The Magnetic Susceptibilities of the Metal Ketyls Citation Doescher, Russell Nimtz (1934) The Magnetic Susceptibilities of the Metal Ketyls. Master's thesis, California Institute of Technology. doi:10.7907/3rap-7h11. https://resolver.caltech.edu/CaltechTHESIS:09182024-174614735 Abstract No abstract. Item Type: Thesis (Master's thesis) Subject Keywords: (Chemistry) Degree Grantor: California Institute of Technology Division: Chemistry and Chemical Engineering Major Option: Chemistry Thesis Availability: Public (worldwide access) Research Advisor(s): Unknown, Unknown Thesis Committee: Unknown, Unknown Defense Date: 1 January 1934 Record Number: CaltechTHESIS:09182024-174614735 Persistent URL: https://resolver.caltech.edu/CaltechTHESIS:09182024-174614735 DOI: 10.7907/3rap-7h11 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 16741 Collection: CaltechTHESIS Deposited By: Benjamin Perez Deposited On: 18 Sep 2024 18:19 Last Modified: 18 Sep 2024 18:21 Thesis Files PDF - Final Version See Usage Policy. 5MB Repository Staff Only: item control page

TM MAGNETIC SUSCEPTIBlLITlIDS OF THE METAL KETYLS Thesis by Russell Doee~h~r In partial fulfillment of the requirements for the degree of Master of Science in Chemistry California Institu t8 of Te~hn ology Pasade~a, California 1934 CONTENTS The dis s oc iations of several metal ketyle into free radicals are found by measurang , their magnetic susceptibilities in various solvents. It is found that the solvent has a very great effect upon the degree of dissociation. The metal ketyls are found to be only~·slightly dieeociated,, for the most cases, in benzene and diethyl ether. The preparatiom of several ketones is discussi:!d briefly. Evid~nce of non-dissociation is brought out from boilingpoint4 measurements. The Magnetic susceptibilities of tl e Met ,,, 1_ Ke t;yls Introduction I t does not seem that the question as tc whethe r the ~et .1 Ketyls sx! st- aa fr ee radicals has yet been settled~ Schlenk and his cc-workers seemed s a t isftea. thst they are completely dissocia•ted. ( in particule.r ln diethyl ether s olution) when they left the problem about 1914 . However, recently , Bachmann ha s come out with evidence of a chemical nature showing that no dissociation at all sxi sto . I ~ h E..S in a.n effort to throw some light on the wi.:) wer to this question by the ap pli cation of a new me th od, that of dete rmini ng the magnetic s u.s capt i b illti e s of '~he free :ralie:uls in va.ri m.: ~ sol- uti ons , that the present problem was undertaken. Apparatus The meta l ::e ·~y ls are very m:i. stab l e substances with re- spect to or.1gen or any ionizable hydrogen , s uch as the hydro- gen on t he - OH group of a l cohols and the encl f or ms of ketones . The,y- a.re very highly c olored, and traces of ox;:;g en or r;10 'is~ :.::-- ~ a ir coming in contact with them is evidencal by im- in figure 1 was used. Bis a gr ound glass ~oint, s o tha.t the Figure l A B Fig-u re~ 2 2 reaction flask A and. :i. ts 3-way stopcock J ma.y be detached from the rest of the apparatus. The two passages in the core of J meet each other at right angles . C is the t ub e which is :p l a c eo. be t ween the poles of the magnet. It ls dl ~ • vlded by a glass partition, the lower half being open to the rneasu:·emBnts of magneti c susceptibility are thus made against air a t one n tmc aphe:co . .i.:1.is tube is closed wl th s.. .stop cock as shown in the d.iagrs.m. • G is a three-way stopcock s.rrn.nge.i ,30 ·tha t '--he tnbe C can e:1 ther be evacua t ed or f:t.lled with dry nitrogen from the absorption train . H is a mercury seal f or obtaining small pressures in the system ! The gas wash-bottles :E1 ai·e fi.lled with a potassium hydr;,oxide soluti on of potassium pyrogallate to re move any traces of oxygen left in the nitrog en from the "tB.l-ik. D is fillec. ,v i th a.."1.h;d.ro:is calcium chloride and. .E r.ith phosphorus pentoxide so that all traces of moisture are removed from the gas. The apparatus is used a.s follows : The reaction fla.sk A with its stop-cock J is detached and partially filled with the ketone solution , after wh ich sodiurr: wire is introduced. The flask is then closed and placed in a sha3.d.ng 1.::achine for the proper length of time. AnJ oxygen initi al ly 5-:n the fl a sk is soon remo ved by the reaction with the metal ketyl, the ketone being regenerated a..."l.d reacting vfi th more sodium U...""1 til a ll the oxygen :!.~ gone . .After the re a c ·tlon is complete, the reaction flask is connected ', i th th e tube C a..'1.d 'i1 11e st c,:p c ock G, and the who l e is allowed to stand until any sediment has ..,,,; settled. The upper p art o f the ·~uba C is then filled with nitrogen by al terna.tely e.racu.a. t ing and. filling wi t1 gen until all traces of oxygen .1re ra moved. .:.i tro - ·:r.he ~ube c is then evH-cu-taJ. once more and the me ·t al kety l soh:t i o:n o.:rawn in from t;he upp'3r ~.xc f a ce of the liquid in the reaction flask . C is f'illed to a ce:i:tain mark , and the space between the liquid. and the stopcock is f ill ed with ni tre - Figure 2 is a diagram of the a;paratus for rr;easuring tue me,gnetic susc eptib ili t ;,;r Jf ~he r:1etal ~~etyl solut:tons by the Gouy method . A rod is suspended from the 1 ft sti~ - rup of the balance carrying a h ook Bat its lower end . ':::he ro d is broken at A and a.'11 arrangement made to use the pan rest. The tube containing the metal katyl is suspende d from the hook B between the poles of ·the magnet in such a way that the partition in half-way between the upper a..~d lower edg es of the pole-pieces. The whole suspended sys- tem is enclosed in a careful ly constructed wooden box to eliminate aii~ currents. ,.6. thermometer is p l aced in 'the box.for record ing temperatures . The door of this box is provided ~ith a window opposite the pole-pieces for obsei·ving t he tub e. The magnet is provided with coo l ing coils through whic h water flows. The curren·t is furnished by a di rect-current generator and is controlled by means of rheostats. The core is demagnetized b y z:ieans of c, ;.;;1::a.11. gradua lly-d.ecreased, a lternating current. The same cycle 4 of current changes is repeated each time, so that the sar:1e fie ld i s reproduced very closely~ The field is found to be about 16,500 gauss at 14 .0 amperes. Chemic a ls The concentration of metal ketyl in the solution is found by treating the contents of Lh tu.be with water and. titrating the sodium hydroxida formed with dilute hydrochloric e.cid . The volume ~f the tube up to the mark was found by seventeen meas urements of the ciifference in ·the weight of the tube when filled with water or benzene and when empty , the density of t he liquid at the temperature involved being knowri. The ree.otions are illustrated for the case of benzophenone below: + Na~ Na. 0I OyO or free rod ieal N~ 0 I rJ-Oco I 0 OC-ON<.t or \I. , ~QC-ONa 0 11he hyclrochloric acid was stan dardi zed. with carefully prepared an.hydr ous sodium carbonate and with carefullf prepared sodium tetraborata. Agreement within 0.2% was cb- ·tained between ·the tv.io r:1e thods. The benzophanone used had. a pure whit e u olor s.nd melted at 48 . 0°-~e.5°,the literature value. The p-meth- oxybenzophenone had. a melting-point c.f c1° -c 2° ( lU;er.'.lt u.:re The p-biphenyl phenyl kstone was ~repared on the basis of the very brief directions in "Beilstein's Handbuch d.er O:rganischen Chemie" ( vierte aufl age), band VII, s. 521. It was crysta llized 3 t imes from alcohol, after which l t had. a good white color and s. me1 ·; .1ng-poin t of 101°-102° (literature value 102°). The yield was about 1~. Three attempts were made to prepare naphthyl phenyl ketone before a successful method was obta ined. I:he f i rs t was from benzoyl chloride and naphthalene, anhydrous zinc chloride being v.sed as a catalyst .. :~e second was from ct'- naphthyl magnesium bromide and benzoyl chloride in anhydrous ether. The sueseesfu l r:i ethod was to oxid ize «-benzy1 naphthalene by refluxing it thirty hours with nitric acid o:f 36 0 Be' gravity. References tci a ll these methods wf>ll be fou."ld in "Beilstein' s Handbuch der Organischen Chemie " {vier·te aufla.ge), band VII, s.511. The tt-benzyl naphthalene was prepared accordi ng to directions f ound in the same reference, band Y, s.689,. The !!-naphthyl phenyl ketone had. a f a i:r- r:hi te color and. a, rnel t ing point of ( li tera ·t ~rs ·al :.:.s '75 .fs°) . The yield over ·t he t ·.vo processes ( the p .~ epara,i.ion of ·:;he hyd. ~·oca i·bon from n aphthalene and benzyl chloride with zinc d. ust as a catalyst and its oxidation with nitric acid) was about 2%. 6 t:anipulation Tile ketone solutions were shaken with sodium wi re f or a period of four to four and one-half hours , this time being sufficient for complete conversion of the ketone t o the meta l ketyl in the case of benzophenone , according to Schlenk. It was found by making the ketone solutions up in a definite known concentration and analyzing them for sodium afterwards that this time was sufficient. It was observed that th& rate of reaction was much greater in the cas e of p-biphenyl phenyl ketone. However, t he reacti vity of p ~methox;,vbenzophenone was much lower than that of benzophenona , there being a certa in , {relatively high) concentration of the ketone necessary for the reaction to go a t all .. This effect was not observed with eithe r of the other two ketonas. It was observed tha t th ere ware smB-11 changes in the :field of the ~::a.gnet ( at 14 .. 0 amp eres) fro.:n day to day , a l though .the same cyc l e o:f current changes had been repeated in all the measure~ents . To eliminate err or fr om this source it was nec essary to find the force on the t ube filled with the standar d s ubstance {benzene) just te f ore or j ust a fter each measurement on a me t a l ketyl solution . ;For some cf the measurements a sta:.- idard t ub a fill d with benzene was used; this tube was ca li- brated. against the origin.al tube fill ed with benzena . ,.., ' age a t t he ·vo ltug e a t vvLich it was necessary to operat e it, and this added t o the difficulty of the measurements. The balance was not as fine an instrwnent as should have been used for t .i.1.is work, but by repe a ting the weighings three t o six times, it was possible to eliminate ei""rors. A weighing with '' no field" was ma.de both before and after each weighing with the field corresponding to 14.0 amperes, and if there was any difference between tha iino :field" weighings, the average was taken . vvi1en the electromagnet had not been used for some time, a "'iri:fting"effect was observed, that is, the value for the force on the tube would start high and. gradually diminish to a steady value as more weighings were made. The steady valu.e was the one taken. By making measurements with both water and benzene, it was shown that the steady value was the correct one, the susceptibilities for water and for benzene being known. Formulas The theory of the Gou1 method for magnetic susceptability gives the following equation: XA 8 where XAis the volume susceptibility to be found X8 i s the volume susceptibility for the substance in the lower half of the tube (air , in this case) Xois the volume susceptibility of the substance used as a standard FA is the force when the tube is filled with the substance the susceptibility of which is desired F0 is the force when the tube is filled with the standard substance. To find the ruclnl susceptibility of the solute , ·we must :first calculate its contribution to the volume susceptibility of the solution and then use the following equation: x' A where XM la tte molal suscept.i.bili ty of the solute I X.4. is the contribution of ·t he s olute to the vo lume susceptibility of the solution. C is the concentration of the solut0 in mols per cubic uentimeter. From the Laugevin theory we hav·e the follov;ing relation: 2 Nu XM = Na: + 3kT where o: is the d.ia.-:iagnetic s~sueptibility per molecule u iu tne "permanent" magnetic moment per molecule k is Bol tzman's constant N is Avogadro's number . T is the absol ute tempera t1.:.1~e .. If we so lve this l a s 't equation f e r u, ·;e obtain u = 2. 8 3 l T(.XN,- Nft} in Bohr magnetons. The percentage dissociation into free radicals is given by u'4 D = 100-r- • The dissociation is of the foll owing form: Na R;!:oNa ~ R-e-oNa. ~ 9 2R I -9-R l., t R2.. where R , ~ R~, and R),I are a.ryl radicals e 1 10 Da t 8, in anal~s i s Na.0H Found. B0.9(26°) 0.000147 0.000383 Solvent 18.6 (26°) 21.1(25°) Metal Ketyl oenz ene 3.848 20.9 (25°) tube 0. 001o5 mc ls in p-biphenyl phenyl ketone sodium dio:x:ane 3.330 0.00026 7 vVt. o.f Force on sol- Force on solution utiontrng) 0 Benzenet~) 3.334 gm 20.2 (2'1) 2o.1m) p-biphenyl phenyl ketone sodium benzene 20.4(24°) # p-methoxybenzophenone sodium 20.5 (25) 0 3.247 ar~ophenone sodium Volume-of tube {to mark} -173 x 10 141 -124 -6 cgsm. uni ts The soluti on conta ined excess ketone; this was necessary t o make the reaction go at all. ,.. C, cgsm; uni ts, Specific susceptibility for benzene ("Interna tional Critical Tables") -0.712 x 10 benzophanone sodium p-methoxybenzophenone sodium p-bipher..yl phenyl ketone sodium rule from values in the "International Critical Tables" are as follows: The value of Na for thw various metal ketyls as calculated according to Pascal's 3:,'l cc. benzene # -6 cgsm. uni ts. The molal susoeptibility of p-methoxybenzophenone as ca lculated according The analysis of the c ontents of the t u.be showe d. 0.1965 gm. t otal ke ·tor.i.e, or C.1625 gm. excess ketone. t o Pascal •s rule from ·values in t he "Internati onal Critical Ta blestt is -11 7 x 10 11 Results of Magnetic Measurements The table below gives the calculated percentages of dissociation for the various solutions. In making the calcula tions, the values for N~ were found according to Pasoalws rule from values in the "International Critical Tables" . The calculated values for Ns are accurate to within one or two percent. Percentage Dissociation Solute solvent cone. T~. Min. uax. p-biphenyl phenyl ketone sodium benzene~ 'Z'1 C 1.7% 3.0% p-biphenyl phenyl ketone sodium dioxane 3.0% 26° 36% 41% 45% p- methox:ybenzophenone sodium benzene 1.0% 25'.}& 17% 33% benzophenone sodium benzene 1.7% 25% 42% In the first column under ".Pe1~oantage Lissoc i.at ion" are given values calculated U..'lder the assumption that the values for the force on the solution and f or the force on the standard substance (benzene) are each in error by one-tenth of a milligra~. the error being taken with such a sign that the percentage dissociation will be a minim~. In the third column under this same head are given values calculated under the same assumption but with the errors taken with the sign that gives the maximum value for the percentage dissociation. Pe1~haps too much weight should. not be placed on the val- ues for p-methoxybenzophenone sodium since the solution was necessarily very dilute a.n.d the seusi tiveness of the method decreases very rapidly with decreasing concentration. 12 A very rems1·kabla solvent effect is noted in the case of p-biphenyl phenyl ketone sodium. This great increase in dis- sociation does not seem to be due to the fact tha dioxane is an ether since diethyl ether does not have thie effect (as we shall see later), nor to tha fact that dioxa.ne has a ring structure since benzene also possesses a six-membered ring structure. When it was observed that benzophenone sodium appeared to be almost comp letely associated in benzene solution, it was thought profitable to try replacing one of the phenyl groups with a p-biphenyl·group since Lhe latter group greatly enhances the dissociation in the case of the ordinary carbon free radicals. For this reason, also,~-naphthyl phenyl ketone was tried, but this ketone resisted all efforts to cause it to form a metal ketyl. Sodium, potassium, and sodium amalgam were all tried. but without effect. This effect may be due to steric hinders.nee. It could hardly be due to impur- ities, although impurities ao have a gre at effect , since a metal ketyl was always obtained with the other ketones when in moderate concentration, although the metal ketyl might n ot go into solution. p-methoxybenzophenone was t ried bec ause the p-methoxy group g1"eatly increases the dissociati on in the oase of the nitrogen f ree radicals . It was thought possible that the failure to detect dissociation of the p-biphenyl phenyl ketone sodium in benzene 13 solution might be due to the diluteness of the solution. A measurement on some of the solid precipitated in the bottom of the tube was accordingly made (the solid is colored), but it was found to be diamagnetic. It is remarkable how traces of certain impurities will reduce the reactivity of the ketone and especially how they will cause the metal ketyls to coagulate { the s olutions are always supersaturated). One of the impurities is thipphene, but there are others. In one case,the t hiophene was care- fully removed from a certain sample of benzene and the benzene distilled, the portion boiling at a constant temperature being taken. Yet there was some impurity present which ca.used the metal ketyl to coagulate. Results from Boiling Point Data Schle:nk perfor med "the following experiment: ("Berichte der Deu.tshen Chemishan Ges ellschaft" , v. 46,p.2840 (1913)). He observed the boiling point of a certain sample of anhydrous diethyl ether; he then introduced s. weighed quantity of p -biphenyl phenyl ketone a...-1.d again observed the boiling point ; now he introduced a_ piece of potassium metal and watched the boiling-point. No chu..-ige in the b oiling point of the solution after the second addition was observed , according to SohlenK. contact with an atmosphere of nitrogen. The solution was in Schlenk usas this as evi dence that the metal ketyl is completely diss ociated .. However, 14 upon repeating t his ex~eriment the fol l owing results were obtained; I II 1 . 545 1. 500 soluti on 1.998 1 . 585 Reading after steady value wa s reached following addition of potassium 1 . 737 1 . 540 wt . ketone used l.202gm . 0.293gm* wt. ether used {approximate) 2J .. ,6 gm . 21 ...6 gm. Reading a t belling-p oint of ether Rea.ding a.-t; boiling=point of ket one I n each case the deep green color of the metal ketyl began to appear a.bout five minu.-tes after · the potassium metal had been added. The temperature then gradually fell until the steady value rec or ded was reached . In I the steady value was reached in a.bout an h cru.~ and in II it was reach e d in about two hours . In each case , to make certain that no :further change would take place. the boiling was allowed to go -011 at the s ·teaiy val ue f er a bout an hour. The slowness of the fall in temperat ure s eems to be du.a t o ·t he slowness of the reaction . The solution was such a deep green at t he end of both e~--per iments that light could not be seen thr ou.gh it; there was an eicess of potassi~~ metal in each cas e . The fact that the thi£d :readin6 in ea.ch exp eril..ent was not half-way betwe en the other two is probably due t o the fact that a very s mall ring o:f p1·ecipi ta.tad metal ketyl formed on the gla ss j ust above 15 the level of the liquid. ~his was due to spattering (when the metal ketyl i s once precipitated from solution , it wi ll not dissolve). calculations from the above figures show that the ketone exists as single molecules in the ether solution. These experiments show tha·t the metal ketyl is ..not probabiy,,._dias ociated at all in the ether solution. Certain 01·ga.nic chemists feel that tco much reliance can not be placed up on zuch boiling point evidex.we. It was i1· om a desire to fiuu ou t whet1j,e:c such evidence is relia.:ble that ox.J)eriments similai· to tha s.ocve -;;ere t1~1ecl using dioxane as the solvent. I't appears, however, that o.t the • 0 boiling-point of dioxane {101) the metal ketyl is tlmost completely coagula·t e d.. There is evidence , also , that at thia ·I ;emperatu.r6 ther'a lt:» 1,; 0nsi<ier at.le breaki ng up of the ketone into smaller molecules by t.he alkc:..li metal. Evid.en.ce of a Chemiua.l :N a. tu.rat) Bachmann has pointed ou·l:; ( 11 Jou.rnal of the AG1erican Chemical society\", "" .55,:p.lris ( 1933)) that if the ketone is ~ompletsly associated , a pinacol should be obtained upon the reaction of the .wetal ketyl with ·,iu.1.er , according to the following eq_.;.a.tion: R,. R,.. J1 C -r:'li R,:-C -mra . -. I +2H O ➔' I + 2NaOH :z ..., ,.. . o -a ,L\ v .._, .ll lt-C -O!'Ta ' n,. I "."> ...... 2. 16 where R, and R~axe aryl radicals. However. ha also points out that these pina cols are uns ·t a.ble in alkaline solution. He repeats that he Owtai neci good yields of some of the pinacols by pouring the metal ketyl solutions into acid solutions. He closes the argument ·by pointing out that if a pinacol could be found whic.h is stable in alkaline solution, the corresponding metal ketyl sho uld give a 1ooi yield of the pinacol . such a pinacol is 9 .10 di- phenyldih~drophenantlu·enediol , anu.. .Sachmann says that he obtained a lC0;0 yield of this pinacol. It yet remains to expla in the color of these metal ketyls i:u solu ti ens in vvhioh they h~ve been f cu.:ad to be a. l most c ,Jrr.pletely associa.•t.ed.. We mi ght say 'that t..he free pre~ .::.r..t g i v-·es rise t o the observed cvlor. '.:.'hi B is sup- ported. 'L ~· -:he ob s e rvation i n t.he c ase oi the p-biphenyl phenyl ketor. . ~ s cd.i um that t he d.ioxa.ne s ulu.tior.. , al though only a bout one-thL.' d 8.8 0011cen tre.ted a::: the ·benzene solution , ,,as far d.s.:r::e:: than thl., latter; tho dl$socia.tion is quite high in cii oxru1e lrn.t c:nlJ vei-y l ov. :.r.. benzene. I. A .nev, method., that of magnetic s u s,J:=iptibili ty, is ai:,plled tc the determ ina t:i on o:f the d.i.ss ocia ti ou cf the meta l ketyls. 2. The preparation of oertait ar o~~ti c kotones is disc i2 ~ed . 17 3 .. The f ormation ancl properti es of the metal ketyls is d.iscussed. 4. A very remarkab le specific solvent effect is note d. 5. Evidence fr or:i boil ing points in agreement wi th the magnetic evi.clonce but in d1sagreet1ent with cer tain previous wcrk i s gi ven. Literature Re~erence Sugden ("Transactions of the Faraday Societyr Jan. 1934, p.23) gives the results of some measurements of the magnetic susceptibilities of benzophenone potassium and p-biphenyl ph€nyl ketone potaesiunr,-; in dioxane. He gives no particulars but say9 that these are about 85% dissociates in solutions whi~h are more concentrated than those employed here.