 1984, 67, 467. (b) Müller, P.; Bernardinelli, G.; Gadoy-Nguyen Thi, H. C. HCA 1989, 72, 1627.
22. (a) Seyferth, D.; Mui, J. Y.-P.; Burlitch, J. M. JACS 1967, 89, 4953. (b) Seyferth, D.; Mui, J. Y.-P.; Damrauer, R. JACS 1968, 90, 6182.
23. Mal'tsev, A. K.; Mikaelyan, R. G.; Nefedov, O. M. IZV 1971, 199 (CA 1971, 75, 19 592q).
24. Mal'tsev, A. K.; Mikaelyan, R. G.; Nefedov, O. M.; Hauge, R. H.; Margrave, J. L. PNA 1971, 68, 3238.
25. (a) Saraf, S. D. S 1971, 264. (b) Ebine, S.; Hoshino, M.; Machiguchi, T. BCJ 1971, 44, 3480.
26. Seyferth, D.; Damrauer, R. JOC 1966, 31, 1660.
27. Cohen, H. M.; Keough, A. H. JOC 1966, 31, 3428.
28. Mayr, H.; Heigl, U. W. AG(E) 1985, 24, 579.
29. Seyferth, D.; Gordon, M. E.; Damrauer, R. JOC 1967, 32, 469.
30. Parham, W. E.; Potoski, J. R. JOC 1967, 32, 278.
31. Saraf, S. D. CJC 1969, 47, 1173.
32. Meilahn, M. K.; Olsen, D. K.; Brittain, W. J.; Anders, R. T. JOC 1978, 43, 1346.
33. Hassner, A.; Currie, J. O., Jr.; Steinfeld, A. S.; Atkinson, R. F. JACS 1973, 95, 2982.
34. Seyferth, D.; Mai, V. A.; Mui, J. Y.-P.; Darragh, K. V. JOC 1966, 31, 4079.
35. Seyferth, D.; Mui, J. Y.-P. JACS 1966, 88, 4672.
36. Martin, C. W.; Lund, P. R.; Rapp, E.; Landgrebe, J. A. JOC 1978, 43, 1071.
37. Seyferth, D.; Tronich, W.; Smith, W. E.; Hopper, S. P. JOM 1974, 67, 341.
38. Berkoz, B.; Lewis, G. S.; Edwards, J. A. JOC 1970, 35, 1060.
39. Schweizer, E. E.; O'Neill, G. J. JOC 1963, 28, 2460.
40. Iqbal, J.; Rahman, W. JOM 1979, 169, 141.
41. (a) Kang, J.; Im, W. B.; Choi, S.-g.; Lim, D.; Choi, Y. R.; Cho, H. G.; Lee, J. H. H 1989, 29, 209. (b) Kang, J.; Lim, D. S. SL 1990, 611.
42. (a) Weidenbruch, M.; Peter, W.; Pierrard, C. AG(E) 1976, 15, 43. (b) Lukevics, E.; Sturkovich, R.; Goldberg, Yu.; Gaukham, A. JOM 1988, 345, 19.
43. (a) Sommer, L. H.; Ulland, L. A.; Ritter, A. JACS 1968, 90, 4486. (b) Sommer, L. H.; Ulland, L. A.; Parker, G. A. JACS 1972, 94, 3469.
44. (a) Weidenbruch, M.; Pierrard, C. JOM 1974, 71, C29. (b) Weidenbruch, M.; Pierrard, C. CB 1977, 110, 1545.

José Barluenga, Miguel Tomás & José M. González

Universidad de Oviedo, Spain



Copyright 1995-2000 by John Wiley & Sons, Ltd. All rights reserved.