Introduction

Biopolymers encompass a plethora of diverse materials whose chemistry and uses have occupied considerable research attention over the past ~ 70 years. IUPAC defines biopolymers as “macromolecules (including proteins, nucleic acids and polysaccharides) formed by living organisms” [1]. This does leave some scope for inclusion of petroleum-based polymers, derived from the catagenesis of organic material. One could also include bio-derived polymers such as polylactic acid or vegetable oil based polymers. However, as a general definition, this review will be concerned with biopolymers in the sense of those materials which are biosynthesized in the cells of living organisms either through natural or man-made processes. Broadly, biopolymers of interest to industrial and academic materials researchers can be divided into four categories, based on their primary structure: polysaccharides (e.g., starch, [2,3,4,5] cellulose, [6,7,8,9,10,11] alginate, [12, 13] chitin [14,15,16,17,18] and its derivative chitosan, [19,20,21,22,23,24,25] Fig. 1a), structural proteins (collagen, [26,27,28,29] fibroin, [30,31,32,33] zein, [34,35,36,37,38] Fig. 1b), heterogenous structural polymers in plants (lignin, [39,40,41,42] Fig. 1c) and polyisoprenes (rubber, [43, 44] Fig. 1d) all occur naturally. Several polyesters (polycaprolactone, [45,46,47,48,49] polylactic acid [50,51,52,53,54,55,56] and polyhydroxyalkanoates, [57,58,59,60,61,62,63,64,65] (Fig. 1e) occupy key roles in biomedical science and can be obtained via biosynthesis from various microorganisms, earning them the descriptor of biopolymers despite not occurring naturally.

Fig. 1
figure 1

Representative examples from 5 classes of biopolymer, including a polysaccharides, b proteins – note, fibroin and other structural polymers are typically non-repeating; most common fibroin sub-unit shown, c lignin – lignins are highly heterogeneous

Full size image

Biopolymers have been exploited in several roles (Fig. 2); their origins render biopolymers attractive “green” alternatives to polymer materials ultimately derived from petrochemicals, while their synthesis in living systems results in consequent increased biocompatibility in many cases. Polycaprolactone (PCL) nanoparticles used in drug delivery exhibit large variety of degradation properties, a feature which can be exploited to allow tuneable timing of drug of drug release up to several months; [66, 67] physical or chemical modification of the polymer can lengthen this timeframe to up to 1 year [68] Zhou and Medel (among other researchers) have both reported the use of PLA nanoparticles for delivery of drugs including the anti-cancer agents doxorubicin and bortezomib [69, 70]. Similarly, control of biodegradation to tune the release of antibiotics from silk fibroin scaffold containing gelatin microspheres loaded with vancomycin has been reported by Lan et al. [71] Liu and co-workers have also demonstrated the use of PLA as a drug-carrier material for paclitaxel-eluting stents [72]. Chitosan hydrogels have been explored as carriers for drug delivery of a range of therapeutics, including antibiotics, anti-cancer agents and anaesthetics. A comprehensive review of the area has recently been published [21].

Fig. 2
figure 2

Examples of the uses of biopolymers outlined in the introduction

Full size image

Biopolymers have also been modified to exploit intrinsic therapeutic activities as well as releasing small-molecule drugs. For example, at low pH the amino groups borne by chitosan result in electrostatic interaction with bacterial cell walls (which are typically negatively charged) ultimately causing cell death; [73] at higher pH values chitosan maintains bactericidal activity through other mechanisms [74, 75]. Combined with adhesive properties and permeability to oxygen, chitosan is an attractive material for burn and wound dressings [74, 76]. Inspired by this antibacterial mechanism, Ḉalabak and co-workers reported the bactericidal activity of electrospun silk fibroin/polyethylenimine nanofibers against S. aureus and P. aeruginosa [77]. Similarly, DermaSilk, silk fibre bearing the positively charged AEM 5772/5 antimicrobial is commercially available as clothing for treatment of a variety of skin conditions.

Among the polyesters, polylactic acid (PLA) has been exploited as 3D-printed implants for facilitating bone and tissue growth [78,79,80]. Subsequent dissolution of PLA in blood eliminates the need for removal or complications from long-term implantation. Similarly, the predictable biodegradation kinetics, ease of fabrication and biocompatibility of PCL make it ideal for tissue engineering scaffolds [81]. Polyhydroxyalkanoates have also received considerable medical attention; for example, poly-3-hydroxybutyrate slowly hydrolyses to (R)-3-hydroxybutyric acid, which naturally occurs in blood at concentrations of 0.3 – 0.13 mM, making it an attractive option for biodegradable drug delivery systems as well as sutures [82, 83]. Notably Tephaflex®, the first PHA to receive FDA approval is an absorbable surgical floss composed of poly-4-hydroxybutanoate which biodegrades to 4-hydroxybutyraic acid, another decomposition product which naturally occurs in the human body and is readily excreted and metabolised [84].

Depolymerisation to non-toxic sub-units already present in living systems also facilitates the use of biopolymers as food additives, of which zein is the best-known example. The food applications of zein have been recently reviewed [37]. Enjoying generally recognised as safe (GRAS) FDA classification, zein has been exploited to make edible, water-resistant films, and coatings as well as for stabilising emulsions and as the basis for vegan alternatives to meat and cheese products [85, 86].

Many biopolymers are readily biodegraded and are subsequently employed in a range of applications as biodegradable materials. The excellent dielectric and mechanical properties of PLA have led to a variety of applications in disposable electronics, e.g., inkjet-printed organic electrochemical transistors (OECTs) and organic field-effect transistors (OFETs) [87]. An analogous inkjet-printing approach had led to PLA-based thin-film transistors (TFTs) and light emitting diodes (LEDs) [88]. The ready degradation of PHAs to comparatively innocuous and non-toxic products (vide supra) has led to interest in use for both rigid and flexible disposable containers [89, 90]. Increasing awareness of microplastic pollution has led to the development of PHA microbeads for use in cosmetics and skincare products by Bio-on SpA, an Italian-based company.

As renewable resources, biopolymers are also attractive sources of chemical feedstocks. Though possessed of an intrinsically disordered structure, biosynthesis of lignin starts from 3 phenolic precursors (sinapyl alcohol, coniferyl alcohol and p-coumaryl alcohol [40]) and valorisation of lignin to diverse phenolic products prior to further functionalisation or reduction allows access to valuable feedstocks for synthesis otherwise derived from petrochemicals. As the second most abundant biopolymer on Earth, [91] considerable research effort has focussed on lignin as a source of substrates for the future green chemical industry. Starch has also been employed as a source of diverse feedstocks on an industrial scale (Scheme 1) – enzymatic production of dihydrogen is well known, [92] as is production of 5-hydroxymethylfurfural [93]. Additionally, levulinic acid, [94] lactic acid [95] and acetone-butanol-ethanol (ABE) mix [96] may be obtained from starch in bulk. N-Heterocycles, including oxazolines, pyrroles, pyridines and indoles commonly used as pharmaceutical synthesis substrates may also be obtained from reaction of starch in the presence of a nitrogen source [97, 98].

Scheme 1
scheme 1

Chemical feedstocks from reactions of starch

Full size image

Continuous reaction monitoring

In terms of literature, the research into synthetic polymers far outweighs that of biopolymers (Fig. 3). This abundance focussing on synthetic polymers can be used as a valuable resource to highlight areas of polymer analysis to accelerate progress towards more sustainable materials. This is true with regards to the growing demand for continuous analysis of polymers, which can provide several advantages over traditional batch analysis such as automation, continuous reaction optimisation and the potential for the application of machine learning. Herein we will provide a brief representative summary of the current state of the art of continuous analysis to highlight potential applications to the field of biopolymers.

Fig. 3
figure 3

Comparison of research output between synthetic and biopolymers from Web of Science Core Collection data base from 1931 to 2022

Full size image

Defining reaction monitoring: inline, online, atline and offline

When describing methods of continuous reaction monitoring, it is common to use a qualifying prefix to denote the type of reaction monitoring taking place. These prefixes are described in a visual aid in Fig. 4 to show their relationships with regards to continuous reaction monitoring. Highlighting the differences between each type of continuous reaction monitoring is crucial to demonstrate their respective advantages and disadvantages. This allows for selecting the most appropriate method for a given set of reaction conditions, while also identifying the limitations of others. With particularly reference to the field of polymer analysis, several reviews have succeeded in differentiating each type of continuous reaction monitoring [99,100,101,102,103,104].

Fig. 4
figure 4

Graphic summary of definitions for online, inline, atline and offline analysis as relates to a continuous flow process

Full size image

In 2020 a review of enabling technologies in polymer synthesis by Knox et al., the authors state the defining feature of ‘online’ and ‘inline’ analysis as that which is performed on a chemical process in-situ, whereas offline and atline and offline processes refer to measurements taken either by manual or automated sampling and then subsequent analysis ex-situ [101]. This statement echoes an earlier definition laid out by Dagge et al.: “Online and inline analyses differ essentially from the offline and atline methods in that the time in which information about process or material properties is obtained is shorter than the time in which these properties change” [105]. These definitions separate the ‘real-time’ analysis achieved by online/inline processes from the delayed analysis of atline and offline methods. However, a clear, standardised definition between the types of analysis in each category is often lacking. For inline/online analysis, the definitive separation between the two methods must lie in the degree of disconnection from the initial process conditions. Inline analysis is a continuous process for measuring a reaction stream without changing any process conditions. For example, inline analysis of reaction temperature can be simply and immediately measured by a temperature probe directly included in the reaction stream in a continuous, closed loop manner. Online analysis also allows for continuous processing, but fundamentally, the analysis does not take place directly in the reaction stream. Instead, a branch or bypass is included in the reaction stream through which substances from the reaction stream are automatically fed. This is done, for example in cases where the conditions of the reaction stream e.g., pressure, temperature or concentration are outside of the operating parameter of the analytical instrument. Atline or offline measurements are not generally fed back into the reaction stream after analysis, either because a significant period elapses during the analysis or the aliquot is processed in a way as to make it impossible to reintroduce to reaction stream without significantly changing the outcome of a reaction. An atline measurement is one where a sample is taken from the reaction stream and analysed with minimal sample processing but is not then returned to the reaction stream. In contrast, an offline measurement also takes a sample from the reaction stream, but a significant level of sample processing is required before the relevant analysis can take place. This sample processing also prevents the sample being returned into the reaction stream.

Each type of reaction monitoring described above has its own sets of advantages and limitations. Online or inline analysis allow for close to real-time analysis of reaction progress or conditions with continuous data logging. However, there may be circumstances where these types of analysis may not be achievable or desirable, when atline or offline analysis may suffice. Regardless of the mode of analysis, the application of continuous flow to a chemical system gives the widest range of possibilities for reaction monitoring. It is important to note that an effort has been made here to provide a base definition of the four types of continuous analysis. It is true however, that previous publications may not have used the same definitions and a caveat has been added in the following literature examples where the description of the type of analysis contradicts what we have stated above.

In polymer chemistry, the use of flow technology is a burgeoning field with many advantages over traditional batch processes [106,107,108]. There are myriad analysis techniques which have been successfully adapted for continuous analysis of polymer systems including differential scanning calorimetry in conjunction with Fourier-transform infrared spectroscopy [109] and inline mass spectrometry [110,111,112]. Whilst these techniques clearly show promise for polymer synthesis, for biopolymers and processes which are concerned with the environmental impact of polymeric materials, molecular weight is a key parameter to analyse in a continuous fashion. Two techniques which show great promise regarding continuous monitoring of monomer conversion and molecular weight are nuclear magnetic resonance (NMR) and size exclusion chromatography (SEC).

Nuclear magnetic resonance (NMR) spectroscopy

Nuclear magnetic resonance (NMR) spectroscopy as a means of monitoring a chemical reaction such as a polymerisation is a standard practise in research. However, the expense associated with integrating flow-chemistry systems to conventional NMR instruments has limited its application to dedicated research facilities [113, 114].

As an initial example of the suitability of NMR in polymer synthesis without flow technology, Fillbrook et al. utilised time-resolved diffusion ordered nuclear magnetic resonance (DOSY) to study the kinetics and molecular weight evolution during a photoinduced electron transfer reversible addition fragmentation chain transfer (PET-RAFT) polymerisation of methyl acrylate (MA) activated by zinc tetraphenylporphyrin [115]. The system provided a good estimate of polymer molecular weight when compared to batch size exclusion chromatography (SEC) analysis, with the important advantage of continuous monitoring during the reaction.

An example of adaptation into continuous flow is seen in Vrijsen et al. who used high resolution flow NMR spectroscopy coupled to a continuous flow reactor to monitor monomer conversion and molecular weight distributions of a methyl acrylate reversible addition fragmentation chain transfer (RAFT) polymerisation (Scheme 2). Using DOSY NMR the group were able to monitor decreasing diffusion coefficients of the translational movement of polymer chains during the continuous flow reaction and their analysis determined that measurement of polymer molecular weight by this process was more accurate than tradition SEC analysis, with relative errors around 5% as opposed to the 20% expected by the IUPAC working party on kinetics and mechanisms in radical polymerisations [116, 117].

Scheme 2
scheme 2

Schematic of the experimental setup used by Vrijsen et al. showing a scematic for the continuous flow RAFT polymerisation of methyl acrylate using inline DOSY 1H NMR to determine evolving polymer molecular weight [118]. Reproduced with permission from the publisher

Full size image

The development of lower-field ‘benchtop’ NMR instruments has opened the door for accessible online monitoring of reactions using NMR, [119,120,121] given its smaller size, lower cost and low maintenance requirements when compared to conventional NMR [122]. Knox et al. provided the first example in use of a benchtop NMR spectrometer to monitor both controlled and free radical polymerisations in batch and continuous flow [123]. The group were able to monitor the synthesis of homopolymers and diblock copolymer nanoparticles using the system, noting that they were able to achieve a correlation in reaction kinetics for a methyl methacrylate polymerisation monitored in batch and continuous flow.

As mentioned previously, reactions in continuous flow are uniquely accessible to automation and the pairing with benchtop NMR provides an ideal method for feedback loop control in polymer synthesis. Rubens et al. designed a process for the thermally initiated RAFT polymerisation of methyl acrylate [120]. The group’s goal in using the inline NMR system was to generate a highly accurate means of targeting a specific monomer conversion, which could be pre-programmed prior to starting the reaction. This was achieved by means of a benchtop NMR spectrometer coupled to the outlet of the flow reactor measuring spectra every 15 s, following monomer conversion with high time-resolution. By performing the reaction in four short-range time sweeps whilst monitoring monomer conversion during the reaction, the group were able to match the % monomer conversion of batch polymerisations, quenched at random time intervals with an error of ±3%. The group then were able to construct an algorithm to run consecutive time-sweeps of polymerisations at varying temperatures which allowed for minimal operator contact and the ability to pre-set monomer conversion after the algorithm had been trained. As stated by the group, although the example is given for RAFT polymerisation, there is a wide scope of other polymerisations, to which this system would be generally applicable.

Another powerful demonstration of autonomous optimisation using benchtop NMR is given by Knox et al. who presented a closed-loop (i.e., user-free) method of RAFT polymerisation incorporating inline benchtop NMR and atline size exclusion chromatography (SEC, referred to as gel permeation chromatography, GPC by the authors) (Scheme 3) [121]. The synthesis platform comprised a computer-controlled flow reactor, inline benchtop NMR and atline SEC. The RAFT polymerisation of tert-butyl acrylamide (tBuAm), n-butyl acrylate (BuA) and methyl methacrylate (MMA) were optimised using the Thompson sampling efficient multi-objective optimisation (TSEMO) algorithm which explored the trade-off between molar mass dispersity and monomer conversion without user interaction [121]. Again, the scope in applying this system to other polymerisations or polymer-degradation studies has great potential.

Scheme 3
scheme 3

(a) Generalised scheme for the RAFT synthesis of P(tBuAm)200. Example (b) gel permeation chromatography (GPC) chromatograms and (c)1H nuclear magnetic resonance (NMR) spectra from the automated continuous flow platform used in this work. (d) Schematic of that automated platform and (e) an overview of the structure of the machine learning directed (Thompson-sampling efficient multi-objective optimisation (TSEMO) algorithm) experiments used in this work [121]. Reproduced with permission from the publisher

Full size image

Size exclusion chromatography

Size exclusion chromatography (SEC, also referred to interchangeably as gel permeation chromatography, GPC) relies inherently on a time resolved process in its methodology, making genuine real-time analysis during a reaction a difficult prospect [101]. However, it can still provide useful information about the course of a polymerisation or depolymerisation by providing accurate measurements of polymer molecular weight at pre-programmed time intervals as a tool to measure reaction kinetics [124]. The use of so called, ‘rapid’ chromatography columns have also enabled a reduction of the interval between samples to just a few minutes, at some cost to spectral resolution [125,126,127,128]. Initial pioneering work in the use of rapid SEC columns came from Hadziioannou and co-workers with their continuous online rapid size exclusion chromatography monitored polymerisation (CORSEMP) system [125]. This self-contained system allows for close to real-time analysis of polymerisations using SEC with a sample interval of 12 min. In their original 2007 paper, the group used polymerisations of polystyrene and block copolymerisations of butyl acrylate and styrene to showcase how the CORSEMP system was able to track the progress of each polymersation. This gave detailed information of reaction progress by monitoring molecular weight and molecular weight distribution. In a further example, Haddleton and co-workers used rapid online SEC analysis to monitor Cu(0) catalysed polymerisations of methyl acrylate in toluene. This provided a viable means obtaining to obtain data on conversion and molecular weight distribution in real-time compared to conventional analytical techniques [127].

SEC is also not limited to measurements of polymer molecular weight but can also provide useful information about polymer architecture. For instance, Sherrington and co-workers used SEC analysis to distinguish between linear and branched polymers by exploiting changes in their hydrodynamic radius [129]. Several copolymers of methyl methacrylate and ethyl acrylate were formed with the same reaction composition, save for the inclusion of a branching comonomer, ethylene glycol diacrylate (EDGA) to form a branched copolymer. This would ostensibly produce copolymers with identical molecular weight. However, as Fig. 5 shows, the branched copolymer, B2 shows a higher molecular weight at any given elution volume than the linear copolymer, L9. A further increase in the degree of branching resulted in a further shift towards the top right corner of the molar mass vs. elution volume graph.

Fig. 5
figure 5

Graph of molar mass vs. elution volume for linear copolymer L9 and branched copolymer B2. The two copolymerisations were conducted with identical reaction compositions, save for the inclusion of the branching comonomer EGDA for B2. The presence of branching resulted in a shift to higher molecular weight at each elution volume [129]. Reproduced with permission from the publisher

Full size image

One of the first online examples of real-time analysis of a polymerisation using SEC was given by Rubens et al. with their autonomous platform for polymer synthesis (Fig. 6) [120]. The group were able to implement an autonomous synthesis platform for the RAFT polymerisation of n-butyl acrylate to achieve predefined molecular weights with less than 2.5% deviation from the goal molecular weight. The platform was designed around modular flow reactors whereby the flowrates of the individual reactant streams are controlled through pumps and the final product stream is analysed in real-time with online SEC. A key point is made by the group that the system is not limited to RAFT polymerisations but should also apply to any system which is similar to RAFT in terms of molecular weight evolution and dependency of chain length on monomer concentration (i.e., anionic, cationic, or other reversible deactivation radical polymerizations).

Fig. 6
figure 6

The Experimental set up used by Rubens et al. The product stream is injected into an online SEC system and subsequently analysed and processed by the optimization algorithm [120]. Reproduced with permission from the publisher

Full size image

As detailed previously, the group of Knox et al. Integrated both online benchtop NMR and SEC to their autonomous optimisation of RAFT polymerisation system [121]. The conjunction of these two techniques provided a multi-objective route to optimisation in measuring both degree of polymerisation by benchtop NMR and the evolution of molecular weight by online SEC.

There remains a great potential to integrate SEC into both polymer synthesis and polymer degradation studies. Exploiting the advantages of atline analysis, continuous SEC can be readily applied to an experiment initially relying on batch processing. For example, almost 20 years ago Osanai et al. developed a system for the enzymatic degradation of biodegradable polyesters in both batch and continuous flow using an enzyme packed column (Scheme 4) [130].

Scheme 4
scheme 4

Conceptual scheme of the continuous degradation equipment using the enzyme column by Osanai et al.. [130] (a) Batch injection through loop injector. (b) Continuous flow injection, reproduced with permission from the publishers

Full size image

The degradation of typical biodegradable polyesters, poly butylene adipate (PBA) and poly(ε-caprolactone) (PCL) was achieved to yield polymerizable cyclic oligomers by using immobilised lipase from Candida antarctica at 40℃ in a toluene solution (Fig. 7). Currently there is a shift in research focus towards maximising the useful lifespan of polymers or providing effective means of recycling [131,132,133]. This highlights the great importance of developing systems for monitoring the effectiveness of polymer degradation. Given the solvent and temperature stability of the enzymes, the process described by Osanai et al. only be connected to an automated valve and SEC analysis equipment to achieve online monitoring. To demonstrate this, we have modified the experiment given by Osanai to include a simple method of continuous atline monitoring of polymer molecular weight by SEC, in the next section.

Fig. 7
figure 7

Left: SEC profiles of PBA degradation products by continuous flow injection method. 1.0 wt.% PBA in toluene was continuously injected into the enzyme column at a flow rate of 0.2 mL min−1 at 40℃. Right: SEC profiles of the degradation products of PCL by continuous flow injection method. 0.25 wt.% PCL in toluene was continuously injected into the enzyme column at a flow rate of 0.2 mL min−1 at 40℃ [130]. Reproduced with permission from the publisher

Full size image

A perspective system for atline SEC analysis

Herein we describe an example of a simple method for integrating SEC with commercially available flow reactors for continuous atline analysis. The method relies on an automated 6-port valve which can be programmed to change between a default position and an active position at programmed time intervals, allowing for direct automated sampling of a reaction stream as shown in Scheme 5.

Scheme 5
scheme 5

A perspective system for atline SEC using a commercial flow reactor connected to an SEC instrument via a programable six-port valve fitted with a 20μL sample loop. The valve’s two position are shown in the dashed highlight bubble: the default and active positions

Full size image

In brief, the reaction stream is continuously fed through a sample loop on the automated valve and in the default state, this is recirculated. Also, whilst in the default position, eluent is pumped by the SEC system through an input of the automated valve and directly out again into the SEC eluent stream. At a pre-programed time, the automated valve changes to the active position for a set time; sufficient to flush the sample loop with SEC eluent before changing back to the default position. Thus, an aliquot of the reaction stream is diverted to SEC analysis. This process is repeated at time intervals which allow for collection of a complete chromatogram within calibration limits before another aliquot is taken. As the reaction stream is not recycled after SEC analysis, according to the definitions given above, this method of analysis would be considered ‘atline’. However, a reclassification to ‘online’ analysis can be achieved simply by plumbing the waste stream back into the reaction stream. In the setup described the analyte in the reaction stream must remain at a concentration that is within the operating parameters of the SEC instrument but keeping pressure and temperature constant is not required as the automated valve provides total separation between the reaction and eluent streams. As a proof of principle study, the system described by Osanai et al. was adapted to incorporate the continuous analysis described above [130]. A solution of PCL in toluene was continuously flowed through a fluidised bed of candida antarctica lipase B, immobilised on acrylic resin at 1 mL min−1 for 12 h at 40℃ with a 30-min sample window (Scheme 5). The SEC chromatograms were then processed into a single plot shown in Fig. 8A. Figure 8B compares the initial chromatogram of PCL, before being introduced to the fluidised bed, with the final chromatogram taken after 12 h.

Fig. 8
figure 8

A Chromatograms from 12 h of sampling at 30 min time intervals, showing the gradual degradation of PCL at ~ 10.5 min and gradual increase of an oligomer peak at ~ 19 min, outside of the calibration limits. B a comparison of the first sample taken of PCL in comparison to the final sample after 12 h showing complete degradation of the polymer to oligomers

Full size image

The results of the analysis show that the original PCL peak reduces from a molecular weight of approximately 75,000 g mol−1 and affords total degradation with the formation of oligomers which were outside the limits of our calibration. In the original source paper, a reaction time of just 2 h was required to afford the complete degradation of the polymer. We chose to alter the reaction conditions, moving from a fixed bed to a fluidised bed of immobilised enzymes and using an increased flow rate to reduce retention time within the fluidised bed to slow the degradation and hence increase the resolution of the SEC analysis. Coupled with a rapid analysis SEC column, as discussed earlier to allow for a decreased analysis widow, this system would allow optimisation of the reaction conditions, whilst still collecting useful information on the kinetics of polymer degradation.

Conclusion

To conclude, we have described the current state of the art in online analysis of polymer systems with a focus on NMR and SEC in continuous flow. Although the examples reviewed have focussed on non-biobased polymers, they do describe the methods and hardware which would be required to adapt these systems to studies of polymer biodegradation, leeching, changes in chain architecture and optimisation potential for biopolymer synthesis. We have also described a simple system for online SEC analysis and shown its potential as a method to measure the rate of biodegradation of PCL. With advances in flow technology, we can expect to see more sophisticated monitoring techniques that can provide greater accuracy and real-time data analysis. This will enable researchers and manufacturers to make more informed decisions about their materials, leading to better quality control and improved product performance. Allowing for abundant data collection through continuous analysis also paves the way towards integration of machine learning algorithms and artificial intelligence to enhance material generation. Furthermore, the ability to monitor polymers continuously will play an important role in the development of more sustainable and environmentally friendly materials.

Materials

Lipase B Candida antarctica immobilized on Immobead 150, recombinant from Aspergillus oryzae was purchased from Sigma Aldrich and was soaked in toluene for 24 h at 40℃ to remove any soluble material before filtering and drying under vacuum. Polycaprolactone (PCL) 80,000 Da was purchased from Sigma Aldrich and used without further purification. The continuous flow system was achieved using a Vapourtec E-series flow chemistry system. Aliquots of the reaction mixture were taken at programmed intervals using an FCV-32AH 2-position 6-port high pressure flow switching valve, kindly supplied by Shimadzu. SEC samples were analysed using a Shimadzu High Performance Liquid Chromatograph fitted with a 7.5 mm internal diameter Agilent SEC column. The detector used was a Shimadzu SPD-20A UV–Vis detector set to 254 nm. HPLC grade tetrahydrofuran (THF, 99.8%, Acros Organics) was utilized as the eluent with flow rate of 1 mL min−1 with an oven temperature of 40℃. The measurement was calibrated against 10 polystyrene standards in the range of 162–364,000 g mol−1.

Experimental

300 mg of the immobilised enzyme resin was introduced into a 3.5 mL reactor bed and was equilibrated under a continuous flow of toluene at 40℃ at 1 mL min−1 for 20 min. A 100 mL 0.5 w/v% solution of PCL in toluene was prepared and pumped into flow system, initially bypassing the fluidised bed reactor, to displace the pure solvent. At this point the analysis began and the PCL solution was pumped continuously through the fluidised after the initial aliquot was taken through the automated valve sample loop. Samples were then collected at 30-min intervals for 12 h.