CA1171377A - Purification of phenol with reduced energy consumption - Google Patents

Abstract

ABSTRACT
PURIFICATION OF PHENOL WITH REDUCED ENERGY CONSUMPTION
Phenol (10) produced by the cleavage of cumene hydroperoxide is chemically treated with a base (12) such as a polyamine, is optionally then treated with an acid or acid anhydride (15) such as phthalic anhydride, is then steamed distilled (12) to remove a lites fraction as a water azeotrope (21) and the bottoms (34) of the steam distillation are vacuum distilled (35) to recover high purity phenol (36) as an overhead, The overheads (21) of the steam distillation are condensed (22) and may be phase separated (23) into an aqueous phase (25) and organic phase (24), with the aqueous phase (25) mixed (26) with an organic solvent (27) which preferentially dissolves benzofuran impurities. This mixture is phase separated into an organic solvent phase (30, 130) and a recycle aqueous phase (31). Alterna-tively, the entire condensed overheads (21) may be ex-tracted (146-151 and 153) with organic solvent. In some forms, the organic solvent phase (130) is washed with aqueous base (142) to remove phenol and impurities and the organic solvent (127) is regenerated for again mixing with the aqueous phase. In other forms the organic solvent is a portion (152) of the high purity phenol product (36). This treatment and careful control of the chemical treatment to minimize formation of benzofuran impurities reduces the steam consumption of the steam distillation step, thereby producing high purity phenol with reduced energy consumption.

Description

~7~377 DESCRIPTION
PURIFICATION OF PHENOL WITH REDUCED ENERGY CONSUMPTION
. _ _ . _ _ _ _ . .
BACKGROUND OF THÉ INVENTION
The present invention relates to the purifica-tion of phenol and particularly to the production of high purity phenol from technical or refined grade phenol which has been produced by the cleavage of cumene hydroperoxide.
A major source of phenol for many applications is the oxidation of cumene to cumene hydroperoxide and the acid cleavage of cumene hydroperoxide to a mixture of phenol and acetone which typically also contains unreacted cumene and side products such as alpha-methyl-styrene, various carbonyl compounds and various benzo-furan compounds. The mixture is generally subjected to a series of distillations to recover various fractions including a crude phenol fraction. Fractional distil-lation of the crude phenol fraction, under various conditions, produces a "~technical", "refined" or waler-white grade which is sufficiently pure for many pur-poses. Such phenol is unsatisfactory, however, for certain purposes and, in particular, when chlorinated or sulfonated with sulfuric acid, the phenol becomes red.
Additionally, such phenol, itself initially gray, gener-- ally discolors on aging, becoming darker and eventually black. These effects may be due to either ketonic im-purities such as mesityl oxide, acetol and ace~ophenone or to benzofuran impurities such as methylbenzofuran which may be formed by reaction of certain ketonic im-.-.': , ~137~

purities such as acetol with phenol.
The purification of such technical refined grade phenol by chemical treatment is described, for example, in U.S. Patent 3,692,845 to Cheema et al.
(September 19, 1972), U.K. Patent 883,746 to National Distillers Company (December 6, 1961); U.S. Patent

2,864,869 to Crocker et al. (December 16, 1958); U.K.
Patent 1,108,327 to Universal Oil Products Company (April 3, lg68), U.l~. Patent 1,148,907 to Imperial Chemical Industries Limited (April 16, 1969) and U.S.
i Patent 2,971,893 to Hood (February 14, 1961). The present invention is applicable to processes of puri-fication which include treatment of the phenol with a base such as alkali metal hydroxides or carbonates or monoamines or diamines as described in U.K. Patent 883,746 or hydrazines as described in U.S. Patent 2,864,869 or al]~aline hydrogen peroxide as described in U.S. Patent 2,971,~393. The present invention is especially applicable to processes of purification which include treatment with polyamines followed by the addi-tion of acid or acid anhydride followed by distillation as described in U.S. Patent 3,692,845. The present invention also includes steam and produc~ distillations as separate steps, which are also described as separate steps in U.K. Patent 883,746 and U.S. Patent 2,971,893.
The object of the present invention is to achieve phenol of high purity from technical or refined grade phenol or from impure phenol streams with a minimal use of energy, and especially a minimal use of steam in the steam distillation. This is accomplished by careful control of the chemical treatment to minimize the formation of certain hard to separate impurities during the treatment and by extraction of the overheads of the steam distillation.
BRIÉF DESCRIPTION OF THE INVENTION
The present invention includes an improvement in the process of purifying phenol produced by the cleavage of cumene hydroperoxide wherein phenol is ~ ~97~37~

3-treated with a base, optionally an acid or acid anhy-dride is added, the treated phenol is steam distilled to remove a lites fraction as a water azeotrope and the bottoms of the steam distillation are distilled to re-cover high purity phenol as an overheadr In the im-prOVelllent r the overheads of the steam distillation are condensed and phase separated into an aqueous condensate phase and an organic condensate phase, organics are ex-tracted from the aqueous condensate phase with an or~
ganic solvent which preferentially dissolves benzofuran impurities to produce a reflux aqueous phase, and the reflux aqueous phase is returned to the steam distilla-tion.
In one form of the present invention the or-ganic solvent is a material other than phenol, such as cumenej and the organic solvent after extraction is washed with aqueous base to remove phenol. The organic solvent is separated from the impurities (e.g. by dis-tillation) and used to extract additional organics from the aqueous condensate phase. In another form of the invention, the organic solvent is a portion of the high purity phenol.
DET~ILED ~
The phenol used as feed in the present process is produced by the cleavage of cumene hydroperoxide. It may he the technical or refined grade phenol normally produced as one distillation product from such a pro-cess. It may also be one or more impure phenol streams found in plants for the produc~ion of phenol containing water, containing coproduct acetone, containing unoxi dized cumene, containing alpha methyl styrene (a by-product), containing various high boiling impurities such as dimethylphenylcarbinol. The phenol may contain certain difficult to separate impurities r which are principally methyl benzofuran and ketonic impurities such as acetol~ mesityl oxide and acetophenone. In general the normal refined grade phenol is 10~7 in over-all impurities (under 5000 ppm~ but has relatively high 31 17!137~
~ -4-levels of the impurities hardest to separate methyl- -benzofuran (50-150 ppm), acetophenone (75-2000 ppm), acetol (1000 -4000) and other ketonics (100-500).
Various phenol-containing streams are produced (for example in residue reEorming and dephenolization of waste water) which have substantial water contents (5-20%) and high total organic impurity levels (10,000~30,000 ppm) but may have lower levels of methylbenzofuran (generally under 100 ppm).
As illustrated in the examples that follow, various such phenol streams can be used in the present process, thereby not only producing high purity phenol of better quality than refined grade phenol~ but also reducing the requirements for batch stills and the like to purify various other phenol streams. It should be appreciated that many of these other streams are currently purified by distillation to produce a product lower in overall quality than refined grade phenol.
If water is present in the feed, it is easily removed by evaporation from the phenol during the chem-ical treatment with the base. Some phenol is removed thereby as the water.phenol aæeotrope. This phenol may be recovered by combining the azeotrope with the azeo stripper overheads for solvent extraction.
It might be possibIe to construct a purifica-tion system from the prior art that followed chemical :- treatment of refined grade phenol with azeotropic distillation with water and then vacuum distillation to recover the product high purity phenol as an overhead stream. Such a process, assuming that the aqueous condensate (containing methylbenzofuran) was returned to the first (azeotropic distillation) column, would require large amounts of water (and thus energy for re-boiling) compared to phenol feed to the first column.
For example, one such experiment required 2.5-3.0 parts of water by weight for each part of refined phenol. If the aqueous condensate (containing methylbenzofuran) is not returned to the columnr a smaller fresh water ratio ~17~37~

could be used, such as about 1 part water per part of refined phenol in comparable experiments. Such a fresh water process ~ould generate undesirable amounts of phenol-contaminated water which must be treated to re-move organics.
In the present process, by extracting organicsfrom -the aqueous condensate or from the entire conden-sate tcontaining methylbenzofuran) in either case before returning the aqueous condensate to the column, a ratio close to that achieved by feeding fresh water to the column has been achieved. Thus, with cumene as the organic extractant, the same quality phenol was ob-tained using a water to phenol feed ratio of about 1.2:1. With high purity phenol as the extractant, a ratio of about 1:1 was achieved in compara~le experiments, Of course it should be appreciated that the level of certain impurities such as methylbenzofuran is directly linked for each system to the water:phenol feed ratio in the column, such that experiments can be considered comparable only if they achieve equivalent low levels of such impurities in the stream removed from the bottom of the first still.
The phenol removed with impurities during solvent extraction or the phenol from high purity phenol extractions can be recovered by treating with aqueous sodium hydroxide. For the sol~ent extraction process, the impurities are concentrated in the solvent phase.
For the high purity phenol extractions, solvent extrac-tion of the aqueous sodium hydroxide-phenol solution is required.
In addition to the use of extraction before returning the aqueous condensate, two other improvements are preferably made in the overall process. First the azeotropic distillation column is preferably controlled (as by the sensor control on the reboiler illustrated in Figure 1) so that the control temperature is at or near the boiling point of phenol. As a result, this portion of the column dehydrates the phenol since a separation l~7a37~

between phenol-water azeotrope and phenol vapor can occur. Second, the chemical treatment is preferably conducted with a polyamine followed by phthalic acid acidification to a pH of about 4.n to about 5.5 to prevent ormation of methylbenzofuran during chemical treatment.
DESCRIPTION OF THE PREFERP~ED E~lBODIMENmS
.
Referring to Figure 1, phenol from a cumene hydroperoxide cleavage process in stream 10 is charged to a treatment appa~atus 11 equipped with agitation along with a base in stream 12. Exemplary of the base is either 5 molar sodium hydroxide, 100 weight percent hexamethylene diamine or 70 weight percent aqueous polymeric hexamethylene diamine such as is sold by E.I.
du Pont de Nemours as Du Pont Amine ~48. The tempera-ture and residence time in apparatus 11 should be suffi-cient to complete the desired chemical treatment and tie up various impurities. A residence time in apparatus 11 of 1 hour to 5 hours is generally sufficient, with shorter residence times as small as a few minutes being satisfactory for certain polyamines. Temperatures of above 100C such as 150 180C are preferred.
If the phenol fed in stream 10 contains sig-nificant amounts of ~ater or low boiling impurities such as cumene, acetone and alpha methyl styrene, they will evaporate off in treatment apparatus 11 and can be con-veniently removed, preferably after passing through a simple still to minimize entrained organics, through stream 13 for treatment as with other low boiling mixture produced in phenol manufacture.
From appara~us 11~ the mixture is conducted to acidification vessel 14 where an acid or acid anhydride such as phthalic anhydride is added in stream 15. Such acid treatment may not be required in the case of many bases introduced through stream 12, but is generally desirable, especially when monoamines, diamines, or other polyamines are used as the base. It is preferred that the acid or anhydride in stream 15 lower the pH of ~171~

the liquid in vessel 14 to a pH no lower than a'~out 4.0 and pre~erably to a pH o between about 5.0 and about 5.5 to prevent formation of additional methylbenzofuran.
In the Examples, phthalic anhydride was fed into the acidification vessel as a solution dissolved in high purity phenol.
From acidification vessel 14, the mixture is passed through filter 16a, preheater 16 and stream 17 into the first still 18 in which it is steam distilled.
The recovered azeo water after extraction, stream 32, can be combined with the chemically treated phenol fed to preheater 16 and fed through stream 17 to the first still 18. Heat is supplied to reboiler 19 of still 18 in a conventional manner. Reboiler 19 is controlled by a thermal sensor 20 so as to maintain a constant temperature at a level in still 18 slightly above the boiling point of phenol, such as 185-195C , preferably 186-190C (assuming that still 18 is at about atmospheric pressure). The efect is to cause still 18 to perform two functions: Azeotropic distillation in section 18a above sensor 20 and dehydration in section 18b below sensor 20.
An overhead stream 21 is continuous~y with-drawn from the still 18 above the top plate or tray, and a bottoms stream 34 is continuously withdrawn from the reboiler 19 of still 18.
The overhead stream 21 of the first still 18 is condensed in condenser 22 where it is cooled to a temperature such as 25-50C and may then be fed to a separation vessel 23 where it is permitted to separate into an organic condensate phase and an aqueous phase.
The organic condensate phase is withdrawn from the bottom of vessel 23 in stream 24 since phenol which comprises the major portion o the organic phase is denser than -~ater. The aqueous condensate phase (con-taining both methylbenzofuran and phenol dissolved in water) from vessel 23 is withdrawn in stream 25 to a mixing vessel 26 where it is mixed with an organic ~7~3~

solvent, stream 27, to extract organics from stream 25. Alternatively, the entire condensed overheads in stream 21 may be passed to mixing vessel 26, especially if the solvent in stream 27 is high purity phenol as illustrated in Figure 3, discussed below. The mixture in vessel 26 is then fed in stream 28 to a second separation vessel 29 where it is again separated into an organic phase (now called the organic solvent phase) in stream 30 and an aqueous phase (now called the aqueous recycle phase) in stream 31. As illustrated, the organic solvent phase is assumed to be less dense than the aqueous phase (as is the case with cumene) such that it can be withdrawn from the to~ of vessel 29 in stream 30 and the aqueous recycle phase can be withdrawn from the bottom of vessel 29 in stream 31. If the organic solvent phase is denser~ however, the locations of streams 30 and 31 would be reversed. The aqueous recycle phase in stream 31 is fed back through stream 32 into the still 18.
It is desirable to reheat the aqueous recycle before introducing it back into still 18. Therefore stream 32 may be introduced into preheater 16 along with the mixture from acidification vessel 14. Alternative-ly, the recycle water (in stream 32), the mixture (from vessel 14) or both may be passed through condensor 22 in heat exchange with overheads so as to perform a part of the processes of preheating these streams and of cooling and condensing the overheads.
A stream 33 is provided to either add or with-draw water from stream 32 as is required to maintain a controlled or constant water content in still 18. It should be appreciated that the amount of water charged into the system through streams 10, 12 and 15 and the amount of water withdrawn from the system in streams 13, 24, 30 and 34 are sufficiently variable that, in some cases, make up water through stream 33 will be required while, in other cases, a portion of stream 32 must be diverted out of the system through stream 33 to avoid ';:

~7137~

water build up.
The bottoms from the reboiler 19 of still 18 in stream 34 are fed to a vacuum product still 35 which may be operated at 50-200 torr pressure (about 0.06-0.26 atmospheres). Product high purity phenol isremoved in stream 36 from vacuum still 35 near, but preEerably not at the top. A small pasteurizing cut is preferably taken ~rom ~he condensed overhead vapors 37 taken from the top of still 35, which are returned as reflux 38. A portion of the pasteurizing cut in stream 37 is recycled in stream 39 to treatment vessel 11.
Material in stream 39 may alternatively be returned to the first still 18 directly or indirectly, by introduc-tion into neutralization vessel 14 or preheater 16.
I'he bottoms from vacuum still 35 are withdrawn in stream 40. Some phenol of various grades may be recovered from these bottoms in a conventional manner.
Figure 2 shows a modified form of the conden-sate extraction represented by elements 26-31 in Figure 1. Elements 21 and 23-26 in Figure 2 are identical to the corresponding elements in Figure 1.
Thus the aqueous condensate (containing both methylbenzofuran and phenol) in separation vessel 23 is conveyed in stream 25 to mixing vessel 26 where it is mixed with an organic solvent (for removal of these r organics) introduced now through stream 127. From mixing vessel 26, the mixture passes in stream 128 to separation vessel 129 where it is separated into an aqueous recycle phase (illustrated as the more dense phase) and an organic solvent phase (illustrated as the less dense phase). The aqueous recycle phase is returned to the azeotropic still 18 via streams 31 and 32 wi~h water added or removed through stream 33 as in Figure 1.
The organic solvent phase in separation vessel 129 is conveyed through stream 130 to mixing vessel 141 where it is mixed with strong aqueous base such as sodium hydroxide introduced through stream 1~2. The 37~
mixture is then removed through stream 143 to se~aration vessel 144 where it is separated into a solvent phase (illustrated as less dense than water) and an aqueous phase ~removed in stream 145) that will contain most of the phenol in the organic solvent phase of stream 130, now as a water soluble phenate such as sodium phenate~
This phenol can be sprung by adding concentrated sul-furic acid, separating phases and returning the water-saturated phenol to stream 17.
The organic layer in separation vessel 144 con-tains impurities which must be removed from the solvent.
The organic phase is conveyed through stream i46 to distillation column 153. Pure organic solvent is re-moved overhead and returned via stream 127 to mixing vessel 26. Some make-up solvent can be added if needed.
Organic impurities are removed from the process through stream 155.
The embodiment of Figure 2 utilizes a solvent such as cumene or toluene to extract phenol from the aqueous condensate layer before returning it to the azeo still. This phenol is removed from the solvent by conversion to a water-soluble phenate. Pure solven~
is recovered for recycle by distillation.
In both Figures 1 and 2, various impurities such as methylbenzofuran, acetone, mesit~1 oxide and cumene are removed from the system in stream 24 and also by extrac-tion of the aqueous condensate.
Referring now to Figure 3 a further rnodified system is shown in which the condensed overheads of stream 21 from the azeotropic still 18 are charged directly to mixing vessel 1~6. Alternatively, phase separation in vessel 23 shown in Figs. 1 and 2 could be performed and the aqueous layer in stream 25 would then be charged to mixing vessel 146.
This condensate (or aqueous layer) is not mixed directly with the organic solvent, which in this case is high purity phenol from the product stream 36, but is rather mixed in mixing vessel 1~6 with the organic phase .

~17X37~

in stream 147 from a separation vessel 148. Mixture from mixing vessel 146 is char~ed to another separation vessel 149 through stream ].46a. The organic layer in vessel 149 is withdrawn through stream 149a and (like the organic condensate layer in stream 24 if vessel 23 is used) withdrawn.
The aqueous layer from separation vessel 149 is conveyed in stream 150 to mixing vessel 151. A
prtion oE product high purity phenol from product still 35 in stream 36 is diverted in stream 152 to mixing vessel 151. The mixture from mixing vessel 151 is con-veyed in stream 153 to separation vessel 148, where it is separated into an organic phase (conveyed to mixing vessel 146 in stream 147) and an aqueous recycle phase.
The aqueous recycle phase is conveyed in streams 31 and 32 back to the first or azeotropic still 18, with water added or removed if needed through stream 33, as in Figures 1 and 2.
The system of Figure 3 utilizes high purity phenol to extract impurities in two stages from the aqueous condensate of stream 25. Depending upon resi-dence times, impurity levels and other factors, more or less than two stages may be required. The efficiency of the system is based upon the preferential solubility of certain hard to remove impurities in the high purity phenol (such that they migrate to the organic phases in separation vessels 148 and 149) compared to their lesser solubility in the aqueous layer in separation vessels 148 and 149 ~and 23 if used). The advantages of the high purity phenol extraction, particularly of the entire condensed overheads, are that most of the phenol is not lost with a solvent such as curnene and that the relatively slow phase separation in vessel 23 can be avoided.
EXAMPLES
Feed phenols - Aliquots of the phenol batches used as feed materials in the chemical treatment step of the following Examples were analyzed by gas chromato--, ~', . ' ', 37~

graphy for impurities. Some 10-15 peaks were identi-fied, with nine assigned definite structurGs. The first 5 samples are considered technical grade phenol and generally have 1000-5000 ppm total impurities. Analyses produced the results (in ppm) shown in Table 1.
Table 1 - Refined Phenol Starting Materials Impurity Acetone MO C~mene Acetol A~ M~F AP DP Other Phenol Sample 10(total impurities) (1844) 15(2622) (3326) t5066) (5765) = mesityl oxide ~ Acetol = hydroxyacetone AMS = Alpha methyl styrene MBF = methylbenzofuran AP = Acetophenone or methylphenyl ketone DP = dimethylphenylcarbinol Other materials used were a plant process stream known as "pink phenol", a plant process stream known as "sprung phenol", a 70/30 blend of the two by volume and the product of dehydration of the 70/30 blend by removal of 15.5 weight ~ by evaporation at 160C and one atmosphere pressure. Analyses of these other mate-rials (in ppm except for water) are shown in Table 2.

-~ ~7137~

Table 2 - Other Startin~ ~aterials Acetol Impurity ~lO & AMS ~P MBF Other ~later Phenol Sample (total impurities except water) Pink Phenol 30 1279 2,283 ND1,108 2.15%
(4700) Sprun~ Phenol 72 1850 10,600 189 13,289 16.4%
(26,000) 70/30 84 212610,400 12413,96612.13%
(26,700) - 15 Dehydrated 44 216810,900 13n2,058 NM
70/30 (15,300) ND = not detectable NM = not measured Example 1 In a 22 L round bottom flask, phenol sample B (having 2622 ppm total impurities of which 2194 ppm were carbonyls and 51 ppm were MBF) was treated by heatin~ to 160C and adding 0.2194~ hexamethylenediamine (HMDA) by weight of phenol (0.1% per 10~0 ppm of total carbonyl impurities). Treatment was con-tinued for 4 h at 160C under a nitrogen atmosphere with agitation.
Phthalic anhydride (PAA) was then added in small incre-ments until the pH was down to 5.5. Phenol aliquots dissolved in water (5% solutions) were used to measure pH. 0.185% PAA hy original weight of phenol was added.
Gas chromatographic analysis of the treated phenol before and after acid addition showed 455 ppm total impurities (36 ppm MBF) before acidification and 470 ppm total impurities (46 ppm MBF) after acidification.
lhe treated phenol as described above, and phenol frorn similarly treated 3-5 L batches, were a~eo-tropically distilled in a 1 inch t2.5 cm) by 51 tray Oldershaw column. The column was assembled from 1, 5, ~A

~17~377`

10 and 20 tray sections with the treated phenol and water fed at the 50th tray. Azeotrope overheads were removed by a total condenser above the 51st tray and still bottoms removed from the bottom of the reboiler.
The reboiler had an electrical heating element controlled by a thermosensor to sense and maintain a selected temperature in the range of 180C to 186C at a locationr in most runs, above the fifteenth tray. The conditions for these runs are displayed in Table 3.
10 Table 3 - Azeot i Distillations With Fresh Water Azeo -~ Bottoms Phenol MBF Water r Phenol Overheads (Total ' Conc. After Feed Ratio (% phenol Impurities/
Azeo Chem. (total removed MBF in Run Treatment g~g) overhead) ppm) A 1.07;1 874.3 g 660.2 g (51 ppm)* (800,:750) (12.0-12.6%) (110/12) B #4 0.95.1 1415 g 1208 g (46 ppm) (1278 1345) (10.2%) (116/21) C #4 1.09 1 1435 g 1054 g (46 ppm) (1300:1197) (12.0- (110/21) 11.3%) D 1.01:1 1681 g 1303 g (37 ppm)* (1510 1493) (12.7- (58/3.0) 11.5%) *HMDA treated phenol only. No acidification with PAA.
This material had a poor shelf life, developing a pink color rapidly.
- The % phenol in the overheads was calculated from analyses of the bottoms and of the overheads.
Where these values differ, they are both shown in Table 3.
The bottoms from these four azeo stripper runs were separately batch distilled in a one inch (2.5 cm) diameter by 20 plate Oldershaw column. A controlled vacuum source set at 100 mm of mercury (13.3 kPa) and a water cooled condenser (at 40C) ahove the reflux split-ter were used. A lites or pasteurizing cut was taken prior to collecting the high purity phenol fraction.
These product distillations are summarized in Table 4, Runs A-D.

.; '`

~713~7 Table 4 - Product Distillations Azeo Bottom~s ~Main Main Cut Impurities %Lites as High Impurities Product Total/MBF Pasteur- Purity (in ppm) Run (in ppm) izing Cut Phenol Total MeF
A 110/12 2.94 92.7 26 10 B 116/21 S.2 91.0 20 10 C 110/21 3.5 94.9017 10 D 58/3.0 2.5 94.1 17 2 F 80/16 5.~* 93 1~ 4 G 74/14 2.5* 97 8 5 *In runs E, F & G, a continuous distillation column was used which included a second reflu~ splitter separated from the first by a 10 inch (25 cm) Vigereux section.
Pasturizing cuts were removed from the reflux splitter installed above the Vigereux section in runs F & G.
Main (product) cuts were removed using the reflux splitter installed below the Vigereux section.
Example 2 - Production of ~i~h Purity Phenol ~sing Extracted Recycle Water The azeotropic and product distillations of Example 1 were repeated using, for the azeotropic dis-tillation, cumene extracted aqueous condensates from previous runs. Thus overheads from a run were collected and allowed to phase separate. The aqueous layer was extracted three times using a cumene azeo water weight ratio of 0.5:1 (or volume ratio of 0.43:1). After phase separation, an aliquot of the aqueous layer from the third separation was analyzed for phenol and impurities. All of the MsF was selectively removed by the cumene extractions. Approximately 80% of khe dissolved phenol and other organics were also removed.
A larger sample of this aqueous layer was then used as feed water for the following distillations. The azeo-tropic distillations are shown in Table 5.

` ~713~7 Table 5 - ~zeotropic Distillations With Cumene-extracted Recycle Water Phenol-MBF
Concentra- ~ater: Overheads Bottoms tion After Phenol t%phenol ~total Azeo Chemical(Totalremoved imp/MBF
Run Treatmentg:g)overhead) in ppm) 1:1 1042 g 899 g E 227 ppm(973: (7-1%) ~214/74) 96~) 1.01:1 824 g ~64 g F 88 ppm (10.3-10.0) ~172/31) 1.19:1 1710 g 1198 g G 45 ppm(1603:(11.0-8.0) (76/14) 1346) 1.18:1 1347 g 927 g 15 H 45 ppm(1239:(11.7-10.3) (101/17) 1050) 1.05:1 1365 g 1055 g J 45 ppm(1239:(11.9-9.3) (79/16) 1050 ) 1.03:1 1174 g 940 g 20 J 45 ppm(1070: (9.96) (78/10) 1044) 1:1 1811 g 1458 g R 45 ppm(1639:(11.5-10.5) (78/15) 1647) :.
1.14:1 2261 g 1584 g 25 L 45 ppm(2045:(11.83-11.96) (81/18) . 1797) : The azeo bottoms fractions from the above runs were subjected to batch distillation as in Example 1.
The azeo bottoms sample from Run G was used for two separate product distillations, Runs I and H (with and without pasteurizing cut). A pasteurizing or lites cut was also not obtained for azeo bottoms from aæeo dis-tillation Run J in batch product distillation Run L.
The results are shown in Table 6.

~7~3~

Table 6 - Product Distillations Azeo Bottoms Impurities % Main Cut ImPurities Product (in ppm) Distilled (in ppm) 5Run Total/MBF as Lites ~Main Total MBF
H 76/14 0 96.1 12 5 I 76/14 7.6 91.1 14 7 J 101/17 3.9 93.0 10 4 K 79/16 3.0 90,0 20 8 L 7~/10 0 97.0 21 9 In the next azeotropic distillation (azeo Run r~ M) phenol with 104 ppm MBF impurity was used; a (high purity phenol extracted) water: phenol feed ratio of 0.99: (1344 grams to 1356 grams) was em~loyed; 1478 g of material was removed overhead (containing 9.0% of the phenol feed); and 1195 g of material was removed as bottoms (containing 195 ppm total impurities including 3 ppm M~F). In the subsequent batch product distillation (Product Run M) 4.3% of the material distilled was removed as a pasteurizing cut and 89.7% as the main or product cut. The product cut contained 10 pprn total impurities (5 ppm MBF) compared to 1~5 ppm total impurities (3 ppm MBF) in the azeo bottoms fed to the product distillation column.
Example 3- Production of High Purity Phenol Using Low Grade Starting Materlals .~ A 1513 g sample of the dehydrated 70/30 mixture described above in ~able 2 was chemically trea-ted at 160C with ~MDA for ten hours. HMDA was added slowly over ten hours, with 2.5 g HMDA having been added by two hours (pH 5.45), 4.95 g EIMDA having been added by four hours (p~ 7.1) and 12.3 g HMDA having been added after ten hours (pH 7.5). Each pH measurement was obtained by making a 5% aqueous solution of phenol from a sample from the chemical treatment. PAA was then added to adjust the pH down to 5. Aliquots were taken before any HMDA addition and at 2 h, 4 h and 10 h and after PAA addition. The results of analyses (in ppm) of ~171377.

these aliquots are displayed in Table 7.
Table 7 - Chemical Treatment of Low Grade Phenol Total ~cetol Ali~uot Impurities MO ~ AMSAP MBF
5 Initial 15,300 44 216810,900 130 2 h 15,300 19 010,900 133

4 h 13,200 ND 010,700 105 10 h 12,800 ND 010,700 69 After PAA 12,700 10 0 10,700 77 10 A larger batch was then chemically treated in like manner, except that after HMDA addition to a pH of of 7.1, the treatment was limited to 4 h at 160C. PAA
was then added to adjust the pH to 5. Four samples of the resultant treated material were then subjected to azeotropic distillation as in Example 1 using fresh water feed. The results are displayed in Table 8.
Table 8 Phenol MBF
After Overheads Azeo Chemical Water:Phenol (% Phenol Bottoms 20 Azeo Treatment (total Removed As (Total Imp Run (ppm) g:g) Overheads in ppm)*
N 64 1.14:1 467 g 411.2 g : (450:415) ( 11) (9870) P 64 1.03:1 873 g 6~8 g (760:739) ( 14-15) (9756) Q 64 2.13:1 648.2 g 227.2 g (580:296) (23.1) (9579) :~ R 64 1.91:1 1285 g 460.0 g (1146:600) (23.3) (8541) : 30 *M~F concentration of azeo bottoms could not be established due to interference of unknown impurities.
Total impurities represent an average of several samples.
The 648 g bottoms of azeo Run P and the 227.2 + 460.0 g combined bottoms of azeo Runs Q and R were batch distilled as in Example 1. The results are s'nown in Table 9.

~ ~137Z

Table 9 - Product nistillations % Lites Feed as Pas-Product Impurities teurizing Main Cut Impurities Runs Total Cut %Main Total MBF
5N 9756 4.0 86.5 16 2.4 P 8884* 3.0 87.0 18 0 * calculated as (227-2 x~927729)+ 46040- x 8591) = 8884 The product main cut from Product run N contained 16 ppm total impurities (2.4 ppM MBF). The pasteurizing cut from Product Run ~ contained 1010 ppm total impurities including 27 ppm of a contaminant having the same elution time as MBF. Previous studies have indicated that MBF cannot be separated from phenol by a simple vacuum distillation. The peak eluting as MBF in the pasteurizing cut is probably some other impurity. The results of these experiments indicates that azeotropic distillation using a 1:1 feed ratio of chemically treat-ed low grade phenol(s) to Eresh water is adequate to obtain a high purity phenol product.
Example 4 Using the dehydrated 70/30 mixture and the chemical treatment of Example 3, four azeotropic dis-tillations (with fresh water) were conducted as described in Table 10. The chemically treated phenol feed had 14,570 ppm total impurities including 64 ppm M~F. The impurity levels of the azeo bottoms has not yet been determined.
Table 10 Overheads Water:Phenol (% phenol 30Azeo ~eed Ratio feed removed Azeo Run (total g_g) overhead Bottoms S 0.95:1 922.8 g 773.5 g (833:~80) (12.1-9.7) T 0.97:1 945.0 g 760.0 g (840:~62.5) (11.9-11.1) U 0.96:1 932.0 g 7~3.0 g (831.4:870) (14.6-10.8) V 0.95:1 ~68.0 g 794.n g (880:924) (14.1-9.1) . ., .

~137~

The azeotropic still bottoms were combined into two samples and each of the two ~as batch distilled by the process used in Examples l and 4. These dis-tillations are summarized in Table ll. Product Run Q
used azeo bottoms from ~zeo P~uns S and T. Product Run R used azeo bottoms from Azeo Runs U and V.
Table ll Main Cut Feed ~ Lites% Main Impurities Product Impurities as Pasteur- HP in ppm lQ Run Total _zing CutPhenol Total MBF
Q N.A. 4.2 87.1 22.8 1.0 R N.A. 3.0 87.8 20.4 1.6 The overheads from azeotropic distillations S, T, U and V shown in Table lO were phase se~arated and the aqueous layers composited. Approximately 3500 g of azeo water was collected. Analysis of the phenol saturated aqueous layer indicated llO0 ppm of dissolved impurities. 400.2 g of this aqueous condensate were ex-tracted three times at 40C with 20.6 g each of high purity phenol containing 16 ppm total impurities ~9 ppm acetone, 7 ppm othersj. A 16.0 g phenol layer and 404.~
; g aqueous layer were recovered from the first extraction (with a 2.0 g sample of the aqueous layer removed before ~ the second extraction). A 30.5 g phenol layer and a `~ 25 392.3 g aqueous layer were recovered from the second extraction (with a 2.0 g sample removed before tne third extraction). A 33.7 g pheDol layer and a 379.4 g aqueous layer were recovered from the third extraction (again a 2.0 g sample~was removedj. Analyses of the three recovered phenol layers and three aqueous ~layers (impur-ity levels in ppm) are displayed ln Table I2.
, : ~ :~ ::

, ~7~13~

Table 12 After Total Extraction Phase Impurities Acetone MBF AP Others 1 Phenol 7564 4433 49.6 2576 505 2 Phenol 3873 3200 12 539 122 3 Phenol 2342 2195 ND 112 35 1 Aqueous* 713 656 1 45 11 2 Aqueous* 485 473 0.3 10 2 3 Aqueous* 300 297 0.4 2 *also saturated with phenol A second 400.13 g of the aqueous condensatè
was extracted, first with 28.2 g of the phenol layer from the second stage of the extraction shown in Table 12, then with 29.22 g of the phenol layer from the third stage of the extraction shown in Table 12 and finally with 20.60 g of fresh high purity phenol. Three 2.2 g samples of the aqueous layers were removed for analysis after the first, second and third extractions.
The size and impurity levels (in ppm) on analyses of layers removed frorn each extraction are shown in Table 13.
Table 13 Extrac- Size Tbtal tion Layer (g) Impurities Acetone AP MBF Others 1 Phenol 13.912,612 71204610 89.2 793 ~j 25 2 Phenol 23.9 5,583 4471945 14.1 153 3 Phenol 30~3 3,303 3079 183 ND 41 1 Aqueous* 414.4 897 829 56 ND 12 2 Aqueous* 417.6 695 680 10 ND 5 3 Aqueous* 405.6 495 490 2 ND 3 *also saturated with phenol This second series of extraction simulates counter-current extraction in three stages. It is apparent from these runs that extraction with small quantities of high purity phenol is an effective method for reducing the 35 impurity content of the aqueous condensate. According-ly, return of the condensate to the azeotropic distilla-tion column should be about as effective as fresh water feed, as shown in Example 3.
' , 3~7 Example 5 - Chlorination Colors Chlorination colors were taken of the high purity phenol products from Product Runs D-L, J~ and P, above. Chlorination colors were measured by the tests described in U.S. Patent 2,992,169 (July 11, 1961).
Runs E, F and G were continuous product distillation runs; all others were batch.
The absorption spectrum showed a peak at 490 nm (millimicrons) which was higher than values conventionally taken at 510 or 540 nm . The optical densities ranged from 0.137 to 0.341 for these runs.
Optical densities compared to total impurity and MBF
concentration for these runs are shown in Table 14.
Table 14 Impurity Content:
Product Total vs.O.D. at 0.D. at O.D. at Run MBF (ppm)490 nm 510 nm 540 nm D* 17/2 0.171 0.141 0.107 E* 31/10 0.341 0.263 0.165 F* 14/4 0.341 0.294 0.212 G* 8/5 0.202 0.166 0.108 - H** 12/5 0.178 0.151 0.121 I** 14/7 0.147 0.120 0.086 J** 10/4 0.137 0.109 0.081 K** 20/8 0.149 0.127 0.105 L** 21/9 0,222 0.181 0.129 N*** 16/2 0.243 0.168 0.079 P*** 18/0 0.19A 0.138 0.068 * Refined phenol as starting material followed by azeo distillation with fresh water.
** Refine~ phenol as starting material followed by azeo distillation using recycled cumene extracted azeo water.
*** Low grade phenol as starting material followed by azeo distillation with fresh water.

Claims (9)

What is Claimed is:

1. In the process of purifying phenol pro-duced by the cleavage of cumene hydroperoxide wherein the feed phenol is treated with a base, optionally an acid anhydride is added, the treated phenol is steam distilled to remove a lites fraction as a water azeotrope and the bottoms of the steam distillation are distilled to recover high purity phenol as an overhead;
the improvement wherein the overheads of the steam distillation are condensed and organics are extracted with an organic solvent which preferentially dissolves benzofuran impurities to produce a reflux aqueous phase and the reflux aqueous phase is returned to the steam distillation.

2. The process of claim 1 wherein the con-densed overheads are phase separated into an aqueous layer and an organic layer and only the aqueous layer is extracted with the organic solvent.

3. The process of claim 1 wherein the entire condensed overheads are extracted with the organic solvent to produce the reflux aqueous phase.

4. The process of claims 1 or 2 or 3 wherein the extraction is multiple stage.

5. The process of claim 1 or 2 or 3 wherein the organic solvent is cumene, toluene or a xylene.

6. The process of claim 5 wherein the organic solvent after extraction is washed with an aqueous base to remove phenol and impurities and the organic solvent recovered is used again to extract organics.

7. The process of claim 1 or 3 wherein said organic solvent is high purity phenol.

8. The process of claim 1 wherein said treated phenol has between about 30 and about 150 ppm methylbenzofuran by weight as an impurity and wherein sufficient steam is distilled with said treated phenol to reduce methylbenzofuran to below about 25 ppm in said high purity phenol.

9. The process of claim 1 or 8 wherein said feed phenol is treated with a polyamine as base and then sufficient phthalic anhydride per mol of polyamine is added prior to steam distillation to lower the pH of said chemically treated phenol to between about 4.0 and about 5.5.