Report Committee on Examination

REPORT of COMMITTEE ON EXAMINATION

This is to certify that we the undersigned , as a Committee of the Graduate

School , have given - ~;.l.l.. ~.E!-.l'L .~l·. ~.@ ... $.~Yf.r. ~ -~-4. . .. final oral examinaticn for the degree of Master of Scienco.... -.. . We recommend that the

degree of aster of ~ci~~~~ -- be conferred

upon the candidate .

Minneapolis , innesota

...... ~...... : .... -~-:J Chairman of Committee on Thesis

The undersi gn d, acting as a Committee of the Graduate School, have r Gl\Cl the. accompanying t. o i s submitted by ..~!-~-~ - ~-8.:l?'.. ...~~ - ;-~- - ~-~.Y.J.1:".; .~~---· · ········· for the dogreo of ·-···· .....~~~~~! --~-~ --- ~-~-~~-~~-! ...... -··-··

Th ey approve it as a th~sis m e ting the ~equir"- m•nt s or t he Gradu~te Scl ool of th University or

Yj nno ota. and recommend that it be acceptDd in partial f ulfil lment of th req lirem nt for the degree of ...... 1laster...... of... · -·-··Science...... ·····-··

.... ~~'---~Chairman

...... ,......

...... [ ___ _,. _...... ········ ..... ·-·········· ...... Y..~ (}_:.. ~ .... :... J&----·-··tr-····· ·· ·······-·-···· .. ·-····-····-·-··-···-·······-·················-· THE cmroN OF IODINE ON SODTIP~ TRICHLORPBEio TE

A THESIS

SUB ITTED r."10 THE FACULTY OJ!' THE GRAD :ATE SC OOL

OF THE

UNIVERSITY OF Ii '..SOmA

by

LILL .. • SEYJ.f RIED

P R.. L ULFI.!JL~ TT 0 THE RE IRE 'TS FOR THE DEGREE

,,{AS ~R 0 SC IE. CE

June 1917

I I I I I I I I '1 I I I I I I I 1 t I I I I .. I . I . II.... I

THEORETICAL P.ART

During the study of the catalytic decompositions of the silver salts of the symmetrically trihalogenated phenols, it was shown that there resulted a white amorphous compound, the empirical for­ mula of which corresponded to the formula of the original silver salt, less a molecule of silver halide. For instance, from the sil-

l ver salt of tribromophenol1 , C6H2 Br3 0Ag, there resulted a compound (CsH2Br20) , and analogous compounds were obtained from the silver n salt of trichlorphenol, which yielded (CsH2Cl20)n• and from the silver salt of tribromresoroinmonomethyl ether, which gave (CsH(OCH3)Br20)n2• It was shown that this reaction occurred by a splitting out of silver halide from the molecule, leaving an un­

saturated residue of the general formula C6H2X2 0, which then poly­ merized to form the above compounds. Further study of this reaction in this laboratory has shown that such unsaturated residues must be formed in the following re­ actions: First, thermal decompositions of the silver salts of the above !!named phenols. In this, the silver salts were heated in benzene,

splitting out silv~r halide and leaving the residues, which could be isolated from the benzene by precipitation with alcohol, and proved to be identical with those obtained by the oatalytio decomposition. Second, the action of silver or mercury on tribromphenolbromide

This action was orked out by Mr. Woollett3 , who found that bromine ------1 Torrey and Hunter, Am . Soc . 33, 196 (1911) l. H. Hunter and others, .Am . Soc . 38, 1761 (1916) Unpublished work

•) •) .) ) -2- as split out from t he tribromphenolbromide hen a solution in an indifferent solvent ae shaken up 1th mercury or finely divided silver. It is interesting th t the first recorded formation of the hite amorphous compound, which e obtained from the tribrom­ phenol, ie the formation of a compound from tribro~phenolbromide by the action of light to which Kastle gav~ the erroneous formula oollett as able to show that these compounds re both identical 1th the one obtained from tribromphenol silver salt. The action of the mercury may then be represented by the f ollo ing equ tion: CeH2BraO + 2HgBr n CaH Br20 • (CeH BraO)n Third, the otion of iodine n the s i lver salt of triiodo- phenol• . oollett ob erved that hen the silver salt as armed with ethyl iodide , a reaction exactly like that shOW'?l by the tr bro ophenol silTer salt occurred, yielding silver halide and an uns turated residue which after ard polymerized to the white or­ phoue oxide.

CeH3!30Ag Ag! + C4H I 0 CaH I 0 (CaH3I O)n n th silver salt as treate with a very little 1o ine in ater in the oold red orphous substance as for ed in laoe of the hite oxide. This red substance a th t fir t described by L te , and kno ae Laute ann's Red ." It has th identio l co position of the white oxide. On the other hand, hen the salt ------1 27, 34. 1902 2 • Che • J., • Soc., 38, 2 74 1 1916) Ann., 120, 309 (1861 ~)======-U t th v ry littl od ne benze h h1te 0 1 a in obt ne .

hi oti n led to the 0 1 0 t 0 th :to rth otion o iodine on the pot 881 8 lt 0 triiodo henol t r sol ti on , hi ch led to the orm t on of ute n' e r e Of 10 ne 8 ici nt to br ing bout th s tr ns-

ion, hi h roceede ver pidly. t s e pro ble th t otion 1 c in he follo r: io in n- to i lo 1 h r - 0 n th 8 t r e n at te , n t r th h ot i 0 oth ol o 1 . n t t 00 tin o ttin 0 t 0 r lo n bo nd ot 1 l 00 t l n e to th t r idu 0 t ho 0 c . h r c ion nn r e 0'

H

1 a o or or ho

1 n c - 5- present in both oases , it would not be possible to follow separate­ ly the fate of both the ''inside" and "outside" halogen. It was decided to carry out the decompositions using to dif­ ferent concentrations of iodbie, a 2/5 gram atom equivalent and the gram atom equivalent, in order to show whether less than one atom equivalent could bring about the decomposition of a whole equiva- lent of sodium salt. The reaction was studied in three solvents; acetone, water and benzene.

Reaction in Acetone

en sodium trichlorphenolate was heated with a 1/10 gram atom

equivalent of iodine in acetone, only a small part ( 270) of the so- di um salt was transformed into the oxide . Heating under the same conditions and for approximately the same time with 1/6 equivalent

of iodine increased the salt transformation to 10. 61 , while with 2/6 equivalent , 66.0f- of the sodium salt was converted to the amor­ phous oxide. Increasing the iodine concentration to a gram equi­

valent resulted in 71. 61 of the salt transformed to the oxide . The oxides ere all of the colorless type, and appeared identical ith those prepared from analogous salts in the other ways . Analysis for chlorine showed entire agreement with the theoretical value for the oxide . When an attempt as ma.de to trace the ring halogen and the iodine in this reaction, we discovered that our choice of solvent was an unhappy one . In the presence of halogen, acetone was far from being an indifferent solvent. Our attention was called to this by the very sharp odor of iodo- acetone given off from the reaction mixture. From this we concluded an attemnt to -6- trace quantitatively the action of the two halogens would be unsuc­ cessful. From qualitative experiments, however, we found the io­ dine present as free iodine, and combined with sodium as the iodide, in addition to that in the acetone derivatives. We were also able to show that the ring halogen, at least in part, existed as the

chloride combined with sodium. ~he results of the acetone reaction were in accordance with our ideas~but for the reason mentioned, ir.t ~o further work was carried on A this solvent.

Reactions in ater

The sodium salt when suspended in water with a 2/5 equivalent weight of iodine for 10 days, keeping the reaction flask in a dark place, yielded an amorphous substance slightl pink in color. cept for its color, the oxide was identical with that obtained in

the acetone deco~position . The pinx coloration here s extremely interestin , appearing to .indicate the formation of the chlorine analogue of 'Lautemann's Red" , obtained when the potassium salt of triiodophenol was allowed to react with a trace of iodine in ater. In the purification of the oxide by solution in chloroform and re­ preoipi tation by alcohol, most of this color as removed, leaving the pure oxide light tan color. At first it as thot the color might have been due to the adsorption of iodine on the colorless oxide, but no test for free iodine oould be obtained from the al­ cohol used in reprecinitating the purified amorphous substance. Out conclusion here was that a small amount of the colored oxide, analogous to the Lautemann red) had te~porarily formed in the crude oxide. The yield of the amorphous oxide was 63.4 of the original -7- sodium salt. en the same reaction was oarried on using a gram equivalent of iodine the oxide formation inoreased to 79%, which showed that the 2/5 equivalent was sufficient to bring about a practically complete transformation of the sodium salt over to the oxide. A procedure of analysis was then worked out, and a oomplete examination was made on reaction mixture resulting when the salt was allowed to react with a 2/5 equivalent of iodine under the conditions observed in the preliminary experiments. The yield of

oxide was 64 .210 of the sodium salt taken, leaving 35.8% to be ac­ counted for. 'Unchanged salt" was determined by conversion into

the phenol , and l~ . 4,o was accounted for in this Wf!Y· No other pro­

ducts could be demonstrated to cheok up the remaining 22.36 , ~ . Ho - ever, it was found that our fig res for "unchanged salt" were in­

accurate, owing to the solubility of the pha~ol in water. This as determined, and was found to be more than sufficient to account for the missing 22.36,o. 76.28 of the iodine added, that is of the 2/5 equivalent, was found as free iodine. The 22.72%, occurring in a non free st te and apparently not used in the decomposition of the iodine salt, had no doubt been converted into hydriodic acid or even iodic. This conversion might easily have taken place in the resence of water and chlorine) especially when it as demonstrated that ioniz d chlorine appeared in the reaction.

e were able to demonstrate that the ring halogen behaved ex- actly in accordance with our ideas held of the reaction. or every Olecule of the salt decomposed to form the oxide, one chlorine atom ust have been set free, hich then displaced the iodine attached to -8- the sodium to form sodium chloride. The ionized chlorine was de­ termined as the silver salt. In this way the chlorine recovered was 62 . 910 of that available from the whole sample introduced. Ho - ever, only 64 .2 of the sodium salt was transformed into the amor­ phous oxide, and only this fraction could really provide the ionized chlorine . 97 .8 ~ of the actually available chlorine was t erefore recovered. That is in agreement with the theory and shows that the view taken of the reaction is correct.

~he formation of a quinone in this deco~position as held not improbable. We were not able to demonstrate the presence of any such compound at this time, although a study of that side of the reaction will be undertaken later. When the oxide from the water decomposition waa washed with warm alcohol, the washings ere pink in color. This deepened considerably on exposure to air , and the solution left on evaporation a slight amount of a purplish-red amorphous residue . This residue resembled the colored product ob­ tained from a similar decomposition of the potassium salt of tri­ bromophenolate , but was not nearly so intensely colored. It is our belief, at present, that this colored substance may be an alcohol­ soluble amorphous prodact, a lower polymer of the unsaturated resi­ due than the colorless va!iety. The ease ith which the analogous triiodo and tribromo salts deoonposed as compared with the slug ish action of the trichlor salt as also observed. Previous work in this laborator has shown that the transformation of the triiodophenolate over to the red oxide is almost instantaneous and that the tribrom salt undergoes practically complete decomposition, yielding a brownish- ed oxide - 9 - after standing two to three hours , hile our sodium salt showed ap­ parently no change under such conditions. It was only after stand­ ing for a day that any formation of the oxide was apparent.

Reactions in Benzene

Preliminary experiments with the sodium salt in benzene sho~d that with the 2/5 iodine equivalent just as complete a transforma­ tion of the sodium salt could be obtained as when using the full equivalent of iodine . A complete examination was then made on a reaction mixture resulting when the salt was heated with a 2/5 equivalent of iodine . Analysis showed that 92. 2,a of the sodium salt ent to form a colorless oxide, identical with that obtained in the previous decompositions. This when analyzed for halogen was found to contain nearly lv more chJo ine than that calculated. It is

believed that this was du~ to some of the ioni~ed chlorine substi­ tuting in the ring of nart of the oxide molec les , in the presence of the iodine as carrier. Dur belief for this was substantiated hen an analysis for the ionized halogen was made . This, hioh represents the chlorine removed by the reaction,and hich, according

to 0 r theory should appear in the nascent state at first, proved to be less in amount than exoected . Only 85, o the av ilable chlorine eared in the reaction mixture as ionized chlorine. he only thing th t could prevent this from ppearing s ionized chlorin uld be its disap earanoe by substitution. rt ill be remembered that in the ater decomposition 97 . 80 of the chlorine appeared s ionized chlorine , and here the amorphous oxide had the caloul ted omposition. -10-

The iodine was traced to two positions 83 . 6,~ was found as free iodine, and 16.04,o as sodium iodide. The presence of the iodide

lends further proof to our th~ory. No doubt some of the iodine was still in the ionized form because the nascent chlorine was not available to replace it, as it had disappeared into the amorphous molecule as stated in the above section. That the iodine in this

reaction has been fully traced is shown by the reQovery of 99 . 64% of that introduced. The brownish tar residue found when the alcohol used in the recovering the purified oxide was evaporated, although of no ap­ preciable yield here, aroused considerable interest. rt appeared to be identical with the tar residues formed in preliminary work when higher concentrations of iodine were used . With 1- 1/2 gram equivalent or more this tar residue was obtained in considerable ' amount. At that time, no particular stress was laid on the reac­ tion, because it did not appear to hsve any relation with the forma- tion of the amorphous oxide . Later, however, its appearance in the lo er concentration of iodine decomposition warranted further study. I It is our plan to attempt to isolate a sufficient amount of the tar for study. The occurrence of the tar with the amorphous oxide, ~iffering only, so far as we know, by its solubility in alcohol, brings forth the suggestion that it may be another polymer of the unsaturated residue. The results of analysis for the decompositions with a 2/6 gram ~quivalent of iodine in water and benzene is graphically shown by the followi:rg scales:

-11-

Swmnary

When sodium trichlorphenolate is allowed to react with less than a gram equivalent of iodine, in acetone, water, and in benzene, a Practically complete transformation into the colorless amorphous oxide occurs.

The oxide has the formula (C 5H2 Cl2 0)n, and is analogous to the amorphous substances formed from salts of the type

B in other modes of decomposition. The outside halogen acts indirectly as a catalyst, at the end of the reaction remaining practically unchanged, but forming an in­ termediate product with the sodium. I The ring halogen leaves the ring in the nascent state, replac- ing the iodine of the sodium iodide to form sodium chloride, and leaving the iodine free. The velocity of the reaction with the trichlor salt is much less than that of the reactions with the analogous triiodo and tri­ brom salts, both of which proceed easily. The reaction has been shown to take place in the following llle.nner: -12-

Ne. + T dddt:d -r Nar + NaI I 0 0 0 c / c I / / Cl) 001 Cl~ 001 010 CCl II I - 1 11 H C ~ H H 16 lH H C H c/ +Cl c/ I I { Cl Cl

The original free iodine is again set free: Cl+ ar : aCl+ I L

r 1 n of 0

t h nol 0 1 0 br on, n to

0

n 0 0 -14- 6 ,a excess of sodium hydroxide solution was tried. Thia plan proved unaucoessful because it waa extremely hard to recover an appreci­ able amount of the sodium phenolate on account of its great solu­ bility in water, and in any liquid m:tsoible with water.

A slight modification of the first method of preparation was finally used for making the salt. As in the first, the trichlor­ phenol dissolved in alcohol, was added to a water solution of ao. equivalent amount in slight excess of sodium hydroxide, and the mixture evaporated to dryness on a water bath. The residue of crude sodium salt was then heated in the oven at 150°0 for three hours, and taken up with a minimum amount of water-free acetone. After filtering free from inxoluble carbonate and excess alkali, the ace­ tone solution was poured into fifteen volumes Of toluene and the mixture boiled on the electric plate until the sodium salt precipi­ tated . The mixture was allo ed to cool to insure comulete se~ara- tion of th alt, and filtered on a Buechner. he salt s shed free from toluen with benzene nd sucked dry. allowing this math- od of preuaration, a satisfactory yield of pure and practically an­ ydroue sodium salt as obtained.

Properties of Sodium ~riohlorphenolate

The odium salt crystallizes in white flakes . In contr st, the potassium salt of this ph nol h a decided yello color.. .:.he sodium salt is easily soluble in ater, alcohol, ether, and acetone; . but is practically insoluble in chloroform, toluene nd benzene. The salt may exist in a hydrated form having one molecule of ter of crystallization. This point was determined by heating to looo a -15- weighed sample previously heated to 120°. It is a remarkably stable salt, from 11 npe r noes rem in­ ing unchanged and displaying the same solubilities fter heating

to a temperature of 220°.

The Aation of Iodine on Sodium Triohlorphenolate in the Following Solvents: Aoetone, Water, and Benzene. 1th Acetone as the Solvent. The first work was done using a 1/10 gram atom equivalent of

iodine. Two grams of the anhydrous sodium salt ere dissolved with . 115 gram of iodine in 20 o.c. of aoetone. This was heated gently under a reflux condenser on a water bath at 45°-50° for ab:>ut thirty minutes . To precipitate any amorphous compound founi, 50 o.c. alcohol was added. A slight amount of colorless amorphous sub stance was thrown down, which represented but 21 of the original sodium salt. On examining the f il tra te. it as found to c en ta in mostly unchanged sodium triohlorphenolate, traces of sodium chlorid and iodide. There was also a distinot odor of iodo- oetone to the filtrate. This oompound might easily have been formed in the pre­ aenc of free iodine and acetone. Although no determination o~ the amount of free iodine remaining in the solution was made, the aae­ tone-aloohol filtrate was deeply colored ith iodine. A seoond 2 gram sample of sodium salt was treated in a similar manner, only this time in plaoe of using a 1/10 gram atom equivalent of iodine, a 1/5 gram atom equivalent s t ken. The amount of amorphous sub stance similar to that previously obtained in this oase increased the yield representing 20.7% of the original salt for the same time 0

1

0 0

0

0

0

0

0 10 0 0 0 0 -17-

0 0 in io 1 0 n t 0 . h 0 -

d 0

. 2 l or C C . 07 .

o 1t1on 1 10 .

0 1

I t io 1 1

0 on olo

0

1 0

1 ( ) . c .. ( 16 o n 0 0 d n o o i on d b 110 0 no , l -1 -

i • • tr 0 00 0 1 0 0

1 h

th d 00 l

ho 0 id 1

1 b e n n

0 t t 0 .

ch 0 n t 0 0 0 c

0 id

0 0 0

on t

0 0 h 0 0

l 1 b h 0

0 c io

0 -20-

t . o c oibl oxide trlohlo henol 0 e . . o er ible o tde

t . of trlchlorphenol 77

t . of e c bl o ide . 22 m . . of er c ble

t. 0 ----.o------0 -

77 g - . 28 • ph no 1. 0 c . o. 0 t he ori o i m lt. g - . 1 g 8 . o 1 e . ( . ini 1 1 reei itate - . o pho o ide o 0 th t . 6 - ."'' 0 th oco or . r tt ti on th n d 0 e o the h lo

coo 1 to ct on 0 0 0 d 00

0 on 0 ' chlo in ho b r on - t ird o th oh lo ne d od b oco n ed r . 0 ohlorin d b 0 th 0- d th io d d ho in c or 0 0 h dr o le

i h 0 t t r re t or y h

d n r 0 1 c . o.c . ol - 0 i h o ul h

0 h in d l 0 le ( r 1 to ). iYe s 1 ve sli Of 11 1 • 0 0 n

0

n

1 01 n 0

0 0 -22- Swnmar y of the Analysis Reaction ixture , filtered. precipitate filtrate amorphous oxide iodine weight . 23 grams sodium t r iohlorphenolate sodium chloride . 23 grams . 196 g . sodium iodide = .426 grams of mad up to 600 c . o . total oxide 63 .4% sodium salt. 60 c . c . portions titrated for free iodine ound for entire e ple . 17 grams 92, in terms of gram atom equivalent of iodine in original sample

To 100 o . o . added dilute acetic Cid filter precipitate iltrate trichlorphenol sodium ahloride amorphous oxide sodi iodide eight . 0967 grams Ad Ag 0 3 , sh with alcohol Filtered ilter Re idue iltr ta Preci i te morphous contained silver chloride oxide trichlor­ a 11 ver iodide eight . 039 pheno l. Sh With 2 grams weight by ammoni . 196 gr a difference hydroxide for entire : . 0577 60 c . c . gra Reaid Solution o . 2885 r ail ver silver in the 0 iodi chloride 0 . c . aoidified Represents ith • nitric sodium salt preoi i ­ ta te eilv r ohlo i -23- The Aotion------of One--- Gram---- Equivalent------of ------Iodine --in ater. One gram of the sodium salt was dissolved in about 50 c.o. water, to which had been previously added .63 gram iodine dis­ solved in potassium iodide. As before, the mixture was placed in a glass stoppered flask and allowed to stand in the dark for a week An analysis, the procedure of which was similar to the one first completed, was undertaken. At this time a rather hurried analysis was made, as it was decided to run one mere determination on a quantitative basis, using a larger sample. This first sample was examined simply for comparison with th& sample run with a lower

iodine concentration for oxide and iodine. · ith this sample, the reaction was practically identical with the forroor determination,

/ the only difference being int he greater quantity of the smorphoua residue formed, which, like the other, a colored in the crude form, but when purified by solution in chloroform and repreoipita­ tion With alcohol remained as a colorless product. The reauits from this determination are given in the summary belo •

l -24-

Reaction mixture resulting after 1 gr am sodium salt, dis­

solved in 50 c . c . H2 0 was allowed to stand for 1 week in the dark ith . 63 grams iodine. Filter Precipitate Filtrate ma.de to 500 c . c . : oxide contains iodire weight . 52 sodium salt of trichlorphenol grams sodium chloride 7910 sodium sodium iodide salt Titration on 50 o . c . portions for free iodine resulted in aocounting for . 666 g . iodine or 89 . 8~ To 100 c . o . dilute acetic added very small precipitation of trichlorphenol. eight neglected here . 'Filter iltrate conta ins sodium chloride sodium iodide Add Ag10s , filter Precipitate : AgCl Ag I ash i th 2, !i140H Residue : Filtrate = A Cl Ag I Add HNO no eight, reoipitate but positive AgCl test no eight taken but subst ntial precipit te formed •

• - 25- A third determin tion was de in this water- iodine deoom­ po ition, the pur ose of which wa to obtain a complete quantita­ tive na.lysis of all the reaotion pr oducts . Thie time, in pl oe

of using the olid sodium salt, 4 . 5 1 gram of the triohlorphenol,

which s exaotly equiv lent to 6 gr ma of the anhydrous so ium

salt, as added to 6 o. c . of . 912 /2 aodi hydro ide (this bei the calculated amount neoes r to re ct ith • 01 r ms of the

triohlor henol) . o this mixture s then dded 2 . 79 9 gr a of

iodine di olved in a am 11 qu tity of pot seium iodide. This

solution, 3 0 o . o . in vol e , stood in d rk nl ce :for 1 aye.

At the end of th t time , a heavy prec ini t te o th oxi e had

out. Tnis was fil ere itn suction, orki quickl to lose of iodine thru e anor ion nd 8 th ter, until th shings ho ed no iodine color. is mar hous ot s a very light bro i h 1 • re e blin e color d ct obt i ed fro a m·1 eoo oei ti on of he tribromo- Ol te, b t not ne 80 n en el col d I s tra err to atoh lass nd ried in the ova • t 0 s rved that on rais the temn r 0 re to 1 th x· er d At this pain h ti 1 ediately i con i ued , the o 001 d in a deeio- c tor and eighe . The color in the 0 d d re tly to 11 ht tan in the dryi in e tor r oolo no doubt arti lly ue to iod n d or ion. weig t Of e ho us nroduc t, 2. 3 9 ra oorree on ed to ' . 21 ra sodium or 6 • 2,~ Of the S ple . • he filtrate f r om the morphous ubetanc • lua the Shings -26- was oollected in a liter flask and mad e up to vo lume . Ac cording

to our idea of the reaotion, this filtra t e cont ained any unchanged sodium salt, the halogens , chl orin e and iodine, pr esent aa such, or combined with sodium as the chloride or iodide . Several titra- tions for free iodine were made on 60 c . c . portions of the fil-

trate, and 76 . 281 of the original iodine accounted for as free ,

23 . 721 remaining to be traoed . The unchanged sodium salt was then deter mined . To 260 c . c .

II of the fil tra·te dilute AgN0 3 was added until the solution was just slightly acid . Thia was done to convert the 4os ium salt in solu­ tion to the trichlorphenol. A flocculent precipitate of ph Nol indicating a fair amount of sodium salt trichlorphenol settled out . This s filtered on a Gooch crucible , ashed carefully with dis- tilled water, dried and wei . ed ••he filtrate and the washings

ere saved, and an analysis for chloride made on it. From this eight, 1606 grams or . 6024 gram in the entire sample, the equiva­ lent so4ium salt was calculated . The amount eq ivalent to the . 6024 gr a of phenol s . 669 gra This amount added to the amount eq ivalent to the amorphous product for.ned gave the el t

of sodium salt accounted for as • 2 grams , 77 . 64 of the or inal s m le. The remaining 22 . 6 was et to be traced • • o other pro­ duct of the sodium salt of any annreciable yield could be ound and it as decided that one of the known rod cts must have been

made soluble in the ter th n ntici ted . ither the morphous oxide a somewhat soluble of the parts of the trichlorphenol rep- resenting the unconverted sodium salt might be in solution. On this l -27-

ooount 1 deoide to t ne 0 0 h 1 bl

henol. If th e were 01 b 0 l h of 1 h or-

henol found 0 l be too lo nd c n 0 1 ,. c le 1 ted ro th 0 ld b co 1n lo .

olubilit of richlo h nol

1 t Of dis 111 d ohlo 0 in nd t 1 tirr in h 1 . . ( )

or t 0 h u e 1lt r n h iltr on 0

t th, on 1 n o

2 0 .o . o tion

1 r ion t 0 0 th 0 0 d .

0 n 1

0 0 • 0 •

0

l )

0

l n 0

0 0

0

1 -28- r ult d not oheo • so a eeoond deter 1 ti on h r f t e silv r s lts thin n t r 1 to th or o bl 8, hin wit loohol, d ry n 0 deco. po tion of the silve e 0 ooc h treatment. h r e ei h 0 th ei v 8 hi r -

l ti on t en, nd fro t ei h the e 1 trich or o lo 11 ted . ro 2 c .c ., ilv 1 0 Of • obt 1 ed . h tr ohlor h nol

1 ht ~o d to be • 2 t 0

21 . . 2 t Of 1 bl r 0

1 0 t hen ol

0 to h h nh nol n 0 • 02

n • 6 n • 02 0 0 0

0 0 0 0 r o

n 11 bo e

1 d

0 n h

r 1

0 0 d

0 1 .

0

n i l 22 . I n d 0 a non free st~te, 23.721 equivalent. It was concluded here that

all of the iodine not used in the deoom oaition o~ the sodium s lt had been converted into the acid, hydriodio or iodic. This a - peared all the more evident after the determination of the ohlorine existing as sodium chloride was completed.

Determination of Ionized Chlorine

or the ionized chlorine determination, the filtrate and ash­ ings, remaining after the precipitation of trichlor henol from the 250 o.c. taken for the runchanged sodium salt" deter ination, w s

used . Though already slightly aoid with H 0 , eno h as added at the oint to produce an acid solution. Silver nitrate as then added in excess, hereupon a heavy hite preci itate settled o t.

This a transferred to a Gooch, ashed free from old d nitrate ' dried and weighed. mhe eight of silver ohloride .4 ~ r ms, represented .11 grams ionized chlorine in the 250 o.c. a mnle • • 4416 gra in the total samnle. rom the s ori in 1 sodium

trichlor henol assu~in one-third of the chlorine to b ioniz d ' the eight of ionizable chlorine as oalcul ted s .7 7 r h chlorine recovered then was 62.9 of t at 1lable from the hol sam le introd oed. o ever, t e or ho s com ound el h d cor- 6 .2 res onded only to of the e 1 intro oed, onl t is fraction could really ive ionized chlorine. ere ore, he recover of otu lly available chlorine a 97 . .his is e r ble greement with the th ory nd aho that t e vie ta n of reao- tion is corr ot. oleoule of sodium trichlor benol t ~aoh ch gives amorphous oxide, does actually give p an atom of chlorine -30- also . Summary of the Analysis Reaction mixture resulting when 4 . 501 grams triohlor nhenol as added to . 912 N/2 sodium hydroxide, amounts oaloulated to form 6 grams of the anhydrous sodium salt, and added to this mixture 2 . 799 grams iodine previously diss olved in potassium iodide, the hole made up to 380 o . c . and allo ed to stand in the dar k for 10 days, filtered preoioitate filtrate amorphou oxide iodine weight, 2 . 3586 sodium triohlorphenolate grams sodium chloride 64 . 2, hydriodio and iodio aids sodium salt made uo to 1000 o . c . 50 o . c . portions titr ted for free iodine eight of iodine for entire 1000 c . c . • 2 . l~t> gra 76. 28 amount used

250 o . o. filtr te acidified with dil te 0 3 filtered preoipit te filtrate triohlorphenolat sodium ohlori sh d dried dd dilute H O nd ilv r and w ighed nitrate. iltered weight = . 6 24 . 669 gra Preoipit te , ailv ohlor1d sodi m salt 1 ht .4 6 . 797 bl I -'1-% 6 S ohl rin s l . '11. 0 - 31-

D t min tion o! uinoid Grou in in !ro ter eoomposi ion.

A ei he e mple e t sted for ino d oxy n by n l line solution of hydr zine after the nner of the t e eter- mination or4ed out by oollett (Dootorrs th 1 , blishe ) . The p incipl of the ethod is the e s re ent o nitro en volve

1 oxidize b inoid ox en. n the irst o

deter in tlons, ell ht increase in vol ob d hio probably due to e eriment 1 error, s h co d ter in tion ed no effect.

D o osition n en ene o on

otion of 2/ tom eq~iv len o odi on he

h r t d • 0) l . 9 r 1 dd c. en e oon­ • 621 io in • 8 1 t n len­ 1 on th co

d 00 8

ed in o o c 0 n

he h h 1 on d t

n 1

e 0

0 1

c 1 0 0 th loe of io 11 tion. th n tr er ed an i i 1 0 0 c . c n Ol -32-

he e id e 8 d l t en n n ol ti on de up to vol in 2 c .c . !l he 00 - nletely sol ol in te , d oont ined a no n c sodium s lts, alo itb ny unch 0 t h nol t .

1 o . o . Of thi olut on a j t oidi d 1 h th esu t th t ost no eo it tion o! t c lo h nol occu

Ho ever, ell ht 0 t ttled o tt th ol ion

to t n t 0 d y • h ilt ed on 00 h , h 1th

11 0 t 0 t r , r d d h nl h

nol 0 f o th tir 2 1 l 0 co r to th b lit 0 1

th i 0 th deco

1 id l

0 b 1 0

0 d 10 id 0

0 0

11

0

0 h 1 h 0 r -33-

i od ine added ( .4621 grams) , leaving onl 16 . 11 of the added iodine still unaccounted for .

Determination of Amorphous OXide

200 c . o . of the filtrate was evaporated over a water bath, leaving a brown varnish- like residue , exactly like the residue o - tained from evaporation of solutions of the pol dibromphenylene oxide. Thia as taken up ith chloroform, whiah on adding produced a red coloration unlike iodine color in chloroform. The color soon disappeared leaving the solution a light brown . The oxide as ex­ tracted from this with alcohol . It as colorless and a neared identical with that formed in the acetone and ater decompositions .

The amorohous oxide as filter~d on a Buechner funnel , aahe ith

alcohol and sucKed dry. ~ne lconolic fil r te and shin a ere saved for the phenol determination. he oxide e then tr nsf rred o a ei hed atch glass and heated at a emperature of 80° to con­ tant eight. The mount of oxide, recovered from the 2 O o. c . ample weigned • 336 rams . The oxide in the entire s ple a then . 34 gr ms, reoreaenting a 91. 6% transform tion of the sodium salt. he alcoholic filtrate and shings from the oxide determination ere evapor ted on a ater bath. A sl1 htl colored resid e re­ mained. This residue was extracted with ter and on the ter soluble ortion affir tive ualitative tests ere obt ned for hloride and iodide. The ter insoluble ortion apoe red to con­ sist mainly of trichlorphenol. Summ ry of the An lyais rom 2/5 ram tom e uiv lent in benzene. Benzene Solution, Filter residue Solution contai iodine a trichlorphenolate amornhoue oxide Na ohloride triohlorphenol la iodide wash 1th benzene 50 c . c . portions titrated for free dry and take up iodin th H20 make up Found . or e t re ea 1 to 25 c . c . . 3 76 gr. 3 . a iodine added

To 100 c . c . add d. HN 3 200 o . o. eva orated to dryness. until slightly acid ken un th chloro orm and ex­ filter tr oted ith alcohol. iltered precipitate iltrate filtrate , evanor ted Re idue : trichlorphenol ~a chloride sli ht residue con­ a orpho a ei~ht a iodide tained trichlor­ oxide .00079 add 3 phenol o d for .ooo g . and Ag 0 so i chloride entire sample sodium trichlor­ filter sodium iodide 1. '° -' rams henolate traote th to 91. preci it te ter so salt. • AgCl gr . eeid e Solution eight - trichlor­ sodi m . 1 1 r h nol ohlo ide 0 l iodide -35- Deoom osition in Benzene using a Gram Atom Equivalent of Iodine .

1 . 9941 grams of sodium salt re allo ed to react ith 1 . 1571 grams odine der the oonditiona of the 2/6 gr atom deoo osi- tion, heating in 100 c .o . benzene until the action eared oom- nlete. The same procedure of analysis w s undert en in this case , and pr otioally the same results were attained the onl difference being in the relative weights of the recovered roducta. o deter- mination as de for ionized chlorine or iodine. -36- The Analysis of Reaotion ixture from uivalent Amounts of Iodine Benzene Solution, Filter resid e Solution sodium trichlorphenolate iodine sodium ohloride amorphous oxide sodium iodide triohlorphenol washed with benzene, dry and taKen up with water • Titrated 50 c.c. portions for ._ade up to volume in free iodine 260 o.o. flask Found for entire sample 1. 5 grams, 90.7 Of the 100 c.o. made sli htly acid amount of iodine added. with dilute H 03 ilter 200 o.o. evaporated to dryness. Ta en un 1th chloroform and "Oree ipi ta te filtrate con­ extracted ith lcohol. ichlor"Ohenol tain il tered eight for sodium chloride e tire sample sodium iodide precipitate 11 tr te : .116 g . Added HN03 nd amorphous ev porated • 13 grams so­ Ag 03 oxide found ium salt iltered eight for aodi m ch lo ride .64 ; entire so 1 iodide reoipitate a nle trich lor he no 1 silver chloride l.2975 gr m silver iodide • I Of no eights sodium a lt t en ta en. -37-

A third determination in benzene s made, this ti e s1 8 larger sample of sodium salt 1th a 2/ gram atom equivalent o! iodine. Thi was for the urpose of obt ni complete nalys1 of all the products. The solution of 65 r s of the s l t w1 th .66 2 grams of iodine in 150 o.c. benzene s made in a flas ground to fit the reflux. In this wa.y the use of a connecting cor

s done a y th, in th t y reventing any los of h lo en in- curring thru ad orption on the connecting cor • To detect, nd me sure if necresaary, loss of iodine and chlorine by volat·11zation a U-tube containing glass beads moistened ith silver nitr te sol1- t1on, s ttaohed to the reflux. In this ay, ny esoa 1 h lo- gen ould be taken u by the nitrate an converted o t co es ponding silver salt. The flas s h ted t or t ee h ur hen the U-tube a re oved as no oh e in the nitr te app rent. The heating s then contin d until he odi lt no longer a eared e resid e. n 1 sis of t re otion ix e

as de follo in the ener 1 ooe r aed in h n the other deco positions de n benzen . he r ot1 filtered o a Bu chner funnel c ion b

s d 1th benzene 11 free :trom 10 1 nd

re tr n f e red u nti t ive 0 0 l nd a to vol e. of R id

'.!: e abov res due 8 s c ed d d en 1 h e a ter solution s de slightl cid i th itric oid 0 ci- pitate any •unchan ed sodium salt as t ichlor henol. 0 oi 1- tate as obtained here, but an absence of henol a hie Oint did -39- not indicate a complete transformation of the sodium s lt, because the solubility of the phenol in water, as already determined, would prevent the detection of even appreciable quantities of sodium salt, a saturated solution containing the equivalent of .03 parts sodium salt to 25 parts water. The solution ae then made distinctly acid with dilute nitric acid and silver nitrate added in excess. A vol­ uminous precipitate consisting of silver chloride and silver iodide settled out. The silver ealte ere transferred to a Gooch crucible and washed free from silver nitrate and aoid. o o/o onium hydroxide was run through the Goooh until all of the silver chloride had been dissolved out, leaving the silver iodide ae a residue. The eight of silver iodide obtained as .1708 grams. The iodine equivalent to this ae .0921 grams whioh represented 10.61 o/o of the amount of iodine added recovered as iodide. The silver chloride as recovered from the ammonia solution by aoidi! ing 1th nitric The mi ure as allo ed to stand several hours before filt order that a comp ete and coagulated precipitate of eilv r chloride be obtained. The silver salt as then f i ltered on a ooch nd after ashing and drying, as ighed. he 1 ht of the chlo­ ide was 1.1026 grams, the chlorine content of hich a oalou at d

~o be .2728 grams. This as 67.18 o/o of av ilable 1on1 ble chlorine, .5435 gr s. An analysis of the benzene soluble products as then und rt en.

Determination of r e Iodine

The same method of titration was used here as in the ot r ti­ trat1ons made in a benzene solution. Three portions of the benzene filtrate made to 500 o.c., re­ ferred to in the above section, were titrated ith sodium thiosul­ phate for iodine, and the average of the ratios taken.

Benzene filtrate Thiosulphate Ratio Titration l 49.23 c.c. 6.89 8.368

Titration 2 46.68 6.63 8.262 Titration 3 41.44 5.02 8.264 Average ratio 8.291 1 c.c. Ba2S203 = .01203 grams iodine. The iodine determined for the total filtrate was .7260 grams. This was 83.6 o/o of the sample added, .8682 grams.

Determination of Amorphous Oxide

~60 c.c. of the benzene filtrate was evaporated to dryness on a water bath. A brown varnish-like residue identical in appearance with those obtained in the previous benzene decompositions re ined Thia was taken up 1th chloroform and as before, a deep red oolora- tion was observed hich gradually disappeared leaving a light brown ~ solution. The oxide as extracted by the addition of alcohol in great excess, and transferred to a Gooch, ashed with alcohol and sucked dry. The alcoholic filtrate and washings were heated on the electric plate to drive off any re aining chlorofor hioh ould tend to hold oxide in solution. This was found to be the oae because after gentle heating for one-half hour a second alight pre­ cipitation of oxide as observed. This was added to the first and the hole again ashed thoroughly ith alcohol and dried, first by suction and later, after removing the oxide to a eighed atch -40- glaee , by heating at 80• in the oven. The eight of the oxid 1 . 1334 grams for 260 o. o. , or 2. 2668 grams for the hole a ple. This ae equivalent to 3 . 088 grams of sodium salt nd repre ented a 92. 2 o/o recovery of the sample taken.

An analysis of the oxide for halogen after the method of Carius as made . Substance . 1664; A.gel , . 3007 Calculated for (CaH ClaO) 44. 07 2 n ound 4. 97

Analysis of .Alcohol iltrate

The filtrate and washings fr om the oxide det r in tion er evaporated to dryness on a ater bath. A sli t residue brownish in color, re ined and appeared to consist ainl of i or nio salts and a brown tar. In preli inary experi en a in he rl part of this ork, br o tar similar to this w Si l d

in appreciable amounts when iodine in e ter q iti atom equivalent was used. This r prod o The residue as extr c ed se eral ti es 1 r soluble portion as added ni ric aoi and heavy hite preoinita.te of B er ea t a 0 of chloride and iodide. Thi a allo e 0 00 1 l out and after several hour a e tr d 0 h.

i th 2 o /o onium hydroxide until th h 0 d a b e oe of silver chloride hen aoidi in 1th 1 rio 0 d r pr cipi- ta ting 1th nitric aoid, it as possibl to ep r e chloride from the iodide. he eight of silver iodide 88 0 6 8 or -41- .0872 grams for the entire sample. The weight of iodine oorree­ ponding to this was .0471 grams. That is, 5.43 o/o of the iodin added was reoovered as iodide from the benzene filtr te. The sil­ ver chloride found was .3007 grams, the chlorine content of which was .0744 grams. The percent of chlorine recovered from the ben­ zene as ohloride was 27.6 o/o of that available as ionized chlo­ rine. It was not expected to find this inorganio salt present in benzene solution. Either sodium chloride and iodide ere soluble in benzene to that extent or a small amount of ater was present. Our conclusion here was that if water as present, it as there thru the sodium salt as great care had been taken to have the ap- paratus thoroughly dry for this ork. determination for ater I as made on the sodium salt by heating a eighed a ple for t o hours at 150•. The salt as found to contain a very a 11 amoun of water, 77 centigrams of the salt containing 2 milligr e of ater. From the 3.3366 gr of salt introduced 9 illigrams of ater ould be freed. In vie of the eat solubility of the sodium chloride and iodide in ater, this ount as judged euffi- oient solvent. Summary of Analyeie Reaction mixture resulting when sodium salt was heated ith 2/5 I gram atom equivalent of iodine in b~nzene. Filtered. esidue IFiltrate sodium trichlorphenolate amorphous oxide sodium chloride iodine sodium iodide alcohol residue Taken up with water acidified made up to 500 c.c. with nitric acid No precipitation of phenol Portions titrated for free iodine. Found for sample .7260 grams. light amount of nitric acid 83.6 o/o iodine added. added ilver nitrate in excess addc 250 c.o. evaporated. Filtered Residue chiefly amorphous oxide. Taken up with chloroform and extracted ith reoipitate Filtrate alcohol. Filtered. silver chloride thrown silver iodide away Precipitate Filtrate evaporated washed with 2 o/o oxide brown residue ammonium hydroxide found for taken up with ater sample esidue Solution 2.2668 Residue Solution acidified silver iodide silver grams brown 1th nitrig. eight .17089 chloride 3 .088 tar Silver nitrate .0921 g. acidified grams eight added. Filtered iodine with sodium .0398 10.61 o/o dilute salt .0642 precipitate iodine added nitric 92.2 o/o sodium silver chloride acid of that salt(?) silver iodide filtered I added 1 .6 ashed i th 2 o /o onium hydroxide precipitate silver residue Solution chloride eilv r Silver weight iodi e chloride 1.10269 wash Cidified .2728 grams and 1th nitric chlorine dried acid. 67.18 o/o of wei t il tered. chlorine for available i n s ple precinitate sample 0.0872 silver gr B chloride .o 71 ei t for gr a B ple .300 iod ne .0744 5.43 chlorine or o/o of .1487 for iodine total sampl added 27.6 o/o of po ibie ion ze. Methods for Determining Free Iodine in Benzene Solution

When the first benzene decomposition mixture as analyzed for free iodine, the method of titrating the solution directly with standard thiosulphate was used. The titration was done in an ordinary beaker, stirring the two nonmiscible liquids together 1th a glass rod. The amounts of iodine caloulated ere so lo , only 50 o/o of what was expected, and it as evident another method would be needed. Titrating directly a benzene solution of iodine with sodium thiosulphate appeared to be impossible.

Determination of Partition Coefficient of Iodine Bet een Benzene and Potassium Iodide

As a second method of determining free iodine in benzene, the use of the partition coefficient of iodine bet een benzene and I potassium iodide was considered. Our idea was to determine this partition for a range of iodine conoentr tiona, the minim and maximum of which ere to correspond 1th the oonoentrationa used in the benzene deoo positions. Having determined the partition coef­ ficient, e ould then shake together equal volumes of the unkno iodine solution and potassium iodide of the e e str ngth used in I the partition determinations, for a definite length of ti e. After the t o liquids had separated, t e potassium iodide solution as then to be poured a ay from benzene and titrated 1th thiosulphate for iodine. Having the partition value, it ould be a simple at- ter to calculate the free iodine in the benzene solution. or the partition determinations, the follo ing s plea of iodine ere used: -44- 1. .3832 grams 2. .4309 grams 3. .4589 grams 4. .4621 grams 5. .5138 grams These were all dissolved individually in benzene and made up to 100 c.c.; 50 c.c. portions were then shaken with the small amount of approximately N/10 potassium iodide in tightly stoppered sepa­ ratory funnels for the same periods of time. This was done at room temperature. After the two liquids had separated into two well de­ fined layers, the potassium iodide as poured away, transferred to a burette and titrated with thiosulphate for iodine, using starch solution as an indicator. The amount of iodine remaining in the benzene was found by the difference bet een the iodine dissolved and iodine determined in the potassium iodide solution. The ratio between these two values was taken as the partition coefficient. The following values were obtained: Sample 1. Weight of iodine .3832 grams.

Kl3 Ba2S2 03 (factor 1 c.c. = .01212 I) Titration 1 25.26 c.c. 5.37 Titration 2 23.32 c.c. 4.97 Partition from l = 2.049, from 2 2.068 Average 2.06 Sample 2. Weight of iodine .4309 grams. Partition from first titration 2.207 Partition from second titration 2.202

Kls 26.44 -46- added in small quantities, and the flask vigorously shaken after each addition. The amount of iodine determined checked with the amount used in making up the sample. A solution was then made containing a known amount of i odine to hich was added some of the amorphous oxide; the point here was to ascertain whether or not the presence of the oxide would interfere with the titration. The average of several titr ations was taken and the iodine oaloulated agreed with that used. Iodine taken • 5331., iodine found • 5332. PARTITION C 0 E FF I CI ENI ~ ~ ()) N 0 0 - 0

.38

.39

.40

< .41 -("'l (j) I .4Z -I

0 .43 "'l - 0 .44 0- z rl .45 -z .46 (i) :0 )> .47 ~ (J)

Jf8

.49

.50

.51

.5Z -45- Sample 3. Weight of iodine .4589 grams.

KI3 Na2S203 1 25.33 6.43 2 24.16 6.20 3 9.96 2.58 Coefficient 1 2.035 2 2.103 3 2.167 Average 2.101 Sample 4. Weight of iodine .4621 grams.

KI3 Na2S203 1 18.93 4.83 2 18.45 4.77 Average 18.69 4.8 Coefficient = 2.04 Sample 5. Weight of iodine .5138

KI3 NaaS203 c. c. 1 24.25 c.c. 7.78

2 26.13 c.c. 6.76 Average 25.19 7.02 Coefficient = 1.360

Xt was decided to work out a method of direct titration. A known am~unt of iodine was dissolved in benzene and an aliquot part titrated with the thiosulphate. The titration was carried out in an Erlenmeyer fitted with a ground stopper. The thiosulphate was i