Method of Making Trichloromethoxybenzene

This invention relates to a method of making trichloromethoxybenzene by reacting anisole with a source of chlorine free radicals in the presence of actinic radiation. In particular, it relates to the use of solvents based on benzotrifluoride in that reaction.

Trichloromethoxybenzene has been made by reacting anisole with gaseous chlorine in the presence of ultraviolet light. This reaction was performed in the presence of carbon tetrachloride. However, carbon tetrachloride has been found to be toxic and an ozone depleter and its use is now heavily taxed and regulated. Efforts are being made to find a substitute solvent.

We have discovered that anisole can be photochlorinated without significant chlorination of the aromatic ring if the solvent used is based on benzotrifluoride (BTF), .alpha.,.alpha.,.alpha.-trifluoromethylbenzene. While the use of no solvent or other common solvents results in significant chlorination of the aromatic ring, when the reaction is performed using a solvent according to this invention, less than about 5 mole % of the aromatic ring is are chlorinated.

We have further discovered that chlorination of the anisole aromatic ring can be still further reduced by adding the anisole and the chlorine separately to the solvent. We have no explanation for this unexpected improvement.

In the method of this invention, anisole is reacted with chlorine free radicals in a BTF solvent to produce trichloromethoxybenzene (TCMB). If chlorine is used, the reaction is: ##STR1## Anisole is a liquid which can be mixed with about 10 to about 70 wt % of the BTF solvent in order to control the rate of reaction. At least about 10 wt % anisole (based on total solvent plus anisole weight) should be used for an economical process, and if the weight % of anisole is greater than about 60, ring chlorination may begin to occur. Preferably, the concentration of anisole is about 30 to about 50 wt %.

The source of chlorine free radicals can be, for example, elemental gaseous or liquified chlorine or liquid sulfuryl chloride (SO.sub.2 Cl.sub.2). Gaseous chlorine is preferred as it results in fewer byproducts, it is inexpensive, and it works well. At least a stoichiometric amount of the source of chlorine free radicals is needed (i.e., 3 moles Cl.sub.2 per mole of anisole), but a slight (1 to 5 mole %) excess is preferred to insure a complete reaction and reduce ring chlorination.

BTF solvents that can be used in this invention include BTF, orthochlorobenzotrifluoride, metachlorobenzotrifluoride, parachlorobenzotrifluoride (PCBTF), and dichlorobenzotrifluoride. BTF and PCBTF are preferred as they have been found to work well. The use of these solvents is essential to reducing ring chlorination.

The anisole, solvent, and chlorine free radical source can be mixed together in any fashion such as, for example, adding the chlorine free radical source to a mixture of the anisole and the solvent, adding the anisole and the chlorine free radical source separately to the solvent, or mixing some of the solvent with the anisole first, then adding that mixture and the chlorine free radical source separately to a solvent. For some unknown reason, it has been found that if the anisole and the chlorine free radical source are kept apart until they are mixed with the solvent, ring chlorination is reduced.

It has also been found that ring chlorination increases at lower temperatures and therefore it is preferable to perform the reaction at as high a temperature as is practical. Generally, therefore, it is preferable to reflux the solvent during the reaction; this reduces ring chlorination by removing hydrogen chloride. For benzotrifluoride, the reaction would be performed at about 105.degree. C. and for parachlorobenzotrifluoride, the reaction would be performed at about 136.degree. C.

Any actinic radiation, i.e., radiation that will produce chlorine free radicals, such as by the reaction ##STR2## can be used. Examples of actinic radiation include, for example, ultraviolet light, radio frequency, x-rays, or free radical initiators. Ultraviolet light is preferred as it is convenient and easy to use. An ultraviolet wavelength of about 320 to about 340 nm is suitable. Since the light does not penetrate deeply into the mixture, the light source should be placed as close to the mixture as possible. This can be accomplished, for example, by placing the light in a well which is inside the reactor. The mixture should be stirred to expose all portions of the mixture to the light to ensure continuous generation of chlorine free radicals.

If the reaction mixture is in contact with a metal, it may be desirable to add about 5 to about 500 ppm (based on mixture weight) of a metal scavenger to the mixture to prevent the metal ions from catalyzing the production of byproducts. Examples of suitable metal scavengers include N,N-dialkylalkyl amides (sold as "Hallcomid" by the C. P. Hall company) and ethylenediaminetetraacetic acid (EDTA).

The reaction can be performed as a batch, continuous, or semi-continuous process, but a continuous process is preferred as it is more efficient and is more likely to result in less ring chlorination. It is also possible to partially chlorinate the anisole to a mixture of mono-, di- and trichloromethoxybenzenes in a continuous process, and then transfer the mixture to a batch reactor to finish off the chlorination. In a preferred continuous process, the anisole and the chlorinating agent are added separately to a stream moving past a source of ultraviolet light. The rate of addition to the stream in a continuous process should be selected to optimize the reaction.

The trichloromethoxybenzene product is useful as a chemical intermediate. For example, it can be reacted with hydrogen fluoride to make trifluoromethoxybenzene, which can be used to make herbicides and pecticides. The following examples further illustrate this invention. Unless otherwise specified, the reactor used was a 350 mL photochlorination apparatus equipped with a 100 W medium pressure Hanovia UV light (air cooled), a reflux condenser, thermometer and an inlet for chlorine.

Into the reactor was placed 71 g of anisole and 400 mL of 1,2-dichlorotrifluroethane, sold by Dupont as "Freon 113" (the solvent), along with 500 ppm of a mixture of 5 wt % N,N-dimethyl caproamide, 50 wt % N,N-caprylamide, 40 wt % N,N-capramide, and 5 wt % N,N-dimethyl lauramide, sold by The C. P. Hall Company as "Hallcomid M 8-10" (a metal sequesterer). The reactor was heated to reflux (50.degree. C.), the light was turned on, and chlorine was sparged in. Gas chromatography (GC) analysis indicated the formation of 36.1 wt % undesired by-products, consisting predominantly of ring chlorinated materials. Side chain chlorinated materials were formed in 63.9 wt % yield, as a mixture of mono-, di-, and trichloromethoxybenzenes.

Into the reactor was placed 53.1 g of anisole and 250 mL of glacial acetic acid, along with 500 ppm of "Hallcomide M8-10." The reactor was heated to reflux (122.degree. C.), the light was turned on, and chlorine was sparged in. GC analysis indicated the formation of 36.1 wt % undesired by-products, consisting predominantly of ring chlorinated materials. No desired product was formed.

Into the reactor was placed 292 g of anisole, along with 500 ppm of "Hallcomide M8-10." The reactor was heated to reflux (133.degree. C.), the light was turned on, and chlorine was sparged in. GC analysis indicated the formation of 44.9 wt % ring chlorinated material, with 24.6% side chain chlorinated materials as a mixture of mono- and dichloromethoxybenzenes along with 30.5% unreacted anisole.

Into the reactor was placed 284 g of anisole, along with 500 ppm of "Hallcomide M8-10" and 1.0 g of phosphorus trichloride. The reactor was heated to reflux (136.degree. C.), the light was turned on, and chlorine was sparged in. GC analysis indicated the formation of 27.3 wt % ring chlorinated material, a 26.5 wt % yield of side chain chlorinated materials, as a mixture of mono- and dichloromethoxybenzenes, and 46.2 wt % unreacted anisole.

Into the reactor was placed 53 g of anisole and 250 mL of benzotrichloride (.alpha.,.alpha.,.alpha.-trichloromethylbenzene), along with 500 ppm of "Hallcomide M8-10." The reactor was heated to reflux (136.degree. C.), the light was turned on, and chlorine was sparged in. The solution became dark upon photolysis, and the reaction was stopped immediately.

Comparative Examples 1 to 5 show that certain solvents are unacceptable in the photochlorination of anisole because they lead to high levels of ring chlorinated product along with other byproducts. Photochlorination reactions run in the absence of solvent also lead to unacceptable levels of ring chlorination.

Reactions in which the anisole and solvent are added together at the beginning of the reaction.

Into the reactor was placed 53 g of anisole and 250 mL of BTF along with 500 ppm of "Hallcomide M8-10." The reactor was heated to reflux (108.degree. C.), the light was turned on, and chlorine was sparged in at 120 mL/min for 5 hours. GC analysis indicated the complete consumption of anisole and the formation of trichloromethoxybenzene in an 83.9 wt % yield. There was 15.7 wt % ring chlorinated products, and 0.4 wt % of dichloromethoxybenzene remaining.

Into the reactor was placed 53 g of anisole and 250 mL of parachlorobenzotrifluoride (4-chloro-.alpha.,.alpha.,.alpha.-trifloromethylbenzene) along with 500 ppm of "Hallcomide M8-10." The reactor was heated to reflux (135.degree. C.), the light was turned on, and chlorine was sparged in at 120 mL/min for 5 hours. GC analysis indicated the complete consumption of anisole and the formation of trichloromethoxybenzene in an 92.8 wt % yield. There was 7.2 wt % ring chlorinated products present.

Examples 1 and 2 show that anisole pre-mixed in BTF solvents prior to chlorination under free radical conditions leads to high yields of the desired trichloromethoxybenzene product.

In Examples 3 to 12, anisole and the solvent (in all cases PCBTF) were added together at the beginning of the reaction. The chlorine addition time is the time in which the theoretical amount of chlorine (3 moles per mole of anisole) had been added.

Examples 3 to 12 show that higher temperatures (Examples 3, 5, 6, and 12) lead to higher levels of side chain chlorinated products than reactions run at lower temperatures.

Reactions in which anisole is added concomitantly with chlorine to the solvent.

Into the reactor was placed 124 mL of PCBTF along with 600 ppm of "Hallcomide M8-10." The reactor was heated to reflux (135.degree. C.), the light was turned on, and chlorine and anisole were added together. The chlorine was added at a rate of 42 g/hr. The anisole (125 g, 50% by weight of PCBTF) was also added at a rate of 42 g/hr. Once all of the anisole had been added, additional chlorine was sparged in until the reaction was complete. An assay by GC showed 94.2 wt % desired trichloromethoxybenzene, with 5.7 wt % ring chlorinated material. No anisole was present.

In a 3 L photochlorination apparatus equipped with a 100 W medium pressure Hanovia UV light (air cooled, Pyrex filter inside the light well), a reflux condenser, a funnel for the addition of anisole, a thermometer and an inlet for chlorine was placed 1500 g of BTF. The reactor was heated to reflux (105.degree. C.), the light was turned on, and chlorine and anisole were added together. The chlorine was added at a rate of 1200 mL/min. The anisole (1500 g, 50% by weight of PCBTF) was added at a rate of 150-200 g/hr. When all of the anisole had been added, additional chlorine was added to complete the reaction. An assay by GC showed 87.8 wt % desired trichloromethoxybenzene, 1.9 wt % dichloromethoxybenzene, and 10.3 wt % ring chlorinated material. No anisole was present.

The following experiments summarize results in which anisole and chlorine were added at the same time to the reaction at the reflux temperature of the solvent (105.degree. C. for BTF, 135.degree. C. for PCBTF). In the following table, the percent solution is weight/weight percent anisole/solvent, the anisole add time is the time it took to completely add the anisole into the solvent, the chlorine rate is in standard cubic centimeters per minute, the light used is either a 100 or 400 watt medium pressure Hanovia lamp, the % side and ring products were analyzed by GC, and the reaction scale was the size of the photochemical reactor used.

Examples 13 to 44 show that co-addition of anisole and chlorine to a refluxing solution of BTF solvents under free radical generating conditions under different process variables such as chlorine flow, temperature, and concentration leads to good selectivities and yields of trichloromethoxybenzene.