Poly(ethylene glycol) (PEG) is a biocompatible, flexible, and hydrophilic polymer that is widely applied in numerous fields. Especially in pharmaceutical research, PEG is used as a bioconjugate agent for PEG-ylated drugs. A well-defined structure is crucial, since dispersity affects biological activity (e.g., toxicity and efficacy). Thus, intensive efforts to develop synthetic protocols approaching uniformity have been made in recent decades. Different approaches utilizing iterative step-by-step synthesis procedures have yielded promising results, and improvement is still ongoing. In this comparative study, we adopted several procedures for the preparation of uniform PEGs in combination with careful characterization, including size exclusion chromatography (SEC) analysis, which has yet to be reported. Oligo(ethylene glycol)s up to the dodecamer were synthesized. The results obtained were compared in terms of yield and purity with those previously reported in the literature. We clearly show the importance of SEC analysis with high separation capacity in the oligomer range for the synthesis of short-chain oligo(ethylene glycol)s.
Poly(ethylene glycol)s (PEGs) are versatile, biocompatible, chemically stable, flexible, relatively nontoxic, and water-soluble polymers. They are most frequently used in biopharmaceutical research as well as for everyday detergent applications. PEG-ylation  of biomacromolecules, such as proteins and peptides, or small therapeutic molecules has been shown to improve the pharmacological properties of these molecules, for instance by increasing their solubility  and stability, and influencing their pharmacokinetics and mode of action (for instance, the oral bioavailability is enhanced) [1, 3,4,5]. PEG increases the size of the conjugate and acts as a steric shield to protect counterpart active ingredients against recognition by the immune system . By increasing the half-life of drugs in vivo, the dosing frequency can be reduced [7,8,9,10,11]. Furthermore, PEGs are applied in fields such as bionanoparticles , electrolytes , nanocomposite films , and organic–inorganic hybrid materials . Particularly in medicine, where heterogeneity influences the biological activity of PEG-ylated drugs (e.g., their toxicity and efficacy), it is essential to use uniform PEGs to obtain distinct structure–property relationships and to precisely adjust the aforementioned functions. PEGs with low dispersity but more than four PEG units, i.e., tetra(ethylene glycol), are commercially not readily available and are rather expensive (up to 100 EUR/g for PEG8 ≥ 95% ). Separation, via either distillation or preparative size exclusion chromatography (SEC), of a nonuniform (polydisperse) polymer mixture would be a Sisyphean task. Thus, over recent decades, several working groups have developed methods to synthesize uniform PEGs (Fig. 1) [17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33].
These approaches to the synthesis of uniform macromolecules are based on iterative methods. Hence, the preparation of high molecular weight macromolecules that retain their uniform nature requires particular synthetic effort, including the use of protecting groups and excess of starting materials to achieve maximum yield for monofunctionalization. Since the Williamson ether synthesis, where elimination is the most significant side reaction under the often employed harsh conditions (i.e., >100 °C for several hours), is the reaction mechanism of choice , tedious purification after each step to achieve highly pure products is crucial .
In 1936, Hibbert and coworkers reported a dodeca(ethylene glycol) by reacting a dichloro- and a mono-deprotonated ethylene glycol, and an oligo(ethylene glycol) with 42 repeat units was reported in 1939 [17, 32]. Kamachi et al. adopted this approach in 1994 using a ditosylate instead of the dichloro compound to obtain a PEG28 . In the last three decades, four different concepts applying monofunctionalized building blocks (to prevent polymerization) have been developed: iterative coupling of a mono-protected ethylene glycol to one end (a) or to both ends (bidirectional growth) (b) of a growing polymer chain, chain doubling (iterative exponential growth (IEG)) (c) and chain tripling (d) (Fig. 2).
Jenneskens and coworkers described the synthesis of a dodeca(ethylene glycol) via bidirectional growth of a mono-protected and a ditosylated tetra(ethylene glycol) . The product was obtained in 95% yield over two steps on a multigram scale. By reacting two equivalents of a monobenzyl tetra(ethylene glycol) tosylate with a PEG36 diol, a uniform PEG44 was prepared by Tanaka and Ahmed with an overall yield of 17% in nine steps on a 1.6-g scale . In 2008, Springer et al. obtained a 29-mer in an overall yield of 36% (490 mg) in six steps via bidirectional growth starting from hexa(ethylene glycol) . Based on the procedure of Tanaka and Ahmed, the group of Bruce achieved a bis-methyl-protected 24-mer on a multigram scale with the highest purity reported to date, as determined by MALDI-MS (>98% after one ether coupling and >95% after three couplings) . More recently, Jiang and coworkers described a new strategy for the synthesis of uniform PEGs, taking advantage of a macrocyclic sulfate (MCS), which circumvents the protection and activation steps. In this method, a 64-mer was prepared via the bidirectional growth strategy using a diol as a nucleophile in the ring opening of the MCS .
The disadvantage of concepts (b) and (d) (Fig. 2) relates to the fact that only symmetric products can be achieved via bidirectional growth. Since single post-functionalization of PEGs longer than 12 repeat units is challenging due to the lack of selectivity , unidirectional approaches (a) and (c) (Fig. 2) are more versatile. Furthermore, full conversion via approaches (b) and (d) is difficult to achieve. For example, chain tripling of PEG16 with α-benzyl-ω-tosyl hexadeca(ethylene glycol) yielded a mixture containing 31% monobenzyl PEG32 and 8% bis-dibenzyl PEG48 . In the iterative exponential growth strategy (c), orthogonal protecting groups are necessary. Repetition of a coupling reaction and subsequent selective deprotection leads to the elongation of the PEG chain. In 1999, Burns et al. prepared hexa(ethylene glycol) on a 17.0-g scale in 80% yield . Uniform dodeca(ethylene glycol)s were synthesized by Hill et al. on multigram scales with overall yields between 49 and 55%, starting from hexa(ethylene glycol) building blocks. Furthermore, 1.65 g of a PEG24 was obtained in an overall yield of 43% . Davis and coworkers described the syntheses of PEG32 and PEG48 in purities >98%. In addition, hydroxyl-, azido-, and amino hexadeca(ethylene glycol) monomethyl ethers were prepared on gram scales with purities >99.8% . To avoid tedious purification steps, Baker et al. recently reported a chromatography-free synthesis using tetrahydropyranyl (THP) and benzyl (Bn) protecting groups and taking advantage of selective extraction. A practical and scalable method for the synthesis of asymmetric and well-defined PEGs was developed. In this way, Bn octa(ethylene glycol) was obtained on a 100-g scale .
Livingston, Gaffney and coworkers applied linear synthesis (a) towards heterofunctional uniform oligo(ethylene glycol)s by using a three-armed Bn hub. Iterative chain extension with mono-protected octa(ethylene glycol) in combination with simplified purification, end-group functionalization, and subsequent cleavage from the core unit was conducted. In this way, PEG24 monomethyl ether (yield 37%) and PEG56-acetoxy-monomethyl ether in an overall yield of 13% were obtained on a milligram scale [25, 33]. In addition, Jiang et al. prepared well-defined oligo(ethylene glycol)s in a step-by-step manner on a fluorous tag, which allows simple purification via fluorous solid-phase extraction (FSPE) and solid-phase extraction (SPE). PEG20 was obtained in an overall yield of 13% in 13 steps on a gram scale .
Recently, Fang reported the synthesis of PEG12 on a 36 mg scale with an overall yield of 81% in five steps by the stepwise addition of a dimethoxytrityl tetra(ethylene glycol) tosylate on a Wang resin and monobenzyl PEG12 on a gram scale with an overall yield of 80% in five steps using Bn-protected building blocks. By applying this approach, column chromatography was completely avoided, milder reaction conditions could be applied, and post-functionalization after cleavage was possible. The purity of the products was determined with ESI and MALDI-TOF MS and was described as “close to monodispersity.” Further chain elongation to PEG16 and PEG20 led to mixtures with different chain lengths .
Unfortunately, purification methods are not always reported, and characterization using multiple techniques, including at least one chromatographic analysis, is not necessarily applied in all reports. We herein discuss the importance of complete characterization, including SEC, nuclear magnetic resonance (NMR), and mass spectrometry (MS), to confirm uniformity. We find that SEC measurements using columns optimized for separation in the oligomer range reveal currently unreported selectivity issues and allow comparison and optimization of the reported routes. SEC, compared with other chromatographic methods such as HPLC, offers the advantage of running isocratically and typically using RI instead of UV detectors, thus allowing a straightforward routine analysis without the necessary gradient optimization and allowing all present species (contaminations) to be detected. Our purity values are based on a simple symmetry peak analysis (Figs. S38 and S39, Supporting information). Thus, this paper objectively compares reported synthetic routes towards uniform PEGs by using a set of characterization methods that allow the establishment of an unbiased data set for comparison. It is important to clarify that we compare different synthetic methods and approaches to highlight advantages and disadvantages, whereas it was not our intention to exactly reproduce the procedures described in the literature, since this is often not possible practically (i.e., availability of different grades of reagent, same type of silica, and so on).
Results and discussion
PEGs with a low dispersity of Ð ≤ 1.04 are prepared via well-controlled anionic polymerization of 2-(benzyloxy)ethanol, potassium hydride, and ethylene oxide [36, 37]. To achieve uniformity (i.e., Ð = 1.00), an iterative synthesis approach must be followed. Therefore, desymmetrization by introduction of protecting groups is indispensable, representing a synthetic bottleneck due to the formation of double-protected ethylene glycols as side products. Hill et al. reported the synthesis of monofunctionalized PEGs by adopting the Bouzide procedure of using stoichiometric amounts of the protecting group and silver(I) oxide (Ag2O) [21, 38]. In this way, monobenzyl-, monotrityl-, mono-p-methoxybenzyl-, and monotosyl ethylene glycols with three or six repeating units were prepared in yields between 43 and 92%. To avoid stoichiometric amounts of Ag2O, an excess of less expensive tri- or tetra(ethylene glycol) can be used, which can be easily removed by washing with water. The combination with a slow addition of the protecting group statistically favors monofunctionalization, resulting in yields comparable with those of the silver(I) oxide approach. We therefore started our investigations by adopting different reported procedures for the synthesis of THP-, trityl (Trt)-, and Bn-protected tri- or tetra(ethylene glycol), as summarized in Scheme 1.
Monofunctionalization of PEG-diols
Thus, we were able to compare yields and purity directly for the different approaches. SEC analysis using columns that offer high resolution in the oligomer range proved to be the most important technique for assessing the efficiency of the different approaches in terms of uniformity, especially that of the crude reaction mixtures. The investigated protecting groups were chosen due to the simple purification required after the deprotection step via filtration, solvent evaporation, and/or extraction. The products are numbered consecutively, where the letter is related to the chain length of the oligo(ethylene glycol) and the numbering in brackets refers to the applied synthetic protocol (see Scheme 1).
The THP protecting group was introduced by applying the chromatography-free reaction protocol of Baker et al.  under acidic conditions using p-toluenesulfonic acid (0.10 eq.) in dry dichloromethane (Scheme 1, method A). Tri(ethylene glycol) 1a was used in an excess of five equivalents. Although traces of doubly protected ethylene glycol (THP2(EG)3) would not influence the subsequent reaction steps, product 2a(i) was purified via column chromatography, affording a yield of 60.9%. The same reaction was conducted with tetra(ethylene glycol) 1b, affording the crude mono(THP)-protected PEG4 2b(i) in a yield of 2.30 g (74.4%). In addition, a procedure for the synthesis of 2b(ii) described by Ahmed and Tanaka was performed . Here, the reaction time was decreased considerably to 30 min. Product 2b(ii) was obtained in 51.0% yield after purification via column chromatography.
Figure 3a shows the SEC traces of the THP-protected ethylene glycols. A significant shift towards a lower retention time was observed for 2b(i) and 2b(ii) compared with tetra(ethylene glycol) 1b. Interestingly, the crude product 2b(i) obtained via method A(i) exhibits tailing towards higher hydrodynamic volumes, which was ascribed to the doubly protected ethylene glycol (turquoise, dotted trace) via NMR analysis. For the purified products 2a(i) (Fig. S2, Supporting information) and 2b(ii), symmetric and narrow SEC traces with dispersity indices of Đ = 1.00 were obtained, indicating high purity.
Monotrityl tetra(ethylene glycol) 3b(i) (Scheme 1, method B) was prepared from 1b using triethylamine, 4-dimethylaminopyridine (DMAP), and trityl chloride . The reaction was refluxed for 6 h, and purification of the product via column chromatography afforded 3b(i) in 64.6% yield. In another chromatography-free approach adopted from Kinbara et al., the reaction was performed using DMAP without any additional solvent. The reaction time and temperature were decreased to 3 h and room temperature, respectively . SEC analysis showed a contamination of 8% with symmetric tetra(ethylene glycol) bis-trityl ether and 3% of the starting material, even after several additional washing steps (Fig. 3b, 3b(ii)). The yield of the pure product 3b(ii) was calculated via SEC (88.4%, product not separated). The mixture was used for the next step without any further purification. In a third approach, an increase of 28.6 percentage points in yield compared with that of 3b(i) was obtained when following the procedure of Davis et al. . Here, pyridine was used as the base, the reaction was conducted at 45 °C for 12 h, and toluene was used for the extraction instead of DCM. The narrow and monomodal SEC trace of 3b(i) (Fig. 3b) and 3b(iii) (Fig. S6, Supporting information) with a dispersity of Đ = 1.00 confirms the uniformity.
Bn ether was chosen as an orthogonal protecting group for a trityl- or THP functionality (Scheme 1, method C). It was introduced via a nucleophilic substitution of an alkali alkoxide and Bn bromide. Deprotonation was accomplished either with sodium hydroxide under aqueous conditions (Schemes 1, 4a(i), 4b(i))  or with sodium hydride in dry tetrahydrofuran (THF)  (Schemes 1, 4b(ii)). The monobenzylated tetra(ethylene glycol)s 4b(i) and 4b(ii) were obtained in comparable yields of 77–83% after purification via column chromatography. Hence, our initial comparative study revealed that method C(i) is preferred because of its more practicable performance and the avoidance of hydrogen formation. In this way, tri(ethylene glycol) monobenzyl ether 4a(i) was synthesized in 72.9% yield. The products were again analyzed with SEC (Fig. 3c) in addition to NMR spectroscopy and MS to confirm their purity. A comparison of the monobenzyl tri- (4a(i)) and tetra(ethylene glycol) (4b(i)) is shown in Fig. 3d. A difference of 0.45 min in the retention time was observed.
In the literature, the electrophilicity of the mono-protected ethylene glycol is shown to be improved by chlorination with thionyl chloride, which unfortunately leads to bond cleavage (depolymerization) affording a mixture of different chain lengths [20, 35]. Therefore, sulfonate esters (mesylate and tosylate) are more suitable for the activation of alcohol functions [17, 32, 39]. Although the mesylate shows marginally better results in ether coupling than the tosylate, Ahmed and Tanaka showed that the tosylation in aqueous THF is more reasonable than the introduction of mesylate in pyridine. In addition, mono-tosylation with subsequent protection is not appropriate because traces of the formed bis-tosylate must be carefully removed to prevent undesired side products in the subsequent reaction process. Elimination of the mono(THP)- and monotrityl-protecting groups was observed under the tosylation conditions . Here, we adopted the reaction procedure of Bruce et al. (Scheme 2). Bifunctionalized tri- and tetra(ethylene glycol)s (5a and 5b) were obtained after reacting 4a or 4b for 15 h in basic aqueous THF with p-tosyl chloride. Purification via column chromatography afforded the monobenzyl ethylene glycol tosylates in 55.7% (5a) and 96.2% (5b) yield, respectively. Careful characterization via NMR spectroscopy, SEC (Fig. 4) and MS revealed the purity of the products. To prevent the degradation of the tosylates, they were stored under argon and shielded from light.
A comparison of the SEC traces of the monobenzyl ethylene glycols 4a and 4b with the corresponding tosylated products 5a and 5b is shown in Fig. 4. Narrow peaks with a dispersity of Đ = 1.00 and a shift towards lower retention times and thus a higher hydrodynamic volume were observed.
Chain elongation for the synthesis of uniform PEGs is conducted via iterative Williamson’s ether synthesis, where elimination under harsh reaction conditions constitutes the most significant side reaction and is thus related directly to the purity of the product . Tanaka et al. described the synthesis of PEG44 by applying sodium hydride in THF. Unfortunately, chain degradation was observed, induced by the formation of PEG-alkoxides under basic conditions . Baker et al. adopted the Tanaka procedure and added sodium iodide as a catalyst, which undergoes tosylate–iodide exchange in a Finkelstein-type reaction to improve the reactivity of the alkylating agent. Well-defined PEGs were obtained by a chromatography-free method, affording the products of ether coupling in quantitative yields . Substitution of NaH in THF with KOtBu in DMF and 18-crown-6 was investigated by Davis et al. The base was added slowly to the reactants to keep the alkoxide concentration as low as possible and thus to prevent chain scission. To simplify the purification by column chromatography, orthogonal protecting groups (Bn, tert-butyl, and trityl) were used. Since the trityl- and Bn ether protecting groups are both cleavable via reductive hydrogenolysis, the yields using these protecting groups were significantly lower than those with tert-butyl ether. Nevertheless, PEG32 and PEG48 derivatives were obtained with 98.9 and 98.0% purity, as indicated by MALDI-MS, which allowed the first exceptional insight into the 3D-PEG morphology and an extended helical secondary structure by X-ray crystallography . Further optimization studies were performed by Bruce et al. with the aim of avoiding the addition of 18-crown-6 to improve the solubility of KOtBu in DMF, as this additive was difficult to separate. To overcome the solubility issue, DMF was replaced by less toxic THF, which is also easier to evaporate. In addition, the reaction temperature was decreased, and the addition of the base was changed due to crystallization issues. In this way, PEGs with a purity >95% after three coupling steps, as indicated by MALDI-MS, were prepared on a multigram scale .
We adopted the above-discussed procedures for the synthesis of uniform PEG oligomers, which were all carried out under an argon atmosphere. Bis-Bn dodeca(ethylene glycol) 6e was prepared according to the procedure of Baker et al. (Scheme 3 6e(i)) using sodium iodide as a catalyst . In a second approach, the reaction was performed with KOtBu instead of NaH (Scheme 3 6e(ii)), according to the synthesis protocol of Bruce et al. .
Based on the findings of Tanaka et al., tetra(ethylene glycol) 1b and monobenzyl tetra(ethylene glycol) tosylate 5b were used as coupling reagents instead of bis-tosylate and monobenzyl tetra(ethylene glycol), since the elimination product of the monosubstituted intermediate is more difficult to separate from 6e. Unfortunately, we observed various byproducts by SEC measurements of the crude reaction mixtures (Fig. 5a). The SEC traces of 6e(i) and 6e(ii) show the formation of the bis-Bn dodeca(ethylene glycol) 6e, as indicated by a significant shift towards a lower retention time of 18.78 min compared with that of the starting materials. Additional peaks beside the product peak were successfully assigned via SEC coupled to electrospray ionization-mass spectrometry (SEC-ESI-MS) analysis for 6e(ii) and are summarized in Table S2 (Supporting information). Several mono- and bifunctionalized oligo(ethylene glycol)s ranging from monobenzyl octa(ethylene glycol) to bis-Bn icosa(ethylene glycol), as well as the elimination product of 5b, were observed. Due to the structural similarity of the formed compound mixture, separation via column chromatography was challenging, and the products were obtained in rather low yields of 12.6% (6e(i)) and 36.5% (6e(ii)) with a purity of ≥98% determined by SEC (Fig. 5b). These results demonstrate the importance of SEC analysis and clearly show that this synthetic approach leads to unfavorable results.
Orthogonal protected PEGs via IEG
Although the chain growth for the first two ether couplings is faster in the case of chain tripling/bidirectional growth when compared with the iterative exponential growth, we could not obtain the dodeca(ethylene glycol) 6e in reasonable yields via bidirectional growth (Scheme 3, approaches (i) and (ii)). This might be related to improved analytical protocols (i.e., SEC with high resolution). Therefore, attempts to prepare orthogonally protected oligo(ethylene glycol)s by applying iterative exponential growth were conducted. The monobenzyl oligo(ethylene glycol) tosylate 5a or 5b was coupled with either mono(THP)- (2a, 2b) or monotrityl tetra(ethylene glycol) 4b, affording bifunctionalized hepta- and octa(ethylene glycol)s 7c, 7d, and 8d. The reaction conditions were adopted from Baker et al. (Scheme 4 (i))  and Bruce et al. (Scheme 4 (ii)) , which were already applied for the chain-tripling approach.
SEC traces of the coupling products before and after purification via column chromatography are shown in Fig. 6a–c. A peak corresponding to 5b was observed at a retention time of 20 min in the SEC of method (ii), since it was used in an excess of 1.30 equivalents (Fig. 6a, b). However, both reaction methods proceeded in a similar fashion. A clear shift to a lower retention time was observed for the formation of products 7c, 7d, and 8d. Smaller impurities were found at a higher hydrodynamic volume and at a retention time of 20.8 min in the case of method (i) (Fig. 6a, b). Several masses of byproducts up to the mono-protected hexadecamers were observed by SEC–ESI–MS analysis for the synthesis of 7d(ii), which are summarized in Table S3 (Supporting information). Different solvent systems (DCM:MeOH, DCM:acetone, pure ethyl acetate) were tested, but the separation of the products via normal phase column chromatography was difficult, resulting in a decrease in yield. Interestingly, Baker et al. claimed a quantitative yield for the synthesis of 7d(i) after drying under high vacuum . In contrast, we obtained a yield of 68.4% for 7d(i) and 34.5% for 7c(i) after purification via column chromatography. Furthermore, the yields of the products prepared via method (ii) were not constant in repeated trials but were 37.0–47.6% for the octamer and 71.3% for the heptamer. Since the THP protecting group is unstable under acidic conditions, we changed the reported purification protocol and used water instead of 1 M HCl for the washing step. Unfortunately, no considerable increase in yield was observed. Using monotrityl tetra(ethylene glycol) 4b in the ether coupling (method ii), we obtained α-benzyl-ω-trityl octa(ethylene glycol) 8d (Fig. 6c) in 49.6% yield, which is lower than that described in the literature . However, the high purity of the bifunctionalized ethylene glycols was indicated by NMR spectroscopy, MS, and SEC analysis. In summary, we can reveal that it is possible to distinguish PEG7 and PEG8 via SEC (Fig. 6d), again pointing out the necessity to report SEC traces when working with uniform macromolecules.
To avoid tedious purification via column chromatography, Kinbara et al. established a chromatography-free approach for the synthesis of well-defined asymmetric PEGs. Making use of the functionalization-dependent distribution of PEGs between the organic and aqueous phases, an iterative monofunctionalization of tetra(ethylene glycol) is described (Scheme 5) . A trityl protecting group was used as a hydrophobic tag, as well as a p-toluenesulfonyl moiety, which also acted as a leaving group in Williamson ether coupling. Bis-trityl-protected PEG byproducts, which did not interfere in further reactions, were transferred to the subsequent step and could be easily removed after the final deprotection step via liquid–liquid extraction. Since products that differ in one ethylene glycol unit, resulting either from the base-induced depolymerization or from impurities in the starting material, cannot be separated during the extraction step, a suitable analytical method was crucial to verify uniformity. Taking advantage of the p-toluenesulfonyl group as a chromophore for UV detection, reverse-phase HPLC (RP-HPLC) previously showed a higher resolution in comparison with MALDI measurements, demonstrating that MS is not a suitable method for quantitative dispersity analysis .
In this way, PEG8-Ts (72% yield over five steps, 98.7% RP-HPLC purity), PEG12-Ts (63% over seven steps, 98.2% RP-HPLC purity), and PEG16-Ts (62% yield over nine steps, 97.0% RP-HPLC purity) were prepared on multigram scales. The limitation of this procedure was investigated, since PEG tosylate with a certain chain length prefers the aqueous phase during the extraction step, but even the PEG24-tosylate remained quantitatively in the organic layer, as indicated by HPLC analysis. Here, monotrityl tetra(ethylene glycol) 3b(ii) (contaminated with 8% of bis-trityl tetra(ethylene glycol), as indicated by SEC analysis) was activated via tosylation, and 9b was obtained in quantitative yield, still contaminated with 8% bis-trityl tetra(ethylene glycol), which was used in the subsequent step without further purification. In the coupling step, NaH was used as the base, and tetra(ethylene glycol) 1b was added in an excess of 7.32 equivalents, leading to monotrityl octa(ethylene glycol) 3d in quantitative yield. Another tosylation was performed, affording 9d in 95.9% yield. In the last two steps, an additional impurity of bis-trityl dodeca(ethylene glycol) (2%) was observed in the SEC chromatogram. In the final reaction step, the trityl protecting group was cleaved under acidic conditions, affording the crude octa(ethylene glycol) monotosylate 10d in an overall yield of 66.3% in five steps, which is 5–10% lower than that reported in the literature . However, SEC indicated a contamination with 12% impurities at a lower retention time (Fig. 7a); thus, purification of product 10d via column chromatography was necessary after the final step, resulting in a 2.5% loss of the product (63.8% final yield). The crude and isolated yields are summarized in Table 1. The results demonstrate that a chromatography-free approach with an optional final purification step is a practical synthetic option.
SEC chromatograms of each step are shown in Fig. 7 (before (a) and after purification via column chromatography (b)). The contamination of the bis-trityl tetra(ethylene glycol) at 19.5 min is no longer visible in 9b(i), since the retention time is similar. Narrow and monomodal peaks with a dispersity of Đ = 1.00 were obtained for the single products after purification.
Separate deprotection of the α-benzyl-ω-tetrahydropyranyl octa(ethylene glycol)
Nevertheless, orthogonal deprotection of α-benzyl-ω-tetrahydropyranyl octa(ethylene glycol) 7d was performed separately (Scheme 6). The THP-ether was deprotected under acidic conditions according to the procedure of Baker et al. , affording the monobenzyl octa(ethylene glycol) 4d in 97.7% yield.
Reductive hydrogenation under reflux conditions was conducted to cleave the Bn protecting group, resulting in mono(THP) octa(ethylene glycol) 2d in 98.9% yield. A comparison of SEC chromatograms of products 4d and 2d with starting material 7d is shown in Fig. 8. Significant shifts towards higher retention times were observed due to a decrease in the hydrodynamic volume. A narrow peak was observed for product 2d, whereas peak broadening occurred after Bn deprotection, which could be the result of complete deprotection towards the octa(ethylene glycol), since the mass was determined from the ESI-MS spectra as well. Furthermore, the mass of bis-Bn octa(ethylene glycol) was found in product 4d and the mass of bis-tetra(hydropyranyl) octa(ethylene glycol) in product 2d.
Macrocyclization of ethylene glycols
Recently, Jiang et al. reported the synthesis of uniform PEG derivatives via nucleophilic ring opening of an MCS. Macrocyclization was performed with several diols and thionyl chloride at a rather high concentration of 0.04 M, followed by in situ oxidation of the cyclic sulfite with ruthenium tetroxide (RuO4) (Scheme 7). A variety of different nucleophiles were used for the nucleophilic ring opening, giving PEG derivatives in yields of 34–99% . Since this method avoids the use of protection and activation steps, it is an adequate alternative to previously described procedures. Furthermore, Jiang et al. reported the scalability and versatility of this method, e.g., for the synthesis of dual-functional PEGs , as well as the preparation of an α-amino-ω-methoxyl dodeca(ethylene glycol) on a 53-g scale, high purity determined by 1H-NMR and an overall yield of 61% in eight steps .
Unfortunately, even at lower concentrations of 0.01 M, we observed the formation of larger macrocycles in SEC traces for the macrocyclization of tetra(ethylene glycol) 1b, up to the cyclic pentamer (Fig. 9), which was further confirmed by SEC-ESI-MS analysis (Tables S3 and S4, Supporting information). Interestingly, the approach at a concentration of 0.02 M showed the lowest side product formation and was therefore used for in situ oxidation (Fig. 9a). The MCS was purified via column chromatography, affording product 12b in 59.7% yield, which was used in a nucleophilic ring opening with monobenzyl-protected tetra(ethylene glycol) 4b, leading to the monobenzyl octa(ethylene glycol) 4d. Due to ring formation, the hydrodynamic volume decreases, resulting in a shift of the product peak of 11b towards a higher retention time, whereas cyclic oligomers were observed at a lower retention time compared with tetra(ethylene glycol) 1b (Fig. 9). As a result of the nucleophilic ring opening with monobenzyl tetra(ethylene glycol) 4b, a clear shift towards a lower retention time was observed. Unfortunately, we were not able to reproduce the results described by Jiang et al. ESI–MS analysis confirmed the formation of the desired product 4d, but we also observed a side product at a lower retention time in SEC, which we could not assign to products of the ring opening of larger macrocycles.
In summary, different synthetic strategies to achieve mono-protected octa(ethylene glycol)s were investigated according to described literature procedures [22, 24, 27,28,29,30]. Most importantly, all reactions were analyzed using an SEC system with high resolution in the oligomer range, thus allowing an unambiguous comparison of the different procedures in terms of practicality, selectivity, purity of the final product, and yield. In contrast, the literature indicates that SEC is not suitable for verifying the purity of PEGs . Nonetheless, it is quite possible to distinguish oligomers with only one additional repeat unit for the described oligo(ethylene glycol)s, as monodisperse species should be observed as highly symmetric peaks in SEC. As shown in the Supporting information (Figs. S38 and S39), contamination of a PEG8 with only 2% PEG7 can be clearly identified by SEC using a simple symmetry peak analysis. Furthermore, for most side products, we observed a difference of at least four repeating units, thus supporting our hypothesis that SEC is a powerful analytical tool to monitor the reaction process. The results are summarized in Tables 2 and 3 and reveal that there is no difficulty in mono- and difunctionalization (Table 2: entries 1–7, Table 3: entries 1–3), but the yields are often lower than those reported if SEC is used to measure the purity. This also accounts for the separate deprotection of oligo(ethylene glycol)s 4d and 2d, while problems reproducing the ether coupling described in the literature arose (Table 2: entries 8 and 9, Table 3: entries 4–9). The purities calculated via SEC analysis are comparable with those reported, which were estimated mostly just by MS analysis, but one has to carefully consider whether MS is generally suitable for purity calculations. Chromatography, which detects all species and does not have a bias towards ionization of different species, is a better choice and gives more trustworthy data. The best yield was obtained for the synthesis of the double-protected heptamer 7c(ii) according to the procedure of Bruce et al. , whereas the results for the chain tripling method did not match those described in the literature (6e(i) and 6e(ii)). Due to the formation of several side products, identified via SEC–ESI–MS (Tables S2 and S3, Supporting information), the isolation of the products via normal phase column chromatography was rather challenging, resulting in low and nonconsistent yields. Such side products were not identified before, suggesting that the previously reported samples were contaminated. Therefore, investigations into and optimizations of the ether synthesis in combination with careful characterization must be carried out in the future, and we are currently working on novel strategies for the preparation of uniform PEGss. Nevertheless, the results described herein provide a generalized overview of previously reported procedures as well as their limitations.
J Whelan, C Pan. (Bayer Pharmaceuticals Corporation, USA), PCT Application WO. 2005-US20469, 2005.
Brocchini S, Godwin A, Balan S, Choi J-W, Zloh M, Shaunak S. Disulfide bridge based PEGylation of proteins. Adv Drug Deliv Rev. 2008;60:3–12.
Haag R, Kratz F. Polymer therapeutics: concepts and applications. Angew Chem Int Ed Engl. 2006;118:1218–37.
Zalipsky S. Chemistry of polyethylene glycol conjugates with biologically active molecules. Adv Drug Deliv Rev. 1995;16:157–82.
Greenwald RB, Choe YH, McGuire J, Conover CD. Effective drug delivery by PEGylated drug conjugates. Adv Drug Deliv Rev. 2003;55:217–50.
Roberts MJ, Bentley MD, Harris JM. Chemistry for peptide and protein PEGylation. Adv Drug Deliv Rev. 2012;64:116–27.
Caliceti P, Veronese FM. Pharmacokinetic and biodistribution properties of poly(ethylene glycol)–protein conjugates. Adv Drug Deliv Rev. 2003;55:1261–77.
Harris JM, Chess RB. Effect of pegylation on pharmaceuticals. Nat Rev Drug Discov. 2003;2:214–21.
Veronese FM, Pasut G. PEGylation, successful approach to drug delivery. Drug Discov Today. 2005;10:1451–8.
Fishburn CS. The Pharmacology of PEGylation: Balancing PD with PK to generate novel therapeutics. J Pharm Sci. 2008;97:4167–83.
Knop K, Hoogenboom R, Fischer D, Schubert US. Poly(ethylene glycol) in drug delivery: pros and cons as well as potential alternatives. Angew Chem Int Ed. 2010;49:6288–308.
Furusho H, Kitano K, Hamaguchi S, Nagasaki Y. Preparation of stable water-dispersible PEGylated gold nanoparticles assisted by nonequilibrium atmospheric-pressure plasma jets. Chem Mater. 2009;21:3526–35.
Ganapatibhotla LVNR, Zheng J, Roy D, Krishnan S. PEGylated imidazolium ionic liquid electrolytes: thermophysical and electrochemical properties. Chem Mater. 2010;22:6347–60.
Shah SAS, Nag M, Kalagara T, Singh S, Manorama SV. Silver on PEG-PU-TiO2 polymer nanocomposite films: An excellent system for antibacterial applications. Chem Mater. 2008;20:2455–60.
Elimelech H, Avnir D. Sodium-Silicate Route to Submicrometer Hybrid PEG@Silica Particles. Chem Mater. 2008;20:2224–7.
Sigma-Aldrich prices 2019. https://www.sigmaaldrich.com. Accessed 08. Oct 2019.
Perry SZ, Hibbert H. Studies on reactions relating to carbohydrates and polysaccharides: XLVIII. Ethylene Oxide and related compounds: Synthesis of the Polyethylene Glycols. Can J Res. 1936;14B:77–83.
Keegstra EMD, Zwikker JW, Roest MR, Jenneskens LW. A highly selective synthesis of monodisperse oligo(ethylene glycols). J Org Chem. 1992;57:6678–80.
Harada A, Li J, Kamachi M. Preparation and characterization of a polyrotaxane consisting of monodisperse Poly(ethylene glycol) and α-Cyclodextrins. J Am Chem Soc. 1994;116:3192–6.
Burns CJ, Field LD, Hashimoto K, Petteys BJ, Ridley DD, Sandanayake KRAS. A convenient synthetic route to differentially functionalized long chain polyethylene glycols. Synth Commun. 1999;29:2337–47.
Loiseau FA, Hii KK, Hill AM. Multigram synthesis of well-defined extended bifunctional polyethylene glycol (PEG) Chains. J Org Chem. 2004;69:639–47.
Ahmed SA, Tanaka M. Synthesis of oligo(ethylene glycol) toward 44-mer. J Org Chem. 2006;71:9884–6.
Niculescu-Duvaz D, Getaz J, Springer CJ. Long functionalized poly(ethylene glycol)s of defined molecular weight: synthesis and application in solid-phase synthesis of conjugates. Bioconj Chem. 2008;19:973–81.
French AC, Thompson AL, Davis BG. High purity discrete PEG oligomer crystals allow structural insight. Angew Chem Int Ed. 2009;48:1248–52.
Székely G, Schaepertoens M, Gaffney RRJ, Livingston AG. Iterative synthesis of monodisperse PEG homostars and linear heterobifunctional PEG. Polym Chem. 2014;5:694–7.
Li Y, Guo Q, Li X-F, Zhang H, Yu F-H, Yu W-J, Xia G-Q, Fu M-Y, Yang Z-G, Jiang Z-X. Fluorous synthesis of mono-dispersed poly(ethylene glycols). Tetrahedron Lett. 2014;55:2110–3.
Maranski K, Andreev YG, Bruce PG. Synthesis of Poly(ethylene oxide) approaching monodispersity. Angew Chem Int Ed. 2014;53:6411–3.
Zhang Q, Ren H, Baker GL. A practical and scalable process to selectively monofunctionalized water-soluble α,ω-diols. Tetrahedron Lett. 2014;55:3384–6.
Zhang H, Li X, Shi Q, Li Y, Xia G, Chen L, Yang Z, Jiang Z-X. Highly efficient synthesis of monodisperse Poly(ethylene glycols) and derivatives through macrocyclization of oligo(ethylene glycols). Angew Chem Int Ed. 2015;54:3763–7.
Wawro AM, Muraoka T, Kato M, Kinbara K. Multigram chromatography-free synthesis of octa(ethylene glycol) p-toluenesulfonate. Org Chem Front. 2016;3:1524–34.
Khanal A, Fang S. Solid phase stepwise synthesis of polyethylene glycols. Chem Eur J. 2017;23:15133–42.
Fordyce R, Lovell EL, Hibbert H. Studies on reactions relating to carbohydrates and polysaccharides. LVI. The synthesis of the higher polyoxyethylene glycols. J Am Chem Soc. 1939;61:1905–10.
Székely G, Schaepertoens M, Gaffney PRJ, Livingston AG. Beyond PEG2000: Synthesis and functionalisation of monodisperse PEGylated homostars and clickable bivalent polyethyleneglycols. Chem Eur J. 2014;20:10038–51.
Williamson A. Theory of aetherification. Philos Mag. 1850;37:350–6.
Boden N, Bushby RJ, Clarkson S, Evans SD, Knowles PF, Marsh A. The design and synthesis of simple molecular tethers for binding biomembranes to a gold surface. Tetrahedron. 1997;53:10939–52.
Reed NN, Janda KD. A one-step synthesis of monoprotected polyethylene glycol ethers. J Org Chem. 2000;65:5843–5.
Zhang J, Zhao Y-J, Su Z-G, Ma G-H. Synthesis of monomethoxy poly(ethylene glycol) without diol poly(ethylene glycol). J Appl Polym Sci. 2007;105:3782–6.
Bouzide A, Sauvé G. Silver(I) oxide mediated highly selective monotosylation of symmetrical diols. application to the synthesis of polysubstituted cyclic ethers. Org Lett. 2002;4:2329–32.
Coudert G, Mpassi M, Guillaumet G, Selve C. A Novel, Unequivocal synthesis of polyethylene glycols. Synth Commun. 1986;16:19–26.
Wawro AM, Muraoka T, Kinbara K. Chromatography-free synthesis of monodisperse oligo(ethylene glycol) mono-p-toluenesulfonates and quantitative analysis of oligomer purity. Polym Chem. 2016;7:2389–94.
Wan Z, Li Y, Bo S, Gao M, Wang X, Zeng K, Tao X, Li X, Yang Z, Jiang Z-X. Amide bond-containing monodisperse polyethylene glycols beyond 10 000 Da. Org Biomol Chem. 2016;14:7912–9.
Xia G, Li Y, Yang Z, Jiang Z-X. Development of a scalable process for α-Amino-ω-methoxyl-dodecaethylene glycol. Org Process Res Dev. 2015;19:1769–73.
Li Y, Thapa B, Zhang H, Li X, Yu F, Jeong E-K, Yang Z, Jiang Z-X. Synthesis of gemini surfactants with twelve symmetric fluorine atoms and one singlet 19F MR signal as novel 19F MRI agents. Tetrahedron. 2013;69:9586–90.
This work was supported by the DFG within the framework of the collaborative research centre 1176 (SFB 1176, project C3). The authors would like to acknowledge Peter Gödtel, Maximilian Knab, Rebecca Seim, and Fabienne Urbanek for synthetic support; the analytical team from KIT for analytical support; and Prof. Barner-Kowollik and his group for access to SEC-ESI-MS equipment.
Conflict of interest
The authors declare no conflict of interest.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
About this article
Cite this article
Bohn, P., Meier, M.A.R. Uniform poly(ethylene glycol): a comparative study. Polym J 52, 165–178 (2020). https://doi.org/10.1038/s41428-019-0277-1