Synthesis. In addition to DNA and proteins, synthetic polymers are also suitable for data storage, at least in principle. As early as 1986, ▇▇▇▇▇▇▇ ▇▇▇▇▇▇▇ suggested that, at least in theory, any polymer composed of at least two different monomers could be used to store data97. Although the controlled synthesis of polymers with more than two monomers is possible — and although such polymers would potentially provide a more economical solution for data storage — most dataencoding polymers employ only two different monomers (directly represent ing 0 and 1 in the binary code). The main advantages of synthetic polymers are the possibility of having full con trol over their synthesis and the greater flexibility, mean ing that one is no longer restricted to four monomers, as in the case of DNA. Instead, the monomers can be selected and tuned for the purpose of the application. In such synthetic dataencoding copolymers, it is essential to achieve perfect control over the monomer sequence. For example, DNA can be used as a template for the assembly of free nucleotides (including nonnatural ones) before chemical or enzymatic polymerization98–103. The drawbacks of this approach are the low efficiency and difficulty of removing the synthesized polymer from the template. Recently, molecular machines that mimic biological polymerization have been developed, such as the peptide synthesis machine designed by ▇▇▇▇▇ and coworkers104. The machine is a rotaxane system in which the macrocycle sequentially picks up amino acids from the thread to assemble a peptide of known sequence. The current rates and yields of these reactions make practical applications difficult — and thus far, these sys tems are limited to the synthesis of natural polymers, that is, polypeptides104. To work around biological polymeri zation techniques, ▇▇▇ and coworkers designed a DNA translation system to synthesize sequencecontrolled pol ymers not based on natural monomers105. The polymeri zation in this case depends on the hybridization of DNA base pairs to a template. Synthetic building blocks are attached to these DNA base pairs via a cleavable linker. In this system, the DNA base pairs perform a very similar function to tRNA, that is, they serve to bring together the desired building blocks in the correct order. After poly merization, cleavage of the linker results in the release of the synthetic polymer105. Complete chemical polymerization has the advan tage of a much wider range of available building blocks, but achieving perfect sequence control remains chal lenging because classical chaingrowth and stepgrowth polymerizations do not allow for precise control over the Rotaxane A mechanically interlocked molecular architecture in which a macrocycle is kinetically trapped on a thread by the presence of two large ‘stoppers’. was used to convert the protein into a shape repre senting a 1. For reading, a lowpower laser beam was used to detect the conformation of the protein without disturbing the conformation itself. The ability of bR to shift between different states also allows for rewritable data storage95,96. monomer position. Sequence control can be improved by using living chain polymerization methods, in which each polymer chain grows in a more uniform way. For instance, by using controlled radical polymerization techniques, that is, reversible addition−fragmentation chaintransfer (RAFT) polymerization, ▇▇▇▇▇▇▇▇ et al. REVIEWS Monodisperse polymers Polymers composed of uniform molecules with the same structure and mass. Naturally occurring polymers are frequently monodisperse, while synthetic polymers are usually not. constructed new sequencecontrolled macroRAFT agents by inserting two monomers in a sequential man ner106. The low yield of the monomer insertion, however, made this process suboptimal for the synthesis of long chains. Atomtransfer radical polymerization (ATRP) was used by ▇▇▇▇ et al. to construct vinyl polymers by an iterative process of single monomer addition107. However, this method also suffered from low yields (as a result of side reactions), making the synthesis of long chains impossible. In addition to controlled radical polymerization, sequence control in chain polymerization can also be obtained by using living ionic polymerization tech niques. Living cationic polymerization was used by ▇▇▇▇▇▇ et al. to create sequenceregulated polymers by the addition of monomers one by one in order of decreasing reactivity108. Nevertheless, defects occurred during the polymerization, necessitating purification after each monomer addition. Living anionic polymeri zation has also been used to obtain sequencecontrolled chains composed of two different monomers (the choice of one bulky monomer prevented homopoly merization). In this way, an alternating pattern of two different monomers could be obtained109. This type of kinetic control was also applied in radical polymeriza tion to obtain an alternating pattern, that is, the high affinity between two different monomers was used to create regions with a specific sequence in the polymer chain110. ▇▇▇▇ et al. built upon this alternating method by tuning the sequence through timecontrolled mono mer addition111,112. In this strategy, one monomer (an electronrich donor) present in excess is polymerized by a radical reaction, while a second monomer (an acceptor) is added in small amounts at specified times. A highly favourable donor–acceptor interaction between the two monomers results in the incorporation of acceptor monomers in small, welldefined regions of the polymer backbone111,112. Using automated protocols, ▇▇▇▇ et al. were able to construct welldefined polymer chains containing up to eight precisely positioned blocks with a specific sequence113. A similar approach was adopted by O’Reilly et al. in ringopening metathesis polymeri zation, in which the position of four different functional moieties could be relatively well controlled along a grow ing polymer chain114. However, in all these methods, var iations in the length and precise composition of each segment may occur, and some polymer chains might contain defects101,115. To minimize the number of defects, long building blocks in multiblock copolymerization may be used. ▇▇▇▇ et al. employed this strategy in combination with degenerative transfer radical polymerization to con struct welldefined multiblock copolymers116. In addi tion, Engelis et al. used long blocks for the synthesis of welldefined multiblock copolymers by emulsion polymerization, in which monomers and catalysts were separated in micelles to isolate the growing polymers from one another and reduce unwanted side reactions117. Despite the reported improvements in the synthetic procedures, it can be concluded that chain polymer ization always results in polymers with deviations in chain length and composition115. This limitation means that chaingrowth polymerization is, at present, not a good method to prepare welldefined polymer sequences. Nonetheless, it could be employed for easy copying of already synthesized sequences by template polymerization118. In addition to chaingrowth polymerization, step growth polymerization techniques can be used to syn thesize polymer sequences with periodic monomer patterns. Conventional stepgrowth polymerization has been used for the synthesis of polyamides and polyurethanes. Although these methods are relatively straightforward, they do not allow for perfect sequence control. New stepgrowth polymerization techniques using radical polymerization119,120 or click chemis try121, however, do allow for such a sequencecontrolled polymerization. The latter can also be achieved by applying multistepgrowth synthesis, which involves the stepwise chemical attachment of monomers attached to a support122. This procedure results in highly monodisperse polymers. One such method is iter ative solidphase synthesis, similar to the wellknown solidphase peptide synthesis methodology. It employs an insoluble support on which the polymers are grown by the stepwise addition of monomers123. The method is highly efficient but also very time consuming, and fur thermore, the efficiency of the coupling steps decreases with increasing polymer length, making it best suited to the synthesis of short polymers. Despite these disad vantages, solidsupported synthesis remains the most frequently applied and most reliable method for the syn thesis of sequencecontrolled polymers122. An alterna tive is the use of a soluble polymer chain as a support124, which improves the process efficiency, but the synthesis of long sequences is still not possible101. The ▇▇▇▇ group has investigated numerous strategies to exploit this multistepgrowth methodology for the production of sequencecontrolled polymers, including those encod ing data125. The previously described stepgrowth syn thesis of polyurethanes, for instance, could be improved by applying a multistepgrowth approach126. This strat egy relied on a sequence of two chemoselective steps: the reaction of an alcohol with an Nhydroxysuccinimide (NHS) moiety and then the reaction of an amine with NHS126. Data were encoded using two different amino alcohol monomers (serving as 0 and 1), while N,N′ disuccinimidyl carbonate, containing two NHS moie ties, was used as a linker126. Another method developed by ▇▇▇▇ and coworkers is based on phosphoramidite coupling, a method already widely used for oligonucleo tide synthesis127. The synthesis uses a solid support, and the monomers are coupled one by one in three steps (FIG. 4a). First, N,Ndimethyltryptamine (DMT) depro tection of the monomer occurs, allowing the connec tion of the next monomer by phosphoramidite coupling, followed by oxidation of the phosphite to a phosphate. Optimization of this method allows each threestep cycle to be completed within a few minutes128. ▇▇▇▇ et al. used this approach to synthesize a polymer with a controlled sequence from two monomers containing either a propyl moiety (representing 0) or a 2,2dimethylpropyl moiety (representing 1)129. In addition, another monomer con taining a 2,2dipropargylpropyl group (representing 2) O HO N3 3 NH2 R O O O
Appears in 1 contract
Sources: End User Agreement
Synthesis. In addition to DNA and proteins, synthetic polymers are also suitable for data storage, at least in principle. As early as 1986, ▇▇▇▇▇▇▇ ▇▇▇▇▇▇▇ suggested that, at least in theory, any polymer composed of at least two different monomers could be used to store data97. Although the controlled synthesis of polymers with more than two monomers is possible — and although such polymers would potentially provide a more economical solution for data storage — most dataencoding polymers employ only two different monomers (directly represent ing 0 and 1 in the binary code). The main advantages of synthetic polymers are the possibility of having full con trol over their synthesis and the greater flexibility, mean ing that one is no longer restricted to four monomers, as in the case of DNA. Instead, the monomers can be selected and tuned for the purpose of the application. In such synthetic dataencoding copolymers, it is essential to achieve perfect control over the monomer sequence. For example, DNA can be used as a template for the assembly of free nucleotides (including nonnatural ones) before chemical or enzymatic polymerization98–103. The drawbacks of this approach are the low efficiency and difficulty of removing the synthesized polymer from the template. Recently, molecular machines that mimic biological polymerization have been developed, such as the peptide synthesis machine designed by ▇▇▇▇▇ and coworkers104. The machine is a rotaxane system in which the macrocycle sequentially picks up amino acids from the thread to assemble a peptide of known sequence. The current rates and yields of these reactions make practical applications difficult — and thus far, these sys tems are limited to the synthesis of natural polymers, that is, polypeptides104. To work around biological polymeri zation techniques, ▇▇▇ and coworkers designed a DNA translation system to synthesize sequencecontrolled pol ymers not based on natural monomers105. The polymeri zation in this case depends on the hybridization of DNA base pairs to a template. Synthetic building blocks are attached to these DNA base pairs via a cleavable linker. In this system, the DNA base pairs perform a very similar function to tRNA, that is, they serve to bring together the desired building blocks in the correct order. After poly merization, cleavage of the linker results in the release of the synthetic polymer105. Complete chemical polymerization has the advan tage of a much wider range of available building blocks, but achieving perfect sequence control remains chal lenging because classical chaingrowth and stepgrowth polymerizations do not allow for precise control over the Rotaxane A mechanically interlocked molecular architecture in which a macrocycle is kinetically trapped on a thread by the presence of two large ‘stoppers’. was used to convert the protein into a shape repre senting a 1. For reading, a lowpower laser beam was used to detect the conformation of the protein without disturbing the conformation itself. The ability of bR to shift between different states also allows for rewritable data storage95,96. monomer position. Sequence control can be improved by using living chain polymerization methods, in which each polymer chain grows in a more uniform way. For instance, by using controlled radical polymerization techniques, that is, reversible addition−fragmentation chaintransfer (RAFT) polymerization, ▇▇▇▇▇▇▇▇ et al. REVIEWS Monodisperse polymers Polymers composed of uniform molecules with the same structure and mass. Naturally occurring polymers are frequently monodisperse, while synthetic polymers are usually not. constructed new sequencecontrolled macroRAFT agents by inserting two monomers in a sequential man ner106. The low yield of the monomer insertion, however, made this process suboptimal for the synthesis of long chains. Atomtransfer radical polymerization (ATRP) was used by ▇▇▇▇ et al. to construct vinyl polymers by an iterative process of single monomer addition107. However, this method also suffered from low yields (as a result of side reactions), making the synthesis of long chains impossible. In addition to controlled radical polymerization, sequence control in chain polymerization can also be obtained by using living ionic polymerization tech niques. Living cationic polymerization was used by ▇▇▇▇▇▇ et al. to create sequenceregulated polymers by the addition of monomers one by one in order of decreasing reactivity108. Nevertheless, defects occurred during the polymerization, necessitating purification after each monomer addition. Living anionic polymeri zation has also been used to obtain sequencecontrolled chains composed of two different monomers (the choice of one bulky monomer prevented homopoly merization). In this way, an alternating pattern of two different monomers could be obtained109. This type of kinetic control was also applied in radical polymeriza tion to obtain an alternating pattern, that is, the high affinity between two different monomers was used to create regions with a specific sequence in the polymer chain110. ▇▇▇▇ et al. built upon this alternating method by tuning the sequence through timecontrolled mono mer addition111,112. In this strategy, one monomer (an electronrich donor) present in excess is polymerized by a radical reaction, while a second monomer (an acceptor) is added in small amounts at specified times. A highly favourable donor–acceptor interaction between the two monomers results in the incorporation of acceptor monomers in small, welldefined regions of the polymer backbone111,112. Using automated protocols, ▇▇▇▇ et al. were able to construct welldefined polymer chains containing up to eight precisely positioned blocks with a specific sequence113. A similar approach was adopted by O’Reilly ▇’▇▇▇▇▇▇ et al. in ringopening metathesis polymeri zation, in which the position of four different functional moieties could be relatively well controlled along a grow ing polymer chain114. However, in all these methods, var iations in the length and precise composition of each segment may occur, and some polymer chains might contain defects101,115. To minimize the number of defects, long building blocks in multiblock copolymerization may be used. ▇▇▇▇ et al. employed this strategy in combination with degenerative transfer radical polymerization to con struct welldefined multiblock copolymers116. In addi tion, Engelis ▇▇▇▇▇▇▇ et al. used long blocks for the synthesis of welldefined multiblock copolymers by emulsion polymerization, in which monomers and catalysts were separated in micelles to isolate the growing polymers from one another and reduce unwanted side reactions117. Despite the reported improvements in the synthetic procedures, it can be concluded that chain polymer ization always results in polymers with deviations in chain length and composition115. This limitation means that chaingrowth polymerization is, at present, not a good method to prepare welldefined polymer sequences. Nonetheless, it could be employed for easy copying of already synthesized sequences by template polymerization118. In addition to chaingrowth polymerization, step growth polymerization techniques can be used to syn thesize polymer sequences with periodic monomer patterns. Conventional stepgrowth polymerization has been used for the synthesis of polyamides and polyurethanes. Although these methods are relatively straightforward, they do not allow for perfect sequence control. New stepgrowth polymerization techniques using radical polymerization119,120 or click chemis try121, however, do allow for such a sequencecontrolled polymerization. The latter can also be achieved by applying multistepgrowth synthesis, which involves the stepwise chemical attachment of monomers attached to a support122. This procedure results in highly monodisperse polymers. One such method is iter ative solidphase synthesis, similar to the wellknown solidphase peptide synthesis methodology. It employs an insoluble support on which the polymers are grown by the stepwise addition of monomers123. The method is highly efficient but also very time consuming, and fur thermore, the efficiency of the coupling steps decreases with increasing polymer length, making it best suited to the synthesis of short polymers. Despite these disad vantages, solidsupported synthesis remains the most frequently applied and most reliable method for the syn thesis of sequencecontrolled polymers122. An alterna tive is the use of a soluble polymer chain as a support124, which improves the process efficiency, but the synthesis of long sequences is still not possible101. The ▇▇▇▇ group has investigated numerous strategies to exploit this multistepgrowth methodology for the production of sequencecontrolled polymers, including those encod ing data125. The previously described stepgrowth syn thesis of polyurethanes, for instance, could be improved by applying a multistepgrowth approach126. This strat egy relied on a sequence of two chemoselective steps: the reaction of an alcohol with an Nhydroxysuccinimide (NHS) moiety and then the reaction of an amine with NHS126. Data were encoded using two different amino alcohol monomers (serving as 0 and 1), while N,N′ disuccinimidyl carbonate, containing two NHS moie ties, was used as a linker126. Another method developed by ▇▇▇▇ and coworkers is based on phosphoramidite coupling, a method already widely used for oligonucleo tide synthesis127. The synthesis uses a solid support, and the monomers are coupled one by one in three steps (FIG. 4a). First, N,Ndimethyltryptamine (DMT) depro tection of the monomer occurs, allowing the connec tion of the next monomer by phosphoramidite coupling, followed by oxidation of the phosphite to a phosphate. Optimization of this method allows each threestep cycle to be completed within a few minutes128. ▇▇▇▇ et al. used this approach to synthesize a polymer with a controlled sequence from two monomers containing either a propyl moiety (representing 0) or a 2,2dimethylpropyl moiety (representing 1)129. In addition, another monomer con taining a 2,2dipropargylpropyl group (representing 2) O HO N3 3 NH2 R O O O
Appears in 1 contract
Sources: End User Agreement
Synthesis. In addition to DNA and proteins, synthetic polymers are also suitable for data storage, at least in principle. As early as 1986, ▇▇▇▇▇▇▇ ▇▇▇▇▇▇▇ suggested that, at least in theory, any polymer composed of at least two different monomers could be used to store data97. Although the controlled synthesis of polymers with more than two monomers is possible — and although such polymers would potentially provide a more economical solution for data storage — most dataencoding polymers employ only two different monomers (directly represent ing 0 and 1 in the binary code). The main advantages of synthetic polymers are the possibility of having full con trol over their synthesis and the greater flexibility, mean ing that one is no longer restricted to four monomers, as in the case of DNA. Instead, the monomers can be selected and tuned for the purpose of the application. In such synthetic dataencoding copolymers, it is essential to achieve perfect control over the monomer sequence. For example, DNA can be used as a template for the assembly of free nucleotides (including nonnatural ones) before chemical or enzymatic polymerization98–103. The drawbacks of this approach are the low efficiency and difficulty of removing the synthesized polymer from the template. Recently, molecular machines that mimic biological polymerization have been developed, such as the peptide synthesis machine designed by ▇▇▇▇▇ and coworkers104. The machine is a rotaxane system in which the macrocycle sequentially picks up amino acids from the thread to assemble a peptide of known sequence. The current rates and yields of these reactions make practical applications difficult — and thus far, these sys tems are limited to the synthesis of natural polymers, that is, polypeptides104. To work around biological polymeri zation techniques, ▇▇▇ and coworkers designed a DNA translation system to synthesize sequencecontrolled pol ymers not based on natural monomers105. The polymeri zation in this case depends on the hybridization of DNA base pairs to a template. Synthetic building blocks are attached to these DNA base pairs via a cleavable linker. In this system, the DNA base pairs perform a very similar function to tRNA, that is, they serve to bring together the desired building blocks in the correct order. After poly merization, cleavage of the linker results in the release of the synthetic polymer105. Complete chemical polymerization has the advan tage of a much wider range of available building blocks, but achieving perfect sequence control remains chal lenging because classical chaingrowth and stepgrowth polymerizations do not allow for precise control over the Rotaxane A mechanically interlocked molecular architecture in which a macrocycle is kinetically trapped on a thread by the presence of two large ‘stoppers’. was used to convert the protein into a shape repre senting a 1. For reading, a lowpower laser beam was used to detect the conformation of the protein without disturbing the conformation itself. The ability of bR to shift between different states also allows for rewritable data storage95,96. monomer position. Sequence control can be improved by using living chain polymerization methods, in which each polymer chain grows in a more uniform way. For instance, by using controlled radical polymerization techniques, that is, reversible addition−fragmentation chaintransfer (RAFT) polymerization, ▇▇▇▇▇▇▇▇ et al. REVIEWS Reviews Monodisperse polymers Polymers composed of uniform molecules with the same structure and mass. Naturally occurring polymers are frequently monodisperse, while synthetic polymers are usually not. constructed new sequencecontrolled macroRAFT agents by inserting two monomers in a sequential man ner106. The low yield of the monomer insertion, however, made this process suboptimal for the synthesis of long chains. Atomtransfer radical polymerization (ATRP) was used by ▇▇▇▇ et al. to construct vinyl polymers by an iterative process of single monomer addition107. However, this method also suffered from low yields (as a result of side reactions), making the synthesis of long chains impossible. In addition to controlled radical polymerization, sequence control in chain polymerization can also be obtained by using living ionic polymerization tech niques. Living cationic polymerization was used by ▇▇▇▇▇▇ et al. to create sequenceregulated polymers by the addition of monomers one by one in order of decreasing reactivity108. Nevertheless, defects occurred during the polymerization, necessitating purification after each monomer addition. Living anionic polymeri zation has also been used to obtain sequencecontrolled chains composed of two different monomers (the choice of one bulky monomer prevented homopoly merization). In this way, an alternating pattern of two different monomers could be obtained109. This type of kinetic control was also applied in radical polymeriza tion to obtain an alternating pattern, that is, the high affinity between two different monomers was used to create regions with a specific sequence in the polymer chain110. ▇▇▇▇ et al. built upon this alternating method by tuning the sequence through timecontrolled mono mer addition111,112. In this strategy, one monomer (an electronrich donor) present in excess is polymerized by a radical reaction, while a second monomer (an acceptor) is added in small amounts at specified times. A highly favourable donor–acceptor interaction between the two monomers results in the incorporation of acceptor monomers in small, welldefined regions of the polymer backbone111,112. Using automated protocols, ▇▇▇▇ et al. were able to construct welldefined polymer chains containing up to eight precisely positioned blocks with a specific sequence113. A similar approach was adopted by O’Reilly et al. in ringopening metathesis polymeri zation, in which the position of four different functional moieties could be relatively well controlled along a grow ing polymer chain114. However, in all these methods, var iations in the length and precise composition of each segment may occur, and some polymer chains might contain defects101,115. To minimize the number of defects, long building blocks in multiblock copolymerization may be used. ▇▇▇▇ et al. employed this strategy in combination with degenerative transfer radical polymerization to con struct welldefined multiblock copolymers116. In addi tion, Engelis et al. used long blocks for the synthesis of welldefined multiblock copolymers by emulsion polymerization, in which monomers and catalysts were separated in micelles to isolate the growing polymers from one another and reduce unwanted side reactions117. Despite the reported improvements in the synthetic procedures, it can be concluded that chain polymer ization always results in polymers with deviations in chain length and composition115. This limitation means that chaingrowth polymerization is, at present, not a good method to prepare welldefined polymer sequences. Nonetheless, it could be employed for easy copying of already synthesized sequences by template polymerization118. In addition to chaingrowth polymerization, step growth polymerization techniques can be used to syn thesize polymer sequences with periodic monomer patterns. Conventional stepgrowth polymerization has been used for the synthesis of polyamides and polyurethanes. Although these methods are relatively straightforward, they do not allow for perfect sequence control. New stepgrowth polymerization techniques using radical polymerization119,120 or click chemis try121, however, do allow for such a sequencecontrolled polymerization. The latter can also be achieved by applying multistepgrowth synthesis, which involves the stepwise chemical attachment of monomers attached to a support122. This procedure results in highly monodisperse polymers. One such method is iter ative solidphase synthesis, similar to the wellknown solidphase peptide synthesis methodology. It employs an insoluble support on which the polymers are grown by the stepwise addition of monomers123. The method is highly efficient but also very time consuming, and fur thermore, the efficiency of the coupling steps decreases with increasing polymer length, making it best suited to the synthesis of short polymers. Despite these disad vantages, solidsupported synthesis remains the most frequently applied and most reliable method for the syn thesis of sequencecontrolled polymers122. An alterna tive is the use of a soluble polymer chain as a support124, which improves the process efficiency, but the synthesis of long sequences is still not possible101. The ▇▇▇▇ group has investigated numerous strategies to exploit this multistepgrowth methodology for the production of sequencecontrolled polymers, including those encod ing data125. The previously described stepgrowth syn thesis of polyurethanes, for instance, could be improved by applying a multistepgrowth approach126. This strat egy relied on a sequence of two chemoselective steps: the reaction of an alcohol with an Nhydroxysuccinimide (NHS) moiety and then the reaction of an amine with NHS126. Data were encoded using two different amino alcohol monomers (serving as 0 and 1), while N,N′ disuccinimidyl carbonate, containing two NHS moie ties, was used as a linker126. Another method developed by ▇▇▇▇ and coworkers is based on phosphoramidite coupling, a method already widely used for oligonucleo tide synthesis127. The synthesis uses a solid support, and the monomers are coupled one by one in three steps (FIGfig. 4a). First, N,Ndimethyltryptamine (DMT) depro tection of the monomer occurs, allowing the connec tion of the next monomer by phosphoramidite coupling, followed by oxidation of the phosphite to a phosphate. Optimization of this method allows each threestep cycle to be completed within a few minutes128. ▇▇▇▇ et al. used this approach to synthesize a polymer with a controlled sequence from two monomers containing either a propyl moiety (representing 0) or a 2,2dimethylpropyl moiety (representing 1)129. In addition, another monomer con taining a 2,2dipropargylpropyl group (representing 2) a O O R R NC O P N O ODMT R R ODMT (i) (ii) (i) (ii) (iii) (iii) (iv) R = CH3 (1), H (0) R = CH2C CH (2) H H H3C CH3HO O O O – P – O O O 0 1 P O O 2 O – O P O O H 0 H O – O P H3C O O 1 CH3 OH b HO O O CH3 1 O N N N O O HO N3 3 NH2 R (i) (ii) (i) (ii) (iii) Spacer = R = CH3 (1), H (0) O O OO O O N N N N 3 H 3 H H N N CH N N3 0 1 3 N H O CH3 1 N N N O 3 O N H c O O NH2 O O Br Br O N NH2 ▇ ▇ ▇ (i) (ii) (i) (ii) (iii) R = CH3 (1), H (0) Spacer =O HO N O N N H H O H 0 O H 0 O N N H O H C 3 1 O N N H O Br H C 3 1
Appears in 1 contract
Sources: End User Agreement