Organic synthesis design and the computer - American Chemical

we will introduce program ORGSYN which was developed at. Franklin and Marshall College to incorporate the reactions of our introductory organic course...
0 downloads 0 Views 5MB Size
Organic Synthesis Design and the Computer Bonnie Burns Sandel' and Robert W. Solomon Franklin and MarshallCollege Laneaster. PA 17604

The abilitv. to svnthesize new comnounds is crucial to the . research of most organic chemists, and considerable effort is expended in teachina svnthesis skills in most introductorv and advanced organic c&es. It is not surprising, then, t h s t the twentieth ~entury'smost generally accessihle new research tool-the computer-has heen brought to hear on the problems of organic synthesis design and its teaching. Research in the area of computer-assisted synthesis has made important contributions to the fields of generalized problem-solving and artificial intelligence. In this paper, however, we shall emphasize the ways which this research impinges on the way svntheses are desiened and the wav svnthesis is taueht. Then we will intnd~rcepwyrnm 0 ~ ~ ~ i ' ~ ' w wasdeveloped hich at Franklin and Mnrshnll Collree to incorporate the reactions of our introductory organic courses. Finally, we will present our experience with the design of a computer program as a heuristic device in an introductory organic course. Computer-Asslsted Synlhesls in Industry and Academe A symposium on computer-assisted synthesis held in 1976 brought together representatives of all its major researchers from Canada, Germany, Great Britain and the United States. Perusal of the edited transcripts from this symposium ( I ) gives some indication of the impact of the computer on industrial and academic synthesis research. Computer assistance synthetic chemists in industry can be grouped into two hasic cateaories (2).retriever and svnthesi~e;~rograms.The retriever programs are an attempi to organize reactions from the chemical literature in a form useful to synthetic chemists. The major task for the designers of these programs is to devise an appropriate encodingscheme to enable chemists to retrieve reactions as thev are needed. Schemes have been described (2,3) involving classifications bv bonds and rines formed. bv mechanistic t .. w e ., hv reaction conditions, by s t k i n g matekal and by product. These retriever proerams serve as a desk-top reference to the literature. They mayalso serve a current awareness function, helping to ensure that potentially useful reactions are not overlooked. A second type of retriever system had been developed by information scientists a t ICI Pharmaceuticals (3). This system provides information about specific chemicals in order that samples of the chemicals may be obtained expeditiously. This allows rapid acquisition of proposed syntheiic precurkors. Industrial chemists also make use of the sophisticated programs for synthesis design that have been developed in academic settings (videinfra). These "synthesizer" programs may serve the industrial svnthetic chemist in several wavs . (.4.) . ~ y p r o v i d i n ga large number of potential routes to the target compound, the proflams increase the possibilitv of uncoverine a useful path-for example, a common route toieveral desire2 products. Chemists a t DuPont (4) report a benefit in trying to adapt an academic program to industrial synthetic chemistry. T h e adaptation forces the chemists to organize their thoughts about the known and unknown generality of their reactions. In academic settings. has - . work on "svnthesizer" oroerams " led to programs capable of deriving syntheses for complex organic compounds (5).Despite their considerable abilitv to generate synthetic routes, ihe major benefits to synthetic chemists from these programs has not been documented, succesqful lahoratnry syntheses of natural products but rather,

-

.~

'

Present address: Olin Research Center. P.O.Box 30-275. New Haven. CT 06511; author to whom correspondence should be addressed. 798

Journal of Chemical Education

increased definition of the rules that guide synthesis design. For example, work on LHASA (for Logic and Heuristics Applied to Synthetic Analysis) enabled Corey ( 6 )to formulate a list of six characteristics of the bonds in tricyclic or higher polycyclic systems that are "strategic"in the sense that disconnecting them leads to "especially simple or accessihle ring systems of lower cyclic order." These characteristics can easily he utilized by a chemist faced with the task of synthesisdesign. A further example of the henefits of these programs is vided hy Warren's programmed text (7). "Designing Organic Syntheses," which may be used in an advanced organic coune. His texthook derives it.9 organization, symbolism and language from the svnthon aonroach-a reorranization of the reactions of organicchemistryhased on the structure of their products. This reoreanization of reactions was a necessarv precondition to the design of computer programs for organics;nthesis, and its principles and practice are delineated in the early papers of Corey and coworkers (8)on LHASA. Warren's text is a clearly written set of problems and answers that teaches an analytical approach to synthesis design for moderately complicated molecules. In a substantial chapter on strategy he resents the henefits of convereent svnthesis and the ~ r i n c i .. ples for identifying useful disconnections in the target comoound. The svnthon anoruach. with its roots in computerassisted synthesis, maiwell p n k e to he the most fruitiu~analvtical tuol available for oreanic synthesis design. On the mechanistic front, the rapid calculation potential of the computer has hcen utilized to prnvide insight into conformational analysis and dterir fnctor5 influencing nucleophilic additions. As part of SKCS (for Simulat ion and Evaluation of Chemical Svnthrqi