C4vCrw • GO CORNELL UNIVERSITY LIBRARY GIFT OF Dept. of ^heraestry CHEMISTRY LIBRARY Cornell University Library The original of this book is in the Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31924004036962 THE ORGANOMETALLIC COMPOUNDS OF ZINC AND MAGNESIUM BY HENRY WREN, M.A., D.Sc, Ph.D. Head of the Department of Pure and Applied Chemistry at the Municipal Technical Institute, Belfast NEW YORK D. VAN NOSTRAND COMPANY TWENTY-FIVE PARK PLACE 1913 EDITOR'S PREFACE The progress of Chemistry is so rapid that it is becoming a matter of ever-increasing difficulty to keep abreast of the modern developments of the science. The volume of periodical literature is so enormous that few can hope to read, far less assimilate, all that is published. The preparation of summaries has therefore become a necessity, and has led to the publication of various well-known journals devoted to the abstraction of original papers. For obvious reasons, however, these do not fully supply the wants of advanced students and research workers, and it is now generally recognised that monographs on special subjects are also needed. This series of monographs is intended primarily for Advanced and Honours students. As each mono- graph is written by an author with special knowledge of the subject, and copious references are given, it is hoped that the series will prove useful also to those engaged in research. January 19 1 3. CONTENTS SECTION I. PAGE Geneeal Notes on Geignaed's Reaction ... 1 A typical example — The use of Catalysts — Use of Solvents other than Ether — Use of Metals other than Magnesium — Mode of employment of Grignard's Eeagents. SECTION II. Peoducts Foemed by the Aid of Geignaed's Ee- agents 13 Saturated Hydrocarbons — Unsaturated Hydro- carbons — Iodo-Compounds — Alcohols — Second- ary Alcohols — Tertiary Alcohols — Glycols — Phenols, Thiophenols and Selenophenols — Ethers — Aldehydes — Ketones — Acids — Nitriles — Nitro- ' gen Compounds — Formation of Additive Com- pounds — Keducing action of Organomagnesium Compounds — Silicon Compounds — Action of Grignard's Reagents on various Inorganic Sub- stances. SECTION III. Theoeetical 72 Resume of various theories as to the constitution of Grignard's Reagents— Mode of Catalytic action of Ethers and Tertiary Amines. viii CONTENTS SECTION IV. PAGE Zmc Organometallic Compounds .... 80 General — Constitution of Zinc Organometallic Derivatives — Mode of employment of Organo- zinc Compounds — Normal action on Acid Chlo- rides — Formation of Tertiary Alcohols from Acid Chlorides — Formation of Ketones. bibliography 93 Index 99 The Organometallic Compounds of Zinc and Magnesium SECTION I. Geignaed's Reaction — Inteoductoey. In u 899, Barbier 1 found that when methylheptenone, (CH 3 ) 2 . C : CH . CH 2 . CH 2 . CO . CH 3 , was allowed to react with methyl iodide and magnesium in the presence of ether and the product so formed decomposed with dilute acid, dimethylheptenol, (CH 3 ) 2 . C : CH . CH 2 . CH 2 . C(0H)(CH 3 ) 2 , was formed. This observation suggested that the magnesium had reacted with methyl iodide to form magnesium methyl iodide, CH 3 MgI, which had then transformed the ketone into an alcohol analogously to the well-known action of zinc alkyls on ketones. E. Frankland and Wanklyn had previously observed the formation of the compound, Zn(CH 3 ) 2 . (C 2 H 6 ) 2 0, when zinc, methyl iodide, and ether were heated in a sealed tube. Starting from these known facts, Grignard 2 was led to investigate the action of magnesium on methyl iodide and similar com- pounds in the presence of anhydrous ether. Magnesium alkyl halides were thus obtained, which are classed generally under the name " Grignard's Reagents." A 2 GRIGNARD'S REACTION The preparation and use of Grignard's reagents are exemplified in the preparation of trimethyl- carbinol from acetone and magnesium methyl iodide. The following description is typical : — Preparation of the Ethereal Solution of Magnesium Methyl Iodide. — Twenty-five c.c. of a solution of methyl iodide (49 grams) in dry ether (25 c.c.) are allowed to flow on to magnesium (8 '2 grams) prepared by cutting clean magnesium ribbon into pieces of 1-2 cms. in length. A brisk reaction commences after a short period. The reaction mixture is cooled with water, two portions of ether (each. 100 c.c.) added, followed by the remainder of the methyl iodide solution, the latter being added drop by drop, cooling being adopted if necessary. To complete the reaction, the ethereal solution is heated to gentle boiling during thirty minutes. Action of Acetone on Magnesium Methyl Iodide Solution. — A solution of acetone (20 grams) in dry ether (20 grams) is added drop by drop to the cooled solution of magnesium methyl iodide. Each drefj causes a hissing sound and produces a precipitate which, at first, is rcdissolved, but subsequently collects as a greenish-grey viscous mass at the bottom of the flask. Keaction is completed by allowing the mixture to remain overnight at the room temperature. The product is decomposed by the addition of ice in small pieces, accompanied by brisk agitation. As soon as the colour of the precipitate is changed to white, sufficient acetic acid (33 per cent.) is added to brings about its A TYPICAL EXAMPLE 3 solution. The tethereal and the faintly acid aqueous layers are now separated. To isolate the trimethyl- carbinol — which is present in each layer — the ether is distilled off from the ethereal solution and the residual oil united with the acid aqueous layer and distilled with steam until a sample of the distillate does not become cloudy when saturated with potassium carbonate. The trimethylcarbinol is salted out from the distillate by means of potassium car- bonate, dried over solid potassium carbonate, and distilled. To complete the desiccation, the product is left in contact with barium oxide, at first at the room temperature, subsequently at its boiling- point, after which it is again distilled. The reactions involved may be represented by the following equations : — CH 3 I + Mg = CH 3 .Mg.I CH 3V CH 3V /OMgl >C = + CH Mg.I = Xq( CH/ CH/ \CH 3 CH,. /OMgl CH 3X /OH >C< +H.,0 = >C< +Mg(OH)I CH / \CH„ CH/ X?H, 2Mg(OH)I + 2HC 2 H 3 2 = MgI 2 + Mg(C 2 H 3 2 ) 2 + 2H 2 0. The above example is intended to show the most general method of preparation and use of Grignard's reagents. As will be seen later, under certain circumstances, the method needs to be modified by more careful regulation of temperature and by avoid- ance of the use of water or acid. For the prepam- i GRIGNARD'S REACTION tion of small quantities of the reagent, it is generally sufficient to mix the ether, magnesium, and all the alkyl or aryl halide in a flask provided with a reflux condenser. Experimental Details. The following points are of general interest : — " It is essential that all apparatus and materials employed should be thoroughly dry, and that all access of moisture and carbon dioxide to the reacting mixture should be avoided. This may be sufficiently assured by providing the upper end of the condenser with a tube, one half of which is filled with calcium chloride, the other half with soda-lime. It should be further noted that the reagent is not entirely insensitive to the action of oxygen of the air. For this reason, some chemists prefer to exclude the latter by the passage of a slow current of dry hydrogen through the reaction mixture. This pre- caution, however, is by no means universally adopted. Magnesium may be conveniently employed either in the form of turnings or ribbon. The latter should be thoroughly rubbed with fine emery-paper to remove the superficial coating of oxide, wiped clean, and cut into lengths of 1-2 cms. The filings should be cleansed from grease by treatment with alcohol and ether, and dried by heating in the steam-oven. The following method of purifying and drying commercial ether has been found to yield satisfactory results. The ether (500-1000 c.c.) is shaken with three successive portions (each about 100 c.c.) of dilute sulphuric acid, washed with water, and dried USE OF CATALYSTS 5 over plenty of calcium chloride during two or three days, after which it is filtered, allowed to stand in contact with sodium wire until evolution of hydrogen ceases, and then distilled. The distillate is preserved over a small quantity of sodium wire in a bottle provided with a chloride of calcium tube. Im- mediately previous to use it may be distilled over phosphorus pentoxide, but this latter precaution is not always necessary. Dryness of the ether is essential to success. The presence of the minutest trace of water betrays itself by the formation of a White cloudiness at the commencement of a Grignard's reaction. The alkyl (aryl)halide used should be carefully dried. Use of Catalyst to Expedite the Formation of Grignard's Reagents. — It is generally found that' aliphatic bromides and iodides attack magnesium < with great readiness, whereas aryl bromides and iodides as well as aliphatic chlorides of complex structure sometimes react with difficulty or not at all. In the latter case, addition of a trace of iodine or of methyl iodide is frequently sufficient to start the reaction, which then proceeds without difficulty. The influence of iodine in the case of chlorides is attributed by Wohl 3 to the formation of magnesium iodide, which then converts the chlorides into the more reactive iodides according to the following -bromo- dimethylaniline, when the magnesium was attacked by the latter. Baeyer G raised the objection that the reaction was not quantitative and the reagent was contaminated with magnesium ethyl bromide. To avoid this, he proposed to " activate " the magnesium by covering it with a thin film of magnesium iodide, which could be effected by heating magnesium in portions of about 10 grams, with continuous shaking, over a free flame and adding half its weight of iodine in small portions, and in such a manner that a new portion was not added until the whole of the previous portion had been consumed. The temperature must be high, but not sufficient to cause the mass to melt. With the given quantity of magnesium, the operation could be completed in a quarter to half an hour. The " activated " magnesium so obtained was a dull grey powder which became brown in course of time and required to be carefully protected from moisture. It reacted readily with an ethereal solution of o- m- or £>-iodoaniline or -dimethylaniline. Again, Hesse 7 failed to obtain an approximately quantitative yield of the rffagriesium compound of USE OF CATALYSTS 7 pinene hydrochloride when iodine, aluminium chloride, alkyl halides, etc., were used as catalysers. The method of Ehrlich and Sachs (see above) gave, at first, no better results, but when modified in such a manner that the solution of pinene hydrochloride was added to the magnesium whilst it was reacting vigorously with a readily attackable alkyl halide, led to the formation of an 80-85 per cent, yield of the desired magnesium compound. Hesse concludes that, in the reaction studied by him, the magnesium alkyl halides and not the alkyl halides themselves are the actual catalysers. 8 / It has occasionally been observed that the ether may contain substances which retard the formation of Grignard's reagent, i.e. negative catalysers. Bischofi 9 found that the interaction of magnesium powder and ethylene dibromide in ethereal solution was retarded by the presence of phenetol or ethyl acetate, or of small quantities of acetone, acetophenone, benzo- phenone, ethyl oxalate, ethyl malonate, or ethyl succinate, even in the presence of iodine. Larger quantities of the latter six substances prevented reaction. Ahrens and Stapler 10 found that the method of desiccation of the ether exerted consider- able influence on the course of the interaction of the same two substances. This is in accord with the observation of Freundler and Damond, 11 that the violence of the interaction of trioxymethylene and magnesium sec. butyl bromide in ethereal suspension can be diminished by the addition of a few drops of carbon disulphide. JJ.se of Sfolveiits %>ther than Ether. — Although 8 GRIGNARD'S REACTION ether is by far the commonest solvent employed in the preparation of Grignard's reagents, its use is by no means essential. Spencer and Stokes 12 found that when certain aryl halides, such as iodobenzene, ^-iodotoluene, m-bromoaniline, p-bromophenol, or a-bromonaphthalene, are heated at their boiling-points with magnesium powder, an energetic action occurs, and that the products so obtained, according to their reaction with water, can be represented by the general formula R . MgHal. Similarly, Spencer and Crewdson 13 showed that aryl chlorides and all the lower alkyl halides up to the butyl derivatives only react with magnesium when heated to about 270° in a sealed tube during several hours, whilst alkyl halides higher in the series than the butyl deriva- tives react with magnesium when they are heated with it at their boiling-points for a few minutes. The reaction is regarded as taking place along the lines indicated by the equations : R.Hal + Mg = R. MgHal. 2R.Hal + Mg = R-R + MgHal 2 . In addition to these compounds, however, un- saturated hydrocarbons are always found in the gaseous portion of the reaction products obtained from all interactions effected in sealed tubes. These unsaturated hydrocarbons, containing both olefines and acetylenes, are probably due to the action of heat on the original halides or on the initial products of the reaction. 14 Similar experiments are recorded by Spencer. 15 Tschelinzew, 16 in agreement with previous observa- EFFECT OF THE SOLVENT 9 tions of Bruhl 17 and Malmgren, 18 showed quite gener- ally that magnesium alkyl salts can be obtained by heating magnesium with alkyl halides to a sufficiently high temperature in the presence of a variety of ^solvents, notably of xylene. In these cases, however, the reaction between the metal and alkyl halide is greatly facilitated by the addition of a trace of diethyl ether, anisole, or of a tertiary amine such as dimethylanilin©{ which appears to play the part of catalyserV Tschelinzew, indeed, proposed to prepare Grignard's reagents by the interaction of magnesium and alkyl halide in benzene solution in the presence of a trace of dimethylaniline, addition of a minute quantity of iodine being often useful in starting the reaction. In the hands of other investigators, how- ever, the method has not led to satisfactory results. 19 The so-called "individual" magnesium alkyl salts thus obtained (so named because, unlike the Grig- nard's reagents prepared in ethereal solution, they do not contain combined solvent) form loose, white masses, readily soluble in ether or in a mixture of ether and benzene. They decompose at a high temperature without melting. Their isolation is, in general, quite unnecessary when they are to be used for synthetic purposes, since reaction can generally be brought about in ethereal or hydrocarbon solution. This may be regarded, indeed, as one of the great advantages of Grignard's reagents. The Quantitative Formation of Grignard's Re- agents. — It need scarcely be pointed out that calculated quantities of alkyl (aryl) halide and magnesium should be employed, and that the reaction 10 GRIGNARD'S REACTION / should be brought as nearly as possible to completion. Solution of all the magnesium can rarely be secured if the calculated quantity of halide is employed, but, in most cases, the quantity of residual metal should be very small. Replacement of Magnesium by other Metals.-*- Beckmann 20 has shown that finely divided calcium reacts with an ethereal solution of iodobenzene or ethyl iodide, with the formation of compounds of the formula C 2 H 0Xo/ R C„u/ ^Cal which resemble the Grignard reagents, whilst Spencer and Wallace 21 observe that aluminium, thallium, indium, and lithium form complexes, when heated with alkyl or aryl halides, which yield hydrocarbons when treated with water. These complexes are not formed, however, in the presence of ether. Zeltner 22 finds that the capacity of metals to form organo- metallic derivatives, which commences with the metals of the first group of the Periodic System, reaches its maximum in the metals of the second group, and diminishes in the metals of Groups III., IV., and V. The elements of Groups VI. and VII. show no tendency to form organometallic compounds, but this tendency reappears in the transition Group VIII. The stability of organometallic compounds of metals of the same group increases with the atomic weight of the metal and the size of the organic radical. Mode of Employment of Grignard' s Reagents. The usual procedure in employing Grignavrl's reagents DIRECT REACTION 11- is that already exemplified in the preparation of trimethylcarbinol, and divides itself into two parts : (1) preparation of the reagent ; (2) use of the reagent so prepared. Da vies and Kipping 23 point out that this method is not invariably necessary, that it generally involves the use of considerable volumes of solvent, and, frequently, waste of the halogen compound employed on account of the formation of by-products. The method originally employed by Barbier 2i — in which the organomagnesium compound is not separately prepared — frequently gives excellent results, and is, in certain cases, preferable to that usually adopted. Thus, Jaworsky 25 has prepared unsaturated alcohols by treating ketones with magnesium and allyl bromide or iodide in the presence of ether, although magnesium allyl halides of the usual type are unknown. Analogous observa- tions have been made by Oddo, 26 who finds that similar reactions can be carried out in the presence of an inert solvent such as benzene which has been freed from all traces of ether or tertiary base, and under conditions in which magnesium exerts no action on the alkyl iodide alone. Thus, in the presence of benzene, dimethylethylcarbinol is obtained from acetone, ethyl iodide, and magnesium. Scope of the Reaction. — In the following section, various syntheses effected by means of the Grignard reagents will be treated. In the allotted space, it is impossible to enumerate all the compounds pre- pared by means of these reagents, but it is hoped to deal with the more important classes of syntheses which have been accomplished by their aid. The 12 GRIGNARD'S REACTION behaviour of magnesium alkyl compounds will be seen, in general, to strongly resemble that of the \>ine alkyls. Over these latter compounds the former possess the great advantages that they are more readily prepared, are not spontaneously in- flammable, do not usually require to be isolated in the pure state, and have greater reactivity, doubtless due to the more electropositive character of the metal. SECTION II. The Products formed by the Aid of Grignard's Reagents. Saturated Hydrocarbons. When water is added to a Grignard's reagent, an energetic action ensues whereby a saturated hydro- carbon is formed : CH 3 .MgI + H 2 = CH 4 + Mg(OH)I. Dilute acids, 27 ammonia or amines, 28 bring about the same result : C 2 H 5 .MgI + NH 3 = C 2 H 6 + Mg(NH 2 )I, as do also alcohols, and, in general, substances which contain the hydroxyl group : CH 3 MgI + C 2 H 5 OH = CH 4 + Mg(OC 2 H 6 )I. From many points of view, the most convenient agent for bringing about this change is dry powdered ammonium chloride : 29 2CH 3 MgI + NH 4 Cl = 2CH 1 + NH 2 MgI + MgICl. The evolution of methane, which occurs when magnesium methyl iodide is mixed with a substance containing the hydroxyl group, has been employed for the detection and estimation of the latter in organic substances. 30 For this purpose, Zerewitinoff prepares a solution of magnesium methyl iodide in dry amyl ether to which pyridine is added, since amyl ether has insufficient solvent action upon many substances containing the hydroxyl group. A known 13 14 GRIGNARD'S REACTION weight of the substance under examination is treated with an excess of the reagent and the volume of methane evolved is measured. Reaction must be carried out at the ordinary temperature and with all reasonable speed, since pyridine on long standing, more quickly on heating, reacts with evolution of a gas. The same reaction has been applied 31 to the estimation of active hydrogen generally in organic compounds, i.e., to hydrogen contained in the sulphydryl, imido-, and amido-groups, and in certain tautomeric substances, and also 32 to the estimation of small quantities of water in substances such as coal, starch, etc. It is worthy of notice that allelotropic substances frequently behave towards Grignard's reagents as if they were entirely hydroxylic in structure. McKenzie 33 has shown that memthyl acetoacetate reacts according to the formula, CH 3 .C (OH) : CH . COOC 10 H 19 , and not as CH 3 . CO . CH 2 . COOC 10 H 19 . In the. formation of Grignard's reagents a second reaction — analogous to the Wtirtz synthesis of hydro- carbons from sodium and alkyl halides — occurs, by which hydrocarbons, formed by the union of the two alkyl radicles of the alkyl halide, are produced : C 2 H 6 I + Mg + I.C 2 H fi = MgI 2 + C 2 H 6 .C 2 H 5 . This "synthetic" action only takes place to a small extent in the case of halides of the lower members of the paraffin series. The quantity of by-product increases with increasing molecular weight of the halide until, with the hexyl iodides, it has become SATURATED HYDROCARBONS 15 the main product of the reaction. Similarly, diphenyl is invariably formed when magnesium acts upon phenyl bromide or iodide : 2C 6 H 6 Hal + Mg = C e H 5 . C 6 H 5 + MgHal 2 . This " synthetic " action of the magnesium may sometimes be considerably eliminated by careful regulation of the temperature during the formation of the reagent. 34 An interesting study of the action of magnesium on dihalogen compounds has been made by v. Braun and Sobecki. 35 Inspection shows that from a normal dihalogen compound of the type Br(CH 2 ) B Br, a variety of products might be expected; for, on the one hand, if the metal acts merely by the removal of halogen, polymethylenes of the structure L (CH 2 yl or L(CH 2 )„-(CH 2 )J might occur in the final product, whilst, on the other hand, open chain derivatives, such as BrMg(CH 2 ) re MgBr, or BrMg(CH 2 ) ;1 - (CH^MgBr, might result. With a<5-dibromobutane and the higher members of the series, substances of the latter two types are produced, thus : Br(CH 2 ) 4 Br + 2Mg = BrMg(CH 2 ) 4 MgBr BrMg . (CH 2 ) 4 MgBr + 2H 2 = CH 3 .CH 2 .CH 2 .CH 3 +2Mg(OH)Br and Mg + Br(CH 2 ) 4 ; Br + Mg + Br': (CH 2 ) 4 Br + Mg = BrMg(CH 2 ) 4 - (CH 2 ) 4 MgBr + MgBr 2 BrMg . (CH 2 ) 4 - (CH„) 4 . MgBr + 2H 2 = H(CH,) 4 -(CH,) 4 H + 2Mg(OH)Br. 16 GRIGNARD'S REACTION ae-Dibromopentane yielded pentane, Ti-decane, penta- decane, CH 3 (CH 2 ) 13 CH 3 , and higher homologues; a^-dibromoheptane yielded w-heptane and tetra- decane, CH 3 . (CH 2 ) 12 . CH 3 , whilst a /c-di-iododecane yielded decane, eicosane, CH 3 (CH 2 ) 18 CH 3 , and tetra- contane, CH^CH^CHg. Tissier and Grignard, 36 on the other hand, have shown that ethylene is the sole product of the action of ethylene dibromide, and magnesium, whilst Zelinsky and Gutt 37 found that ay-dibromopropane and magnesium yielded almost entirely a mixture of trimethylene and propylene with only minute quantities of the magnesium compound of hexamethylene dibromide, BrMg . (CH 2 ) 6 . MgBr. Gomberg and Cone 38 have succeeded in replacing the chlorine atom in triphenylmethyl chloride by a variety of alkyl groups, by the action of a solution of the former in benzene upon an ethereal solution of the organomagnesium salt. Reaction occurs in accord- ance with the equation : (C H 5 ) 3 C.Cl + R.MgCl = (C 6 H $ ) g C.R + MgCl r Similarly, Werner and Zilkens 39 have obtained hydrocarbons by the interaction of methyl sulphate and magnesium alkyl halides : (CH 3 ) 2 S0 4 + R . MgCl = R . CH 3 + CH 3 (MgCl)S0 4 . And Houben 40 has in certain cases obtained analogous results by the substitution of alkyl halide for methyl sulphate : R.MgBr + R'Br = R.R' + MgBr 2 . Ethyl benzene, toluene, and ^-xylene were obtained in this manner. UNSATURATED HYDROCARBONS 17 Unsaturated Hydrocarbons. Eeference to the production of ethylene and propylene from ethylene dibromide and ay-dibromo- propane respectively has already been made (see above). Alcohols are the normal products of the interaction of Grignard's reagents with aldehydes and ketones. Many of these readily pass by loss of water into unsaturated hydrocarbons, whilst some are so un- stable that they cannot themselves be isolated. Thus, by the action of magnesium benzyl bromide upon benzaldehyde, Hell 41 obtained phenylbenzyl- carbinol, which, when distilled, eliminated water with the formation of stilbene : Mg + C 6 H 5 CH 2 .Br = C fi H 5 CH 2 . MgBr H /^ C 6 H 5 . C-/ + C^H 5 CH 3 . MgBr = C 6 H 5 . C-0 . MgBr CH, . C fi H s J 2 / H , / H C 6 H 5 .C— OMgBr +H 2 = C 6 H 5 .C-OH + Mg(OH)Br \CH 2 .C 6 H 5 X CH 2 .C 6 H 5 C 6 H 5 . CH (OH) . CH 2 . C 8 H 6 = C 6 H 5 . CH : CH . C 6 H 5 + H 2 0. Whilst magnesium benzyl chloride and anisaldehyde yielded j?-methoxystilbene, CH 3 .C 6 H 4 .CH : CH .0 6 H S , as immediate product. Similarly, Hell and Stock- mayer obtained anisylphenylpropene, CrigO . C fi ri 4 ; Cxi . CHj, >• B 18 GRIGNARD'S REACTION as immediate product of the action of magnesium ethyl iodide upon anisylphenylketone. Analogous results are recorded by Hell and Wiegandt 42 and by F. and L. Sachs. 43 A similar interesting application of this reaction is found in the formation of A'' 8|,| -p-menthadiene, which was obtained by Kay and Per kin 44 by the action of excess of magnesium methyl iodide on optically active ethyl- A s -tetrahydro-^-toluate, /CH„ — CH C . COOC 2 H 5 . * >S CH 2 — CH/ As intermediate compound, A 3 -|>menthenol, was formed, CH 2 -CH . / CH 3 ch„ . ch< >c . c— ch 3 \ch 2 -ch/ \ oh which, when left in contact with an ethereal solution of magnesium methyl iodide at the ordinary tempera- ture, lost water, yielding the optically active, un- saturated hydrocarbon, A s ' 8(9) -£>-menthadiene, /CH.,' — CH ^ /CH 3 CH 3 . CH . c \C . C/ ^CH.,— CH/ ^CH„ In certain cases it has been found possible to control similar reactions in such a manner that either alcohols or unsaturated hydrocarbons may be pro- duced at will. Grignard 46 showed that phenyl- dimethylcarbinol, C 6 H 5 . C(OH)(CH 3 ) 2 , is produced when magnesium methyl iodide (1 mol.) reacts with acetophenone (1 mol.). If, however, the reagents are UNSATURATED HYDROCARBONS 19 taken in the proportion of 2 molecules of the former to one of the latter, the ether distilled off after completion of the initial reaction, and the residue then heated at 100° during several hours, metho-(l')-vinyl benzene, C (i H 5\ >C:CH,, CH/ is obtained. Tiffeneau K has succeeded in preparing allyl benzene by the action of magnesium phenyl bromide on allyl bromide, C 6 H 5 MgBr + BrCH, . CH : CH 3 = C 6 H 5 .CH 2 .CH:CH 2 + MgBr 2 , whilst, according to Eesseguier 47 allylcj/cZohexane results from the action of magnesium cycldhexyl bromide on allyl iodide. The general applicability of the reaction has been demonstrated by v. Braun, Deutsch, and Schmatloch, 48 who have obtained unde- cylene, C 9 H 19 . CH : CH 2 , from octyl bromide ; unde- cadiene, CH 2 : CH . (CH 2 ) 7 . CH : CH 2 , from ae-di-iodo- pentane; decadiene, CH 2 : CH . (CH 2 ) 6 . CH : CH 2 , from ct(S-di-iodobutane ; phenylamylene, C H 6 (CH 2 )3 . CH : CH 2 , from phenylethyl bromide ; phenylhexylene, C a H 5 . (CH 2 ) 4 . CH : CH 2 , from phenylpropyl bromide; and phenyloctylene, C 6 H 5 .(CH 2 ) 6 .CH:CH 2 , from phenylamyl bromide. According to Oddo, 49 the interaction of magnesium acetylene bromide and water gives acetylene in small yield : HC- CMgBr + H 2 = HC ; CH + Mg(OH)Br. 20 GRIGNARD'S REACTION Iodo Compounds. v. Braun and Deutsch 50 have succeeded in replacing bromine by iodine by the action of free iodine upon the magnesium alkyl bromide. Thus, < CH/ \MgHal, which, after removal of the solvent, becomes trans- formed into the isomer, R . CH 2 . CH 2 . O . MgHal. The latter, when treated with water, yields the prim- ary alcohol, R . CH 2 . CH 2 . OH. If water be added directly to the primary addition product, ethylene oxide is re-formed, which combines with the magnesium halide simultaneously produced to yield an alkylene-halhydrin : CH„ V CH„Br I j>0 + MgBr 2 CH,/ CH 2 . O . MgBr CH„Br CH,Br I + H 2 <> = | +Mg(OH)Br CH 2 .O.MgBr CH 2 OH ' Similar observations are recorded by Blaise. 57 Henry 58 has extended this method by using propylene SECONDARY ALCOHOLS 23 oxide, from which, by the action of magnesium ethyl bromide, he has obtained the secondary alcohol, methyl ^-propylcarbinol : CH 3 . CH 2 — CH, " + C 2 H 5 .MgBr O CH 3 . CH . CH 3 . C,H 5 CH„ . CH . CH 9 . C„H r ± 3 . v-ii . v^ 2 O . MgBr OH When epichlorhydrin is allowed to react with cold solutions of magnesium alkyl halides, only the hydrogen of the hydroxyl group is replaced by the radical, MgHal. If reaction takes places in warm solution, however, the chlorine atom is also replaced (by alkyl), and decomposition of the reaction-product by ice leads to the formation of primary alcohols : 59 CH 2 C1 CH 2 C1 + E. MgHal =| +R.H CH 2 OH CH 2 . MgHal CH,C1 CH 2 R | +R. MgHal = | +MgCJ(Hal) CH 2 . MgHal CH 2 . O . MgHal CH^R ^^2 • -^ | +H.,0 = | " +Mg(OH)Hal. CH 2 . O . MgHal CH 2 OH Secondary Alcohols. — These have been obtained in large number, and generally in good yield, by the interaction of Grignard's reagents with aldehydes, other than formaldehyde, 60 and with esters of formic 24 GRIGNARD'S REACTION acid. 61 Thus, acetaldehyde and magnesium methyl iodide yielded isopropyl alcohol, CH 8 .C< + CH 3 MgI = CHg.C— OMgl V) \CH 3 H /H CH 3 .C— OMgI + H 2 =-- CH 3 .C— OH + Mg(OH)I \CH S ^CHg Similarly, benzaldehyde and magnesium methyl iodide yielded phenylmethylcarbinol, 6 H 6 . CHOH . CH 3 . isopropyl alcohol was also obtained from ethyl formate and magnesium methyl iodide in accordance with the equations : /OMgl H.Cf +CH 3 MgI = H.C— CH 3 " 0C 2 H !> \oC,H r , .OMgl .OMgl H.C— CH 3 +CH 3 .MgI= H.C— CH 3 +Mg(OC 2 H 5 )I \0C 2 H 6 N \CH 3 .OMgl .OH H.C— CH 3 +H 2 = H.C— CH 3 + Mg(OH)I \CH 3 ^CHg Tertiary Alcohols. — These may be readily obtained by the action of Grignard's reagents on ketones, esters other than those of formic acid, acid chlorides, and anhydrides. Thus, triphenylcarbinol may be obtained by the action of magnesium phenyl bromide TERTIARY ALCOHOLS 25 on benzophenone or ethyl benzoate, according to the schemes : I. C 6 H 5V C 6 H 6X /OMgBi^ >C = + C 6 H 6 .MgBr = W C 6 H/ C e H / \C 6 H 5 C 6 H 6 . /OMgBr C C H 6X y 0H >C< +H 2 = >C< + Mg(OH)Bv C 6 H/ X C 6 H 5 C 6 H/ \C 6 H 5 II. /O / 0M s Br C 6 H 5 . C^ + C c H 6 MgBr = C H 5 . C-C 6 H 5 OC 2 H 5 ^OCsHj /0 MgBr /0 M g Br c „ C 6 H 6 . C-C 6 H 5 + C 6 H 5 MgBr = C fl H 6 . C . C 6 H 5 + Mg/ \0C 2 H 6 \C 6 H 5 Br ^OMgBr /OH QH C 6 H 5 .C.-C 6 H 5 +H 2 = C 6 H 5 .C— C 6 H 5 + Mg/ \C 6 H 6 \C 6 H 6 XBr Tertiary alcohols may likewise be obtained from cyclic ketones : thus, Murat 62 transformed i-menthone (Formula I.) into Z-3-phenyl-l-methyl-4-isopropyl-3- ci/cfohexanol (Formula II.), CH 3 CH, 1 3 1 CH 1 CH /\ „ /\ „ H„C CH„ H 2 C CH 2 1 1 II. 1 1 /OH H 2 C C< H„C C = v CH | CH 1 C 3 H 7 C 3 H y 26 GRIGNARD'S REACTION Similarly, Jermain and Creighton 03 obtained phenyl- borneol, C 10 H 16 . C 6 H 5 . OH, by the action of magnesium phenyl bromide on camphor in equimolecular propor- tions at 60°. An interesting synthesis of terpineol, /.CH — CH ov CH 3 .Cf " >CH . C(OH)(CH 3 ),, \CH,— CH/ has been effected by Perkin 04 by the action of magnesium methyl iodide on ethyl A 3 -tetrahydro-£>- toluate, ^,CH — CHyv . 3 .Cx /( CH, . cf ' >CH . COOC H v ^CH„— CH, J 2 The reactions employed are thus precisely analogous to those used in the preparation of secondary alcohols from aldehydes or formic esters. A modification of the second reaction, in which all three alkyl radicals of the tertiary alcohol are supplied in the form of their organomagnesium compounds, consists in con- verting the compound AlkMgHal by means of dry carbon dioxide into a compound of the type AlkC0 2 MgHal, and treating the latter with a further proportion of the organomagnesium salt (2 mols.) : K R.MgHal + C0 2 = R.C0 2 M g Hal .OMgHal R.cf +R.MgHal = R . C— R \0M g Hal \0M g Hal TERTIARY ALCOHOLS 27 /O.MgHal R R.C— R +R.MgHal = R.C— R + MgO + MgHal 2 X) . MgHal ^OMgHal /E /R R.C— R +H a O = R.C— R +Mg(OH)Hal. ^OMgHal \oH Another method of obtaining tertiary alcohols, which is very similar to that given above, consists in allowing an excess of Grignard's reagent to react with the free acid (see also p. 42). Acid chlorides behave towards magnesium alkyl salts in a manner similar to esters : 60 Q ^OMgHal R.c/ +2R'MgHal = R.C— R' +MgClHal Xci \r< R' 2 .R.C.OMgHal+H 2 0= R' 2 .R.C.OH + Mg(OH)Hal. Carbonyl chloride, when treated with organo- magnesium compounds (2 mols.) yields a mixture of secondary and tertiary alcohols which frequently contains unsaturated hydrocarbons. In the methane and ethane series the action of magnesium alkyl halide (3 mols.) on carbonyl chloride (1 mol.) leads exclusively to tertiary alcohols, whereas in the higher series secondary alcohols are simultaneously pro- duced. 67 Tertiary alcohols may also be obtained in small quantity from carbon oxysulphide, COS, and Grignard's reagent. 68 The behaviour of acid anhydrides towards Grig- 28 GRIGNARD'S REACTION nard's reagents is somewhat similar to that of acid chlorides, 69 and is expressed by the equations : Q ^OMgHal R— C R "< o .OMgHal R— C— R .OMgHal No +2R.MgHal = 2R.«— R + Mg(O.COR) 2 + MgHal, R -0 R . CO R . C— OMgBr \r< R.CO q V> + H 2 = Mg(OH)Br + R . C0 2 H + R . C- R' R . C— OMgBr \R' GLYCOLS 29 Thus, when magnesium ethyl bromide (1 mol.) was slowly added to an ethereal solution of acetic anhydride (1 mol.), cooled in a mixture of ice and salt, and the mixture decomposed after three hours by ice-cold water, methyl ethyl ketone was obtained. The investigation of the interaction of succinic anhydride and magnesium organic halides has been carried out by Houben and Hahn, 71 who found that reaction did not readily take place, but ultimately led to the formation of ditertiary glycols of the type (HO) R 2 C . CH 2 - CH 2 . C (OH) R 2 . Phthalic anhydride behaved similarly towards magnesium p-to\y\ bromide yielding o-di-p-toluoylbenzene, 6 H 4 (CO . C 6 H 4 . CH 3 ) 2 , but, when treated with magnesium methyl, ethyl, or propyl halides, it gave dimethyl-, dietheyl-, and dipropylphthalide respectively, 72 CO yCfCH^ / C(CH 3 ) 2 H / No + 2CH 3 . Mgl _► C B H 4 NoH -> C 6 H ^>0 co \cooh N co whilst camphoric anhydride and [magnesium benzyl chloride yielded the two isomeric campholides, 73 CH 2 CH C(CH 2 C 6 H 6 ) 2 CH — CH CO CMe 2 O and CMe 2 O CH 2 — CMe— CO CH 2 -CMe-C(CH 2 C 6 H 6 ) 2 Primary Tertiary a-Glycola— These may be pre- pared by the action of magnesium alkyl halides on 30 GRIGNARD'S REACTION glycerine a-chlorhydrin. 74 Thus, magnesium amyl bromide yielded the compound CH 3 \c(OH) . CH„OH, together with small quantities of the glycol, C 6 H U . CH 2 . CH(OH) . CH 2 OH. Monoalkyl ethers of primary tertiary glycols have been obtained by Behal and Sommelet 75 from ethoxy- acetic esters or ethoxyketones : OC,H r .CH„.Cf +2R.MgHal X)C.,H 5 .OMgHal = OC 2 H 6 . CH, . C— R + Mg(OC 2 H r )Hal \r XJMgHal oc 2 h, . ch,, . c— r + h„0 \r ,OH = OC 2 H . CH 2 . C— R + Mg(OH) Hal \r \c.CH 2 .OC 2 H 6 + R.MgHal = \C(OMgHal).CH 2 .OC 2 H 5 R/ ^>C (OMgHal) . CH 2 . OC 2 H 5 + H 2 = j>C(OH).CH 2 .OC 2 H s + Mg(OH)Hal. GLYCOLS 31 Secondary Tertiary Glycols. — Compounds of the type R.CHOH.C(OH).R.R have ,been obtained by Acree 78 by the action of Grignard's reagents upon r-benzoin, C 6 H 5 . CH(OH) . CO . C 6 H 6 , and methyl r-mandelate, C 6 H 5 . CHOH . COOCH 3 ." A similar series of optically active glycols have been prepared by McKenzie and Wren 7S from ^-benzoin and methyl i-mandelate. Ditertiary Glycols. — These compounds may be readily obtained according to Valeur 79 by treatment of esters of dibasic acids with excess of an organo- magnesium halide. Thus, ethyl oxalate yielded tetramethylglycol when treated with an excess of magnesium methyl iodide : \;C— C? + 4CH 3 MgI C 2 H 6 0/ X)C 2 H 5 IMg(X /OMgl = CH3-C-C-CH3 + 2Mg(OC 2 H 5 )I ch/ \ch 3 IMgO, X)MgI CH 3 — C— C— CH 3 +2H 2 CH3/ ^CHg = (CH 3 ) 2 C.(OH).C(OH)(CH 3 ) 2 + 2Mg(OH)I Similarly, Valeur 80 and Dilthey and Last 81 found that the final products of the interaction of succinic esters and excess of organomagnesium salts are di- tertiary glycols of the type R— C.CH 2 .CH 2 .C— It oh/ X OH 32 GRIGNARD'S REACTION Ethyl phthalate, on the other hand, 82 appears to yield phthalides in the first place : C (i H/ + 2R'MgHal COOR COOR OMgHal C— R' ,CR', = C e H 4 ^ \R' — > C 6 H 4 / COOR CO Generally, however, the reaction proceeds a step further, the ketone group contained in the dialkyl- phthalide first formed reacting with another molecule of the Grignard reagent and the resulting compound decomposing on contact with water, with the produc- tion of a derivative of phthalan (o-xylylene oxide) : C H / 2 No + R'MgHal + H 2 \co/ CR 2 i CR' 2 = C H / \0 — -> C 6 H / \0 CR'OH C = R" Frankland and Twiss 83 found that the normal action took place when methyl ^-tartrate was treated with excess of magnesium phenyl bromide, optically active tetraphenylerithritol, (C 6 H 5 ) 2 . C(OH) . CH(OH) . CH(OH) . C(OH)(C 6 H 6 ) 2 , being obtained. Acree 84 has also obtained ditertiary glycols from a-diketones. Thus, benzil yielded tetraphenyl glycol DITERTIARY GLYCOLS 33 when acted on by excess of magnesium phenyl bromide : C u H 5 .C = C H 5 .C = O f2C fi H 6 MgBr .OMgBr C H 5 ■ C ~ C H 5 /OH / i C H 5 ■ C C C H 5 > /OH C 6 H 5 .C/ \h 5 /OMgBr C 6 H 5 .C/ X C C H 5 It may be pointed out that the benzil molecule contains two carbonyl groups, each of which — as in the case cited above — may be attacked by Grignard's reagent. By a suitable adjustment of the conditions of reaction and the quantities of material employed, Acree was able to limit the action of the reagent to one of the carbonyl groups and thus to obtain, e.g. phenyl benzoin, C 6 H 5 . CO . C(OH)(C 6 H 6 ) 2 . Bamberger and Blangey 85 have obtained small yields of quinols by the action of Grignard's reagents on p-quinones. The interaction of magnesium ethyl bromide and anthraquinone 86 leads to the production of dihydroxydiethyldihydroanthracene, CEt(OH) C (i H 4 CEt(OH) when excess of the Grignard reagent is employed; when, however, anthraquinone is kept in excess, ethyloxanthranol is produced. Werner and Grob 87 have similarly prepared 9 : 10-dihydroxydiphenyldi- C 34 GRIGNARD'S REACTION hydrophenanthrene by the action of magnesium phenyl bromide upon phenanthrenequinone : .OH CH, . CO C 6 H 4 — C< C 6 H 4 . CO ^C 6 H 5 ,OH 6 H 4 -C/ X C„H, Phenols, Thiophenols, and Selenophenol. — Bodroux 88 found that small quantities of phenols were formed when oxygen was passed through a solution of magnesium aryl halides. This observa- tion is confirmed by Wuyts, 89 who, however, found the action to be rather more complex. Thus, when oxygen was bubbled through an ethereal solution of magnesium phenyl bromide, other phenolic compounds were also formed, together with diphenyl, p-diphenyl- benzene, phenylethylcarbinol, and ethyl alcohol. He suggests that, in all probability, a peroxide is the primary compound of the oxidation. Wuyts 90 found that a thiol was the only product of the action of finely divided sulphur upon organo- niagnesium compounds, if the reaction was carried out in an atmosphere of dry hydrogen and care was taken to avoid excess of sulphur. The yield attained 80 per cent, of the theoretical. By the action of sulphur on the compound formed by the condensation of the thiol with an additional molecule of the reagent, a disulphide was formed : R . S . H + R . MgHal = R . S . MgHal + R . H 2R.S.MgHal + S = R— S— S— R + S(MgHal),, ETHERS 35 whilst, by the interaction of the disulphide and the organomagnesium salt, a monosulphide was obtained : R_S— S— R + R . MgHal = R— S— R + R— S . MgHal. When the same precautions were adopted as in the case of sulphur, selenophenol, C 6 H 5 SeH, could be obtained by the action of selenium on magnesium phenyl bromide. The yield obtained was 81 per cent, of the theoretical. Ethers. Ethers of the type, R . CH 2 . . CH 3 , have been prepared by Hamonet 91 by the interaction of Grignard's reagents and halogen substituted methyl ethers. An ethereal solution of chlorodimethyl ether, C1CH 2 . O . CH 3 , is not attacked by magnesium, and also prevents the solution of the metal by methyl iodide or ethyl bromide. It reacts readily, however, with a solution of magnesium alkyl halide with the production of mixed ethers : CH 3 . O . CH 2 C1 + R . MgBr = CH 3 . O . CH 2 R + MgBrCl. Similar observations have been recorded by Eeychler. 92 Klages 93 has prepared a series of phenol ethers by the action of magnesium alkyl halides on alkoxyalde- hydes ; thus, magnesium propyl iodide and anisaldehyde yielded ^-butenylanisol : (CH 3 0) . C 6 H 4 . CH : CH . CH 2 . CH 3 . Scission of Phenolic Ethers. — Grignard 94 found that organometallic derivatives do not react with phenolic ethers under ordinary conditions. By adding 36 GRIGNARD'S REACTION magnesium, however, to a mixture of equimolecular proportions of an alkyl bromide with anisole or phenetole, using benzene as solvent, reaction may be brought about. If the resultant product is immediately decomposed, the ether is recovered un- changed ; if, however, the solvent is distilled off and the residue heated at 150°-160° under 10-15 mm., only one-half of the ether is recovered, the remainder having undergone conversion into the corresponding phenol. This is explained by supposing that an oxonium complex is first produced by addition and that this undergoes scission, yielding the compound C 6 H 6 . . MgBr, together with, probably, ethylene and a saturated hydrocarbon. Aldehydes. Aldehydes have been prepared by a number of methods which, generally, involve the action of Grignard's reagents upon derivatives of formic acid. Gattermann and Maffezzoli 95 found that when two molecules of an organomagnesium compound are allowed to react with one molecule of ethyl formate, the main product of the reaction is a secondary alcohol; but if, on the other hand, an excess of the formic ester is employed, an aldehyde is formed according to the following equation : .O .R H . Gf + IMgR = H . C{ + Mg(OC„H/)I. \OC 3 H 6 ^O A series of aldehydes has been prepared by Tschitschibatin B6 by the interaction of ethyl orthor ALDEHYDES 37 formate and magnesium alkyl salts. Acetals are first obtained which, when hydrolysed, yield alde- hydes : /OCjHji /R OC H H . C— OC 2 H 5 + R . Mgl = H.C— OC 2 H 5 +Mg/ X OC 2 H 5 X OC 2 H 5 l H . C— OC 2 H 5 + H 2 + [HC1] = H . c/ + 2C 2 H 6 OH + [HC1] \0C 2 H 5 % ° The yields of the various aldehydes so obtained differed greatly. Shdanovitsch, 97 in a rather more complex instance, has obtained ethyl /3-keto-Et COiOEt ; COOEt = 2Mg<^ H . C(OEt) 2 OEt CMe 2 + I Br CO Me 2 C . COOEt In the place of formic esters, the free acid in dilute 38 GRIGNARD'S REACTION ethereal solution, or its copper salt, has been used in the preparation of aldehydes. 98 Bouveault" has investigated the action of Grignard's reagents on disubstituted formamides, and obtained, for example, benzaldehyde from magnesium phenyl bromide and ethyl formamide. The reaction is expressed by the general equation : Q /OMgHal H . cf + R"MgHal = H . C— R" \nrr< \ nrr , .OMgHal ,R" H . C— R" + H 2 = H . C\ Q + NHRR' + Mg(OH)Hal \nrr Isonitriles, according to Sachs and Loevy, 100 form addition compounds with organomagnesium salts (Formula I.), which, when treated with mineral acids, are converted into aldehyde-imide derivatives (Formula II.), and then into aldehydes (Formula III): Aryl /Aryl /Aryl II. R.N:C< III. 0:C/ MgBr \H :C \ Monier- Williams 101 has prepared a series of alde- hydes from ethoxymethyleneaniline, C 6 H 5 . N : CH . OC 2 H 5 . The method leads, in the first place, to the anhydro- compounds of the aldehyde with aniline which, when acted on by mineral acids, are hydrolysed to the aldehydes themselves and aniline. Thus, KETONES 39 a-naphthaldehyde was obtained in 48 per cent, yield from magnesium a-naphthyl bromide according to the equations: C 6 H 5 N : CH . O . C 2 H 6 + Br . Mg . C 10 H 7 = Br . Mg . OC 2 H B + C 6 H 6 N : CH . C 10 H 7 C 6 H D . N : CH . C 10 H 7 + H a O + HC1 = C 10 H 7 . CHO + C 6 H 5 NH 2 . HC1. Ketones. Blaise 102 has succeeded in preparing ketones by the action of Grignard's reagents upon nitriles. Magnesium ethyl iodide and benzonitrile gave an 80 per cent, yield of phenyl ethyl ketone : y S . Mgl C 6 H 5 . C = N + C 2 H 5 . Mgl = C 6 H 5 . C^ C 6 H 3 C = N.MgI + H„0 = C 6 H 5 .C = NH + Mg(OH)I I I C 2 H 6 C 2 H 5 C 6 H 5 .C(:NH).C 2 H 5 + H 2 0= C 6 H 5 . CO . C 2 H 5 + NH 3 . With nitriles of the type of benzyl cyanide, C 6 H 5 .CH 2 CN, much poorer yields were obtained. Similarly, varying yields of ketones were also ob- tained by Beis 103 by prolonged heating of acid amides with an excess of Grignard's reagents. On the probable assumption that the amide reacts in the enolic form, XJH, 40 GRIGNARD'S REACTION the action may be represented in the following manner : NH /NHMgHal R . c/ + 2R'MgHal = R . C— R + R'H 0H ^OMgHal .NHMgHal /NH 2 R.C— R + 2H 2 = R.C— R +2Mg(OH)Hal ^OMgHal ^OH R.RC(OH)NH 2 = R.CO.R + NH 3 . An analogous reaction has been employed by McKenzie and Wren, 104 and by Wren, 105 in preparing T-, 1-, and d- benzoin, C 6 H 6 . CHOH . CO . C 6 H 6 , from r-, 1-, and d- mandelamide, C B H 5 . CHOH . CO . NH 2 , respectively ; and also by Ryan and Nolan, 106 for the production of a series of ketones from palmitamide and stearamide. Busch and Fleischmann 107 found that magnesium phenyl bromide reacted with substituted anilides in the following manner : .O / C 6 H 5 C 6 H 5 .ef +C 6 H 5 .MgBr = C 6 H, . C— OMgBr \NEtC 6 H \NEtQ,H, C 6 H 5 . C— OMgBr + H.,0 \NEtC 6 H 5 = C 6 H 5 .CO.C H 5 + NHEtC H 5 + Mg(OH)Br. If, however, a second molecule of the reagent is used, a stable compound, /C 2 H 5 (C 6 H 5 ) 3 C.N< s C fi H 6 , KETONES 41 is formed. The latter readily decomposes in alcoholic solution, with the formation of triphenylcarbinol and ethylaniline. By the regulated action of Grignard's reagents upon acid chlorides it appears to he theoretically possible to prepare ketones : R . COC1 + R'MgHal = R . CO . R' + MgClHal. In general, however, this reaction does not appear to be easily controllable in such a manner that good yields of ketones are obtained. Acree 108 has prepared phenyl a-naphthyl ketone, C 8 H 5 . CO . O 10 H 7 , by the interaction of benzoyl chloride and magnesium a-naphthyl bromide under conditions which might be expected to lead to the production of phenyl di-a-naphthylcarbinol, whilst Gomberg and Gone 109 obtained a small yield of benzophenone from benzoyl chloride and magnesium phenyl bromide. Further, Oddo 110 found that magnesium pyrryl iodide reacts with acyl chlorides with formation of ketones of the type /CH = CH NH< | \C = CH I COR. Ketones may also be produced under certain circumstances by the action of carbon dioxide on organomagnesium salts. Schroeter 111 thus obtained benzophenone in small yield from magnesium phenyl bromide, and Bodroux 112 showed that ketones are the main product of the action of carbon dioxide on warm solutions of magnesium ^-chlorophenyl bromide and 42 GRIGNARD'S REACTION magnesium p-bromophenyl bromide, v. Braun and Sobecki 113 obtained cycfopentanone by treating magnesium butylene dibromide with carbon dioxide: CH 2 . CH 2 . MgBr CH 2 . CH 2 I +co 2 = ^CO + MgO + MgBr, CH 2 . CH 2 . MgBr CH 2 .CH 2 The action of magnesium organic compounds on free acids has not been very extensively studied. Since the main products formed are tertiary alcohols and ketones, the results may be briefly considered in this section. The first step in the reaction consists in the replacement of the hydrogen of the carboxyl group, R'.COOH + R.MgHal = R'. COOMgHal + RH, and the compound so formed is convertible by excess of the reagent into tertiary alcohols and ketones (compare the action of carbon dioxide on Grignard's reagents, p. 26). Thus, benzoic acid can be converted into diethylphenylcarbinol : 1U C 6 H 5 .C^ +2C 2 H 5 .MgBr .OMgBr = C 6 H 5 .C_C 2 H 5 +C 2 H 6 \0MgBr .OMgBr / C 2 H 5 = C fl H 5 . C— C 2 H 5 + MgO + MgBr 2 C 6 H 6 .C-C 2 H 5 + C 2 H 5 MgBr \3MgBr ^OMgBr / C 2 H 5 C H C .C— C,H 6 +H 2 / C 2 H 5 = C„H 5 .C— C 2 H + Mg(OH)Br. ^OMgBr ^OH KETONES 43 Whilst formic acid 11 ? yields aldehydes under the same conditions : B..CZ +2R.MgI = H.C— OMgl + R.H x OMgI H.C— OMgI + H 2 = H.c/ +2Mg(OH)I. \0MgI ^° The production of ketones from acids has been investigated in the case of phthalic acid by Simonis and Arand, 116 who find that this substance, when acted on by excess of Grignard's reagents, yields two products, viz., a dialkyl phthalide and a ketone. Thus, phthalic acid and magnesium ethyl bromide gave diethylphthalide (Formula I.) and propio- phenone-o-carboxylic acid (Formula II.): C(C 2 H 6 ) 2 ,COC,H r .. C.„<\0 „. C„H,< cooh The analogous compound, naphthoyl-o-benzoic acid, / C OC 10 H 7 c„h/ \COOH, has been prepared by Pickles and Weizmann 117 by the interaction of magnesium a-naphthyl bromide and phthalic anhydride. Acids. Carboxylic acids were prepared by Grignard, 118 who showed that solutions of magnesium organic halides readily absorb carbon dioxide. Thus, when 44 GRIGNARD'S REACTION dry carbon dioxide was passed into an ethereal .solution of magnesium methyl iodide and the product decomposed by dilute acid, acetic acid was obtained : CH 3 .MgI + C0 2 = CH 3 .C0 2 MgI CH 3 . C0 2 MgI + H 2 = CH 3 .COOH + Mg(OH)I. Similarly, Zelinsky obtained a 60 per cent, yield of benzoic acid from magnesium phenyl iodide, 119 and Schmidlin 120 prepared triphenylacetic acid, (C 6 H 6 ) 3 CCOOH, in good yield by the decom- position of magnesium triphenylmethyl chloride, (C<,H 6 ) 3 C . MgCl, by carbon dioxide. Schroeter m substituted magnesium phenyl bromide for the iodide when a more complex change occurred whereby benzoic acid, benzophenone, and triphenyl- carbinol were obtained. The course of the reaction may be represented by the equations : I. C 6 H 5 . MgBr + C0 2 = C 6 H S . COOMgBr C 6 H 6X /OMgBr C 6 H 5 . COOMgBr + C 6 H 5 MgBr = >C< C 6 H/ \OMgBr C 6 H 6X .OMgBr C 6 H /C 6 H 5 >C< +C 6 H 6 MgBr= >C< +MgO + MgBr, C 6 H/ XJMgBr C 6 H/ X)MgBr II. C 6 H 5 . COOMgBr + H 2 = C 6 H 5 .COOH + Mg(OH)Br C 6 H 5X /OMgBr C 6 HU >C< +H 2 = >C = + 2Mg(OH)Br C, ; H/ \OMgBr % C C H/ C (! H S\ / C 6 H 5 - C H 5\ / C 6 H 5 >C< +H,0 =•• >C< +Mg(OH)Br. C C H/ M)MgBr C H/ \qH ACIDS 45 Later, 122 the same author showed that the experi- mental conditions may be so adjusted that no benzoic acid is formed, and that the production of the latter depended upon the quantity of carbon dioxide employed and the temperature at which the gas was passed into the. ethereal solution. 123 In this connec- tion it is interesting to note that Grignard 124 has prepared trialkylcarbinols by saturating a solution of alkyl magnesium halide with carbon dioxide, and protracted heating of the compound so formed with two additional molecules of the organomagnesium salt (see p. 26) ; and further, that Bodroux 125 has observed that carbon dioxide transforms magnesium p-chloro- and p-bromophenyl bromide at a higher temperature chiefly into dichloro- and dibromo- benzophenone, at a lower temperature, however, into the corresponding substituted benzoic acids. Oddo 128 found that a small amount of propiolic acid was formed when magnesio-acetylene bromide was treated with carbon dioxide followed by dilute sulphuric acid : -HC = C.MgBr + C0 2 = H . C=C . C0 2 MgBr 2H . C = C . C0 2 MgBr + H 2 + H 2 S0 4 = 2H . C = C . COOH + MgBr 2 + MgS0 4 + H 2 0. Houben and Pohl 127 have shown that carbithionic acids are formed when carbon disulphide is substi- tuted for carbon dioxide in the above reactions. Thus, when carbon disulphide was added to a cooled solution of magnesium methyl iodide in absolute ether and the product decomposed with ice and cooled 46 GRIGNARD'S REACTION hydrochloric acid, dithioacetic acid, CH 3 . CSSH, was obtained as an unstable, reddish-yellow oil. Hydroxy-acids have been obtained by the action of Grignard's reagents upon ketonic esters followed by saponification of the product so formed. A ketonic ester presents two points of attack to the Grignard reagent, viz., the carbonyl group and the carbalkoxy group, but, by careful regulation of the relative quantities of ester and reagent employed, and by so conducting the experiment that the latter is never in excess, it is possible to limit the action completely, or almost completely, to the carbonyl group. Grignard 128 has prepared a series of hydroxy-acids in this manner, R.CO...CO.OC 9 H 5 + R'MgI -^ >C(OH)...CO.OC,H 5 , R'/ but the reaction cannot be applied to /3-ketonic esters which possess the ability to pass into an enolic form. In a similar manner, the reaction has been exten- sively employed in the investigation of the problem of asymmetric synthesis by McKenzie, 129 the action of various organomagnesium salts on the menthyl, bornyl, and amyl esters of a-, /3-, and y- ketonic acids having been examined. Thus, when i-menthyl benzoylformate was acted on by magnesium methyl iodide, a mixture of unequal amounts of i-mejithyl C B H 5 . CO . COOC 10 H 19 (active) > CH 3 C H 5 . C— COOC 10 H 19 (active) > OH CH, I C 6 H 5 . C— COOH (active). OH Oddo 130 has investigated the behaviour of carbon dioxide towards the iodo-magnesium derivatives of various phenols which can undergo conversion into the corresponding hydroxy -acids (according to Kolbe's synthesis of aromatic hydroxy-acids from sodium aryloxides by means of carbon dioxide). With the phenol and resorcinol derivatives this change only occurred in the absence of solvent, and at a somewhat elevated temperature. With derivatives of the following phenols the reaction proceeded in the presence of a solvent (benzene or toluene) : /3-naphthol, which yielded ^-naphthol-o-carboxylic acid; phloro- glucinol, giving phloroglucinol carboxylic acid; *Jj|iymol, giving o-thymotic acid [CH 8 : C0 2 H : OH : C 8 H 7 = 1:2:3:4]. 48 GRIGNARD'S REACTION An interesting synthesis of aromatic amino-acids by rearrangement has been accomplished by Houben and Schottmiiller l31 by treatment of methylaniline, methyl iodide, and magnesium with dry carbon dioxide. Reaction was found to proceed in accordance with the equation : CH C.HjjNHCHg + CO., + CH 8 I + Mg = CH 4 + C 8 H 6 n/ \COOMgI The latter compound underwent rearrangement with the formation of the substance, CH 8 . NH . C^H, . COOMgl, from which an almost quantitative yield of £>-mono- methylaminobenzoic acid, CH 3 . NH . C 6 H 4 . COOH, was obtained. Methylaniline, in this reaction, may be replaced by a mixture of aniline and dimethyl- aniline, whilst dimethylaniline hydroiodide may be used in place of a mixture of methylaniline and methyl iodide. The action of sulphur dioxide on an ethereal solution of an organomagnesium salt leads, according to Rosenheim and Singer, 132 to the formation of sulphinic acids, the yields being 50 to 60 per cent, of the theoretical. Thus, phenyl magnesium bromide and sulphur dioxide yielded phenyl sulphinic acid : C,H S . MgBr + S0 2 = C 6 H 6 . S0 2 . MgBr C H .SO 2 .MgBr + H 2 O = C G H 6 . S0 2 H + Mg(OH)Br. As by-product, diphenyl sulphoxide, (C 6 H 5 ) 2 SO, was ESTERS 49 obtained. The authors, however, point out that reduction and transformation of the sulphinic acids so formed readily takes place. This fact probably accounts for the circumstance that Oddo 13S finds that phenyl sulphide, small quantities of phenyl sulphoxide and diphenyl, are the products of inter- action of the same two substances. Houben prepared dihydropinenesulph'inic acid by the action of sulphur dioxide on magnesium pinyl chloride, C 10 H l7 . MgCl. The replacement of sulphur dioxide by thionyl chloride in the above reactions led to the production of sulphoxides and sulphides. Thus, magnesium ethyl iodide yielded ethyl sulphide, magnesium phenyl bromide gave phenyl sulphide in addition to small quantities of phenyl sulphoxide and diphenyl. 134 Similarly, Strecker 135 obtained benzyl sulphoxide and benzyl sulphide by the action of magnesium benzyl bromide on thionyl chloride. The former compound was also obtained when magnesium benzyl bromide acted on symmetrical diethyl sulphite, whereas phenylethylsulphone resulted from the inter- action of magnesium phenyl bromide and unsym- metrical diethyl sulphite. Esters. Numerous syntheses of esters are recorded, based upon the action of Grignard's reagents with ethyl carbonate and closely allied compounds. Houben 136 found that when magnesium phenyl bromide was added to ethyl chlorocarbonate, CI . COOC 2 H 5 , in such D 50 GRIGNARD'S REACTION a manner that excess of the former was avoided, ethyl benzoate was obtained: C„H 6 . MgBr + CI . COOC 2 H s = C 8 H 5 . COOC 2 H 5 + MgBrCl. Tschitschibabin 137 obtained varying yields of esters by the interaction of organomagnesium compounds and ethyl carbonate. The reaction appears to proceed in accordance with the equation: / OC 2 H 3 /OMgHal R.MgHal + C = = R-C/ \0C.,H, X (OC 2 H 5 ) 2 ■^"S /OMgHal R.C< +H 2 = R.COOC,H 5 + C 2 H 5 OH ^(OC 2 H 5 ) 2 +Mg(OH)Hal. The yields are stated to be considerably improved if air is excluded during the reaction by means of a current of dry hydrogen. The same author 138 found that the regulated action of orthocarbonic esters on organomagnesium compounds led to the production of ortho-esters, from which, by the action of acids, the normal esters could be obtained : C(OC 2 H 5 ) 4 + R.MgHal = R.C(OC 2 H 6 ) 3 + Mg(OC 2 H 5 )Hal R . C(OC 2 H 5 ) 3 + H 2 = R . COOC 2 H 5 + 2C 2 H 5 OH. An interesting method for the preparation of esters from alcohols and phenols is described by Houben. 139 It depends upon the decomposition of ESTERS 51 a suitable Grignard's reagent by the alcohol (or phenol), R . OH + R'MgCl = R . O . MgCl + R'H, and subsequent addition of acetic anhydride (or, less frequently, of acetyl chloride) to the product so formed : R.O.MgCl + (CH 3 CO) 2 = R.O.COCH 3 + CH 3 .COOMgCl. In this manner a large number of alcohols were con- verted into esters, the yield being generally excellent. Houben concludes that alkyl bromo- and iodo- magnesium alcoholates are only suitable when saturated alcohols and, possibly, phenols are to be esterified, whereas chloromagnesium alcoholates can be employed with satisfactory results even with unstable and unsaturated alcohols and phenols ; and, further, that poly hydroxy and solid, sparingly soluble alcohols can be esterified in this manner. Esters of ketonic acids have been prepared by Meyer and Togel 140 by the action of acid chlorides on the magnesium compounds of halogenated esters. Thus, ethyl benzoylacetate, C 6 H 5 . CO . CH 2 . COOC 2 H 5 , was obtained by adding benzoyl bromide to magnesium ethyl bromoacetate : C 6 H 5 COBr + BrMg . CH 2 . COOC 2 H 6 = C 6 H 5 . CO . CH 2 . COOC 2 H 5 + MgBr 2 . Ethyl acetoacetate, ethyl a-benzoylpropionate, and ethyl /3-benzoylpropionate were formed in an anal- ogous manner. 52 GRIGNARD'S REACTION Zeltner 141 has obtained esters of jg-ketonic acids by the action of magnesium on the esters of a-halogen fatty acids. In this manner, ethyl a-bromopropionate gave a 35 per cent, yield of ethyl a-propionylpro- pionate, CH 3 . CH . (CO . CH 2 . CH 3 ) . COOC 2 H 6 . CH 3 .CHBr.COOC 2 H 6 + Mg = CH 3 .CH(MgBr).COOC 2 H 6 CH 3 . CH(MgBr). cf + CH 3 . CH(MgBr) . COOC 2 H 6 \OC 2 H 5 ,OMgBr = CH 3 .CH(MgBr).C— CHCH 3 .COOC 2 H 5 X OC 2 H 5 XJMgBr CH 3 . CH (MgBr) . C— CHCH 3 . COOC 2 H 5 + 2H 2 Q -OC s H s = CH 3 .CH 2 .C— CHCH 3 .COOC 2 H OH CHC OC 2 H 5 + 2Mg(OH)Br .OH CH 3 . CH. 2 . C— CHCH 3 . COOC 2 H 6 X OC 2 H 6 = CH 3 . CH 2 . CO . CH(CH 3 )COOC 2 H 5 + C 2 H 6 OH. Nitriles. Two methods of preparing nitriles with the aid of magnesium organic salts have recently been proposed by Grignard. 142 The first consists in the cautious NITROGEN COMPOUNDS 53 addition of ethereal solutions of magnesium alkyl halides to an ethereal solution of cyanogen chloride, R.MgHal + Cl.CN = R . CN + Mg(Hal)Cl. The yields obtained are satisfactory. Cyanogen bromide and iodide are not suitable for this purpose; with the latter, reaction proceeds entirely in accordance with the equation, R.MgBr + CNI = RI + MgBi-CN, whilst, with cyanogen bromide, reactions of both types occur, the latter preponderating. The second method consists in substituting cyanogen itself for its halogen derivative. It has been used to prepare benzonitrile, -isohexonitrile, and phenyl- butyronitrile, but the yields are inferior to those obtained with cyanogen chloride. If cyanogen or its chloride is added to the solution of the organo- magnesium compound, ketones are produced in the usual manner. According to Grignard and Bellet, 143 alkyl cyclic nitriles can be prepared by adding the corresponding magnesium alkyl bromide drop by drop to a cold ethereal solution of cyanogen. In this manner, cyanohexamethylene, o-, m-, andp-methylci/c^ohexane- carboxylonitriles have been obtained. Cyanuric chloride and magnesium phenyl bromide react in ethereal solution with the successive pro- duction of dichlorophenyltriazine, C 3 N 3 C1 2 C 6 H 5 , and chlorodiphenyltriazine, C 3 N 3 C1(C 6 H 5 ) 2 . Ui Nitrogen Compounds. Aromatic amides have been obtained by Blaise 145 54 GEIGNARD'S REACTION by the action of organomagnesium compounds on arvl carbimides : /OMgl C 6 H 5 N = C = + R.MgI = C 6 H 5 .N = C/ .OMgl C 6 H 5 . N = C< + H 2 = C 6 H 5 NH . CO . R + Mg(OH) I. ' " \R By a precisely similar reaction, Sachs and Loevy 146 have prepared thioanilides from mustard oils. Phenyl mustard oil and magnesium benzyl bromide yielded the anilide of thiophenylacetic acid, C 6 H 5 .CH 2 .CS.NHC 6 H 6 : C 6 H 5 . N = C = S + C 6 H 5 CH 2 . MgBr = C«H S .N:C< "•SMgBr /CH 2 . C 6 H 5 C 6 H 6 . N = C< + H s O \S . MgBr = C 6 H 6 .NH.CS.CH 2 C 6 H 5 + Mg(OH)Br. Anilides have also been obtained by Bodroux u7 by treatment of the compounds formed by the action of magnesium organic halides on primary amines (of the type R . NH . MgHal) with the ester of a monobasic acid and subsequent decomposition of this product by means of dilute acid : '., R.NH 2 + X. MgHal = R . NH . MgHal + XH .OMgHal 2R.NHMgHal + R'.C^ = R'.C— NHR + Mg(OR")Hal XoR " \nhr /O . MgHal R'.C/ ^ T + HC1 = R-NHo + Rj.CO.NHR + MgCHalJCl. >(NHE), NITROGEN COMPOUNDS 55 The same author 148 found that a substituted urethane was obtained when ethyl carbonate was substituted for the ester of a monobasic acid in the above reaction : NHR / OC 2 H s | .NHR C = +2R.NH.MgI = C< + Mg(OC 2 H 5 )I \0C 2 H 5 P ^ 1 OC 2 H 6 /NHR) 2 /NHR) 2 C— OMgl +H 2 = C— OH +Mg(OH)I \nr.R \f ^OC 2 H 6 \OC 2 H 5 NHR NHR / NHR = C = +NH 2 R. OH \ f K >C 2 H 5 Busch and Einck u6 have succeeded in transforming alkylidene bases into secondary amines by means of Grignard's reagents. Thus, C-ethylbenzylaniline was obtained from magnesium ethyl iodide and benzylidene aniline: C 6 H 5 .N C 6 H 5 .N— CH.C 6 H 5 || +C 2 H .MgI= | | C 6 H 5 .CH Mgl C 2 H 5 C e H 5 .N CH.C 6 H 5 + H 2 Mgl C 2 H 6 = C 6 H 5 .NH.CH(C 2 H 5 ).C 6 H 5 + Mg(OH)I. The interaction of oximes and organomagnesium salts has been investigated by Busch and Hobein 150 56 GRIGNARD'S REACTION in the expectation that addition of the reagents would take place at the =C = N — linkage in the same manner as with alkylidene bases. This was found to be the case, but the process was not limited to this addition, since the hydroxyl group was also replaced by alkyl. In some cases the second reaction alone occurred. The whole process is represented by the equations : R . CH = N . OH + 2R'. MgHal = R.CH— N— R' + Mg(OH)Hal I I R' MgHal R . CH— N— R' + H 2 = R.CH— NHR' + Mg(OH) Hal. R' MgHal K' Thus, a-benzaldoxime and magnesium phenyl bromide yielded diphenylanilidomethane, (C 6 H 5 ) 2 .CH.NHC H 5 . The same products were obtained when O-ethers of oximes were substituted for the simple oximes : C 6 H 5 . CH : N . OCH 3 + 2C 6 H 5 . MgBr = C 6 H 5 . CH N . C 6 H 6 + Mg(OCH s )Br C 6 H 5 . CH N . C H C 6 H 6 MgBr C 6 H 5 MgBr + H 2 C 6 H 5 . CH-NH . C 6 H 5 /OH I +Mg< ^c H 5 \Br. The replacement of oximes or ethers of oximes in the above reaction by /3-phenylhydroxylamine 151 led to unexpected results. This compound was NITROGEN COMPOUNDS 57 converted by magnesium phenyl bromide into triphenylhydrazine, azobenzene being possibly formed as an intermediate product and reacting with the organomagnesium salt, thus : C 6 H 5 .N = N.C 6 H 5 + C 6 H 5 .MgBr = C 6 H 5 .N N.C 6 H 5 C 6 H 5 MgBr + H„0 C 6 H 6 .N N.C H 5 C 6 H 5 MgBr C 6 H 5 . N N . C„H, + Mg(OH)Br C H 5 H Or, more probably, the phenylhydroxylamine is partly transformed into diphenylamine, which then condenses with an additional molecule of phenyl- hydroxylamine, (C 6 H 6 ) 2 NH + OH.NHC 6 H 5 = (C 6 H 5 ) 2 N.NHC 6 H 5 + H 2 0. /3/3-Dialkylhydroxylamines, R 2 NOH, have been obtained by Wieland 152 by the action of nitrogen peroxide upon magnesium alkyl halides, the nitrogen being reduced from the tetravalent to the trivalent condition. An attempt to obtain similar compounds from magnesium aryl halides was unsuccessful. 163 Nitric oxide reacts with magnesium alkyl salts with the formation of nitrosoalkylhydroxylamines, according to the scheme : 154 /O . MgBr NO ^ = N— N = > = N— N< X C 2 H 5 .OH y O = N— N< M 1 H 58 GRIGNARD'S REACTION An interesting case, in which magnesium alkyl salts appear to function as reducing agents, has been examined by Franzen and Diebel, 15 *"who find that a good yield of hydrazobenzene may be obtained by the action of magnesium ethyl bromide on azobenzene : C 6 H 5 . N C 6 H 5 . N . MgBr ||+2C 2 H 5 .MgBr = | +C 4 H 10 C 6 H 6 .N C 6 H 5 .N.MgBr C 8 H 6 .N.MgBr C 6 H 5 . NH | +2H 2 = | +2Mg(OH)Br C 6 H 6 .N.MgBr C 6 H 6 . NH In a somewhat similar manner, benzaldazine is reduced to benzaldehydebenzylhydrazone : C 6 H 5 .CH:N C 6 H 5 . CH (MgBr). N. MgBr | +2C 2 H 5 M g Br = | +C 4 H a0 C 6 H 5 .CH:N C 6 H 6 .CH = N C fi H, . CH (MgBr) . N . MgBr C 6 H 5 . CH„ . NH | +2H 2 0= | +2Mg(OH)Br C 6 H 6 .CH = N C 6 H 5 .CH = N Busch and Fleischmann 166 find that, in addition to this reaction, which may even proceed to the extent of forming dibenzylhydrazine, the normal addition also occurs. Thus, magnesium phenyl bromide and benzaldazine yielded a mixture of benzaldehyde- benzylhydrazone and benzaldehydediphenylmethyl- hydrazone : (C 6 H 6 ) 2 CH-NH C 6 H fl .CH=N NITROGEN COMPOUNDS 59 The action of organomagnesium salts on quaternary ammonium halides, which react with alkali to form pseudobases, has been investigated by Freund and his co-workers, 157 who find that they yield substances which differ from the pseudobases by containing a hydrocarbon residue in place of the hydroxyl group. The action of magnesium ethyl bromide on quinoline methiodide may be considered as typical. It leads to the formation of l-methyl-2-ethyldihydroquinoline : CH„ Addition likewise occurs even if the hydrogen atom in position 2 is replaced by an alkyl group : 2-methylquinoline methiodide and magnesium ethyl bromide yielded 1 : 2-dimethyl-2-ethyldihydroquino- line: ;CH„ N CH„ I Dimroth 168 has utilised organomagnesium salts in the preparation of diazoamino - compounds. Thus, diazoaminomethane (dimethyltriazine) was obtained by leading a current of methylazoimide into a well- cooled, ethereal solution of magnesium methyl iodide 60 GRIGNARD'S REACTION and decomposing the product by means of a concen- trated aqueous solution of ammonium chloride : CH 3 .MgI + N v CH 3 .N = N.N.CH 3 || >N . CH S = | N< Mgl CH„.N = N.N.CH 3 | +H 2 = CH 3 .N = N.NHCH 3 + Mg(OH)I Mgl The interaction of organomagnesium salts and nitro-compounds does not appear to have been very thoroughly examined. Nitrobenzene, ethyl iodide, and magnesium react in benzene solution with the formation of phenylethylamine and azobenzene. Nitroethane and amyl nitrite, according to Mouren, 159 are converted by magnesium ethyl iodide into diethylhydroxylamine. Strecker 160 finds that neither magnesium ethyl iodide nor magnesium phenyl bromide reacts with nitrogen trichloride dissolved in benzene. Formation of Additive Compounds by Means of Grignard's Reagents. In the foregoing section on nitrogen compounds, reference has been made to the addition of Grignard's reagents to the — C = N — and — N = N — groups. The interaction of organomagnesium salts and un- saturated carbon compounds has been extensively investigated by Kohler and his co-workers, 161 who find that, in the case of a/3 unsaturated ketones, the course of the reaction depends upon the nature of the ketone. If the latter contains the methyl group attached to the carbonyl group, it reacts as a ADDITIVE COMPOUNDS 61 saturated ketone, and the final product is a tertiary- alcohol; if, however, a phenyl group is next the carbonyl group, addition of the magnesium alkyl salt occurs in the aS position, and a ketone results : /CH 3 I. C 6 H 6 . CH = CH— C< +R.MgHal ^O / CH 3 = C 6 H 6 . CH = CH— C— OMgHal \r ,CH 3 C 6 H 5 . CH = CH— C— OMgHal + H 2 \r ,CH 3 = C 6 H 5 .CH = CH— C— OH +Mg(OH)Hal. \r c a II. C (i H« i .CH = CH— c/ " 5 +R.MgHal 6 ^O AH 5 C.H,.CH— CH = C<^ | X)MgHal R /C 6 H 6 C.H..CH— CH = C< +H 2 | ^OMgHal K = C 6 H 5 . CH . CH 2 . CO . C G H 5 + Mg(OH) Hal. R Doubly unsaturated ketones containing the group __C = C— C = C— behave similarly. Thus, cinnamyl- ideneacetophenone, C(i H 5 .CH = CH-CH = CH.c/ ° 62 GRIGNARD'S REACTION and magnesium phenyl bromide yield the ketone /3-phenyl-/3-styrylpropiophenone, C 6 H 6 . CH = CH— CH(C 6 H 6 ) . CH 2 . CO . C 8 H 5 , addition again occurring in the a-S position. Further, it has been found that certain ketones can act in both the above ways, and that the relative proportions in which a/3 and aS addition takes place depend upon the nature of the unsaturated compound, the number and arrangement of the hydrocarbon residues, and the character of the magnesium derivatives. A typical case of the addition of organomagnesium compounds to unsaturated esters has been inves- tigated by (Miss) Keynolds, 162 who has studied the interaction of Grignard's reagents with the isomeric methyl esters of cinnamylideneacetic acid, C 6 H 5 — CH = CH— CH = CH— C0 2 CH 3 . The forma- tion of the following compounds is possible : — I. Tertiary alcohols, formed by replacement of the methoxy group and addition of the magnesium com- pound to the carbonyl group : C 6 H 5 . CH = CH— CH = CH— CR 2 OH. II. Unsaturated ketones, C 6 H 5 . CH = CH - CHR . CH 2 . C . O . R, produced by a : <5 addition and simultaneous replace- ment of the methoxy group : /OCH 3 C,3H 6 .CH = CH— CH = CH— c/ + 2R.MgBr ^O = C 6 H 6 .CH = CH.CH— CH = C— R I | +Mg(OCH s )Br R OMgBr ADDITIVE COMPOUNDS 63 C 6 H 5 . CH = CH— CH— CH = C— R I I +H 2 R OMgBr = C 6 H 5 .CH = CH— CH— CH = C— R | | +Mg(OH)Br R OH C e H 5 . CH = CH . CH— CH = C— R I R OH = C„H 5 .CH = CH.CHR.CH 2 .CO.R. III. Unsaturated esters of the type C 6 H 5 . CH = CH . CHR . CH 2 . C0 2 CH 3 , formed by aS addition only : /OCH 3 C r H,.CH = CH— CH = CH— c/ +R.MgHal /OCH 3 = C,.H r .CH.CH— CH— CH = C< | ^OMgHal R C„H. .CH = CH.CH— CH = C< + H 2 R /OCHg ^OMgHal ,OCH 3 C r H-.CH = CH— CH— CH = C< +Mg(OH)Hal 6 ° | X)H R OCH„ C K H, . CH = CH . CH— CH = CSiO + G,H 6 MgBr C 6 H 5 . CH 2 s /C 2 H 5 C 6 H,CH 2S yC 9 H e , Si< C 2 H/ ^OMgBr C 2 H/ x -OH >< 68 GRIGNARD'S REACTION Bygden 17 ° has described similar compounds in addi- tion to hexamethylsilicoethane, (CH 3 ) 3 Si. Si(CH 3 ) 3 , which resulted from the interaction of silicon hexa- chloride with magnesium methyl bromide in ethereal solution. Martin 171 has observed that complex silicon com- pounds containing silicon chains, such as 0=Si— OH C„H,— Si— OH I HO— Si— OH C 2 H 6 Si(OH) 2 , are obtained when a solution of silicon tetrachloride (1 mol.) in dry ether is brought into reaction with magnesium (2 atoms) and ethyl bromide (1 mol.), and the product decomposed by water. Action of Grignard's Reagents on various Inorganic Substances. The action of organomagnesium salts upon sulphur chloride has been examined by Strecker, 175 who found that this substance yielded phenyl disulphide when treated with magnesium phenyl bromide. According to Ferrario, 173 however, a rather more complex change takes place, since the product obtained by him from the same two substances was decomposed by water with the formation of chlorobenzene, bromobenzene, diphenyl, phenyl sulphide, phenyl disulphide, phenyl trisulphidc, and phenyl tetrasulphide. Sulphur INOBGANIC COMPOUNDS 69 dichloride and sulphur tetrachloride furnished similar products when similarly treated. Khotinsky and Melamed m observed that organo- magnesium compounds react with boric esters in much the same manner as with the esters of ortho- carbonic and orthosilicic acids, only one alkoxy group being replaced by a hydrocarbon radical : /OR' R B— OR' + R.MgHal = B— OR' + Mg(OR')Hal. N)R' ^OR' ' The resulting esters of alkylated boric acids were readily hydrolysed by water to the corresponding acids. Aryl boric acids were best prepared by the action of isobutyl borate on aryl magnesium halides ; but, on the other hand, the best yields of akyl boric acids were obtained from methyl borate. It was also noticed that methyl borate has a methylating action on magnesium phenyl bromide, since toluene was obtained in addition to phenyl boric acid. According to Strecker, 175 only one of the chlorine atoms of boron trichloride is replaced when the latter substance reacts with magnesium phenyl bromide. The ultimate product of the action is phenyl boric acid, resulting from the decomposition of phenylborondichloride by water. Pf eiffer and Schnurmann 176 have prepared a series of tin alkyl compounds by the action of Grignard's reagents upon tin tetrahalides. Tin tetraethyl and tin tetraphenyl were obtained from stannic chloride 70 GRIGNARD'S REACTION and magnesium ethyl bromide and magnesium phenyl bromide respectively : SnCl 4 + 4C a H,MgBr = Sn(C (i H 5 ) 4 + 4MgBrC], whilst magnesium benzyl chloride yielded tribenzyl- stannic chloride, Sn(C 6 H s CH 2 ) 3 Cl. The same authors 177 have also converted tin tetraiodide into methylstannic tri-iodide, CH 3 SnI 3 , and trimethylstannic iodide (CH 3 ) 3 SnI, by means of mag- nesium methyl iodide, whilst Da vies and Kipping 178 found that diethylstannic chloride, (C 2 H 6 ) 2 SnCl 2 , was produced when stannic chloride (1 mol.) and ethyl bromide (1 mol.) reacted with magnesium in ethereal solution. 179 In a further series of papers, Pfeiffer and Schnur- mann 180 have extended their researches to a variety of inorganic chlorides. Lead chloride was found to react readily with magnesium phenyl bromide with the production of lead tetraphenyl and deposition of metallic lead : 2PbCl 2 + 4C H 5 MgBr = Pb(C H fi ) 4 + 4MgClBr + Pb. Under similar circumstances, mercuric chloride yielded mercury diphenyl, Hg(C 6 H 6 ) 2 , whilst mercurous chloride also readily yielded the same product, with liberation of metallic mercury. The triphenyl derivatives of phosphorus, arsenic, and bismuth were readily obtained from the corre- sponding trihalides. Antimony triphenyl and tri-p- tolyl were also prepared from antimony trichloride, and magnesium phenyl bromide and magnesium p- tolyl bromide respectively. ARSENIC COMPOUNDS 71 Sachs and Kantorowicz 181 found that finely powdered arsenious oxide readily reacted with ethereal solutions of organomagnesium salts. With a cold solution of magnesium phenyl bromide it yielded diphenylarsenic oxide, (O fl H 6 ) 2 As . . As(C 6 H 5 ) 2 , whilst prolonged heating of the two substances resulted in the formation of triphenylarsine, As(C 6 H 5 ) 3 . SECTION III. Theoeetical. In what has already been written, it has been tacitly assumed that organomagnesium halides may be represented by the general formula, R . Mg . Hal. The use of such formulae does not, in general, involve any inaccuracy, and has been adopted for the sake of simplicity and economy of space. Indeed, such so-called " individual " complexes have been prepared by Tschelinzeff, 182 by the interaction of magnesium and various halides in benzene solution in the presence of a trace of ether or anisole. They form white masses which contain no ether, and which exhibit all the characteristic reactions of Grignard's reagents. The preparation of magnesium organic salts in ethereal solution, however, does not lead to the production of "individual" compounds. Thus, when ethyl iodide reacts with magnesium in the presence of ether and the reaction-product is heated under diminished pressure in a current of dry hydrogen, the ether is not completely expelled, a portion of it being so firmly retained that it can only be partly removed even at 100° to 125°. In the residue there exists the compound, C 2 H fi MgI . (C 2 H 6 ).,0. 72 THEORETICAL 73 This ether was at first regarded by Grignard and Blaise as playing the part of_ether_oi£ry-sMh^ation. v. Baeyer, 183 however, proposed to formulate these substances as oxonium compounds, the oxygen atom being quadrivalent : R/ x Hal An alternative formula was proposed by Grignard : R x ,MgHal W X R, whilst the compounds have also been regarded in accordance with Werner's ideas and formulated : ( \0 . . . MgR J Hal If the formula of v. Baeyer be accepted, isomerides of the type R x /MgR' R/ X Hal R' /MgR and >0< R/ >Hal appear possible. Tschelinzeff', indeed, accords pref- erence to v. Baeyer 's formula, since he claims to have established cases of isomerism of the above type in which R = C 2 H 5 , and R' = C 3 H 7 , C 4 H 9 , C 5 H n or C 6 H 5 . Grignard 184 maintains that the above result does not invalidate the formula proposed by him, since it may be assumed that the two additional valencies of oxygen in oxonium compounds have 74 GRrGNARD'S REACTION not the same value as the normal valencies, so that R 2 OR'MgHal is not necessarily identical with RR'ORMgHal. In general, the ether complexes are somewhat difficult to characterise, since they are, as a rule, uncrystallisable substances. Zerewitinoff 185 has, however, succeeded in preparing a crystalline com- pound, C 5 H 11\ A C,H,/ X MgCH :i , by the interaction of methyl iodide and mag- nesium in the presence of amyl ether, and in, determining its composition by direct analysis. Indirectly, the correctness of the formula for the ether complexes proposed by Blaise has been verified by Tschelinzeff, 186 who was able to show that individual magnesium compounds combine with ether in the presence of benzene, and that the com- plexes so formed are soluble in this solvent. When ether was gradually added to a suspension of an individual magnesium compound in benzene, the latter gradually dissolved, a perfect solution being obtained when the ether had been added in the proportion of one molecule for every molecule of magnesium organic halide. The same author 187 has measured the heat evolved by the combination of individual magnesium com- pounds with ether, and is led to the conclusion that two distinct processes are involved in the preparation of ethereal solutions of Grignard's reagents : (1) the 'formation of magnesium alkyl halides, and (2) the THEORETICAL 75 transformation of these substances into their ether complexes. ' Subsequently Tschelinzeff 188 succeeded in showing, both by analytical and thermochemical methods, that organomagnesium compounds are capable of forming ether complexes which contain 2 molecules of ether in combination with one molecule of "individual" compound. Such a substance was actually isolated when the residue, obtained by evaporating the solvent from an ethereal solution of magnesium amyl iodide, was not too strongly heated. He proposes the general formula : H\ /MgR R/ M = 0< X R It has already been pointed out that tertiary amines, like ethers, are capable of catalysing the action of organic halides on magnesium in the presence of an inert solvent. Tschelinzeff 189 has shown that they are similarly able to unite with "individual" magnesium organic compounds, with the formation of substances which are termed atninates. Such sub- stances, however, never appear to contain more than 1 molecule of the tertiary amine. Their constitution may, in all probability, be expressed by one of the two formulae (which are analogous to those adopted for the ether complexes) : R \ R R \ .MgR R— N< or R— N< r/ W r/ Xi 76 GRIGNARD'S REACTION " Aminates," whilst never able to combine with a second molecule of a tertiary amine, can nevertheless unite with a molecule of ether to yield mixed ether- amine complexes. Replacement of the ether by amine, or conversely, may occur, the change pro- ceeding in one direction or the other according to the relative stability of the resulting additive products, as shown by thermochemical measurements. In the action of ethers on "etherates" and of amines on "aminates," the compound produced is that which has the greater heat of formation. The action of ethers upon aminates is at first additive, and, with aliphatic amines, results in the formation of ether-amine complexes, since ethers cannot replace aliphatic amines. Aromatic amines may be completely dis- placed with the formation of dietherates. The action of an aliphatic amine on an etherate must be, at first, substitutive, but the replaced ether is added on with the formation of a mixed amine-ether complex. It is concluded from these results that addition takes place at two dissimilar positions in the molecule of the magnesium organic halide. Mode of Catalytic Action of Ethers and Tertiary Amines. Attention has already been drawn to the fact that the action of organic halides on magnesium in the presence of neutral solvents is very greatly expedited by the presence of a trace of ether or of a tertiary amine. Tschelinzeff supposes that the activity of these catalysers depends upon their dissociating THEORETICAL 77 action on organic halides, which leads to the forma- tion of oxonium or ammonium compounds : >0+R'I = \o/ W R/ \l R \ R \ A' R_N+R'I = E— N< */ h/ Xi and that these compounds then react with magnesium, with the formation of organomagnesium halides and the regeneration of the ether or tertiary amine : R v /R' R v ,R' >0< +Mg = >0 + Mg< R/ M R/ M R R' R 3 n/ ' + Mg = R 3 N + Mg. ' If this is actually the case, it should be possible to find an ether the oxonium compound of which can dissociate in two directions, with the production of two different organomagnesium compounds : R. R\ .R RV R'I + \o ^— >0< — > >0 + RI. Stadnikoff 190 has discovered such a case in the action of ^-propyl iodide and magnesium in the presence of triphenylmethylethyl ether, (C 6 H 6 ) 3 C . . C 2 H 6 . When the product of this reaction was treated with dry carbon dioxide and subsequently with water, 78 GRIGNARD'S REACTION butyric acid and triphenylmethane were obtained. This points to a fission of the complex molecule (CoW [Mgl >o< C 2 H 5 CgHj into magnesium w-propyl iodide on the one hand and magnesium triphenylmethyl iodide on the other. The action of carbon dioxide on the latter substance would, a priori, be expected to lead to the formation of triphenylmethylacetic acid, (C 6 H 6 ) 3 C . COOH, but it has been shown that reaction only takes place in this particular direction under certain conditions. In a second contribution on the same subject, Stadnikoff 191 has further investigated this interesting point, and has shown that the oxonium compound formed during the interaction of propyl iodide with magnesium in the presence of triphenylmethylethyl ether can be caused to dissociate into triphenylmethyl iodide and ethylpropyl ether, (C.HAC I 6 /\ whilst, with the same ether, -isobutyl iodide in the presence of magnesium forms an oxonium compound which decomposes with the production of triphenyl- methyl iodide and ethyKsobutyl ether. Diphenyl- THEORETICAL 79 methylpropyl ether and isobutyl iodide similarly react to yield the oxonium compound, ( C G H fl)2 CH \ A C 3 H 7 C 4 H 9> which subsequently undergoes decomposition in three ways, giving (1) diphenylmethyl iodide, (C 6 H 5 ) 2 OHI, and propyKsobutyl ether, C 3 H 7 . . C 4 H 9 ; (2) iso- butyl iodide and diphenylmethylpropyl ether, (C 6 H 5 ) 2 CH . . C 3 H 7 ; and, probably (3) propyl iodide and diphenylmethyKsobutyl ether, (C 6 H 6 ) 2 CH.O.C 4 H 9 . SECTION IV. Mixed Organometallic Derivatives of Zinc. The great activity of magnesium organic compounds, to which the latter owe their extended use as synthetic agents, is, under certain circumstances, very disadvantageous, since, when a substance possesses two or more points of attack for the Grignard reagent, it is frequently difficult or even impossible to limit the attack in such a manner that only one group is acted on. Thus, in the case of acid chlorides, the action of organomagnesium derivatives generally takes place in such a manner that not only is the chlorine atom replaced, but, in addition, the carbonyl group is attacked. A reagent of less general activity and of greater ease of control is thus desirable — not to replace the Grignard's reagents, but rather to act as their complement in those cases in which the latter prove themselves too energetic. Such a reagent has been found by Blaise and his co-workers in the mixed organometallic derivatives of zinc. 192 In the formation of zinc alkyls by the interaction of zinc and alkyl iodide, the intermediate formation of substances of the type R-Zn-I has been generally assumed, although such compounds have never been analysed. Bewad 193 has attempted to bring such substances into reaction by the addition of ether — a method, however, which has still much of the 80 ORGANOZINC COMPOUNDS 81 inconvenience of the preparation of zinc alkyls. Michael, on the other hand, has shown that zinc- copper couple is soluble in absolute ethereal solutions of alkyl iodides at the temperature of the water-bath, and from the couple, ethyl iodide and benzoyl chloride, has prepared propiophenone in 30 per cent, yield. There would thus appear to be a striking analogy between the mode of formation of organometallic compounds of zinc and those of magnesium. If, however, the ethereal solution of zinc ethyl iodide is evaporated at the ordinary pressure or under a pressure of 22 mm. of hydrogen, ether and zinc ethyl are obtained. At the ordinary temperature and at a lower pressure of hydrogen a colourless mass is finally left, the analyses of which do not yield concordant results. Since, however, the presence of ether is necessary to the reaction under the experi- mental conditions actually adopted, it seems possible that compounds of the type c 2 h 5Xq/ i C 2 H 5 / ^ZnR are actually formed, by which the ether is only loosely retained. On repetition of Michael's experiment (see above), Blaise made the remarkable discovery that the activity of the organozinc derivatives is not exhausted by the addition of an equimolecular amount of benzoyl chloride, as would be expected, but that nearly five molecular proportions of the latter were required before action ceased. Among the products of the change were methane, ethyl chloride, ethyl iodide, F 82 ORGANOZINC COMPOUNDS ethyl benzoate, a small quantity of acetophenone, and a brown resin. The latter substance evidently resulted from the condensation of acetophenone with loss of water, which accounted also for the evolution of methane. The formation of the other substances is explained by the scheme : •I C,H„ -+ C 6 H 6 .CO.OC 2 H 5 + C,H s I CH 3 — Zn- Cl- - C 2 H 5 O- — CO . C, ; H, + CH,ZnCl CKLZn- 0(C 2 H 5 ) 2 , Br- Zn-C e H 5 , CI CI Mg or (C 2 H 5 ) 2 o/ \o(C 2 H 6 ) 2 , Cl-Zn-C 6 H 5 . CI Br 86 ORGANOZINC COMPOUNDS Mode of employment of Mixed Organozinc Compounds. — These compounds have been chiefly- employed in conjunction with acid chlorid es of varying types, and with compounds containing loosely held halogen atoms. The acid chlorides should be purified by distillation. If this is impos- sible, they should be prepared preferably by means of thionyl chloride, since excess of the latter may be readily removed by gentle heating under diminished pressure. For a given quantity of acid chloride, an excess of 25 to 33 per cent, of the calculated quantity of alkyl zinc iodide should be employed in the case of primary alkyl iodides; with secondary alkyl iodides double the calculated quantity may be employed, whilst with zinc aryl halides an excess of 50 per cent, is generally sufficient. The acid chloride, dissolved in toluene, is added drop by drop with constant shaking to the solution of the reagent, the temperature being maintained at about 0°. Eeaction occurs readily. The temperature is then allowed to rise somewhat, and as soon as all odour of acid chloride has disappeared, the solution is again cooled and decomposed by the addition of successive small quantities of water and a little dilute sulphuric acid. The solution separates into two layers. All the iodine is present in the aqueous portion. The toluene solution contains the required product, together with small quantities of zinc, which may be eliminated by agitation with ammonia or ammonium sulphate solution. Finally, the toluene solution is shaken with very dilute sulphuric acid A TYPICAL EXAMPLE 87 and then with a solution containing potassium hydrogen carbonate and sodium thiosulphate, after which it is dried over sodium sulphate. In general, organozinc compounds do not react in the cold with the halogen, ketonic, carbalkoxy, ethylenic, or alkoxy groups, and herein lies their chief interest. They are capable, however, of react- ing with halogen atoms attached to a carbon atom to which an oxygen atom is also united, and with other mobile halogen atoms. 194 Normal Action on Acid Chlorides. — The normal action between an acid chloride and a zinc organo- salt results in the formation of a ketone : R.CO.Cl + I.ZnR' = R . CO . R' + ZnlCl. The yields are generally excellent (75 to 90 per cent.), and the products frequently pure after one distillation. A small quantity of the ethyl ester derived from the acid chloride employed is sometimes present. This can often be removed by treatment of the product with potassium hydroxide, or by formation of a solid derivative of the ketone {e.g. the semicarbazone), and subsequent decomposition of the latter. Formation of Tertiary Alcohols from Acid Chlorides. — Acid chlorides or esters are readily trans- formable into tertiary alcohols by the action of cold solutions of organozinc salts when an electro- negative group or atom is contained in the molecule in the immediate neighbourhood of the — COOl or COOAlk group. Thus, ethyl oxalate, in which one carbethoxy group accentuates the negative 88 ORGANOZINC COMPOUNDS character of the other, is quantitatively transformed into the ethyl ester of a dialkylglycollic acid : COOC 2 H 5 CR 2 (OH) I — ■> I COOC 2 H 6 COOC 2 H 5 whilst ethoxyoxalyl chloride, COOEt.COCl, under- goes a similar change, in which, however, the tertiary alcohol at first produced is esterified by the action of a further portion of the acid chloride : COOC 2 H 5 CO.OC 2 H 6 COOC 2 H 6 I — > I — > I COC1 R 2 :C— OZnl R 2 .C.O.CO . C0 2 C 2 H 5 . Similarly, chloroacetyl chloride and ethyl zinc iodide give, as main product, the chloroacetic ester of chloro- methyldiethylcarbinol : CH 2 C1 . C(Et) 2 . O . CO . CH 2 C1. Also, the chlorides of a/3 unsaturated acids yield mainly ketones together with small quantities of the ester derived from the tertiary alcohol : 195 R.CH = CH.COCl > R.CH = CH.C(R') 2 .O.CO.CH = CH.R but, as the electronegative influence of the ethylenic linking is lessened by substitution the amount of ketone formed increases, so that the latter is practically the only compound obtained from acid chlorides of the types : k )>C = CH.COCl and R \ V^CRj.COCl R'/ KETONES 89 The action of organic zinc halides on the chlorides of a-acetoxy acids is of particular interest. 196 Ap- parently the carbonyl group of the acetyl radical is attacked preferably to the chlorine atom of the acid chloride. Cyclic acetals are thus obtained : /COC1 .CO— CI R.CH< ^R.CH< /OZnl \0— CO— CH 3 \0— C< I X CH 3 R' CO — o >■ R.CH.< | /CH, \ C< \R' which, on hydrolysis, yield ketones : CO— o R.CH/ I /CH3 +H 2 = R.CH(OH).COOH NO C< +CH..CO.R. One group of the ketone is thus derived from the acetyl radical, the other from the organozinc derivative. A means is thus provided of obtaining ketones from those acid chlorides which, with zinc alkyl halides, normally yield tertiary alcohols. 197 The acid chloride is first allowed to react with an a-hydroxy acid, the new acid converted into its chloride, which is then subjected to the action of the requisite organozinc derivative. The cyclic acetal so obtained yields, on hydrolysis, the required ketone. The preparation of chloromethylethylketone may be considered as typical. Chloroacetyl chloride and zinc ethyl iodide yield mainly the chloroacetic ester of chloromethyldiethylcarbinol (see above). If, 90 ORGANOZINC COMPOUNDS however, chloroacetyl chloride is allowed to react with lactic acid, a-chloroacetoxypropionic acid is formed, the chloride of which reacts with ethyl zinc iodide to yield the corresponding cycloa,cetaA. This, when hydrolysed by acid, is resolved into chloro- methylethyl ketone and lactic acid : CH 3 .CHOH + CI .CO .CH 2 C1 -> CH 3 .CH . O .CO .CH 2 C1 COOH COOH CO . CI CO O I -^11/ CH„.CH.O.CO.CH„Cl CH,.CH— O— 0/ '2"6 CH 2 C1 CH 3 . CHOH . COOH + CH 2 C1 . CO . C 2 H 5 . In certain cases, catalytic decomposition of acid chlorides may he occasioned by the action of zinc organic salts — especially in the case of a-alkoxy acid chlorides. 198 Under these circumstances carbon monoxide is evolved, and the residual a-chloroether then reacts normally with replacement of the halogen atom : R.CH(OC 2 H 5 ).COCl >■ R.CHCl(OC 2 H 5 ) + CO > R.CHR'(OC 2 H 6 ). Reducing action of mixed organozinc derivatives has only been observed in the case of sulphonic chlorides. Thus, benzene sulphonic chloride and ethyl zinc iodide give a small quantity of the sulphone, but the main product of the reaction is found in the form of zinc benzene sulphinate. VARIOUS PRODUCTS 91 Mixed organozinc compounds readily react with substances containing mobile halogen atoms — such compounds generally containing the halogen atom united to carbon, which is itself attached to oxygen. Such compounds comprise a-chloroethers, acid chlorides of the succinic and glutaric series, ethyl chloroethoxy- acetate, OHCl(OC 2 H 5 ). COOC 2 H 6 , and ethyl dichloro- ethoxyacetate, CC1 2 (0C 2 H 5 ). COOC 2 H 5 . The use of a-chloroethers possesses no feature of special interest, since these substances react equally well with Grignard's reagents. Acid chlorides of the succinic series yield, practically entirely, y-lactones, whilst those of the glutaric series give a mixture of ^-lactone and (5-keto acid. 199 Ethyl chloroethoxy- acetate 20 ° reacts very readily with alkyl zinc iodides, giving excellent yields of ethyl alkylethoxyacetate, CHCl(OEt).COOEt y R.CH(OEt).COOEt 3 whilst similar compounds are also readily obtained from ethyl dichloroethoxyacetate, 201 CCl 2 (OC 2 H 5 ).COOC 2 H 6 y CR 2 (OC 2 H 6 ).COOC 2 H 5 . BIBLIOGRAPHY 1 Compt. rend., 1899, 128, 110. 2 Compt. rend., 1900, 180, 1322. 3 Ber., 1906, 89, 1952. 4 Ber., 1903, 36, 2775. J Ber., 1903, 86, 4296. s Ber., 1905, 83, 2759. 7 Ber., 1906, 39, 1132. 8 Compare Zelinsky, Chem. Cent., 1903, II, 277. 9 Ber., 1905, 88, 2078. 10 Ber., 1905, 38, 3620. 11 Compt. rend., 1905, 141, 830. 12 J.C.S., 1908, 93, 68. 13 J.C.S., 1908, 93, 1822. 14 Compare Kahan, Trans., 1908, 93, 133. 15 Ber., 1908, 41, 2302. 10 Ber., 1904, 37, 4534. 17 Ber., 1903, 36, 668, 4272 ; ibid., 1904, 37, 746. 18 Ber., 1903, 36, 2608. 19 Compare McKenzie, B.A. Keport, 1907. 20 Ber., 1905, 38, 905. 21 J.C.S., 1908, 93, 1827. 22 J. pr. Chem., 1908, pi], 77, 393; J. Euss, Phys. Chem. Soc, 1908, 40, 381. 23 J.C.S., 1911, 99, 296. 24 Compt. rend., 1899, 128, 110. 26 Ber., 1909, 42, 435. 26 Gazzetta, 1911, 41, i, 273. w Tissier and Grignard, Compt. rend., 1901, 132, 835. 25 Meunier, Compt. rend., 1903, 136, 758. 03 94 BIBLIOGRAPHY 28 Houben, Ber., 1905, 38, 3019. 30 Tschugaeff, Ber., 1902,85.3912 ; Hibbert and Sudborough Proc. Chem. Soc, 1903, 19, 285 ; Zerewitinoff, Ber., 1907, 40, 2023. 31 Ber., 1908, 41, 2233. 32 Zeitsch. anal. Chem., 1911, 50, 680. 33 J.C.S., 1906, 89, 380. 34 v. Braun and Deutsch, Ber., 1912, 45, 2176. 35 Ber., 1911,44, 1918. 30 Compt. rend., 1901, 132, 831. 37 Ber., 1907, 40, 3049. 38 Ber., 1906, 39, 1461, 2957. 39 Ber., 1903, 36, 2116. 40 Ber., 1903, 36, 3083. 41 Ber., 1904, 37, 453. 42 Ber., 1904, 37, 1429. 43 Ber., 1905, 38, 511. 44 J.C.S., 1906, 89, 839. 45 Compt. rend., 1900, ISO, 1322. 40 Compt. rend., 1904, 139, 481. 47 Bull. Soc. chim., 1910, [4], 7, 431. 48 Ber., 1912, 45, 1253. 49 Gazzetta, 1908, 38, i, 625. 60 Ber., 1911, 44, 3062. 51 Ber., 1912, 45, 1250. 52 Bouveault, Bull. Soc. chim., 1903, [3], 29, 1051. 53 Grignard and Tissier, Compt. rend., 1902, 134, lu7. ^ Ber., 1912, 45, 1250. 55 Compt. rend., 1904, 138, 1048. m Compt. rend., 1903, 136, 1260. 57 Compt. rend., 1902, 184, 552. •'« C.B., 1907, ii, 445. ° 3 Grignard, Compt. rend., 1905, 141, 44 ; A. Ch., 1907, [8], 10, 23. 00 Grignard, Compt. rend., 1900, ISO, 1322. 01 Grignard, Compt. rend., 1901, 132, 336. 02 J. Pharm. Chim., 1911, [viii], 4, 294. 03 Trans. Nova Scotia Inst. Sci., 1908, 11, [4], 593. BIBLIOGRAPHY 95 64 J.C.S., 1904, 85, 654. 65 Grignard, Compt. rend., 1904, 138, 152. m Grignard, Compt. rend., 1900, 130, 1322 ; Tissier and Grignard, Compt. rend., 1901, 182, 683. 87 Grignard, Compt. rend., 1903, 136, 815. 68 Weigert, Ber., 1903, 36, 1007. 69 Grignard, Tissier and Grignard, loc. cit. 70 Bull. Soc. chim., 1910, [iv], 7, 836. 71 Ber., 1908, 41, 1582. 72 Bauer, Ber., 1904, 87, 737 ; 1905, 38, 240 ; Arch. Pharm., 1909, 247, 220. 73 Houben and Hahn, loc. cit. 74 Grignard, A. Ch., 1907, 10, 23, 35. 76 D.E.P., 117, 615 ; Abstr. Chem. Soc, 1907, 1, 275. 76 Ber., 1904, 87, 2753 ; Amer. Chem. Journ., 1905, 38, 193. 77 Compare also Tiffeneau and Dorlencourt, Ann. Chim. Phys., 1909, [viii], 16, 237. 78 J.C.S., 1910, 97, 473. 79 Compt. rend., 1901, 182, 833. 80 Bull. Soc. chim., 1903, [iii], 29, 683. 81 Ber., 1904, 37, 3640. 82 Shibata, J.C.S., 1909, 95, 1449. 83 J.C.S., 1904, 85, 1666. 84 Ber., 1904, 37, 2753. Si Ber., 1903, 36, 1625. 80 Clarke and Carleton, J. Amer. Chem. Soc, 1911, 33, 1966. 87 Ber., 1904, 37, 2892. 88 Compt. rend., 1903, 136, 158 ; Bull. Soc. chim., 1904, [3], 31, 33. 80 Compt. rend., 1909, 148, 930. 80 Bull. Soc. chim., 1909, [iv], 5, 405. 91 Compt. rend., 1904, 138, 813, 975 ; ibid., 139, 59. 9a Bull. Soc. chim., 1907, [iv], 1, 1198 ; compare Hamonet, ibid., 1908, [iv], 3, 254; Dionneau, ibid., 1910, [iv], 7, 327; v. Braun and Deutsch, Ber., 1912, 45, 2176. *» Ber., 1904, 37, 3987. » 4 Compt. rend., i910, 151, 322. 05 Ber., 1903, 36, 4152. 96 BIBLIOGRAPHY 90 Ber., 1904, 37, 186; compare Bodroux, Compt. rend., 1904, 138, 92. 97 J. Russ. Phys. Chem. Soc., 1910, 42, 1279. 08 Houben, Chem. Zeit., 1905, 29, 667. 99 Compt. rend., 1903, 137, 987. 100 Ber., 1904, 37, 875. 101 J.C.S., 1906, 89, 273. 102 Compt. rend., 1901, 132, 38 ; 133, 299. 103 Compt. rend., 1903, 137, 575. 104 J.C.S., 1908, 93, 310. 105 Ibid., 1909, 95, 1583. 100 Proc. Irish. Acad., 1912, 30, 1. 107 Ber., 1910, 43, 2553. 108 Ber., 1904, 37, 628. 109 Ber., 1906, 39, 2958. 110 Ber., 1910, 43, 1012. 111 Ber., 1903, 86, 3004 ; ibid., 1907, 40, 1584. 112 Compt. rend., 1903, 137, 710. 113 Ber., 1911, 44, 1922. 114 D.R.P., No. 166,898. "* Grignard, Bull. Soc. chim., [3], 31, 751 ; Houben, Ch. Z., 1905, 29, 667 ; Zelinsky, ibid., 1904, 28, 303. 116 Ber., 1909, 42, 3721. 117 Chem. News, 1904, 90, 276. 118 C, 1901, 11, 622. 119 Compare Spencer and Stokes, J.C.S., 1908, 98, 70. 120 Ber., 1906, 89, 635. 111 Ber., 1903, 36, 3004. 122 Ber., 1907, 40, 1585. 123 Compare E. Meyer and Togel, Ann., 1906, S47, 55. 124 Compt. rend., 1904, 188, 152. 125 Compt. rend., 1903, 187, 710. 126 Gazzetta, 1908, 38, i, 625. 127 Ber., 1907, 40, 1037. 128 Compt. rend., 1902, 185, 627 ; A. ch., 1902, [7], 27, 553. 129 J.C.S., 1906, 89, 365. See also McKenzie and Wren, ibid., 1906, 89, 688 ; McKenzie and Miiller, ibid., 1909, 95, 544. 130 Gazzetta, 1911, 41, 1, 255. BIBLIOGRAPHY 131 Ber., 1909, 42, 3729. 132 Ber., 1904, 87, 2102. 133 Gazzetta, 1911, 41, [i], 11. 134 Oddo, Gazzetta, 1911, 41, [ii], 11. 135 Ber., 1910, 43, 1131. 136 Ber., 1903, 36, 3087. 137 Ber., 1905, 38, 561. 138 Ibid., 563. 139 Ber., 1906, 39, 1736. 140 Ann., 1906, 847, 55. 141 Ber., 1908, 41, 589. 142 Compt. rend., 1911, 152, 388. 143 Compt. rend., 1912, 155, 44. 144 Ostrogovitch, Chem. Zeit., 1912, 36, 738. 145 Compt. rend., 1901, 182, 38, 478, 978. 146 Ber., 1903, 36, 585 ; 1904, 37, 874. 147 Compt. rend., 1904, 138, 1427. 148 Compt. rend., 1905, 140, 1108. 149 Ber., 1905, 38, 1761; Compare Busch, Ber., 1904, 37, 2691. 160 Ber., 1907, 40, 2096. 161 Busch and Hobein, Ber., 1907, 40, 2099. 153 Ber., 1903, 36, 2315. 153 Wieland and Gambarjan, Ber., 1906, 39, 1499. 164 Sand and Singer, Ann., 1903, 329, 190; Compare Ber., 1902, 35, 3186. 155 Ber., 1905, 38, 2716. 165 Ber., 1910, 43, 740. 167 Ber., 1904, 87, 4666 ; ibid., 1909, 42, 1101, 1746. 168 Ber., 1903, 36, 909 ; 1905, 38, 670 ; 1906, 39, 3906. 169 Compt. rend., 1901, 132, 837. 1C0 Ber., 1910, 43, 1131. 161 Amer. Chem. J., 1904, 31, 624, etc. 182 Amer. Chem. J., 1911, 46, 198. 163 Compt. rend., 1905, 140, 370. J« 4 Compt. rend., 1908, 146, 343. i" 5 Ber., 1908, 41, 2946. J* J.C.S., 1907, 91, 209, and subsequent papers. G 98 BIBLIOGRAPHY 107 Ber., 1904, 37, 1139. 168 Ibid., 1905, 88, 4132. 169 J.C.S., 1911, 99, 140. 170 Ber., 1912, 45, 707 ; ibid., 1911, 44, 2640. 171 Ber., 1912, 45, 2997. 172 Ber., 1910, 43, 1131. 17:i Bull. Soc. chim., 1910, [iv], 7, 518. 174 Ber., 1909, 42, 3090. 175 Ber., 1910, 43, 1131. I7G Ber., 1904, 37, 319. 177 Ibid., 4619. 178 J.C.S., 1911, 99, 300. 170 Compare Pfeiffer, Ber.,' 1902, 35, 3306. 180 Ber., 1904, 37, 1125, 4620. 1,1 Ber., 1908, 41, 2767. 132 Ber., 1904, 37, 4534. lsl Ber., 1902, 85, 1201. 184 Bull. Soc. chim., 1907, [iv], 1, 25C. n: > Ber., 1908, 41, 2244. m Ber., 1905, 38, 3664. ls7 Loc. cit. 188 Ber., 1906, 39, 773. 18 ' J Ber., 1907, 40, 1487. 190 Ber., 1911,44, 1157. 191 J. Russ. Phys. Chem. Soc, 1911,43, 1244. -■ 192 Bull. Soc. chim., 1911, [4], 9, 1. 193 Ber., 1908, 40, 3065. 194 Blaise and Herman, Ann. Chim. Phys., 1911, [viii], 23, 522. 196 Blaise and Maire, Compt. rend., 1907, 145, 73. 196 Blaise, Compt. rend., 1911, 154, 596, 1086. 197 Blaise, Compt. rend., 1912, 155, 46. 198 Blaise and Picard, Compt. rend., 1911, 152, 268, 446; Ann. Chim. Phys., 1912, [viii], 25, 253. 199 Blaise and Kcehler, Compt. rend., 1909, 148, 489. 200 Blaise and Picard, Bull. Soc. chim., 1912, [iv], n, 537. 201 Blaise and Picard, Bull. Soc. chim., 1912, [iv], 11, 587. INDEX Acetals, cyclic, 89 Acids, aromatic amino-, 48* carbithionic, 45 carboxylic, 43 hydroxy-, 46 sulphuric, 48 Alcohols, tertiary, 24, 62, 87 tsopropyl, 24 primary, 20 secondary, 23 Aldehydes, 36, 43 Aluminium reagents, 10 Amides, aromatic, 53 Animates, 75 Amines, secondary, 55 Amino-acids, 48 Anilides, 54 Anisylphenylpropene, 1 7 Antimony compounds, 70 Arsenic compounds, 70, 71 Asymmetric synthesis, 46 Benzoic acid, 44 Benzoin, r-, d-, and 1-, 40 Benzyl alcohol, 21 Bismuth compounds, 70 Boron compounds, 69 Campholides, 29 Catalysts, theory of, 76 use of, 5, 83 99 Diazoamino compounds, 59 Diethylphenylcarbinol, 42 Dimethylheptenol, 1 Diphenyl, 15 Ditertiary glycols, 29, 31 ESTERS, 49 unsaturated, 63 Ethers, 35 Ethyl ketone, 29 Glycols, secondary tertiary, 31 a-, primary tertiary, 29 Grignard's reagents, 2 Hydrazobenzene, 58 Hydrocarbons, 16 saturated, 13 unsaturated, 17 Hydroxy-acids, 46 Indium reagents, 10 "Individual "salts, 9 Iodo compounds, 20 Tsopropyl alcohol, 24 Ketones, 28, 39, 61, 87, 89 unsaturated, 62 Ketonic esters, 51 100 INDEX Lead compounds, 70 Lithium compounds, 1 Magnesium alkyl halides, 1 Menthadiene, A s * |9| -je-, 18 Mercury compounds, 70 Metals other than magnesium, 10 Methane, 13 Negative catalysts, 7 Nitriles, 52 Nitrogen compounds, 53 Phenols, 34 Phenolic ethers (preparation), 35 ethers (scission of), 35 Phosphorus compounds, 70 Phthalan, 32 Phthalides, 32 Polymethylenes, 15 Primary alcohols, 20 Propyl alcohol, iso-, 24 Quixols, 33 Saturated hydrocarbons, 13 Secondary alcohols, 23 amines, 55 Selenophenol, 35 Silicon compounds, 65 Solvent, effect of, 7 Stilbene, 17 Sulphides, 49 Sulphinic acids, 48 Sulphoxides, 49 Sulphur compounds, 68 Teepineol, 26 Tertiary alcohols, 24, 62, 87 Thallium reagents, 1 Thioanilides, 54 Thiols, 34 Thiophenols, 34 Tin compounds, 69 Trimethylcarbinol, 3 Triphenylcarbinol, 25 Zinc alkyl halides, 83 Zinc aryl iodides, 84 Zinc compounds, 80 PRINTED BY OLIVER AND BOYD, EDINBURGH. Cornell University Library QD 411. W94 The organometallic compounds of zinc and 3 1924 004 036 962