INDUSTRIAL AND ENGINEERING CHEMISTRY
September, 1946 90
Io
90 --\\
/'
Q.
80
80 t 0 0
E
d
F
a
LL
0
50
0
20
40
60
TEMPERATURE, DEGREES CENTIGRADE
Figure 1. Per Cent Yield and Setting Point of Technical DDT Prepared at Various Temperatures Using 98% Sulfiiric Acid iii a one-liter, three-necked flask, while 178 grams of 104% sulfiiric acid were dropped into the reaction mixture over a period of 2.5 hours. During the addition the reaction IT-asmaintained a t 20" C. by external cooling with cold water. After the addition the reaction mixture was stirred a t the same temperature for 4.5 hours and the crude D D T isolated according to method A . L I m H o D A . To the reaction mixture (0.5-mole run based on chloral), contained in a Pyrex separatory funnel, w r e added 250 inl. of n-hexane (Phillips Petroleum Company, 7 5 O ; ti-hexane, .,-c*. -J ,,c methylcyclopentane, boiling point 68" C.). The crude D D T was cstract,ed by stirring a t 60" C. The sulfuric acid w a s dropped i t r i d the hexane solution washed successively, while maint'ained :it tiO" C., with 125 ml. of water, 125 ml. of 2% sodium carbonate -oliition, and 125 ml. of wat,er. The resultant hexane solution evaporated on the steam bath and the filial solvent removed hy tieating for 15 minutes on the stenni bath under a pressure of 35 nim. or less. The molten D D T was then drained into a porcelain dish to solidify. This gave 149 grams (84% of the thcory based on the chloral taken) of a cream-colored solid which had a setting point (8)of 86.2" C,. This was further crystallized from 500 ml. of hexane to give 84 grams, or 5 7 5 recovery, of a protiuct melting at 106.5-107.5" C. .4 tot'al of more than six%y runs were conducted in order to drfine the conditions reportrd i n I'igures 1 t o 5 ,
917
initial concentration of acid. I n actual practice the final titrated acid concentrations were usually 6 to 12% below the calculated values. It is apparent from Figure 1 that, for 98% sulfuric acid in the laboratory apparatus employed, a temperature of approximately 15'' C. is optimum for the production of DDT. A temperature of 15-20' C. not only gave the highest yield under these specified conditions but also produced a product with the highest setting point. The setting point, an empirical determination conducted essentially as described by Fleck and Preston (8), was adopted by the Army Quartermaster Corps as a criteria of purity of technical DDT. It should be remembered that the pure D D T has a melting point of 108.5-109.0° C. (uncorrected) and a setting point of 108" C., whereas the crude D D T from the reaction mixture is contaminated with approximately 15y0of the o,p'-isomer along with other by-products and has a setting point from 8093' C. The melting point of a technical sample of D D T is somewhat higher than the setting point. Thus a sample of technical D D T with a setting point of 91 O C. softens indefinitely from 65' to 80" C. but does not melt completely until 94" C. is rearhed. The setting points are indicated by the dotted lines in Figures 1 to 5. These and subsequent curves would undoubtedly be different if such conditions as rate and length of stirring, amount of excess chlorobenzene, and weight and concentration of sulfuric acid were varidd. If relatively large amounts of chlorobenzene are present in the crude DDT, this impurity is not conipletely removed by method A , and it is probable that the lower setting points corie-ponding to the lower yields in Figuies 1 to 5 arc due in part to u~iconsumedchlorobenzene contaminating the product. The lower yields at the higher temperatures reported in Figure 1 would not be greater in shorter time, since warm sulfuric acid of concentrations less than 100'3'0 has little effect on D D T . EFFECT O F REACTION TIME
It is poa*ible that the cauw for decrease in yields a t lower temperatures (Figure 1) is incomplete reaction during the 6-how stirring time specified. 'However, the effect of time on the reaction a t 20' C. T\ as studied, and the results (Figure 2) show that the reaction is virtiially complete :tftcbr 1 hours under the condi-
EFFECT O F TE.\IPERA'I'I R E
Figure 1 records the results of esperiments dmigned to slio\v the csrfect of temperature on D D T condensation when 98% sulfuric acid \vas used as the condensing agent. All references in this p;~pt'r to the acid concentration for a condensation mean calcu1atc.d final acid concentration, on the assumption that the D D T ~.oiitlens:xtionwas 100% complete and no sulfonation took place. .ill, or nearly all, of the chlorobenzene was consumed in the rc.tl.tion either by condensation with the chloral or by sulfonation. .5ince the relative extent of each reaction was not known in :idvance, the true final acid concentration could not be calculated. 1~:xperimentally the acid concentration was kept approximately i:oiistant by the initial mixing of chloral, chlorobenzene, and an ~ ~ x c r of s s the concentration of sulfuric acid being studied, and by rhe subsequent dropnise addition of a sufficient quantity of 1 0 4 7 sulfuric acid, as the reaction proceeded, to maintsin the
I
50 2
4
6
8
TIME IN H O U R S
Figure 2.
Effect of Reaction Time on DDT Condensation
918
INDUSTRIAL AND ENGINEERING CHEMISTRY
Vol. 38, No. 9
i n weight of the chlorobenzene layer. The assumption was made that the disappearance of the chlorobenzene layer was a ilieiisnrc of the extent of sulfonation; the validity of this as.uniptiuii ~ra5shown by isolation, in several cases, of a near' liiantitative yield of p-chlorobenzenesulfonic acid as the am:>ionium salt. The results of these sulfonation euperimcwts are .i!c.d in Figurc 6 . The accuracy of the data on this graph is i i i i t inure than -3%; the values given are greatly dependent i i 1 ) o i i t h c eficieiic>- and uniformity of stirring. In every c x c the ;;c.ic! c~~~iice~itratinn is the initial concentration; the find values t \ (1rc mud1 I(>*?, dvpending upon the exten't of sulfonation. It -1iotiIcl 11c rc-eniphnsized that this is the extent of sultunntioli of ~ ~ I i l ~ ~ r o l ~ iwider ~ n z e nspecific e conditions which approsimate those r ' i i i i i i d in the conventional DDT condensation. 11,:t tkpicnl es:mple 56 grams of technical chlorobenzeric were - t i m d rapidly n.ith 200 grams of 10270 sulfuric acid for 5 hours : I T t53 C. After this time the Chlorobenzene layer m-eighcd 13 grnnis (correspondirtg to 77Y0 sulfonation) and the sulfuric acid l:iver nciylicd 241 grams (corresponding to 74% sulfonation). The sulfuric acid layer was diluted with approsimatcly thrce voluinii> of ice and made basic with 50 nil. of concentrated ammonium hydroxide at 20" C. After standing one hour,*the solutioii deposited crystals of the ammonium p-chlorobenzenesulFonate, which weighed 78 grams after filtering, washing with ether, and drying at 120" C. for 3 hours. This is a 97.570 yield on the basis of the loss in weight of the chlorobenzene layer, and a 100.8"; yield o n the basis of the increase in weight of the sulfuric acid lager. On recrystallizingfrom 150 ml. of boilingwater, these 78 grams gave 63 grams of pure ammonium p-chlorobenzenesulfonate (81% yield on the basis of the gain in weight of the sulfuric acid layer). K i t h the help of these sulfonation data, a study w3s undertaken of the D D T condensation. Acid concentrations were d c c t e d from Figure 6, which at the various temperatures possessed the same sulfonating strength toward Chlorobenzene. A concentration of acid was arbitrarily chosen which resulted in 40% sulfonation of the chlorobenzene under the conditions of the D D T condensation but in the absence of chloral. These acid concentrations are represented by the intersections of the dotted line with the curves for the various temperatures in Figure 6. r m : w
I
40
94
96
I 98
I
40 I02
100
PER CENT CONCENTRATION SULFURIC ACID
Figure 3. Per Cent kield aiid Setting-Point 01 Technical DDT Prepared a t 20" C. Using Yarioub Concentrations of Sulfurir k i d
tions employed. The reactions from n.hic.h 1 t i t , were obtained were carried out as folloivs:
ht;i
f r ~ rE'igurr 3
Siiiiultaneuw r u m \rere pcri'i>rnit~du-iiig I Y J l l h t 8 i l ~ ilid stir. ring (800 rcvolutions per minute) wit while the temperature of the reaction C. Each of the 2-liter, three-necked 73.7 grams (0.5 mole) of chloral and chlorobenzene. Over a &minute period 200 grams of 98% sulfuric acid were added to each reaction; then 200 grams of 104% sulfuric acid were added over a 55-minute period. Immediately upon completion of this addition, run 1 was norked up by stirring successively with 800-, 400-, and 300-ml. portions of carbon tetrachloride at 20' C. These extracts were combined, trashed successively with 200 ml. each of water, 2% sodium carbonate solution, and water, and dried over calcium chloride; the solvent was then removed by heating on the steam bath in vii find heating under vacuum n--as continued 15 iriinut the time when the last bit of solvent was o b ~ c r ~ et od d Run 1 produced 107 grams of technical DDT, sc,itirig point 8 5 . 3 85.5" C. The other runs Jrerc identica! esccpt for t l i r , timc. i n . ter-t.&l allow-cii before the extractioii K R liogiin ~
92
I
92
w U
EFFECT OF ACID COKCE>TRAI'IOV
I t was not feasible to study the effect of varying thc acid coilcentration over a wide temperahre range: conseqnently :I teni-. pernture of 20" C. was chosen, and the yields x i d setting poiut.. ,of the DDT obtained were determined when acid conccJrltratioic:. of 95-104% were employed. The results are shown in Figure 3. The yields with 98-99Y0 sulfuric acid were cqu:illy good; but :Isoon as the arid concentration reachcd 1 0 0 7 , tliit yicxldc f ~ ! l < I < I sharplv SULFONATION OF CNLOHOBE.\LEAE
84
-84
6c z
w -?
rr,
ti
w W
Y
0
-76
76
a
d
w>. I-
z 0 W
f, 6 8 a
Since the main limiting factor in t,hci (widensation is the exteilt
of sulfonation, the sulfonation of chlorobenzene alone n-as studied. In forty experiments 56 grams of chlorobenzene and 200 grams of sulfuric acid of varying concentrations were allo~veclt o react under controlled temperature conditions, for 5 hours with uniform stirring. These conditions approximate those found i l l the D D T condensation reaction. The resulting mixturil ~ v a allowed to separate, the sulfuric acid layer drawn, and the extent of sulfonation under these specific conditions measured k)y determining the increase in weight of the sulfuric acid layer : I I I ~&-
6C
,
60
Figure t . Per Cent l i e l d and Setting Point.. of Technical DDT Prepared at \'arious Temperatures with Acid Concentrations of the Same Sulfonate Strength
September, 1946
919
INDUSTRIAL AND ENGINEERING CHEMISTRY
The results of the condensations using these acid strengths and temperatures are represented in Figure 4. The results indicate that the highest crude D D T production with the smallest concomitant formation of p-chlorobenzenesulfonic acid probably occurs a t a temperature in the neighborhood of 20" C. and acid concentrations of 98-99%. The temperature is only optimum lor the 6-hour reaction time allov ed in these experiments. From the standpoint of the economy of time, hov-ever, it is probably not worth while to run thP reaction longer for the sake of a slightly higher yield. I t is thus apparent that. in the condensation mixture oi sulfuric acid, chlorobenzene, and chloral, t n o chief reactions competethe sulfonation of chlor6ben7ene and the formation of DDT. In the absence of chloral the sulfonation of chlorobenzene is extensive a t the same temperatures and concentrations commonly employed in the D D T condensation. In spite of this fact, the yield of D D T is excellent under conditions which, in the absence of chloral, result in a high percentage of sulfonation; this indicates that, of the two competitive reactions, the D D T condensation is much faster under properly chosen conditions. I n addition, the data on the formation of D D T shon. that the temperature effect on the extent of sulfonation of chlorobenzene is larger than that on the extent of D D T formation. Thus the more concentrated acids with lower temperatures seem to favor D D T formation, whereas the higher temperatures with more dilute acids seem to favor sulfonation. EFFECT OF EXCESS CHLOROBENZENE
Lnder controlled conditions of temperature (16" C.), time ( 5 hours), stirring (800 r.p.m. with a Hershberg-type stirrer), and acid concentration (final calculated acid concentration 99yo), six simultaneous condensations were conducted in ivhich the variable was the amount of excess chlorobenzene. These runs were worked up by method d to give the results reported in Figure 5 . The 15-minute heating on the steam bath ordinarily used in method A to remove the solvent was insufficient for the complete removal of excess chlorobenzene from the runs which used a 50% or greater molar excess of chlorobenzene. It was necessary to heat at 105" C. and 2 mni. pressure for 15 minutes to remove the chlorobenzene completely from these runs. The 10% molar excess chlorobenzene used in the experiments represented by Figures 1 to 4 gave an 84% yield in this series. This indicates that the maximum yields in Figures 1 to 4 would all be approximately 10-1270 higher if 100% excess chlorobenzene had been employed. Figure 5 shows that the increased yield makes it profitable to employ an excess chlorobenzene up to approximately 75y0molar excess, but the increased yield of D D T beyond this point does not laompensate for the large amount of chlorobenzene nccessary. 11AXI\lUJl YIELDS OF DDT
A 97-987; yield of crude D D T of setting point 90" C. was obtained when the condensation with chloral was carried out in a large volume of 98% sulfuric acid and 4-molar excess of chlorobenzene at 5-10' C. for 8'hours. as described in the following experiment :
A mixture of 7-1 grams (0.5 niole) of pure chloral, 250 grams ;2.2 moles) of technical chlorobenzene, and 1100 granis of 98% julfuric acid was placed in a 2-liter, three-necked flask equipped with a dropping funnel, thermometer, and efficient stirrer. Since the reaction is heterogeneous, the efficiency of the stirring is important and should be such that the reaction is practically emulsiEed. While the reaction mixture was maintained a t a temperature betneen 5" and 10' C., 1100 grams of 10170 sulfuric acid were added over a period of 4 hours. The reaction mixture was stirred 4 more hours a t the same temperature and then worked up by extracting with three 500-ml. portions of hexane solvent a t 60" C. The sulfuric acid layer was discarded and the hexane layer washed with a 4%sodium carbonate solution and then with water. The solvent was removed from the product by heating on the st,eam brtt,h, and the last tracw were taken off by heating in a
I
I
80
2
3
4
5
6"
MOLES OF CHLOROBENZENE PER MOLE OF CHLORAL
Figure 5 .
Erect of Excess Chlorobenzene on Yield of DDT
vacuuui oven a t 10 nini. pressure for 2 hours a t 80" (2. Th: yield vas 171.9 grams or 97.0%; the setting point was 89.9 , 90.0" C. A duplicate run gave 174.3 grams (98.4%), setting point 90.2', 90.5" C. Although these proportions are not economically practical, they indicate the trends which lead to the best yields. There is obviously a point at which the increased cost of materials and decreased size of run (due to the increased volume of sulfuric acid used) outyeighs the advantages of higher yields. This depends upon the particular conditions and equipment employed. COMPOSITION OF TECHNICAL DDT
The composition of technical D D T has been extensively studied by Haller et al. (12). The approximate composition of the technical D D T obtained in the expeiiment described above was determined prior to their publication. The results are of interest because of the specifically defined conditions under which tho technical D D T was prepared. The crude DDT, which was obtained in yield and had a setting point of 90.0" C., was ' purified by a systematic fractional crystallization from hexane and s. distillation of the residual oils, followed by further crystallizations to give approximately 77% pure p,p'-DDT, 15% o,p'isomer of DDT, 1.5% of an oil, probably 2-trichloro-1-p-chlorophenylethanol; the remaining 6.5% represents losses and unidentified residues which undoubtedly contain some more p,p'D D T and its o,p'-isomer. It should be emphasized that the sample of technical D D T used here was prepared from chloral, purified through the crystalline hydrate; no impurity such as I-dichloro-2,2-bis-(p-chlorophenyl)-ethane,DDD, should be present. The 171.9 grams of D D T of setting point 90" C. were put through a seven-step systematic fractional crystallization from hexane. Fractions 71, 72, and 71 gave a total of 130 grams of purified D D T with a melting point above 105" C. Fractions 7s, 78, and 77 gave a total of 34.5 grams of an oil which would not crystallize. These last three fractions were combined and distilled to give the following: fraction 1, 2.5 grams, boiling point 120-125" C. (1 mm.); fraction 2, 24.6 grams, b.p. 172-175" C. (0.5 mm.); fraction 3, 2.5 grams, b.p. 175-185" C. (0.5 mm.); S positively and a residue of 3.2 grams. The first fraction W ~ not identified but was soluble in concentrated sulfuric acid, and its
INDUSTRIAL A N D ENGINEERING CHEMISTRY
920
voi. 38.
NO.
9
If it is a mixture, the other component cannot be simply p,p'DDT, for the eutectic between p,p'-DDT and its o,p'-isomer melts a t 59-40' C., as Figure 7 shows. Thus from the 171.9 grams of crude D D T obtained in 97% yield, the authors obtained 113 grams (77%) of pure p,p'-DDT, m.p. above 105' C.; 26.1 grams (15%) of crude o,p'-isomer of D D T , m.p. 63-70" C.; or 17.2 grams (10%) of purified o,p'-isomer; 2.5 grams (1.5%) of a crude material, probably 2-trichloro1-p-chloiophenylethanol;5 grams of oils and residues; and 0.2 gram of an unidentified material melting at 55.0-55.5' C. These findings are in complete agreement with those of Haller et al. (18)but differ somewhat from those of Gunther (11) reported for the material obtained by the Brothman process (4).
t W 2 0
a P W
90
92
94
96
98
io0
102
INITIAL SULFURIC ACID CONCENTRATION
Figure 6. Extent of Sulfonation of Chlorobenzene with Various Acid Concentrations and Temperatures under the Conditions of DDT Condensation
boiling point was the same i s that reported ( l e )for 2-trichloro-lp-chlorophenylethano1. A small sample of fraction 2 was finally obtained in crystalline form by cooling its hexane solution to -40' C. Once the crystals were obtained, it was possible to crystallize many of the oils from previous runs. Fractions 2 and 3 solidified when seeded with the above crystals and melted a t 6370" C. These two cuts were crystallized from hexane, in which they were very soluble, by cooling to -40"; they were then crystallized from ethanol and methanol to give a total of 17.2 grama melting. at 73-74" C. (analysis of Cl4HsCls for GI: calculated, 60.0; found, 49.3). This material was identified by oxidation to 0,p'dichlorobenzophenone in a similar manner to that reported by Haller et al.(If). The remaining fraction, 74, from the systematic crystallization was Forked up to give 200 mg. of a solid which, after successive recrystallizations from hexane, ethanol, methanol, and acetone, still melted sharply a t 55.0-55.5" C.
PER CENT DDT
Figure 7. TemperatureConiposition Diagram for the System DDT and Its o,p'-Isomer
"his same material had been obtained from prixvious fractional erystalliaations. An attempt to prove the structure by oxidizing 800 mg. of this material in acetic acid with chromic anhydride reaulted in the isolation of 92 mg. of the o,p'-isomer of D D T (melting point Z3-4" C., mixed m.p. 72-73" C.), I t appears that the material melting a t 55.0-55.5" C. is either a different crystalline form of the o,p'-isomer or a mixture containing this isomer.
SYNTHETIC MIXTURES OF p,p'-DDT AND o,p'-DDT
Synthetic mixtures of pure D D T and its o,p'-isomer were prepared, and. the melting points were taken in capillary tubes in an electrically heated aluminum block. The point at which the first liquid formed determined the lower curve, and the lowest tempersture a t which a completely clear melt could be obtained defined the upper line. The latter value was determined by allowing the temperature of the block, t o rise a t a rate of not more than 1 degree every 3 minutes. The results are given in Figure 7. Similar mixtures, in which the setting points were determined, were studied by Fleck and Preston (8). RELATIVE SOLUBILITIES OF DDT
The solubility of D D T in various solvents was reported by Gunther (IO), Campbell and West (6),and Fleck and Haller ( 7 ) . Figure 8 shows additional solubilities determined in two ways: ( a ) A known weight of p,p'-DDT wm dissolved in the solvent, well above the saturation temperature being determined, and the solution allowed to cool slowly until crystallization commenced, as noted by the first separation of crystals and the leveling out of the cooling curve. The solubilities were carried out in a jacketed, tared test tube 2.5 X 18 cm., with constant speed mechanical agitation (180 2-inch strokes per minute) to prevent supercooling. After each determination the weight of the solvent was immediately determined. (6) This method was not applicable a t or near the boiling points and, in these cases a solution of a known weight of p,p'-DDT in the desired solvent was adjusted to the saturation point in a tared flask. The flask was cooled immediately by immersion in a cold bath, the gross weight determined, and the weight of solvent determined by difference. The results of these solubility determinations are shown in Figure 8. The solubility of D D T in sulfuric acid was also studied. D D T was found to be inappreciably soluble in anything less than 1 0 0 ~ o sulfuric acid. Thus when 5.02 grams of D D T were heated a t $IOo C. for 10 hours with 390 grams of 98% sulfuric acid, 4.2 grams of t,he original D D T were recovered by filtration and washing. At room temperature all of the D D T was removed; but when 5.02 grams of D D T were heated for 10 hours a t 90" C. with 390 grams of 100Tosulfuric acid, only 0.4 gram was recovered and the sulfuric acid solution had turned black. The o,p'-isomer of D D T is approximately twice as soluble as p,p'-DDT in 95yo ethanol and three times as soluble in hexane a t room temperature, I t is apparent, then, as to why the crystallization from such a solvent greatly purifies the D D T and leaves most of the o,p'-isomer (the main impurity) in the mother liquor. Chlorobenzene and carbon tetrachloride, especially the former, show a high solubility for D D T a t room temperature and are, therefore, good solvents for the cold extraction of DDT. If it is desirable to raise the melting point of the crude product a certain definite amount without the losses inherept in crystallization, partial purification can be achieved either by pouring the molten D D T into a definite amount of ethyl alcohol (or other solvents) or of a predetermined concentration of aqueous ethyl alcohol
INDUSTRIAL AND ENGINEERING CHEMISTRY
September, 1946
921
METHODS O F WORKING UP TECHNICAL DDT
TABLE I. Methods Coinpared
Run No.
A
c- 1
B .-I
H A B d (
.'I D
A E A
E
c1 c35 C35 C32 c32
e39 c'35 C3.5 C3.5 e35 C35 C32 c32
CO?dPAPISOS O F >fET&iODS O F LF'ORKING CP
D D T CONDENSATION ~IIXTCRE Yi$d,
7,
82 63 86 78 89 84 86 83 86 84 86 81 89 85
Setting Point, ' C. 83 94 88 90 89 91 88 89 88 89 88 88 89 87
Methods Cornpared A
F
A F
.4 C
A
H
I"A J
Getting
Run No.
Yield,
C32 C32 C18 C18 C35 C35 e35 C35 c34 c34
89 53 86 53 86 87 86 82 84 83
E!
Yo
Point, C. 87 103 84
90
81
104
88 88 88 87
90
89 89 91
with vigorous stirring, or by \yet grinding the product with such solvents. This produces, in addition to purification, a desirable granulation of the final product. l-DICHLORO-2,2-BIS-(p-CHLOROPHENY L)-ETHANE
The preparation of chloral by the chlorination of ethanol according to the method outlined in Thorpe (18) usually produces a low boiling impurity because oi incomplete chlorination. By fractionating this low boiling material through a sixty-plate column, the dichloroacetaldehyde (14).wa&separated which had a boiling point of 88" C. and an oxime melting point of 177-178" (analysis of CtH,ClZ for C1: calculated, 62.78; found, 62.8663.32). When thi8 material was condensed with chlorobenzene in the presence of concentrated sulfuric acid, ldichloro-2,2-bis-(p-chlorophenyl)ethane,DDD, was obtained. This was accomplished by mixing 113 grams of pure dichloroacetaldehyde, 300 grams of chlorobenzene, and 500 ml. of 98% sulfuric acid in a 2-liter, three-necked flask at 20" C. while 294 grams of 104% sulfuric acid were added to the stirred mixture over a 1hour period. The reaction was stirred 12 hours and the product extracted with ligroin a t 60" C.; the solvent layer was washed twice with 2y0 potassium carbonate solution and once with water; the solvcnt was evaporated to 500 ml., and the product was allowed to crystallize (weight 196 grams or 6 l % , m.p. 104-106" C,). This product was crystallized from ligroin (155 grams, m.p. 108.5-109.5D C.) and recrystallized from 95% ethanol (135grams, m.p. 110.5-111.0" '2,). Further recrystallizations did not change this melting point. A mixed melting point dptermination between this D D D and pure D D T was approximately 85-90" C. (analysis of C14H&14 for C1: calculated, 44.2; found, 43.8). Haller et nl. ( 2 2 ) have shown that this is an impurity in technical DDT
Ten different methods of working up the D D T condensat,ion mixture are compared in Table I. Method A is the standard against which all other variat,ions were compared. I n comparing any two methods, the same reaction mixture was used. This w a ~ accomplished by filtering the crude reaction mixture through glass cloth and dividing both the filtrate and precipitate into equal portions, recombining the separate portions of filtrate and precipitate, and working up one part by method A and the other part by the method being studied. I n this way the starting reaction mixtures of thc t>est and the standard methods were identical. h I w H o D B. The product was filtered directly iron1 the reaction mixture through glass cloth or a sintered glass filter. I t was then washed by mechanical stirring with water, sodium carbonate solution, and water. The product was finally dricd For six hours at 60" c. METHOD C. Thv reaction mixture \vas diluted with an equal voldme of water, and care was taken not to allow the temperature to rise above 80" C. The creamy reaction product'was filtered and washed by mechanical stirring with sodium carbonate solution and water. It, was finally dried as in method B . METHODD. Except that the product mas subjected to a stoaming process for 5 minutes in the molten state this method was identical with the standard process. All of tho water was removed by heating a t 100" C. under vacuum until a clear melt was obtained; the melt was then drained onto a porcelain dish t,o crystallize. METHODE . This was identical with standard method A , except that chlorobenzene instead of hexane was used as the solvent. F . The standard procedure LTas followed to the point METHOD where the solvent was evaporated. Here the total volume wiw reduced t o 200-250 ml. for a one-mole run and the D D T allowed to crystallize; it was then filtered and dried. METHODG . The reaction mixture was poured onto 1 kg. of ice for each half mole of chloral used. The solut,ion was stirred
EFFECT OF IMPURE CHLORAL Oh TECHNICAL DDT
Tlic following two runs were identical n ith the rsccption of the starting chloral. In run C18 the chloral mas pure; in run C19 the chloral was crude material used just as it was obtained by distilling a batch of chloral alcoholate from sulfuric acid under varuuni. (This batch of chloral alcoholate had been chlorinated to a density of 1.5 at 20' C. The distillation of a sample of chloral from this chloral alcoholate gave 178 grams boiling below 96.5' C. and 498 grams boiling between 96.5' and 97.5' C.in an ordinary distilling flask.) The two condensations nere conducted essentially as indicated in the description of a typical D D T condensation and were worked up as in method A . The results were: run C18 (pure chloral), 87% yield, 84.4" C.setting point; run C19 (crude chloral), 86% yield, 72.7" C. setting point. It is a p parent that the low boiling impurity does not loner the crude yield appreciably but contributes greatly t o the lower setting point of the crude product.
TEMPERATURE, DEGREES CENTIGRADE
Figure 8.
Solubility of DDT in Various Solvents
A . Chlorobenzene B . Carbon tetrachloride C, D . Hexane and l i p o i n (boiling range 75-8S0 C.) E . Kerosene (boilinu ranue 181-207° C . ) F. n-Propanol C. Absolute ethanol E. 95%ethanol
INDUSTRIAL AND ENGINEERING CHEMISTRY
922
Vol. 38, No. 9
PLIIFICATION OF TECHMCAL DD'I
TABLE 11. COMPARISON OF METHOD5 O F Setting Point,
c.
Method
No. IO
la lb Ib
Crude Treated Recov- Method D D T D D T ery, % No. 77" 85 b 77a 85 b 770 85b 77' 83s 77' 826 88 b 775 88)
102 94 102 92 92 88
4 4 5
5 6 7 8 8 8 9 10 10
PURIFYIXG
S e t t i y Point, C. Treated DDT
Crude DDT 77a 85) 77a 96 77O 85b 83) 85b 87b 91 86 b .03
102 96 100 106 89 101 89 91 89 9;
DDT Kecovery, % 58
82 45 68 63
In addition to the various methods of working up t'he D D T condensation mixture, there are other obvious methods of treating the crude product in order to purify it. -4description of some of these methods follows; the results obtained by the use of these methods are compared in Table 11. The percentage yields and temperatures \%-ererounded off to the nearest degree,
1 I t . T r i c i D l o . A 50-gram portion of crude D D T \vas tritiiriited tliorouglily in a mortar with 100 ml. of absolute ethanol a t room temperature. This quantity of absolute ethanol should not dis2a 99 -01vi' more than 4 grams of p,p'-DDT a t this temperaturc. 26 91 2b 91 IO? 1tt:Tiiou l b . This was identical viith l a except that 100 ml. of 2C 94 !tf?% i~tli;molivcre employed. This amount of 95% ethanol dip30 107 85 79 3b 11 91 88 87 08 .olvc,c oiily about 1 gram of p,p'-DDT a t this temperature. \ I b c r H o D IC. This method was identical with l a except 100 This sample of crude D D T with setting puir.t of 7 7 c C. was obtained L,, r i i l , rit'!lOcl, ethanol were used. combining t h e crude D D T from several typical runs and the residues fror;, iome previous crystallizations. This sample therefore has a n abnormall: 1 I ~ : ~ i i i oId. o This was identical Tyith l a except 200 nil. of 80% low setting point, and the percentage recover;. on treatAent is always lower V l l l ~ r l l o lI\ ('re usod. than normal whereas the increase in setting point is alwal-s higher t h a , normal. All Letting points were rounded off t o the nearest degree. 1 I I ; T I I O D 2n. Crude molten U D T (280 grams) was poured with b These are samples of D D T from typical condensations. \ I ~ I I ~ O U Fmcchnnical stirring into 300 ml. of 95YG ethanol w h i d Decomposed, in attempting t o distill this sample of D D T , wheu rhe temperature reached about 190' C., the amount of gas evolved --as sucL Irad betii n.armed to 60" C'. The ethanol was cooled to 15' C.. t h a t the vacuum could not be maintained and the distillation had t o be :&nil thc. pIYJdUi.t filtered and dried. abandoned. 1 1 I m i o u 2 b . Crude molten D D T (60 grams) was poured wit1 'VIF(JrOllb mechanical stirring into 200 nil. of 90% ethanol a t 50" c' 11awoD 2c. Crude molten D D T was poured with mechanical TABLE 111. SETTING POINTS OF MIXTURESOF stirring into 100 ml. of 80% ethanol at room temperature. PURIFIED AND TECHNICAL DDT 11aTIioD 3a. Crude molten DDT was poured with vigorour -Component ria--Component BbSettinx nicchanical stirring into 200 ml. of a 1%Triton B solution. The Grams Grama % Point, C producat formed a taffylike mass which solidified on cooling and was readily broken up into a fine granular form. It \vas theti 0.0 0 30.0 LOO 83.4 3.0 10 27.0 90 84.8 6ltercd, xashed, and dried a t 60" C. for 6 hours. 6.0 24.0 20 89.9 80 METHOD3b. This was identical with method 3a except that 10.5 35 19.5 65 95.8 15.0 50 15.0 98.9 50 a 2% Dreft (sodium lauryl sulfate) solution was used. 22.5 75 7.5 25 103.2 hIETHoD 4. Crude D D T (50 grams) was recrystallized fronl 30.0 100 0.0 00 107.5 200 ml. of 05% ethanol, cooled to 15" C., filtered, and dried. Component A is puriEed D D T with a setting point of 107.5' C. METHOD5. Crude DDT (50 grams) was recrystallized from b Component B is crude D D T with a setting point of 83.4O C. 100 grams of hexane, cooled to 15' C., filtered, and dried. METHOD6. Crude D D T (50 grams) was recrystallized from 125 grams of hexane mother liquors from a previous DDT c r y talliz at ion. mechanically uiltil the ice had nivitd, 1 tic, ivater tiec.aritc,d, a1111 METHOD7. Crude DDT (50 grams) \vas recrystallized frorri thesolid D D T washed first with a 2% wliuin carbonatr uoiu100 ml. of methanol. tion and next with water. The D D T w:is finally f i l t t v t i an
