UvA-DARE (Digital Academic Repository) Degradation and ... · Proefschrift – Degradation and Analysis of Synthetic Polymeric Materials for Biomedical Applications by Abdul Ghaffar

UvA-DARE is a service provided by the library of the University of Amsterdam (http://dare.uva.nl)

UvA-DARE (Digital Academic Repository)

Degradation and analysis of synthetic polymeric materials for biomedical applications

Ghaffar, A.

Link to publication

Citation for published version (APA):Ghaffar, A. (2011). Degradation and analysis of synthetic polymeric materials for biomedical applications.

General rightsIt is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s),other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).

Disclaimer/Complaints regulationsIf you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, statingyour reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Askthe Library: https://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam,The Netherlands. You will be contacted as soon as possible.

Download date: 08 Jul 2020

https://dare.uva.nl/personal/pure/en/publications/degradation-and-analysis-of-synthetic-polymeric-materials-for-biomedical-applications(c665b3ef-7644-4e3c-ac73-139ae4fd4610).html

Degradation and Analysis of Synthetic Polymeric Materials for Biomedical Applications

Abdul Ghaffar

The front cover image (courtesy of Bozhi Tian) reprinted with permission (published in T.

Dvir et al., Nature Nanotechnology 6, 13-22 (2011). The foreground of this image represents

polymeric fibres (purple). The background of the image shows a scanning electron

micrograph of an electrospun polymeric fibre mesh. The cells are shown in light blue.

Proefschrift – Degradation and Analysis of Synthetic Polymeric Materials for Biomedical Applications by Abdul Ghaffar ISBN: 978-90-5776-231-4 90-5776-231-5

This research was conducted at Analytical Chemistry Group (former Polymer-Analysis Group), Van’t Hoff Institute for Molecular Sciences, FNWI, University of Amsterdam.

and was supported by

Higher Education Commission of Pakistan

Degradation and Analysis of Synthetic Polymeric Materials for Biomedical Applications

ACADEMISCH PROEFSCHRIFT

ter verkrijging van de graad van doctor

aan de Universiteit van Amsterdam

op gezag van de Rector Magnificus

prof. dr. D. C. van den Boom

ten overstaan van een door het college voor promoties

ingestelde commissie,

in het openbaar te verdedigen in de Agnietenkapel

op donderdag 27 oktober 2011, te 14:00 uur

door

Abdul Ghaffar

geboren te Lahore, Pakistan

Promotiecommissie:

Promotor: Prof. Dr. S. van der Wal

Co-promotor: Prof. Dr. Ir. P.J. Schoenmakers

Overige leden: Prof. Dr. Ir. W. E. Hennink

Prof. Dr. D. W. Grijpma

Prof. Dr. Ir. J. G. M. Janssen

Prof. Dr. C. G. de Koster

Dr. W. Th. Kok

Dr. A. A. Dias

Faculteit der Natuurwetenschappen, Wiskunde en Informatica

Dedicated to my beautiful children NAVERA and SARIM

Table of Contents Chapter 1 ................................................................................................................................1

1. Methods for the chemical analysis of degradable synthetic polymeric biomaterials*......1 1 Introduction .................................................................................................................2 2 Degradable biomaterials ..............................................................................................3 3 Analytical strategies.....................................................................................................5 4 Degradation methods ...................................................................................................7

4.1 Degradation under non-physiological conditions ..................................................7 4.2 Degradation under physiological conditions .........................................................8

5 Chromatographic methods for degradable polymeric biomaterials ............................10 5.1 Size-exclusion chromatography ..........................................................................11 5.2 Adsorption liquid chromatography .....................................................................14 5.3 Liquid chromatography at critical conditions......................................................17 5.4 Two-dimensional liquid chromatography ...........................................................18

6 Gas chromatography..................................................................................................20 7 Direct mass-spectrometric analyses ...........................................................................22 8 Nuclear-magnetic-resonance spectroscopy ................................................................25 9 Conclusions ...............................................................................................................26 10 Scope of the thesis ...................................................................................................27 11 References ...............................................................................................................28

Chapter 2 ..............................................................................................................................33 2. Fast in vitro hydrolytic degradation of polyester urethane acrylate biomaterials: structure elucidation, separation, and quantification of degradation products...................33

1 Introduction ...............................................................................................................34 2 Experimental..............................................................................................................36

2.1 Materials .............................................................................................................36 2.2 Procedure of hydrolysis ......................................................................................39 2.3 1H NMR spectroscopy of hydrolysate.................................................................40 2.4 Size-exclusion chromatography (SEC) analysis..................................................40 2.5 HPLC-ESI-ToF-MS analysis of hydrolysate.......................................................42

3 Results and discussion ...............................................................................................43 3.1 Optimization of hydrolysis method.....................................................................43 3.2 Product identification..........................................................................................44 3.3 Molar mass characterization and quantification of PMAA .................................48 3.4 Quantification of monomeric products by HPLC-ToF-MS.................................50

4 Conclusions ...............................................................................................................56 5 References .................................................................................................................57

Chapter 3 ..............................................................................................................................59 3. Monitoring the in vitro enzyme-mediated degradation of degradable poly(ester amide) for controlled drug delivery by LC-ToF-MS ....................................................................59

1 Introduction ...............................................................................................................60 2 Materials and methods ...............................................................................................62

2.1 Materials .............................................................................................................62 2.2 Solubility ............................................................................................................62 2.3 Enzyme activity ..................................................................................................62 2.4 In vitro enzyme-mediated degradation................................................................63 2.5 Size-exclusion chromatography ..........................................................................64 2.6 LC-ToF-MS study...............................................................................................65

1

3 Results and discussion ...............................................................................................66 3.1 Solubility ............................................................................................................66 3.2 Overall effectiveness of in vitro enzyme-mediated degradation.........................66 3.3 Molecular-weight of remaining material.............................................................68 3.4 LC-ToF-MS analysis following enzymatic degradation......................................68 3.5 Factors affecting enzyme activities .....................................................................71

4 Conclusions ...............................................................................................................75 5 References .................................................................................................................76 6 Supporting information..............................................................................................77

6.1 Two-dimensional H,C-correlated spectrum (HSQC) of PEA..............................77 6.2 ESI-ToF-MS spectra of the identified peaks – enzymatic degradation................77 6.3 Chemical Degradation – optimization of the LC-ToF-MS method .....................81

Chapter 4 ..............................................................................................................................85 4. A versatile system for studying the enzymatic degradation of multi-block poly(ester amide)s*............................................................................................................................85

1 Introduction ...............................................................................................................86 2 Experimental..............................................................................................................88

2.1 Materials .............................................................................................................88 2.2 Dynamic coating of stainless-steel capillaries.....................................................89 2.3 On-line LC-ToF-MS analysis .............................................................................90

3 Results and discussion ...............................................................................................92 3.1 Continuous-feed mode ........................................................................................92 3.2 Pulse-feed mode ..................................................................................................98 3.3 Comparison of pulse-feed mode and continuous-feed mode..............................101 3.4 Application to tri-block PEA coatings ..............................................................102

4 Conclusions .............................................................................................................104 5 References ...............................................................................................................105 6 Supporting information............................................................................................106

6.1 Solubility ..........................................................................................................106 6.2 Molecular weight (Mw) and dispersity of tri-block PEA ...................................106 6.3 NMR experiments.............................................................................................108 6.4 ESI-ToF-MS spectrum of α-chymotrypsin........................................................110

Summary ........................................................................................................................111 Samenvatting ..................................................................................................................113 Acknowledgments ..........................................................................................................115 Bibliography...................................................................................................................118

*A. Ghaffar, P. J. Schoenmakers, Sj. Van der Wal, to be submitted.

Chapter 1

1. Methods for the chemical analysis of degradable synthetic

polymeric biomaterials*

The performance of biodegradable polymeric systems strongly depends on their physical, as

well as on their chemical properties. Therefore, the detailed chemical analysis of such

systems is essential. Enzymatic and chemical hydrolysis are the primary biodegradation

mechanisms for these materials. This review provides an overview of the strategies and

analytical methods used for the structural and compositional chemical analysis of non-

degraded, partially degraded and fully degraded synthetic polymeric biomaterials with an

emphasis on modern solution-based techniques that yield large amounts of information. The

degradation methods that facilitate the study of polymeric networks are also described.

Chapter 1

2

1 Introduction

A biomaterial is a substance that has been engineered to take a form which, alone or as part

of a complex system, is used to direct, by control of interactions with components of living

systems, the course of any therapeutic or diagnostic procedure, in human or veterinary

medicine [1]. Synthetic polymeric biomaterials are of great importance in the medical field

due to an aging population and because of their potential to improve the quality of life [2].

There is a clear trend to replace non-degradable by degradable materials [3]. Biodegradable

polymeric implants are intended to degrade gradually and their degradation products are

meant to be excreted benignly by the body, so that they do not need to be surgically removed

after their functional role (e.g. as drug-delivery carrier) has expired [4]. This causes

increasingly strict demands on the design and synthesis of biodegradable polymeric

materials for applications in drug-delivery devices, gene transfer, regenerative medicine,

scaffolds for tissue engineering, and surgical implants, such as rods, sutures, pins and screws

for fixation devices [5,6]. Degradation of biomaterials has many biological, physical, and

chemical facets. Biological assessment involves cell tests or implantation. Morphology is

important to understand degradation behaviour. For example, crystallinity plays a crucial

role in the degradation of poly(lactic acid) [7].

This review is limited to the in vitro chemical analysis of synthetic polymeric biomaterials.

The suitability of synthetic polymeric biomaterials for medical devices can be inferred from

their chemical structure, mechanical properties, degradation kinetics, and the

biocompatibility (tissue response) of the polymers and their degradation products [8]. The

molecular weight, hydrophilic or hydrophobic nature, fractional composition sequence and

(stereo-) regularity of the monomers in multi-block co-polymers, length of kinetic chains in

photo-polymerized networks, nature and concentration of additives, shape and morphology

of the specimen, and incubating media can all influence the degradation rate and mechanism

in terms of surface erosion or bulk degradation. The biodegradation mechanism for such

materials primarily involves enzymatic and chemical hydrolysis. Highly reactive species,

such as peroxides, are produced in reaction of the human body to the biomaterial (foreign-

body response). Such species may also degrade the polymer chain and contribute to the

overall degradation of biomaterial [9].

Various reviews have been published on the synthesis and application of synthetic

biomaterials. However, despite their increasing use in the biomedical industry, very few

articles have reviewed selective characterization techniques. No review has been published

Methods for the chemical analysis of degradable synthetic polymeric biomaterials

3

that summarizes in detail the analysis methods that lead to the identification and structural

analysis of degradable polymeric biomaterials. Therefore, we set out to review different

analytical methods used for the analysis of degradable polymeric biomaterials as a starting

material, after partial hydrolysis under physiological conditions, and after complete

hydrolysis. Methods involving chromatographic separation followed by spectroscopic or

mass-spectrometric detection are discussed here. Degradation methods required to bring the

complex copolymer and insoluble networks within the realm of chromatographic techniques

and direct mass-spectrometric analysis are also emphasized.

2 Degradable biomaterials

Synthetic degradable polymeric biomaterials contain one or more functional groups, such as

an ester, ether, amide, imide, thioester, anhydride, etc., in their chemical structure. This

enables such materials to degrade gradually, either through chemical stress or through

biological processes. The polymeric chains in a degradable polymer can differ in terms of

their length, chemical structure, architecture, etc. On the basis of their chemical composition,

they can be divided in homopolymers and copolymers. The sequence of the different

monomers in polymeric chains further differentiates copolymers into block copolymers,

alternating copolymers, random copolymers, graft copolymer, etc. The architecture of

polymer molecules can be linear, branched, hyperbranched, or dendrimers. Polymers may

also form three-dimensional chemically or physically cross-linked network. The arrangement

of different fragments in polymeric chains not only determines their configuration (stereo-

heterogeneity, such as isotactic, syndiotactic and actactic), but also their ability to rotate

around a single bond (so-called conformational heterogeneity). All these parameters may

directly (e.g. through chemical stability) or indirectly (e.g. through the crystallinity)

influence the rate of degradation of biomaterials. Hence, all of them need to be investigated.

Poly(2-hydroxyethyl methacrylate) (pHEMA), poly(glycolic acid) (PGA), poly(lactic acid)

(PLA), poly (lactide-co-glycolide) (PLGA), polycaprolactone (PCL), and functionalized

cross-linked polyacrylates are the most extensively studied polyesters for biomaterials [10].

Polyurethanes have been investigated in the biomedical industry and their properties have

been tailored by incorporation of ester and ether components to generate poly(ester

urethane)s or poly(ether urethane)s. Poly(ester amides)s (PEAs), prefereably with natural

amino acids, are attractive for biomedical applications such as drug-eluting stent coating

[11]. Ulery et al. recently published a comprehensive review describing in detail the

Chapter 1

4

biomedical applications of synthetic and natural biomaterials [12]. A few examples of

different functionalities of synthetic polymeric biomaterials are tabulated in Table 1.

Table 1 Selected synthetic biodegradable polymers and copolymers.

Polymer types

Structure Reference

Polyester

[13]

Polyether

[14]

Polyamide

[15]

Polyimide

[16]

Polyurea

NH

NH

O

R2R1NH

NH

O

n

[17]

Polyurethane

[18]

Polyanhydride

[19]

Polythioester

[20]

Polyphosphoesters

[21]

Polysiloxane

Si

R1

O

R2n

[22]

Poly (ester amide)

NH

R2

OR1O

O

n

[11,23]

Poly (ester urethane)

NH

O

O

R2R1O

O

n

[10]

Poly (ester urea)

[24]


5

Poly (ester ether)

[14,25]

Poly (ether urethane)

[26]

Polycarbonates

[12]

3 Analytical strategies

Many procedures and techniques can be applied to study the properties and degradation of

biomaterials. Water-uptake (swelling-ratio) measurements provide useful information on the

hydrophilic or hydrophobic nature of the materials. The results can be related to the degree

of crystallinity of the structure. Monitoring the changes in the pH of media as a function of

degradation indicates the acidic or basic nature of the released degradation products and their

ultimate effect on the surrounding environment (cells, tissues, etc.). Weight-loss studies are

almost universally performed to estimate any change in the mass of biomaterials during

degradation. Changes in the specimen dimensions and surface morphology, such as crack, or

micro channels, and changes inside the material can be highlighted by microscopy

techniques, such as scanning electron microscopy (SEM), transmission electron microscopy

(TEM) or atomic-force microscopy (AFM). The surface chemistry of the biomaterial may

alter or influence proteins and cells and may affect biocompatibility. The common methods

to characterize the surface chemistry include contact-angle measurements, Fourier-transform

infrared – attenuated-total-reflectance (FTIR-ATR) spectroscopy, X-ray photo-electron

spectroscopy (XPS) and secondary-ion mass spectrometry (SIMS) [27]. Differential

scanning calorimetry (DSC) and wide-angle x-ray diffraction (WAXS) are commonly used

techniques to estimate changes in the crystallinity of a biomaterial during the degradation

[28].

However, to tailor the properties of biomaterials, including physically or chemical cross-

linked networks for a specific biomedical application and to estimate the compatibility of

their degradation products with the surrounding biological environment, an in-depth

knowledge of their chemical structure is mandatory [29]. This includes characterization of

the starting material and the degradation products. Chromatographic separations, mass

Chapter 1

6

spectrometry (MS), and FTIR and NMR spectroscopy can provide more insight in the nature

and chemical structure of the degradation products.

There are three fundamentally different approaches to the structural characterization of

biomaterials. The first approach is the analysis of biomaterials without degradation. If the

biomaterials are soluble their identity and average molecular weights can be determined by

NMR spectroscopy and and by size-exclusion chromatography (SEC), respectively. The

compositional analysis of oligomers and low-molecular-weight polymers can be achieved by

mass-spectrometric techniques, such as liquid chromatography coupled through an

electrospray-ionization interface to a tandem mass-spectrometer (LC-ESI-MS/MS or LC-

ESI-MSn) or to a time-of-flight mass spectrometer (LC-ESI-ToF-MS), or matrix-assisted

laser-desorption/ionization (MALDI) ToF-MS. However, all MS techniques have their

limitations for high-molecular-weight synthetic polymers. When using ESI multiply charged

ions swamp the spectrum, amplifying the number of different ions arising from the

molecular-weight distribution. In MALDI both statistics and charge affinity may cause low-

molecular-weight oligomers to dominate the spectrum.

In the second approach the polymer can be degraded at harsh conditions, such as high

temperature or extreme pH, to complete degradation. This approach is suitable for the

characterization of networks that lack solubility and thus cannot be subjected directly to

chromatographic analysis [10]. When a polymer is being hydrolysed the degree of

degradation can be monitored by NMR spectroscopy. The degradation products can be

separated and quantified by, for example, LC with MS or UV-vis detection. The kinetic

chain length of poly-addition backbones (-C-C-) can be determined by SEC.

The third approach to study prospective biomaterials involves a chemical or a specific

enzymatic degradation under physiological conditions. This allows one to study the kinetics

of degradation. First degradation under physiologically relevant conditions is performed,

resulting in partially degraded material, the constituents of which may be identified [11].

Then complete and fast degradation of the products of the first step (oligomers, intermediates

and other products) is performed, followed by quantitative analysis [30].

The collected information is helpful (i) to ascertain the composition of the original networks,

(ii) to evaluate and optimize the synthesis of functional materials, (iii) to evaluate the

toxicological of the degradation products at an early stage, (iv) to determine the rate of

hydrolysis at different sites prone to attack, and (iv) for the rational design of new materials.


7

Figure 1 Schematics for the degradation and analysis of synthetic polymeric biomaterials. This review concerns the green part of this scheme.

4 Degradation methods

Degradable polymeric materials contain moieties that are prone to chemical or enzymatic

degradation (cf. Table 1). The degradation of such materials can be divided as follows.

4.1 Degradation under non-physiological conditions

This type of degradation involves harsh conditions, such as extreme pH values (both acidic

and alkaline) and/or elevated temperatures. Such degradation methods can be used (i) to

measure the kinetic chain length in photo-crosslinked polymeric networks, which lack

solubility [31], (ii) to reduce the size of polymeric chains in complex multi-block

copolymers to allow chromatographic separations, followed by sequence analysis with mass

spectrometry [32], (iii) to estimate the composition of complex polymeric networks by

quantifying each completely hydrolyzed building block [30], or (iv) to investigate the

Chapter 1

8

stability of different chemical bonds in a copolymer under extreme conditions. Haken et al.

reviewed the importance of vigorous chemical degradation of condensation polymers prior

to their chromatographic analysis [33]. Matsubara et al. designed a novel set-up to study

degradation, using supercritical methanol at 300oC and 8.1 MPa in a stainless-steel autoclave

(placed in a GC oven (Figure 2a). They could selectively decompose the ester linkages in

UV-cured acrylic esters [31]. Later, the developed set-up was successfully applied for the

characterization of the network structure of radiation-cured resins of poly-functional acrylic

ester and N-vinylpyrrolidone [34].

Peters et al. investigated the hydrolytic degradation of poly(D,L-lactide-co-glycolide 50:50)-

di-acrylate network coatings. They followed a two-step degradation process, first degrading

the coatings in PBS at 37oC and then further hydrolyzing the released products at 90oC in

10-M sodium hydroxide [30].

To tailor the performance of degradable synthetic polymeric biomaterials, it is important to

understand their structure. High-molecular-weight polymers or complex copolymers are

difficult to analyze by routine SEC or HPLC methods. The analyses usually involve time-

consuming sample-preparation steps especially in case of networks. To assess the structure

of such synthetic polymers, thermal degradation or pyrolysis can be a useful tool. In

pyrolysis, the polymer samples (introduced in the form of a solution or as a solid) break

down into small fragments (e.g. monomers or oligomers) by supplying thermal energy in an

inert atmosphere or vacuum [35]. The small fragments can be separated and analyzed by

chromatographic techniques such as GC or GC-MS [36].

4.2 Degradation under physiological conditions

The degradation of biomaterials under physiological conditions is studied to estimate their

degradation rate and to investigate phenomena involving surface degradation or bulk

degradation. Various other aspects, such as pH changes, degree of swelling, weight loss,

surface chemistry and morphology, and toxicity of the released products can also be studied.

These kinds of degradations are performed at 37oC, with different incubation media, such as

phosphate-buffered-saline (PBS) solution, enzyme-containing buffer, serum, or simulated

body fluids (SBF) at a suitable pH [28]. In conventional batch-mode analysis, the

biodegradable polymers (films, coatings, 3D scaffolds, etc.) are immersed in the respective

media followed by incubation at 37oC [40]. In vitro degradation conditions cannot mimic

real physiological conditions. However, the selection of appropriate enzymes, incubation


9

media, ratio’s of surface-to-mass of the specimen and surface-to volume of the medium,

duration of the experiment, and dynamic or static conditions during the degradation with

respect to the site where the biomaterial is implanted may help to find conditions closer to

the physiological ones [28].

Figure 2 (a) Schematic diagram of fast-degradation apparatus used for supercritical methanolysis (reprinted with permission from ref. [31]), (b) Schematic diagram of the reaction vessel used in the dynamic encrustation of urinary-tract devices based on polyurethanes, percuflex and silicone (reprinted with permission from ref. [37]), (c) Schematic showing the apparatus for studying the degradation of biodegradable scaffolds under dynamic conditions. Using a peristaltic pump, the scaffolds were subjected to a continuous flow (250 µl/min) of phosphate-buffered-saline (PBS) solution (pH 7.4) at 37oC (reprinted with permission from ref. [38]), (d) Schematic of dynamic flow simulation system used to study the effect of fluid flow on the degradation of poly(lactide-co-glycolide acid) (PLGA) for in vitro degradation of PLGA/b-TCP composite scaffolds (reprinted with permission from ref. [39]).

At the anatomical sites where there is minimal fluid flow, such as articular cartilage tissues,

the mass-to-surface ratio may strongly influence the degradation kinetics [41]. The level and

type of agitation (rotation, vibration, flow) during degradation may not only affect the

Chapter 1

10

release of the degradation products from the bulk or the surface of the material to the

surrounding media but also influence the contact between soluble reactants (e.g. enzyme)

and the insoluble substrate [28].

Agrawal et al. demonstrated the effect of static and dynamic conditions on the degradation

of scaffolds, fabricated from a copolymer of poly(lactic acid) and poly(glycolic acid), in PBS

at 37oC for up to six weeks. Figure 2d illustrates the apparatus used by the authors to achieve

dynamic conditions. They found that fluid flow decreased the degradation rate significantly

[38]. Gorman et al. investigated the encrustation of urinary-tract devices based on

polyurethanes, Percuflex and silicone in artificial urine under dynamic conditions. The same

level of encrustation was observed under static and dynamic conditions and significantly

higher levels of calcium and magnesium were found under static conditions [37]. In another

study, the effect of fluid flow on the in vitro degradation of poly(L-lactic acid)/β-tricalcium

phosphate (PLLA/ β-TCP) composite in PBS was investigated. Significantly faster

degradation was observed with a dynamic flow-simulation system [39].

Hooper et al. investigated the effects of SBF and PBS on the degradation of tyrosine-derived

polymers. They noticed a good similarity between the in vitro degradation kinetics of the

polymers in PBS and SBF and their in vivo results [42].

5 Chromatographic methods for degradable polymeric biomaterials

Novel degradable polymeric biomedical devices are developed using more-complex

polymers, i.e. random, block and graft copolymers or polymer blends. The characterization

of such polymers requires the use of chromatography. This involves the determination of the

molar-mass distribution, which reflects the length distribution (dispersity) of the polymeric

chain. Another important application of chromatographic systems is the separation of

polymers on the basis of their chemical heterogeneity, functionality type and sequence

lengths [43]. The size and the chemical nature of the degradation products determine the

adoptability of degradable polymers by the in vivo environment. To estimate the

toxicological nature of the degradation products sensitive and selective chromatographic

methods are required [29]. Biomaterials that are soluble in water or common organic

solvents can be analysed with common liquid-chromatographic (LC) methods. Some (but

not all) degradable polymeric materials designed for biomedical applications are of very high

molecular weight or based on insoluble polymeric networks. Such polymeric systems need

to be degraded prior to their chromatographic analysis. For structural analysis, the


11

degradation methods involve chemical hydrolysis at harsh conditions, methanolysis, or

partial degradation under mild conditions [10,30]. However, to estimate the degradation rate

and release of degradation products, degradation experiments are carried out under

physiological conditions, such as in PBS or enzyme-containing buffer at 37oC (see section

4.2).

5.1 Size-exclusion chromatography

During the polymerization process in which biodegradable polymers are formed, a large

number of chains are grown. The length of the resulting chains may vary. Therefore, it is

important to determine the molar-mass distribution (MMD; or molecular-weight distribution,

MWD). Size-exclusion chromatography (SEC) (also called gel-permeation chromatography,

GPC), is a popular analytical technique to separate polymer chains based on their size

(hydrodynamic volume). Unlike other LC methods, entropic effects are dominant in SEC

(T∆S >> ∆H) [43]. Mobile phases and packing materials are selected that minimize the

enthalpic interactions of the polymeric chains, so that the partition equilibrium is essentially

governed by the conformational entropy differences among the polymeric chains in the two

phases [44]. The information related to peak-average, number-average, weight-average, and

z-average molar masses (Mp, Mn, Mw, and Mz respectively) can be deducted from the position

and shape of the peak. Differential refractometry (dRI), UV-visible spectrometry, and – to a

lesser extent – evaporative laser-light scattering (ELSD) are concentration-sensitive

detection methods that are widely used in SEC experiments. In such experiments the MMD

and molar masses are typically calculated from a calibration curve, constructed using a set of

narrowly dispersed polymer standards. Light-scattering detection methods, such as multi-

angle laser-light scattering (MALLS) may provide useful information on the molecular size

of polymers, as well as on chain branching, conformation, and aggregation [45]. A change in

the shape and size of polymer molecules in solution influences the viscosity. Therefore,

viscometric detection methods are also used to determine the MMD of polymers [43].

“Triple detection-methods” (typically dRI, light-scattering and viscometry) are used to

determine “absolute” (i.e. accurate) molecular weights of branched and star-shaped

polymers. Absorbance or fluorescence detection and MS may – often in combination with

dRI detection – provide useful information on the distribution of specific fragments within

the chains or end groups in a polymer [46].

Chapter 1

12

Burdick et al. used aqueous SEC-dRI to characterize kinetic-chain-length distributions of

poly(methacrylic acid) (PMAA) in the hydrolysates of highly cross-linked systems based on

methacrylated sebacic acid, designed for orthopaedic applications [47]. The authors

investigated the relationship between kinetic chain length and the structural evolution of the

network. Themistou et al. determined the molecular weights and the molecular-weight

distributions of the hydrolysis products and precursors of cross-linked star polymer model

networks (CSPMNs) [48]. The linear and star polymers and their extractables were

determined by SEC-dRI with tetrahydrofuran (THF) as an eluent. The CSPMSs studied were

based on methyl methacrylate and the diacetal-based dimethacrylate cross-linker bis[(2-

methacryloyloxy)ethoxymethyl] ether and designed for biomedical applications [48].

Mojsiewicz-Piénkowska et al. reviewed the applications of SEC-ELSD for determining the

molecular weights of linear polydimethylsiloxanes (PDMSs). These authors also highlighted

the experimental conditions, such as calibration curve, mobile phase, flow rate and columns

used to characterize the PDMSs and the precision and accuracy of the developed methods

[22].

Peters et al. calculated the kinetic chain length of poly(acrylic acid) (PAA) backbone and the

average lengths of chains between cross-links in UV-cured networks prepared from mixtures

of di-functional (polyethylene–glycol di-acrylate) and mono-functional (2-ethylhexyl

acrylate) acrylates after hydrolysis. They used aqueous SEC coupled on-line to reversed-

phase LC with dRI and mass-spectrometric detection. The results were used to express the

chemical network structure for the different UV-cured acrylate polymers in network

parameters, such as the degree of cross-linking, the number of PAA units which were cross-

linked and the network density [49]. In another study, the same group of authors used

aqueous SEC-dRI to monitor the release of PAA chains during the hydrolytic degradation of

cross-linked poly-(D,L-lactide-coglycolide 50:50)-di-acrylate film. An increase in the

molecular weight with degradation time indicated that the release of these polyacrylate

chains was controlled by the number and type of ester groups that had to be degraded

hydrolytically to dissolve the chains [30].

Lin et al. used SEC with triple detection in chloroform to determine absolute molecular

weights. They confirmed that the star architecture in their biodegradable star polymers

consisted of hydrophilic hyperbranched poly-(ester amide) as core and hydrophobic PCL as

shell [50].

SEC is often used in off-line combinations with information-rich detectors, such as MS, or

FTIR or NMR spectroscopy. Rizzarelli et al. used matrix-assisted laser-desorption/ionization


13

time-of-flight mass spectrometry (MALDI-ToF-MS) as an off-line detection method for the

detailed structural characterization of complex polydisperse copolyesters, such as

poly[(R,S)-3-hydroxybutyrate-co-L-lactic acid] and poly[(R,S)-3-hydroxybutyrate-co--

caprolactone]. The results of compositional analysis were in good agreement with NMR

results [51]. Montaudo et al. demonstrated the use of NMR as an off-line detection method

for the compositional analysis of random copolymers with units of methyl methacrylate,

styrene, butyl acrylate and maleic-anhydride. They calculated the polydispersity index of the

copolymers by off-line MALDI-MS of the SEC fractions [52]. Nielen et al. explored the use

of electrospray-ionization – time-of-flight – mass spectrometry (ESI-ToF-MS) as a potential

detector for SEC analysis of polyesters. The absolute mass calibration of the SEC system

based on the polymer itself and determination of monomers and end groups from the mass

spectra were achieved [53].

BA

Figure 3 (A) On-line SEC-1H NMR traces obtained by monitoring the methoxy proton resonance at 3.59 ppm (a) and the α-methyl proton resonances at 0.86 ppm (….........) and 1.20 ppm (———) due to rr- and mm-triads, respectively (b); NMR signals due to α-methyl protons of the PMMA eluted in the elution periods F1 (c), F2 (d) and F3 (e) are also shown (reprinted with permission from ref. [55] ). (B) On-line SEC-NMR analysis of PMMA-block-poly(n-BuMA) prepared with t-C4H9MgBr in toluene at -60oC (reprinted with permission from ref. [56]).

Chapter 1

14

The stereo-regularity of polymeric chains and the chemical composition of copolymers may

affect their degradation rate and processing. On-line coupling of SEC with NMR

(continuous-flow NMR spectroscopy) makes it possible to study directly the chemical

composition and stereochemistry (isotactic, syndiotactic, atactic, etc.) of complex

copolymers separated according to their molecular size [54]. Hatada et al. studied the

tacticity of PMMA with on-line SEC-NMR. The results (Figure 3) showed a higher

concentration of rr-triads (syndiotactic) in fraction F1 of the higher-molecular-weight range

of the SEC chromatogram and a higher concentration of mm-triads (isotactic) in the lower-

molecular-weight fraction F3 [55]. In another study, they investigated the chemical

composition of block (PMMA-block-poly(n-BuMA)) and random (poly(MMA-ran-n-

BuMA)) copolymers of methyl and butyl methacrylates as a function of their MMD by on-

line SEC-NMR [56].

5.2 Adsorption liquid chromatography

Complex degradable polymeric systems are synthesized from (“telechelic”) oligomers and

polymers possessing terminal functional groups. The nature and the number of functional

groups on a chain may vary. Precursors for polymer synthesis, intermediate products, the

produced polymer, and the degradation products after hydrolytic or enzymatic degradation

can be separated based on different numbers of the same functionality or different

functionalities in a polymeric chain by analytical technique, such as adsorption liquid

chromatography (LC). The presence of side-reaction products and chiral impurities in the

degradable synthetic polymer can strongly influence their degradation rate and

biocompatibility [57]. Chromatographic methods with multiple detection methods are

needed for the separation and characterization of such impurities. Adsorption LC involves

enthalpic interactions between the stationary and mobile phases and the analyte molecules

[58,59]. Interactions between flexible polymeric chains in solution and the surface of the

stationary phase depend on the magnitude of the adsorption energy. The higher the

adsorption enthalpy (∆H) the stronger is the adsorption to the packing materials [43].

Adsorption LC is used in the normal-phase (NP) mode (using a polar stationary phase) or in

the reversed-phase (RP) mode (using a non-polar stationary phase).

Vu et al. reported on the use of LC with UV detection at 210 nm for determining the

oligomeric distribution of concentrated lactic-acid solutions [60]. Ding et al. developed an

LC method for the separation and quantification of water-soluble impurities and degradation


15

products in PLGA, to estimate changes in the polymer “micro-climate” (e.g. in pH). The

released products containing ester groups were derivatized with a common chromophore to

produce bromophenacyl esters prior to their gradient elution from a C18 column with UV-

vis detection at 254 nm [61]. Al Samman et al. investigated the influence of the degree of

branching on the retention behaviour of linear and branched aromatic polyesters in LC with

UV and ELSD detection. The branched polyesters showed a stronger adsorptive interaction

with the stationary phase than the corresponding linear molecules [58].

MS has been extensively exploited as an on-line detection method for the identification of

oligomers and low-molecular-weight degradable polymers [18,25,62]. Elliott et al. isolated

the degradation products of L-phenylalanine-based segmented polyester urethane ureas

degraded with chymotrypsin on a solid-phase-extraction cartridge for subsequent LC

separation and identification with LC-MS/MS. They observed the cleavage of urea, ester and

urethane bonds [18]. In an interesting study, Tang et al. investigated the enzyme-mediated

degradation of radio-labeled polycarbonate-polyurethanes (PCNUs). The water-soluble

degradation products were separated by LC with diode-array UV detection. The radioactivity

of the collected fractions was measured by a multi-purpose scintillation counter. The

products were identified by LC-MS/MS. The profile of the released degradation products

was in agreement with the structural analysis of synthesized polymers [62]. Deschamps et al.

simulated the in vivo degradation of segmented poly(ether ester)s block copolymers based on

poly(polyethylene glycol) and poly(butylene terephthalate) by their accelerated in vitro

degradation in PBS. They demonstrated the potential of LC-UV-MS for the detailed analysis

of the soluble degradation products. The results showed high amounts of the PEO fraction in

the soluble degradation products, while a PEOT/PBT fraction was found to be insoluble. The

results were confirmed with NMR [25].

Rizzarelli et al. found evidence for selective hydrolysis of aliphatic copolyesters, such as

poly(butylene succinate-co-butylene adipate), P(BS-co-BA), and poly(butylene succinate-co-

butylene sebacate), P(BS-co-BSe) induced by lipase. The water-soluble products, including

co-oligomers with identical molecular weights, but different sequences, were separated and

identified by on-line LC-ESI-MS/MS). The results showed a preferential cleavage of sebacic

ester bonds in P(BS-co-BSe) and succinic ester bonds in P(BS-co-BA) [32]. Carstens et al.

investigated the in vitro chemical and enzymatic degradation of monodisperse oligo(-

caprolactone) (OCL) and its block copolymer with methoxy poly(ethylene glycol) (mPEG-

b-OCL) by monitoring the water-soluble degradation products with LC-MS. The slow

degradation of OCL ester micelles in phosphate buffer at pH 7.4 was accelerated by lipase

Chapter 1

16

[63]. Pulkkinen et al. reported a fast analysis of the soluble degradation products of 2,2′-

bis(2-oxazoline)-linked poly-ε-caprolactone (PCL-O), degraded in simulated intestinal fluid

by LC-ESI-MS/MS. The polymer degraded primarily by ester hydrolysis, while amide bonds

showed greater stability [23].

Peters et al. demonstrated the use of LC-MS to identify and quantify the various water-

soluble oligomeric and polymeric degradation products released during the hydrolytic

degradation of poly(D,L-lactide-co-glycolide 50:50)-di-acrylate networks. The products

were analyzed directly after release and also after complete hydrolysis of the soluble fraction.

They found a rapid release of residual photo-initiator followed by a gradual release of

lactide/di-ethyleneglycol/glycolide oligomers with varying chain length and composition

[64].

Figure 4 (1) Liquid chromatographic separation of PEG 1000 (chromatograms A and B) and soluble products derived from PEOT:PBT (71:29) copolymer during hydrolytic degradation at 100oC (chromatograms C and D) with UV detection at 251 nm (chromatograms A and C) and mass-spectrometric detection applying atmospheric-pressure chemical ionization in the positive-ion mode (APCI(+)) conditions (chromatograms B and D) recorded in scan mode (m/z=200–1600) (reprinted with permission from ref. [25]). (2) UV-absorbance and radioactivity chromatograms for the degradation products from radio-labeled polycarbonate-polyurethanes: (a) buffer incubation and (b) cholesterol esterase incubation (reprinted with permission from ref. [62]).


17

The hyphenation of the most powerful spectroscopic technique i.e. NMR (containing solvent

suppression features) with HPLC is recently getting more attention for the online chemical

composition and molar mass determination of oligomers and low MW polymers separated

on reversed phase HPLC column according to their chemical structure. Pasch et al.

investigated the chemical structure, molar mass and end group analysis of poly(ethylene

oxide) by an online HPLC-NMR setup [65].

5.3 Liquid chromatography at critical conditions

Liquid chromatography at critical conditions (LCCC) is receiving increased attention for the

separation of complex polymers. LCCC separates the polymers at the so-called critical

conditions, i.e. the chromatographic conditions where the enthalpy and entropy effects

compensate each other (∆H = T∆S) [66]. Under these conditions the retention of polymeric

species becomes independent of their molecular weight [66,67]. LCCC allows the separation

of polymers on the basis of their functionality type distributions (FTDs). It has been applied

for the separation of functional polymers, block copolymers, branched polymers, and

polymer blends and to assess their stereo-regularity [15,58,59,66,68]. The incorporation of

hydrophilic and hydrophobic components and stereo-regularity of degradable copolymers for

biomedical application control their degradation behaviour. Lee et al. resolved oligomeric

PLLA block species of poly(ethylene oxide)-block-poly(L-lactide), (PEO-b-PLLA) by

RPLC at the critical conditions of PEO. They confirmed the composition of each species by

off-line MALDI-MS analysis [67]. In another study, they fractioned the LLA units in tri-

block PLLA-b-PEO-b-PLLA copolymer by RPLC at the critical conditions of PEO and

confirmed the results by off-line MALDI-MS. In tri-block copolymer, unlike in di-block

(PEO-b-PLLA) copolymer, they observed a splitting of the eluted peaks containing the same

number of LLA units. They assigned this peak splitting to the different length distributions

of PLLA blocks at the two ends of the PEO block [14]. Mengerink et al. developed a method

for the separation of linear and cyclic oligomers of polyamide-6 by LCCC-ELSD. ESI-MS

did not allow discrimination between the linear and cyclic products.[15]. Peters et al.

reported the FTD of functional PMMA, obtained by LCCC-ELSD. The mono- and

bifunctional PMMA peaks were identified by ESI- MS [68].

Philips et al. discussed novel developments in water-based LCCC. They varied the buffer

concentration and the proportion of organic modifier in the mobile phase to approach the

critical condition for two polymer systems, viz. poly(styrene sulfonate) and poly(acrylic

Chapter 1

18

acid). The critical condition of poly-(acrylic acid) was then used to study the retention

characteristics of a copolymer containing both acrylic acid and n-vinyl pyrrolidinone [69].

De Geus et al. utilized LCCC, with UV and ELSD detection, to separate PCL polymer

samples with different end groups, in order to gain insight in the initiation process of

enzymatic ring-opening polymerization. PCL chains with three different end groups were

separated, i.e. (i) linear carboxylic-acid end-functionalized species, (ii) linear hydroxyl-ester

species, and (iii) cyclic species. The identity of each peak was confirmed by offline MALDI-

MS [70].

LCCC with on-line NMR analysis can provide detailed information on the end groups and

chemical composition of polymeric chains. Hiller et al. demonstrated the use of on-line

NMR detection for the analysis of complex mixtures of fatty alcohol ethoxylates (FAEs) by

LCCC. The peaks were detected using an ELSD detector [71]. In another study, they

investigated the separation of block copolymers of PS-b-PMMA and blends of PS and

PMMA at the size-exclusion conditions for PS and critical conditions of PMMA with on-line 1H NMR detection [72]. Unfortunately, as the critical conditions are strongly dependent on

the mobile-phase composition, but also on temperature and pressure [73], especially for

high-molecular-weight polymers, LCCC is only rarely applied successfully to polymer

systems with molecular weights exceeding 100 kDa.

5.4 Two-dimensional liquid chromatography

Complex polymers, including degradable synthetic polymers, exhibit several simultaneous

distributions. For example, all functionalized polymers exhibit an MMD and a functionality-

type distribution (FTD) and all copolymers exhibit an MMD and a chemical-composition

distribution (CCD) [43,74]. Moreover, the different distributions in complex polymers tend

to be mutually dependent [74]. SEC or HPLC by themselves may not reveal correct

information on the MMD or the molecular heterogeneity of the polymeric chains [46]. To

gain insight in multiple, mutually dependent distributions analytical techniques such as

multi-dimensional separations are indispensable [74].

Kilz et al. provided a detailed description of the two-dimensional chromatographic

techniques for polymers [75]. Pasch et al. reported on the two-dimensional separation of

PEO-b-PPO block copolymers. In the first dimension, they separated copolymers with

respect to the length of the PEO block by LCCC. The collected fractions were further

separated in the second dimension, either by supercritical-fluid chromatography (SFC) or by


19

SEC based on the length the PEO blocks [76]. Early two-dimensional LC techniques were

based on “heart-cut” or fractionation methods and they were very specialized or time-

consuming [74]. With the advent of modern technology, two-dimensional LC methods are

becoming faster and more comprehensive.

Van der Horst et al. wrote a highly useful review on the advantages of comprehensive

LC×LC for polymers over “heart-cutting” LC-LC [59]. Kok et al. reported on FTIR as an

on-line detection method in comprehensive LC×SEC for the characterization of copolymers

based on styrene and methacrylates [77]. The results were confirmed by UV detection. The

generated functional-group contour plots showed a distinction between UV-active and non-

UV-active groups of the polymer.

Figure 5 LC×LC contour plots of (A) poly(2-ethylhexyl acrylate) P2EHA macro(RAFT agent), (B) copolymer (2-ethylhexyl acrylate and methyl acrylate)-1h, (C) copolymer-2h, (D) copolymer-4h, and (E) copolymer-8h. (F) is a rotation of 90o and inclination of 35o of the LC×LC chromatogram of the sample of copolymer-8h. 1st dimension: gradient LC with 0 to 70% THF in methanol in 200 min at 0.05 mL/min on PLRP-S 5 μm (Y-axis). 2nd dimension: SEC with THF at 1.5 mL/min on PL HTS-C (X-axis). Calibration: PMMA. Detection: ELSD (reprinted with permission from ref. [79]).

Chapter 1

20

The hydrodynamic volume of a branched polymer of certain molecular weight may be

identical to that of a linear polymer of lower molecular weight. Based on this principle Edam

et al. demonstrated the use of comprehensive two-dimensional molecular-topology

fractionation (MTF) × SEC for the separation of branched polymers based on their topology

[78]. Raust et al. performed two-dimensional LC separations with a combination of LCCC

and SEC to gain insight in the polymerization process of copolymers based on 2-ethylhexyl

acrylate and methyl acrylate, (P2EHA-b-PMA), produced by reversible addition-

fragmentation chain transfer (RAFT)-mediated polymerization in organic dispersion (Figure

5). The LC×SEC chromatograms revealed a certain heterogeneity of the polymer and

allowed the precise characterization of the MA block length in the copolymer. For

compositional analysis the results were confirmed by LC-1H NMR [79]. In summary,

LC×LC methods can be useful for the characterization of complex degradable polymers in

terms of several distributions (MMD, CCD, FTD, etc.) simultaneously. LC×LC can provide

an efficient, reliable and comprehensive characterization of biodegradable polymers.

6 Gas chromatography

Gas chromatography (GC) is another powerful analytical tool for the identification and

quantification of impurities, additives, and degradation products of degradable polymeric

biomaterials. GC is most often applied in combination with flame-ionization detection (FID)

and MS for the analysis of oligomers and low-molecular-weight polymers [36]. Barlow et al.

reviewed the applications of GC in combination with pyrolysis for the analysis and

characterization of polymer degradation [80].

Hakkarainen et al. investigated the nature of low-molecular-weight degradation products of

PLA, PGA, and their copolymers. They reported a convenient and rapid solid-phase-

extraction (SPE) – derivatization technique to improve the qualitative and quantitative GC-

MS analysis of hydroxy acids released by the degradation of PLA and PGA in buffer

solution [81-83]. The GC-MS analyses showed a difference in the patterns of degradation

products released in biotic and abiotic media. In another study this group utilized GC-FID to

explore single-drop micro-extraction (SDME) in combination with multiple-headspace

(MHS) extraction for the quantitative determination of lactide in thermally oxidized

polylactide [84]. During a study of the esterification reaction between lactic acid and

different fatty acids, Torres et al. utilized GC-MS to estimate the degree of polymerization in

polymerized fractions of commercial LA [85]. Vu et al. characterized the oligomeric


21

distribution of lactic acid in aqueous media by GC-MS [60]. Urakami et al. reported a rapid,

precise and accurate compositional analysis of co-poly(DL-lactic/glycolic acid) (PLGA).

This was performed by pyrolysis – gas chromatography – mass spectrometry (Py-GC/MS)

combined with one-step hydrolysis and methylation in the presence of

tetramethylammonium hydroxide (TMAH). They found good agreement of the analytical

results with 1H NMR data [86]. Plikk et al. studied the chemical changes in porous scaffolds

based on various L,L-lactide (LLA), 1,5-dioxepane-2-one (DXO) and -caprolactone (CL)

copolymers after sterilization with electron beam and gamma irradiation [87]. The formation

of low-molecular-weight degradation products was studied by GC-MS. Burford et al.

described the rapid qualitative and quantitative analysis of polyester-based polyurethane

elastomers by GC-FID after polymer cleavage into the corresponding glycol, dicarboxylic

acid and diamine fragments by molten alkali fusion at high temperature [88]. All the

caroboxylic-acid products were reacted to dimethyl ester derivatives prior to their GC

analysis [88]. Mallepally et al. investigated the enzymatic degradation of hyperbranched

polyesters (HBPEs). The release of free fatty acids was studied using GC [89].

Figure 6 Pyrograms of UV-curable resins based on bifunctional poly(ethylene glycol)-diacrylate (PEDA). Pyrolysis at 400°C in the presence of TMAH. (a) prepolymer; (b) UV-cured resin. (reprinted with permission from ref. [91]).

Chapter 1

22

Eldsäter et al. studied the degradation of poly (ester amide) and poly(buty1ene adipate-co-

caproamide) in aqueous environment at 37oC, 60oC, and 80oC. GC-MS with SPE was used

to investigate the nature of degradation products at different degradation conditions. Changes

in the polymer composition were investigated by Py-GC-MS [90]. Matsubara et al. used

pyrolysis GC (Py-GC) in the presence of tetramethylammonium hydroxide (TMAH) to

characterize the network structure in UV-cured bifunctional poly(ethylene glycol)-diacrylate

(PEDA) [91].

Kaal et al. developed a fully automated on-line SEC-Py-GC-MS method [92]. The polymer

samples were separated based on molecular size and fractions were transferred on-line. The

SEC solvent was evaporated in a programmable-temperature-vaporizer (PTV) injector prior

to pyrolysis and GC-MS analysis. The scope of the method was extended to include aqueous

SEC and RPLC by introducing a sintered liner, filled with sintered glass beads (60-100 µm)

to approximately half of the cross sectional area [93]. The developed systems provided a

great deal of quantitative insight in the composition of the on-line collected LC or SEC

fractions. Recently, Chojnacka et al. investigated the effect of monomeric ratio of N-vinyl-2-

pyrrolidone (VP) and vinyl acetate (VA) on the dissolution behaviour of their copolymers in

water using Py-GC-MS. The compositional analysis of the fractions, collected at different

time intervals during the dissolution study, revealed that copolymers with higher contents of

VA dissolve considerably slower than the other copolymers [94].

7 Direct mass-spectrometric analyses

Mass spectrometry has emerged as a powerful analytical tool for the characterization of

synthetic polymers and copolymers. A time-of-flight (ToF) mass spectrometer offers high

sensitivity for multi-ion detection, a large mass range, and good mass resolving power.

Therefore, ToF-MS is most commonly used as a mass analyzer for the characterization of

polymers [95]. A ToF-MS can be conveniently combined with ESI or with MALDI.

However, when using ESI multiply charged ions are usually formed, which complicates the

interpretation of the sprectra. In MALDI both statistics and “charge-ability” (a combination

of several parameters, including affinity to charge and efficiency of transfer from solid to

vacuum) may cause low-molecular-weight oligomers to dominate the spectrum. Several

books and reviews have been published that describe the developments in the field of mass

spectrometry of synthetic polymers [95-99].


23

ESI is a soft ionization technique, which produces multiply charged ions and little

fragmentation. The analyte solution emerging from the column is nebulized, ionized and

transferred to the vacuum of the mass analyzer and the MS detector [98]. Andersson et al.

compared the degradation stability of stereo-complex poly(L-lactic acid)/ poly(D-lactic acid)

(PLLA/PDLA) with plain PLLA. The composition of degradation products was estimated

semi-quantitatively by direct ESI-MS. The results showed a shorter hydrolysis time for

PLLA/PDLA and more acidic degradation products [13]. Höglund et al. studied the effect of

plasticizer (acetyl tributyl citrate) on the degradation of PLA. They investigated the water-

soluble products and the plasticizer by ESI-MS [100]. In another study, this group studied

the effect of surface modification on the hydrolytic degradation and investigated the

degradation of PLA grafted with poly(acrylic acid) (PAA). The water-soluble degradation

products were analyzed by ESI-MS [101]. Recently, Rizzarelli et al. developed a convenient

direct ESI-MS method to determine concentrations of sebacic-acid (SA) and terephthalic-

acid (TA) residues in biodegradable copolymers. The obtained results were in agreement

with LC-UV data. The assay was proposed as a fast and sensitive alternative to currently

employed methods for acid quantification [102].

MALDI is also a soft ionization technique. It allows the detection of large, non-volatile and

labile molecules. The compounds of interest are desorbed and ionized by the combined

influence of a laser beam and a chemical matrix, usually under vacuum. The resulting

(predominantly singly charged) ions are directed to a (ToF) mass spectrometer by a

continuous high voltage [99]. Burkoth et al. used MALDI-ToF-MS to characterize the

molecular-weight distribution of (mostly) linear poly(methacrylic acid) degradation products

as a function of the network evolution (i.e. double-bond conversion), rate of initiation, and

monomer size during the degradation of cross-linked polymers based on PMA and sebacic

acid [40]. Rizzarelli et al. employed MALDI with tandem mass spectrometry (MSn) to

investigate the fragmentation pathways of poly(butylene adipate) (PBAd) oligomers [103].

In another study, they applied post-source-decay (PSD) MALDI-ToF-MSn for the sequence

determination of aliphatic poly(ester amide)s synthesized from dimethyl sebacate or sebacic

acid and 2-aminoethanol or 4-amino-1-butanol [104]. Luo et al. found a symmetric

distribution in low-molecular-weight star polymers prepared by grafting poly(ethylene

glycol) (PEG) arms onto a cholic acid core via anionic polymerization [105]. Weidner et al.

performed fragmentation analysis by means of MALDI with collision-induced dissociation

(CID) and MSn to determine sequences and end groups of complex copolyesters based on

hexanediol-neopentylglycol-adipic acid copolyesters [106].

Chapter 1

24

During the course of the degradation process, the surface chemistry of the degradable

polymeric device may influence the biological environment. Thus, it plays an important role

in determining its biocompatibility [27]. Therefore, techniques for characterizing the surface

of the biomaterials are gaining attention. Secondary ion mass spectrometry (SIMS) in

combination with ToF can be used for surface characterization and (quantitative) analysis of

synthetic copolymers and polymeric blends [96,107]. Belu et al. reviewed the application of

ToF-SIMS for the structural characterization of biomaterial surfaces. They described the

technique as a flexible and powerful surface-characterization tool [108]. Chen et al. used

ToF-SIMS to study the in vitro hydrolytic degradation at the surface of different

biodegradable polymers, including PLA, PGA, PLGA, poly(sebacic acid) (PSA), and two

random copolymers of poly(fumaric-co-sebacic) acid (PFS) of different compositions. It was

reported that useful information on reaction kinetics can be obtained from the ToF-SIMS

spectra by analyzing the intensities of the molecular ions in the distribution [107].

Figure 7 (A) The 2 m drift-tube IMS-MS instrument design and operationa and (B) typical output for ions separated in the gas-phase detected by MS in different modes ofoperation (cf. details in ref. [110]). (reprinted with permission from ref. [110]).


25

Recently, the combination of ion-mobility spectroscopy (IMS) through CID with MS has

been gaining recognition for the structural characterization of synthetic polymers [97]. In ion

mobility, ions are separated on the basis of their conformational state (size and shape), as

they drift through a gas (e.g. He, N2) under the influence of an electric field [109]. A major

benefit of including IMS as an intermediate stage in LC with MS detection is the reduction

of “chemical noise” due to the additional selectivity of IMS. This is especially important

when determining trace amounts of compounds in complex mixtures such as body fluids.

Trimpin et al. reported on the use of a multi-dimensional IMS-MS methodology that

rendered a detailed view of molecular components in complex mixtures, based on the

combined analysis of the three-dimensional geometries and masses of polymeric components

adducted with metal cations in the gas phase [110]. The structure of PEGs with different

functionalities, PPG, poly(tetramethylene glycol) (PTMEG), and several poly(alkyl

methacrylate)s (PAMA)s (with alkyl = methyl, ethyl, butyl, etc.) was investigated using

IMS-MS.

8 Nuclear-magnetic-resonance spectroscopy

Nuclear-magnetic-resonance (NMR) spectroscopy is an extensively used analytical

technique in the field of synthetic polymers. The microstructure, region-isomerism,

stereochemical isomerism, geometric isomeric, branches and end groups, copolymer

composition, number-average molecular weight (Mn), chain conformation, and

intermolecular association of the polymers are among the parameters that can be investigated

by high-resolution NMR and 2D NMR experiments [111]. LeMaster et al. studied the effect

of T1 and T2 relaxation on the 2D 1H-13C correlation spectra of linear commercial polymers.

The results were used to estimate the concentration of end groups in polyester urethanes (Mn

40 kDa). They estimated an uncertainty in Mn of 6-7% (r.s.d.) [112]. Two-dimensional

homo-nuclear correlation spectroscopy (1H-1H COSY) was used to confirm the formation of

poly(α-peptide) in the protease-catalyzed polymerization of L-glutamic acid diethyl ester

hydrochloride [113]. Pergal et al. synthesized polyurethanes-siloxane copolymers containing

high contents of PCL-PDMS-PCL segments [114]. The structure of copolymers, the lengths

of hard and soft segments, and the connectivities between homonuclear or heteronuclear

atoms with single or multiple bond were investigated by 1H, 13C NMR and 2D NMR

experiments, such as 1H-1H COSY, HSQC (heteronuclear single quantum coherence), and

HMBC (heteronuclear multiple-bond correlation). In an interesting study, the generation of

Chapter 1

26

hyper-branched poly(amine-ester)s was confirmed by 13C-, DEPT-135 NMR and 2D NMR

techniques [115]. Shaver et al. attached six arms of poly(lactic acid) to dipentaerythritol

cores. 1H NMR experiments provided useful information on the tacticity of the synthesized

star polymer [116]. To investigate branching, Cooper et al. performed 1H-, 13C-, COSY, and

HSQC NMR experiments and SEC to determine the number of end groups and repeating

units in the backbone of poly(lactic acid)-polyurethane functionalized with pendent

carboxylic-acid groups [117].

9 Conclusions

Separation methods based on chromatography are essential analytical tools to estimate the

FTD, CCD, and MMD of complex degradable polymeric systems. The analysis of the

chemical nature of the degradation products not only highlights the stability of different

degradable bonds, but also reflects the toxicological nature and the biocompatibility of

biomaterials. SEC with dRI, UV-vis or ELSD detectors yields molecular-weight

distributions. Light-scattering or viscometry may provide additional information, such as

molecular size and absolute molar masses. LC separates polymers on the basis of their

functionality and chemical composition. LCCC is a method of choice to separate low-

molecular-weight functional polymers, copolymers and polymer blends at critical conditions.

NMR spectroscopy as an on-line detector for SEC, LC, or LCCC provides useful

information about the functionality, chemical composition, and tacticity of the polymeric

chains along their MMD. Coupling SEC on-line with MS also broadens its scope.

Comprehensive two-dimensional LC techniques, such as LC×LC, LC×SEC, LCCC×SEC,

MTF×SEC, etc. are developing into promising analytical tools for the detailed analysis of,

for example copolymers, branched polymers, and polymer blends. Gas chromatography is

extensively used to separate and identify degradation products. For complex polymers and

networks, Py-GC-MS provides more insight in the chemical composition by pyrolyzing the

liquid or solid samples. Direct mass spectrometric methods, such as ESI-ToF-MS, MALDI-

ToF-MS, and SIMS can provide rapid analysis of the chemical composition of oligomers or

low-molecular-weight polymers. MALDI and SIMS allow studying the surface chemistry

before and after degradation. IMS-MS promises to contribute to utilize the 3D structure of

the polymers for additional selectivity. Despite the limitations associated with each

analytical technique, a combination of selective and sensitive methods can usually be

devised for the analysis for different classes of polymeric biomaterials.


27

10 Scope of the thesis

The objective of this thesis is to develop new analytical methods by exploring different

analytical techniques. The thesis deals with the degradation and analysis of synthetic

polymeric biomaterials. The performance of degradable polymeric biomaterials depends on

their chemical structure and on the chemical nature of their degradation products. Therefore,

Chapter 1 reviews the different strategies and analytical techniques for the chemical analysis

of degradable polymeric biomaterials, in particular those based on chromatographic

separations.

In Chapter 2 of this thesis, the quantitative structural analysis of polymeric networks that are

insoluble under normal physiological conditions is described. An in vitro method is

developed that is well suited for the quick and complete hydrolytic degradation of poly(2-

hydroxyethyl methacrylate) (pHEMA), poly(lactide-coglycolide50:50)1550-diol

(PLGA(50:50)1550-diol) and polyester-urethane-acrylate-based networks, using a microwave

set-up. The microwave glass vials are coated internally with Teflon (PTFE) to avoid contact

of the alkaline medium with the inner surface of the glass vessel and thus prevent the

formation of residues. The degree of hydrolysis is monitored by NMR spectroscopy. The

hydrolyzed components can be separated by liquid chromatography and quantified by mass

spectrometry or UV-vis detection. The kinetic chain length of poly-addition backbones (-C-

C-) can be determined by SEC.

Chapter 3 describes an in vitro enzymatic degradation of multi-block PEA with α-

chymotrypsin and proteinase-K at 37oC. The release of different monomeric and oligomeric

products is monitored by LC-ESI-ToF-MS. Semi-quantitative analysis of water-soluble

degradation products reveals the protease and/or amidase activity and provides indications of

the relative fragment (bond) stabilities. The polymer does not degrade by chemical

degradation under physiological conditions. The structure of the polymer is characterized by

NMR spectroscopy.

In Chapter 4 the development of a miniaturized and automated system is reported that can be

used for the fast, on-line investigation of the the in-vitro enzymatic degradation of PEA

coatings. The system can be used under both static and dynamic (flow) conditions and

includes on-line LC-ToF-MS analysis of the hydrolysate (containing enzyme and

degradation products). The system is investigated with respect to different injection volumes

(pulses) of an α-chymotrypsin (α-CT) solution, flow rates of injected α-CT band or α-CT

solution through the coated capillary, concentration of α-CT, and lengths of coated capillary.

Chapter 1

28

The versatility of the system makes it easy to follow the course of degradation and to

differentiate between primary and secondary degradation products.

11 References

[1] D.F. Williams, Biomaterials 30 (2009) 5897. [2] G. Wnek, G.L. Bowlin, Encyclopedia of Biomaterials and Biomedical Engineering,

Informa Health Care, 2004. [3] G. Mabilleau, M.F. Moreau, R. Filmon, M.F. Baslé, D. Chappard, Biomaterials 25

(2004) 5155. [4] Aylvin A. Dias, M. Hendriks, Drug Delivery Technology 10 (2010) 20. [5] P.J.A. in 't Veld, P.J. Dijkstra, J. Feijen, Clinical Materials 13 (1993) 143. [6] H. Sun, F. Meng, A.A. Dias, M. Hendriks, J. Feijen, Z. Zhong, Biomacromolecules

12 (2011) 1937. [7] M. Hakkarainen, in, Springer Berlin / Heidelberg, 2002, p. 113. [8] K.D.F. Vlugt-Wensink, X. Jiang, G. Schotman, G. Kruijtzer, A. Vredenberg, J.T.

Chung, Z. Zhang, C. Versluis, D. Ramos, R. Verrijk, W. Jiskoot, D.J.A. Crommelin, W.E. Hennink, Biomacromolecules 7 (2006) 2983.

[9] R.S. Labow, Y. Tang, C.B. McCloskey, J.P. Santerre, Journal of Biomaterials Science, Polymer Edition 13 (2002) 651.

[10] A. Ghaffar, P.G. Verschuren, J.A.J. Geenevasen, T. Handels, J. Berard, B. Plum, A.A. Dias, P.J. Schoenmakers, S. van der Wal, Journal of Chromatography A 1218 (2011) 449.

[11] A. Ghaffar, G.J.J. Draaisma, G. Mihov, A.A. Dias, P.J. Schoenmakers, S. van der Wal, Biomacromolecules 12 (2011) 3243.

[12] B.D. Ulery, L.S. Nair, C.T. Laurencin, Journal of Polymer Science Part B: Polymer Physics 49 (2011) 832.

[13] S.R. Andersson, M. Hakkarainen, S. Inkinen, A. Södergård, A.-C. Albertsson, Biomacromolecules 11 (2010) 1067.

[14] H. Lee, T. Chang, D. Lee, M.S. Shim, H. Ji, W.K. Nonidez, J.W. Mays, Analytical Chemistry 73 (2001) 1726.

[15] Y. Mengerink, R. Peters, C.G. deKoster, S. van der Wal, H.A. Claessens, C.A. Cramers, Journal of Chromatography A 914 (2001) 131.

[16] B. Rubehn, T. Stieglitz, Biomaterials 31 (2010) 3449. [17] E. Wisse, R.A.E. Renken, J.R. Roosma, A.R.A. Palmans, E.W. Meijer,

Biomacromolecules 8 (2007) 2739. [18] S.L. Elliott, J.D. Fromstein, P.S. J, K.A. Woodhouse, Journal of Biomaterials

Science, Polymer Edition 13 (2002) 691. [19] K.E. Uhrich, A. Gupta, T.T. Thomas, C.T. Laurencin, R. Langer, Macromolecules 28

(1995) 2184. [20] M. Kato, K. Toshima, S. Matsumura, Biomacromolecules 6 (2005) 2275. [21] Z. Zhao, J. Wang, H.-Q. Mao, K.W. Leong, Advanced Drug Delivery Reviews 55

(2003) 483. [22] K. Mojsiewicz-Pienkowska, Journal of Chromatography B 865 (2008) 7. [23] M. Pulkkinen, J.J. Palmgrén, S. Auriola, M. Malin, J. Seppälä, K. Järvinen, Rapid

Communications in Mass Spectrometry 22 (2008) 121. [24] J. Zimmermann, T. Loontjens, B.J.R. Scholtens, R. Mülhaupt, Biomaterials 25

(2004) 2713.


29

[25] A.A. Deschamps, A.A. van Apeldoorn, H. Hayen, J.D. de Bruijn, U. Karst, D.W. Grijpma, J. Feijen, Biomaterials 25 (2004) 247.

[26] E.M. Christenson, S. Patel, J.M. Anderson, A. Hiltner, Biomaterials 27 (2006) 3920. [27] B.D. Ratner, A. Chilkoti, D.G. Castner, Clinical Materials 11 (1992) 25. [28] R.L. Reis, J.S. Román, Biodegradable Systems in Tissue Engineering and

Regenerative Medicine, CRC Press, 2004. [29] A.-C. Albertsson, M. Hakkarainen, in Chromatography for Sustainable Polymeric

Materials -Renewable, Degradable and Recyclable ; Advances in Polymer Science 211, Springer, 2008, p. 85.

[30] R. Peters, J. Lebouille, B. Plum, P. Schoenmakers, S. van der Wal, Polymer Degradation and Stability In Press, Accepted Manuscript (2011).

[31] H. Matsubara, S.-i. Hata, Y. Kondo, Y. Ishida, H. Takigawa, H. Ohtani, Analytical Sciences 22 (2006) 1403.

[32] P. Rizzarelli, G. Impallomeni, G. Montaudo, Biomacromolecules 5 (2004) 433. [33] J.K. Haken, TrAC Trends in Analytical Chemistry 9 (1990) 14. [34] H. Matsubara, H. Kataoka, H. Ohtani, Polymer Chemistry 1 (2010) 1056. [35] W.J. Irwin, Analytical Pyrolysis: A Comprehensive Guide, Marcel Dekker Inc., New

York, 1982. [36] F.C.-Y. Wang, A.D. Burleson, Journal of Chromatography A 833 (1999) 111. [37] S.P. Gorman, C.P. Garvin, F. Quigley, D.S. Jones, Journal of Pharmacy and

Pharmacology 55 (2003) 461. [38] C.M. Agrawal, J.S. McKinney, D. Lanctot, K.A. Athanasiou, Biomaterials 21 (2000)

2443. [39] Y. Yang, Y. Zhao, G. Tang, H. Li, X. Yuan, Y. Fan, Polymer Degradation and

Stability 93 (2008) 1838. [40] A.K. Burkoth, K.S. Anseth, Macromolecules 32 (1999) 1438. [41] A.M.A. Ambrosio, H.R. Allcock, D.S. Katti, C.T. Laurencin, Biomaterials 23 (2002)

1667. [42] K.A. Hooper, N.D. Macon, J. Kohn, Journal of Biomedical Materials Research 41

(1998) 443. [43] H. Pasch, B. Trathnigg, HPLC of Polymers, Springer-verlag Berlin Heidelberg, 1998. [44] T. Chang, Journal of Polymer Science Part B: Polymer Physics 43 (2005) 1591. [45] A. Streigel, W.W. Yau, J.J. Kirkland, D.D. Bly, Modern Size-Exclusion Liquid

Chromatography: Practice of Gel Permeation and Gel Filtration Chromatography, John Wiley & Sons Inc., Hoboken, New Jersey., 2009.

[46] T. Provder, H.G. Barth, M.W. Urban, Chromatographic Characterization of Polymers: Hyphenated and Multidimensional Techniques, American Chemical society, 1995.

[47] J.A. Burdick, T.M. Lovestead, K.S. Anseth, Biomacromolecules 4 (2003) 149. [48] E. Themistou, C.S. Patrickios, Macromolecular Chemistry and Physics 209 (2008)

1021. [49] R. Peters, V.M. Litvinov, P. Steeman, A.A. Dias, Y. Mengerink, R. van Benthem,

C.G. de Koster, S. van der Wal, P. Schoenmakers, Journal of Chromatography A 1156 (2007) 111.

[50] Y. Lin, X. Liu, Z. Dong, B. Li, X. Chen, Y.-S. Li, Biomacromolecules 9 (2008) 2629.

[51] G. Adamus, P. Rizzarelli, M.S. Montaudo, M. Kowalczuk, G. Montaudo, Rapid Communications in Mass Spectrometry 20 (2006) 804.

[52] M.S. Montaudo, Polymer 43 (2002) 1587. [53] M.W.F. Nielen, F.A. Buijtenhuijs, Analytical Chemistry 71 (1999) 1809.

Chapter 1

30

[54] K. Albert, On-Line LC–NMR and Related Techniques, John Wiley & Sons Ltd, 2002.

[55] K. Hatada, K. Ute, T. Kitayama, T. Nishimura, M. Kashiyama, N. Fujimoto, Polymer Bulletin 23 (1990) 549.

[56] K. Hatada, K. Ute, T. Kitayama, M. Yamamoto, T. Nishimura, M. Kashiyama, Polymer Bulletin 21 (1989) 489.

[57] S. Inkinen, M. Hakkarainen, A.-C. Albertsson, A. SoÌˆdergaÌŠrd, Biomacromolecules 12 (2011) 523.

[58] M. Al Samman, W. Radke, A. Khalyavina, A. Lederer, Macromolecules 43 (2010) 3215.

[59] A. van der Horst, P.J. Schoenmakers, Journal of Chromatography A 1000 (2003) 693.

[60] D.T. Vu, A.K. Kolah, N.S. Asthana, L. Peereboom, C.T. Lira, D.J. Miller, Fluid Phase Equilibria 236 (2005) 125.

[61] A.G. Ding, S.P. Schwendeman, Journal of Pharmaceutical Sciences 93 (2004) 322. [62] Y.W. Tang, R.S. Labow, J.P. Santerre, Biomaterials 24 (2003) 2805. [63] M.G. Carstens, C.F. van Nostrum, R. Verrijk, L.G.J. de Leede, D.J.A. Crommelin,

W.E. Hennink, Journal of Pharmaceutical Sciences 97 (2008) 506. [64] R. Peters, J. Lebouille, B. Plum, P. Schoenmakers, S. van der Wal, Polymer

Degradation and Stability In Press, Corrected Proof (2011). [65] H. Pasch, W. Hiller, Macromolecules 29 (1996) 6556. [66] S.G.Entelis, V.V.Evreinov, A.V.Gorshkov, Advances in Polymer Science 76 (1986)

129. [67] H. Lee, W. Lee, T. Chang, S. Choi, D. Lee, H. Ji, W.K. Nonidez, J.W. Mays,

Macromolecules 32 (1999) 4143. [68] R. Peters, Y. Mengerink, S. Langereis, M. Frederix, H. Linssen, J. van Hest, S. van

der Wal, Journal of Chromatography A 949 (2002) 327. [69] S.L. Phillips, L. Ding, M. Stegemiller, S.V. Olesik, Analytical Chemistry 75 (2003)

5539. [70] M. de Geus, R. Peters, C.E. Koning, A. Heise, Biomacromolecules 9 (2008) 752. [71] W. Hiller, A. Brüll, D. Argyropoulos, E. Hoffmann, H. Pasch, Magnetic Resonance

in Chemistry 43 (2005) 729. [72] W. Hiller, P. Sinha, H. Pasch, Macromolecular Chemistry and Physics 210 (2009)

605. [73] Y. Mengerink, R. Peters, S. van der Wal, H.A. Claessens, C.A. Cramers, Journal of

Chromatography A 949 (2002) 337. [74] X. Jiang, A. van der Horst, V. Lima, P.J. Schoenmakers, Journal of Chromatography

A 1076 (2005) 51. [75] P. Kilz, Chromatographia 59 (2004) 3. [76] H. Pasch, C. Brinkmann, H. Much, U. Just, Journal of Chromatography A 623 (1992)

315. [77] S.J. Kok, T. Hankemeier, P.J. Schoenmakers, Journal of Chromatography A 1098

(2005) 104. [78] R. Edam, D.M. Meunier, E.P.C. Mes, F.A. Van Damme, P.J. Schoenmakers, Journal

of Chromatography A 1201 (2008) 208. [79] J.-A. Raust, L. Houillot, M. Save, B. Charleux, C. Moire, C.l. Farcet, H. Pasch,

Macromolecules 43 (2010) 8755. [80] A. Barlow, R.S. Lehrle, J.C. Robb, Polymer 2 (1961) 27. [81] M. Hakkarainen, A.-C. Albertsson, S. Karlsson, Polymer Degradation and Stability

52 (1996) 283.


31

[82] M. Hakkarainen, S. Karlsson, A.-C. Albertsson, Journal of Applied Polymer Science 76 (2000) 228.

[83] M. Hakkarainen, S. Karlsson, A.C. Albertsson, Polymer 41 (2000) 2331. [84] M. Hakkarainen, Polymer Degradation and Stability 95 (2010) 270. [85] C. Torres, C. Otero, Enzyme and Microbial Technology 29 (2001) 3. [86] K. Urakami, A. Higashi, K. Umemoto, M. Godo, C. Watanabe, K. Hashimoto, Chem.

Pharm. Bull. 49 (2001) 203. [87] P. Plikk, K. Odelius, M. Hakkarainen, A.C. Albertsson, Biomaterials 27 (2006) 5335. [88] R.P. Burford, J.K. Haken, P.A.D.T. Vimalasiri, Journal of Chromatography A 321

(1985) 295. [89] R.R. Mallepally, I. Smirnova, W. Arlt, M. Seiler, S.K. Klee-Laquai, G. Hills, Journal

of Applied Polymer Science 112 (2009) 1873. [90] C. Eldsäter, A.-C. Albertsson, S. Karlsson, International Journal of Polymer Analysis

and Characterization 5 (2000) 415. [91] H. Matsubara, A. Yoshida, Y. Kondo, S. Tsuge, H. Ohtani, Macromolecules 36

(2003) 4750. [92] E.R. Kaal, G. Alkema, M. Kurano, M. Geiler, H.-G. Janssen, Journal of

Chromatography A 1147 (2007) 182. [93] E.R. Kaal, M. Kurano, M. Geiler, H.-G. Janssen, Journal of Chromatography A 1186

(2008) 222. [94] A. Chojnacka, A. Ghaffar, A. Feilden, K. Treacher, H.-G. Janssen, P. Schoenmakers,

Analytica Chimica Acta (2011). [95] L. Li, MALDI mass spectrometry for synthetic polymer analysis, John Wiley &

Sons, Inc. Hoboken, New Jersey., 2010. [96] C.N. McEwen, P.M. Peacock, Analytical Chemistry 74 (2002) 2743. [97] S.M. Weidner, S. Trimpin, Analytical Chemistry 82 (2010) 4811. [98] M.C. McMaster, in, John Wiley & Sons, Inc., Hoboken, New Jersey, 2005. [99] F. Hillenkamp, J. Peter-Katalinić, MALDI MS: A practical guide to instrumentation,

method and applications, WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim, 2007.

[100] A. Höglund, M. Hakkarainen, A.-C. Albertsson, Biomacromolecules 11 (2009) 277. [101] A. Höglund, M. Hakkarainen, U. Edlund, A.-C. Albertsson, Langmuir 26 (2010) 378. [102] P. Rizzareli, D. Zampino, L. Ferreri, G. Impallomeni, Anal. Chem. 83 (2011) 654. [103] P. Rizzarelli, C. Puglisi, G. Montaudo, Rapid Communications in Mass Spectrometry

20 (2006) 1683. [104] P. Rizzarelli, C. Puglisi, G. Montaudo, Rapid Communications in Mass Spectrometry

19 (2005) 2407. [105] J. Luo, G. Giguère, X.X. Zhu, Biomacromolecules 10 (2009) 900. [106] S.M. Weidner, J. Falkenhagen, K. Knop, A. Thünemann, Rapid Communications in

Mass Spectrometry 23 (2009) 2768. [107] J. Chen, J.A. Gardella, Macromolecules 32 (1999) 7380. [108] A.M. Belu, D.J. Graham, D.G. Castner, Biomaterials 24 (2003) 3635. [109] T. Gruendling, S. Weidner, J. Falkenhagen, C. Barner-Kowollik, Polymer Chemistry

1 (2010) 599. [110] S. Trimpin, D.E. Clemmer, Analytical Chemistry 80 (2008) 9073. [111] F.A. Bovey, P.A. Mirau, NMR of Polymers, Academic Press, 1996. [112] D.M. LeMaster, G. Hernández, Macromolecules 33 (2000) 3569. [113] H. Uyama, T. Fukuoka, I. Komatsu, T. Watanabe, S. Kobayashi, Biomacromolecules

3 (2002) 318.

Chapter 1

32

[114] M.V. Pergal, V.V. Antić, M.N. Govedarica, D. Goäevac, S. Ostojić, J. Djonlagić, Journal of Applied Polymer Science (2011) n/a.

[115] Y. Pang, Q. Zhu, J. Liu, J. Wu, R. Wang, S. Chen, X. Zhu, D. Yan, W. Huang, B. Zhu, Biomacromolecules 11 (2010) 575.

[116] M.P. Shaver, D.J.A. Cameron, Biomacromolecules 11 (2010) 3673. [117] T.R. Cooper, R.F. Storey, Macromolecules 41 (2008) 655.

A. Ghaffar, P.G. Verschuren, J.A.J. Geenevasen, T. Handels, J. Berard, B. Plum, A.A. Dias, P.J. Schoenmakers, Sj. van der Wal, Journal of Chromatography A, 1218 (2011) 449-458.

Chapter 2

2. Fast in vitro hydrolytic degradation of polyester urethane acrylate

biomaterials: structure elucidation, separation, and quantification of

degradation products

Synthetic biomaterials have evoked extensive interest for applications in the field of health

care. Prior to administration to the body a quantitative study is necessary to evaluate their

composition. An in vitro method was developed for the quick hydrolytic degradation of

poly(2-hydroxyethyl methacrylate) (pHEMA), poly(lactide-co-glycolide50:50)1550-diol

(PLGA(50:50)1550-diol), PLGA(50:50)1550-diol(HEMA)2 and PLGA(50:50)1550-diol(etLDI-

HEMA)2 containing ethyl ester of lysine diisocyanate (etLDI) linkers using a microwave

instrument. Hydrolysis time and temperature were optimized while monitoring the degree of

hydrolysis by 1H NMR spectroscopy. Complete hydrolytic degradation was achieved at

120°C and 3 bar pressure after 24 h. Chemical structure elucidations of the degradation

products were carried out using 1H and 13C NMR spectroscopy. The molecular weight (Mw)

of the polymethacrylic backbone was estimated via size-exclusion chromatography coupled

to refractive index detection (SEC-dRI). A bimodal Mw distribution was found

experimentally, also in the pHEMA starting material. The number average molecular

weights (Mn) of the PLGA-links (PLGA(50:50)1550-diol) were calculated by high pressure

liquid chromatography - time-of-flight mass spectrometry (HPLC-ToF-MS) and 1H NMR.

The amounts of the high and low Mw degradation products were determined by SEC-dRI

and, HPLC-ToF-MS, respectively. The main hydrolysis products poly(methacrylic acid)

(PMAA), ethylene glycol (EG), diethylene glycol (DEG), lactic acid (LA), glycolic acid

(GA), and lysine were recovered almost quantitatively.

The current method leads to the complete hydrolytic degradation of these materials and will

be helpful to study the degradation behavior of these novel cross-linked polymeric

biomaterials.

Chapter 2

34

1 Introduction

Synthetic polymeric biomaterials are of high importance in the medical field due to an aging

population and their potential to improve the quality of life [1]. There is a gradual trend to

replace non-degradable materials with degradable materials mainly because of the need to

avoid reinterventions when complications arise with non-degradable materials [2]. This is

most vividly seen with the move in the stent coating area where stable drug eluting coatings

are being replaced with biodegradable coatings [3]. Such kinds of materials have their

potential use as joint and limb replacements [4], artificial arteries [5], and skin [6], contact

lenses [7], dental implants [8], catheters [9], in tissue engineering [10], and as systems for

controlled delivery of drugs [11] etc. An important class of degradable biomaterials are

chemically cross-linked polymeric networks predominantly based on pHEMA and PLGA

[12,13]. Since its birth in 1936 [14] and first reported application for contact lenses in 1960

[15], pHEMA is one of the most extensively studied polymeric biomaterials in biomedical

applications [16] because of its biocompatibility, hydrophilicity, softness, high water content

and permeability [17], but it has poor mechanical properties [18]. However, numerous

studies reported the modification of the hydroxyl group with poly(ε-caprolactone) (PCL) [3],

poly(3-hydroxybutyrate-co-3-hydroxyvalerate) [19], dextran [11], poly(2-

(dimethylamino)ethyl methacrylate) [20], poly(ethylene oxide) [12],, poly(tetrahydrofurfuryl

methacrylate) [21], poly(ethylene glycol)-methacrylate [22], poly(dimethylsiloxane) [23],

sulfopropyl methacrylate [24], and cross-linker to tune the biomechanical properties of the

pHEMA.

PLGA is an FDA-approved biodegradable and biocompatible polymeric biomaterial [25].

PLGA is widely used as a drug delivery matrix using numerous forms such as microspheres

[26], nanoparticles [27], scaffold [28], microfibers [29], tablets [30], in the field of control

release delivery devices, and tissue engineering. Currently, the focus on synthesis of

copolymers of PLGA with other polymers has been increased such as PLGA-PCL-PLGA

[31], MeO-PEG-PLGA-PEG-OMe [32], PLGA-PEG [33], and PLGA-grafted dextran [34].

Chemical and enzymatic hydrolysis are the primary biodegradation mechanisms for such

materials. Phagocyte-derived oxidants, produced as a result of foreign body response, may

also contribute to the in vivo degradation of aliphatic ether groups in these networks [35].

The suitability of the polymeric biomaterials for medical devices can be inferred from their

chemical structure, the degradation time and the biocompatibility of the polymers and their

degradation products [11]. Swelling ratios (water contents) of the hydrogels [10,12], weight

Fast in vitro hydrolytic degradation of polyester urethane acrylate biomaterials

35

loss [10,23], pH of the medium [36], kinetic chain length [37], and so on are the most

common parameters used to assess the in vitro degradation of material. These parameters

may be insensitive in the early stages of degradation and are not very informative on

toxicology. Chromatographic methods that can give more insight into the structure of these

networks and can be used to predict their properties more accurately are desired. However,

networks lack solubility, a prerequisite for such analysis. This requires a very sensitive

method of analysis, or at least an accelerated in vitro chemical hydrolysis of the novel

biomaterials at extreme pH values or high temperature, possibly avoiding the formation of

any insoluble product, followed by the structural analysis and quantification of their

degradation products. The collected information will be helpful not only (i) to ascertain the

composition of the original networks, but also (ii) to evaluate the biocompatibility of these

polymeric networks and their degradation products and (iii) to modify the existing and to

design new biomaterials for specific applications. Recently, H. Matsubara et al. reported a

supercritical methanolysis to achieve the selective decomposition at ester linkages in a UV-

cured acrylic ester resin to characterize the cross-linking structures, but no quantification of

the decomposition products was done to assess the degree of methanolysis [38].

A more detailed second approach to study these prospective biomaterials is a chemical or a

specific enzymatic degradation during physiological conditions, allowing one to study the

kinetics of degradation. Again, specific and sensitive chromatographic methods will be

needed to draw sound conclusions. In particular a method is needed as the second stage in a

two-step procedure and is reported here. First degradation under physiologically relevant

conditions is performed, resulting in partially degraded material of which the constituents

may be identified. Then complete and fast degradation of the products of the first step

(oligomers, intermediates and other products) is executed for quantification.

In the present study polymeric biomaterials based on pHEMA (backbone) and

PLGA(50:50)1550-diol (PLGA-links) were subjected to fast hydrolytic degradation. One

reason to select these samples is that pHEMA, frequently formed as an intermediate

hydrolysis product in polymeric network biomaterials, is only partially hydrolyzed under

physiologically relevant conditions [11] and no detailed study on the complete hydrolytic

degradation and direct analysis of its degradation products has yet been published to our

knowledge.

In this paper first the development and optimization of a method for the microwave-assisted

in vitro hydrolytic degradation is reported of pHEMA, PLGA(50:50)1550-diol and the photo-

crosslinked polymeric biomaterials such as PLGA(50:50)1550(HEMA)2 and

Chapter 2

36

PLGA(50:50)1550-diol(etLDI-HEMA)2. The hydrolysis of polymeric biomaterials and their

model building blocks, pHEMA (backbone) and PLGA(50:50)1550-diol were performed at up

to 120oC, for different periods of time. The hydrolysis time and the temperature were

optimized while monitoring the degree of hydrolysis of the starting material with 1H NMR

spectroscopy. Then the structure elucidations of the degradation products (Figure 1) were

carried out using 1H and 13C NMR spectroscopy and quantification of high Mw hydrolyzed

polymethacrylic acid backbone by SEC-dRI and LA, GA, EG, DEG, and lysine by HPLC-

ToF-MS in the hydrolyzed sample are reported. The Mw distribution of the hydrolyzed

backbone was estimated via SEC-dRI. The Mn of the PLGA-links was measured by HPLC-

ToF-MS and 1H NMR.

2 Experimental

2.1 Materials

DL-lactide and Glycolide were purchased from PURAC (CSM Biochemicals, Gorinchem,

The Netherlands), ethyl ester of lysine diisocyanate from Kyowa Hakko Europe GmbH

(Dusseldorf, Germany), caprolactone from Solvay, methacryloyl chloride via Fluka. Irganox

1035 was obtained from Ciba-Geigy (Basel, Switzerland). pHEMA [Mv = 300 kDa (192,066)

or 20 kDa (529,265), solvent and temperature conditions of Mv (viscosity average MW)

determination are not known] and all other chemicals were purchased from Sigma- Aldrich

(St Louis, MO, USA). The chemicals were used as such unless otherwise stated. In all the

experiments deionized water was used.

The experimental batches of PLGA(50:50)1550-diol, PLGA(50:50)1550-diol(HEMA)2 and

PLGA(50:50)1550-diol(etLDI-HEMA)2 were synthesized at DSM Biomedical, Geleen,

Netherlands, according to the following procedure:

Preparation of PLGA(50:50)1550-diol : DL-Lactide (51.6 g, 0.358 mol), glycolide (41.5 g,

0.358 mol) and diethyleneglycol (6.85 g, 6.45 mmol) were weighed in the glovebox and

melted at 150oC under nitrogen conditions. 1 mL of a stock solution Tin(II)-ethylhexanoate

(290 mg in 10 mL n-hexane) was added as a catalyst. The reaction was allowed to proceed

for 18 h upon which the reaction mixture was cooled to room temperature to obtain

poly(lactide-co-glycolide50:50)1550-diol [39].

Preparation of PLGA(50:50)1550-diol(HEMA)2: poly(lactide-co-glycolide50:50)1550-diol (100

g, 65 mmol), 200 mg Irganox 1035 and triethylamine (13.05 g, 0.129 mol) were dissolved in


37

150 ml dry tetrahydrofuran (THF). Methacryloylchloride (13.49 g, 0.129 mol) was added

drop wise to the solution at controlled temperature (<5°C). Immediately a white precipitate

was visible (triethylamine.HCl salt). The dropping funnel was rinsed with THF (50 ml). The

reaction mixture was stirred at room temperature for 18 h. The reaction mixture cooled till

5°C and filtered to remove the triethylamine.HCl salt. The THF was removed via

evaporation with a rotavapor. The remainder was dissolved in 200 ml ethyl acetate. The clear

solution was extracted once with 300 ml 0.1 HCl solution, once with 300 ml 5% NaCl-

solution and 300 mL water. The resulting solution was dried with Na3SO4 and evaporated to

dryness. Poly(lactide-co-glycolide50:50)1550-dimethacrylate was obtained as a slightly

coloured yellow oil. 30.49 g poly(lactide-co-Glycolide50:50)1550-dimethacrylate, 13.1 g

HEMA and 0.86 g Darocur 1173 were mixed in a clear formulation [39].

Micro-particles preparation of PLGA(50:50)1550-diol(HEMA)2: 10.52 g of this formulation

was mixed with 39.88 g PEG 35k (40% m/m in water), 30 g water and 5 g aceton. This

mixture was stirred mechanically for 10 min at 800 rpm before polymerization. The

polymerization was allowed to proceed for 60 min under UV light (Macam Flexicure

controller, D-bulb, 200 mW/s/cm2, Livingston, U.K.). After polymerization, the micro-

particles were filtered through a 0.8 µm filter (Supor-800, Gelman Sciences, Ann Arbor, MI,

USA) under vacuum and rinsed with 250 ml water. The morphology was checked with light

microscopy. The methacrylate conversion was >96%. The micro-particles were sieved

afterwards using ethanol as a solvent (Retsch sieves, aperture 63, 125, and 250 µm, Haan,

Germany). The micro-particles were dried via freeze drying [40].

Preparation of PLGA(50:50)1550-diol(etLDI-HEMA)2: Hydroxymethylacrylate (HEMA, 26

g, 0.20 mol) was added drop wise to a solution of the ethyl ester of Lysine diisocyanoate

(etLDI) (45.25 g, 0.2 mol), Tin-(II)-ethylhexanoate (0.080 g, 0.186 mmol), Irganox 1035

(0.260 g) and dry air at controlled temperature (<5°C). Subsequently the reaction mixture

was stirred overnight at 40°C. The etLDI-HEMA was obtained as a slightly yellow oil. The

reaction was monitored with gel-permeation chromatograpgy (GPC). Poly(lactide-co-

glycolide50:50)1550-diol (100 g, 0.064 mmol) was dissolved in 150 ml dry THF. etLDI-

HEMA (46.05 g, 0.129 mol) was added to the reaction mixture at room temperature.

Subsequently the reaction mixture was stirred overnight at 40°C. In the morning the reaction

mixture was analysed with IR (no NCO peak ν = 2260 cm-1 visible). The reaction was

complete, based on IR spectroscopy when all the THF was evaporated. The poly(lactide-co-

glycolide50:50)1550-(etLDI-HEMA)2 was obtained as a yellowish oil [40].

Chapter 2

38

1M

KO

H

2

*

*

OO

H

CH

31

n

xy

5

OH

5

OH

OH

9

OH

O

CH

38

NH

2

15

14

13

12

11

OH

O

NH

2

OH

10

OH

O

OH

6

7

O

7

6

OH

PM

AA

EG

DE

GLA

GA

Lysine

++

++

+

OO

OO

OO

O

O

O

O

O

NH

OO

CH

3

O

NH

O

O

NH

OO

CH

3

O

NH

O

O

CH

3C

H3

O

O**

CH

3

O

O**

CH

3

xy

25

24

O

24

25

OO

21

O23

O23

O21

O

O

O

O

OC

H3

20

CH

320

22

O19

OH

O

O

CH

318

19

O22

OH

O

OC

H3

18

xy

OO

OO

OO

O

O

O

O

OC

H3

CH

3

O

O**

CH

3

O

O**

CH

3

OH

O

3

4

O

**

CH

3

( a)

( b)

( c)

( d)

+O

H

16

CH

3+

CO

2

1M

KO

H

2

*

*

OO

H

CH

31

n

xy

5

OH

5

OH

OH

9

OH

O

CH

38

NH

2

15

14

13

12

11

OH

O

NH

2

OH

10

OH

O

OH

6

7

O

7

6

OH

PM

AA

EG

DE

GLA

GA

Lysine

++

++

+

OO

OO

OO

O

O

O

O

O

NH

OO

CH

3

O

NH

O

O

NH

OO

CH

3

O

NH

O

O

CH

3C

H3

O

O**

CH

3

O

O**

CH

3

1M

KO

H

2

*

*

OO

H

CH

31

n

xy

5

OH

5

OH

OH

9

OH

O

CH

38

NH

2

15

14

13

12

11

OH

O

NH

2

OH

10

OH

O

OH

6

7

O

7

6

OH

PM

AA

EG

DE

GLA

GA

Lysine

++

++

+

OO

OO

OO

O

O

O

O

O

NH

OO

CH

3

O

NH

O

O

NH

OO

CH

3

O

NH

O

O

CH

3C

H3

O

O**

CH

3

O

O**

CH

3

xy

25

24

O

24

25

OO

21

O23

O23

O21

O

O

O

O

OC

H3

20

CH

320

22

O19

OH

O

O

CH

318

19

O22

OH

O

OC

H3

18

xy

25

24

O

24

25

OO

21

O23

O23

O21

O

O

O

O

OC

H3

20

CH

320

22

O19

OH

O

O

CH

318

19

O22

OH

O

OC

H3

18

xy

OO

OO

OO

O

O

O

O

OC

H3

CH

3

O

O**

CH

3

O

O**

CH

3

OH

O

3

4

O

**

CH

3

( a)

( b)

( c)

( d)

+O

H

16

CH

3+

CO

2

Figure

1. Proposed reaction schem

e for the hydrolytic degradation of (a) pHE

MA

(b) PL

GA

(50:50)1550 -diol (c) PL

GA

(50:50)1550 -diol(HE

MA

)2 . and (d) P

LG

A(50:50)1550 -diol(etL

DI-H

EM

A)2 . P

MA

A represents poly(m

ethacrylicacid); E

G, ethylene glycol; D

EG

, diethyleneglycol; L

A,

lactic acid and GA

, glycolic acid. The num

bering corresponds to NM

R peak assignm

ents in figure 6.


39

Micro-particles preparation of PLGA(50:50)1550-diol(etLDI-HEMA)2: 15.58 g PLGA1550-

diol(etLDI-HEMA)2 and 0.31 g Darocure 1173 were mixed together mechanically at 100

rpm in a 250 ml beaker at 50°C, then 62 g PEG35K (40% m/m in deionized water) and 58 g

deionized water were added. This was stirred mechanically for 30 min at 900 rpm. The

polymerization was allowed to proceed for 60 min at 70°C and 900 rpm under UV light

(Macam Flexicure controller, D-bulb, 200 mW/s cm-2). The particles were wet-sieved with

deionized water over a sieving tower (Retsch test sieve Aperture 250, 125, 63, and 45 µm)

and dried under vacuum at room temperature for 18 h. Afterwards methacrylate conversion

was checked: >98% (FT-IR, 1640 cm-1 and 815 cm-1) [40].

pressure sensor

sensortemperature

microwaveenergy

coolingmedium

PTFElining

pressure sensor

sensortemperature

microwaveenergy

coolingmedium

PTFElining

Figure 2 Schematic diagram of CEM Discover microwave apparatus used in this work, with additional PTFE lining of 1 mm thickness.

2.2 Procedure of hydrolysis

20 or 40 mg of each sample was dissolved in 2 mL of 1 M KOH (Merck, Darmstadt,

Germany) in a 10 mL pressurized glass vial (CEM Corporation, NC, USA) using a magnetic

stirrer. The 10 mL pressurized glass vial (i.d. = 12 mm) was internally lined with a PTFE

Chapter 2

40

tube of 1 mm thickness and i.d. = 11 mm (locally made at the mechanical workshop of the

University of Amsterdam, Figure 2). The homogeneous mixture in the glass vessel was

placed in the microwave instrument (Discover BenchMate, CEM) and hydrolysis to PMAA

and EG was carried out at 120oC, 3 bar and for 24, 20, 15, 10, and 5 h. Similarly,

PLGA(50:50)1550-diol , PLGA(50:50)1550-diol(HEMA)2 and PLGA(50:50)1550-diol(etLDI-

HEMA)2 were hydrolyzed at 120oC for 24 h at 3 bar. The mixture was weighed before and

after each hydrolysis.

2.3 1H NMR spectroscopy of hydrolysate

0.5 mL of each hydrolysis solution was acidified by adding carefully a few drops of 37%

HCl with vigorous stirring at 90ºC. The PMAA precipitates and along with supernatant

(containing ethylene glycol (EG), diethylene glycol (DEG), lactic acid (LA), glycolic acid

(GA), lysine etc.) were dried overnight at 40oC with an air flush.

The dried mixtures of the hydrolysates were re-dissolved in d4-methanol (Euriso-top,

France). Samples of un-hydrolyzed pHEMA and PLGA(50:50)1550-diol were also prepared

in d4-methanol. 1H NMR spectra were recorded on a Varian Inova 500 MHz NMR (Varian

Inc., USA) equipped with Probe: 500 5 mm 13C/31P/1H GS. Pulse repetition time: 25 sec,

Pulse: 3.6 µsec, Scans: 63 and temperature: 25oC were used to record 1H NMR spectra.

2.4 Size-exclusion chromatography (SEC) analysis

pH neutralized ( 0.2 mL) hydrolysis solutions were diluted with 0.2 mL aqueous SEC mobile

phase. The SEC experiments were performed on an HPLC system equipped with in-line

degasser, Model 600 pump, 717 plus TRI-SEC auto-sampler and Model 410 differential

refractive index detector (all Waters, Milford, MA, USA). Data were recorded and

chromatographic peaks were treated using Empower 2 software (Waters, Milford, MA,

USA). Calculations for molar mass distribution (MMD) on the chromatographic peaks were

executed using software written in-house in Excel 2003 (Microsoft).

All aqueous SEC separations were performed on the following set of columns used in series:

PL Aquagel-OH Guard (8 µm, 50 mm × 7.5mm i.d.), PL Aquagel-OH 50, 30, and 10 (each 8

µm, 300 mm × 7.5 mm i.d.) columns (Polymer Laboratories, U.K.). For 20 kDa pHEMA

hydrolysates, same set of columns was used except PL Aquagel-OH 50. The mobile phase

was (0.2 M NaNO3, 0.01 M NaH2PO4, pH ≈ 7) pumped at a flow rate of 1 mL min-1.


41

Poly(methacrylic acid) sodium salt (PMA-Na) standards (Table 1) were used to calibrate the

SEC-dRI system. The calibration curves for MMD of PMAA in (300 and 20 kDa) pHEMA

hydrolysates are given by cubic relations of logM and retention time, x: log(M) = –

0.00915x3 + 0.51649x2 – 9.98592x + 70.73912, R2 = 0.999 and log(M) = –0.032x3 + 1.218x2

– 15.62x + 73.11, R2 = 0.998, respectively. To quantify the concentration of hydrolyzed

backbone as PMA-Na in hydrolysates the calibration lines were recorded using PMA-Na

standards with Mp 65.8 kDa (at six concentrations 0.2–2 mg mL-1) and 22.5 kDa (at five

concentrations 1–5 mg mL-1). Highly pure water for mobile phase preparation was obtained

by means of an Arium® 611 Ultrapure (18.2 MΩ*cm) Water System (Sartorius AG,

Goettingen, Germany).

Table 1 Peak molecular weight (Mp), weight average molecular weight (Mw), number average molecular (Mn) and dispersity (PDI) of the Poly (methacrylic acid) sodium salt standards. Data as specified by the supplier.

Standard Mp (D) Mw (D) Mn (D) PDI Supplier

PMA-Na-1 1220 1250 1040 1.197 PSS

PMA-Na-2 1670 1700 1520 1.120 PSS

PMA-Na-3 3180 3150 2700 1.169 PSS

PMA-Na-4 7830 7750 7220 1.073 Fluka

PMA-Na-5 8210 8280 7480 1.108 PSS

PMA-Na-6 22,500 22,100 21,100 1.047 PSS

PMA-Na-7 31,500 31,100 30,400 1.023 Fluka

PMA-Na-8 65,800 62,500 60,600 1.031 PSS

PMA-Na-9 78,300 75,100 73,300 1.025 Fluka

PMA-Na-10 201,000 192,000 186,000 1.029 PSS

PMA-Na-11 480,000 421,000 380,000 1.108 PSS

PMA-Na-12 549,000 483,000 429,000 1.126 Fluka

Size-exclusion chromatography of pHEMA (300 and 20 kDa) was performed on two PL gel

MIXED-C (5 µm, 300 mm × 7.5 mm i.d.) columns with DMF (Acros Organics, NJ, USA)

containing 0.02 M lithium chloride (Acros Organics) as a mobile phase pumped at a flow

rate of 1 mL min-1 via an LC-10AD solvent delivery module coupled with a RID-10A dRI

detector (Shimadzu Corporation, Kyoto, Japan). A Rheodyne 7120 manual injector

(Rheodyne Europe GmbH, Alsbach, Germany) with 20 µL loop was used as an injection

system.

The resolving power of a SEC system can be visualized by an integrity plot, which gives the

integrity index (IIsec) as a function of sample Mw and (Mw/Mn-1) [41]. (IIsec) indicates the

fraction of dispersion of the experimental peak variance that is caused by the polydispersity

of the sample itself and not by dispersion due to the column or extra-column band

Chapter 2

42

broadening [42]. The integrity plot for the used SEC system was constructed using the

polymer standards listed in Table 1 and clearly demonstrates its suitability even for narrowly

distributed polymers in the range of 2–200 kDa (cf. Figure 3).

Figure 3 Experimental SEC-integrity plot as a function of the sample (horizontal axis;Mw/Mn-1) proportional to log(PDI-1) and molecular weight (vertical axis; log M)). System: PL aquagel-OH Guard (8 µm, 50 mm × 7.5 mm i.d.), PL aquagel-OH 50, 30 and 10 ( each 8 µm, 300 mm × 7.5 mm i.d.) columns; mobile phase: 0.2 M NaNO3, 0.01 M NaH2PO4, pH ≈ 7) pumped at a flow rate of 1 mL min-1.

2.5 HPLC-ESI-ToF-MS analysis of hydrolysate

0.1 mL of the hydrolysate was mixed with 4 mL of deionized water and then pH neutralized.

The mixture was filtered with a 0.2 µm pore size PTFE filter (Grace Davison discovery

science, IL, USA), prior to injection. Stock solutions of LA (Fluka), GA (Fluka), EG

(Aldrich), DEG (Fluka) and D-lysine (Sigma) were prepared by dissolving in 1 M KOH

solution and quantification was done with a standard addition method in order to correct for

signal suppression of target analytes by co-eluting compounds. The chromatographic

separations were performed on a Prevail C18 column (250 × 4.6 mm i.d., 5 µm particle size,

Alltech Discovery Sciences, IL, USA) at a temperature of 35ºC. The injection system

consisted of a Rheodyne 7010 manual injector (with 5 or 20 µL loops). The aqueous mobile

phase containing 0.1% (v/v) formic acid (Fluka), 0.03% (w/v) sodium iodide (NaI) Aldrich)

and 1% (v/v) acetonitrile (Biosolve) was pumped via Shimadzu LC-20AD solvent delivery

module at 2 mL min-1 and was split between a waste reservoir and electrospray ionization


43

(ESI) interface by means of a zero dead volume T-piece to enssure a flow of approximately

0.2 mL min-1 into the ESI interface.

In order to evaluate the recovery of the procedure of the method about 10 mg each of PMA-

Na standard (Mp = 78.3 kDa), DEG, EG and D-lysine and approximately 50 mg of LA and

GA standards were dissolved in 5 mL (5.2 g) of 1 M KOH to make control solutions. 2 mL

of the mixture was heated at 120oC and 3 bar in the microwave for 24 h. The concentration

of each analyte was determined before and after heating to calculate the percentage recovery

of the method for each analyte.

The LC system was hyphenated with an Agilent 6210 series ToF-MS (Agilent Technologies,

Waldbronn, Germany) via an ESI interface. The conditions of the ESI-ToF-MS were as

follows: drying gas was nitrogen (N2) at 8 L min-1; and at 300oC; 30 psig of N2 ; capillary

voltage, 3500 V; fragmenter, 140 V; skimmer voltage, 60 V; octopole dc1, 33 V; octopole

radio frequency, 250 V. The data were acquired in the scan mode from m/z 50 to 500 D with

0.88 scans/sec. An Agilent MassHunter Workstation A.02.01 and AnalystTM QS 1.1 software

(Applied Biosystems) were used for data acquisition and data analysis, respectively.

3 Results and discussion

3.1 Optimization of hydrolysis method

Initially the hydrolysis method was optimized by degrading the pHEMA (300 and 20 kDa)

in a 10 mL pressurized glass vial specially designed for the CEM microwave instrument.

After hydrolysis the glass vessel contained the hydrolysis solution and white material, stuck

on the wall of the glass vessel. We were not able to dissolve this material, for further

analysis by NMR, SEC or HPLC, except at very low pH. The residues were originally

considered to be silicates. For further analysis, these residues were washed three times with 5

mL of water, methanol and DMF to wash out possible impurities of hydrolyzed and un-

hydrolyzed pHEMA and dried overnight at 210oC in an oven. The XRF (X-ray fluorescence)

spectrum (Eagle-III Spectrometer, EDAX Inc., Mahwah, NJ, USA) of these residues

confirmed the presence of silicates primarily originating from the glass vessel in alkaline

conditions at high temperature (Figure 4). However, the CHN elemental analysis (Truspec,

Leco, Germany) also revealed the presence of carbon contents in this material. Based on the

percentage of these carbon contents, it can be concluded that up to 35% of the starting

material (pHEMA) is lost by inclusion in the white residue from the hydrolysis solution. It

Chapter 2

44

may be assumed that at high temperature the highly reactive silanol groups present on

silicates react with hydroxyl groups of pHEMA [43].

The formation of white residues during hydrolysis will lead to wrong quantification of the

relative percentage of starting material in the hydrolysis solution, so to avoid contact of the

alkaline solution with the inner surface of the glass vessel (to prevent the formation of white

residues) it was internally lined with PTFE (Figure 2). The hydrolysis was performed

repeatedly after this modification and no formation of white residues was observed. The

hydrolysis time was optimized while monitoring the cleavage of ester linkages in pHEMA

with 1H NMR spectroscopy. Then the hydrolysis of PLGA(50:50)1550-diol,

PLGA(50:50)1550-diol(HEMA)2, and PLGA(50:50)1550-diol(etLDI-HEMA)2 was conducted

at 120oC for 24 h.

Figure 4 XRF spectrum of white residues without lining the pressurized glass vessel with PTFE.

3.2 Product identification

The overlay of 1H NMR spectra of pHEMA hydrolyzed for different times (Figure 5) show

clearly the cleavage of ester groups of pHEMA i.e. the scission of side chains from the

backbone chain and the formation of free ethylene glycol (peak 5 at δ 3.6 ppm) and a small

quantity of diethylene glycol (peak 7 at δ 3.56 ppm and peak 6 at δ 3.68 ppm). The peaks of

free ethylene glycol and diethylene glycol were confirmed by taking the 1H NMR spectrum


45

of sample spiked with EG and DEG standards in d4-methanol. The signals at δ 3.75 ppm

(peak 4) and δ 4.21 ppm (peak 3) corresponding to the side chain of pHEMA almost

vanished after 24 h, indicating the degree of hydrolysis of pHEMA. In the starting material,

before hydrolysis the positions of peaks 3 and 4 were observed at δ 4.08 and δ 3.80 ppm,

respectively. DEG observed in the 1H NMR spectra is present as an impurity in pHEMA.

The 1H-1H gCOSY (two-dimenssional homonuclear H, H gradient-correlated spectroscopy)-

correlated NMR experiment indicates the proton connectivity between signals at δ 3.80 and δ

4.08 ppm in the starting material and at δ 3.75 and δ 4.21 ppm (peaks 3 and 4) for the

hydrolysate affirming the presence of the side chain in both the starting material and the

hydrolysate. Also, the proton connectivity in the signals (peaks 6 and 7) of diethylene glycol

was confirmed. The 13C NMR spectra in DEPT135 mode were recorded to assign the

methyl, methane and methylene group and quaternary carbon in both the starting material

and the hydrolyzed products. The single bond connectivities between 1H and 13C were also

determined by the two-dimensional 13C, 1H-correlated HSQC (heteronuclear single quantum

coherence) NMR experiments.

*

CH3

HO

OCH2

*

HO

H2C

CH2

OH

*

CH3

O

CH2

OCH2

*

H2C

OH

+

n

n

H2C

CH2

OHHOCH2

H2C

O

+

1 2

7

5

6

3

1

2

7

5

6

4

2

3 76

1

5

4

ab

ced

*

CH3

HO

OCH2

*

HO

H2C

CH2

OH

*

CH3

O

CH2

OCH2

*

H2C

OH

+

n

n

H2C

CH2

OHHOCH2

H2C

O

+

1 2

7

5

6

3

1

2

7

5

6

4

2

3 76

1

5

4

ab

ced

Figure 5 An overlay of 1H NMR spectra (CD3OD, 25 oC, 500 MHz) of pHEMA (300 kDa) hydrolyzed after different times of hydrolytic degradation (a) 5 h (b) 10 h (c) 15 h (d) 20 h (e) 24 h at 120 oC in the microwave instrument.

The relative percentage of non-hydrolyzed pHEMA in the hydrolysis solution of pHEMA

(300 kDa) and pHEMA (20 kDa) was determined using the following equation.

Chapter 2

46

100 4

5pHEMA of % Relative

0.20.1

3.47.3

A

A (1)

In which Ax is the peak area for the response at the shift of x ppm.

The relative percentage of un-hydrolyzed pHEMA in the hydrolysate of 300 kDa pHEMA

decreases from 6% at 5 h to less than 0.2% at 24 h. pHEMA with Mv = 20 kDa degraded

much faster .

The 1H NMR spectra of PLGA(50:50)1550-diol before (Figure 6A) and after hydrolysis

(Figure 6B) show the cleavage of ester bonds in PLGA(50:50)1550-diol chains and the

formation of free LA (CH3 = 1.33 ppm and CH = 4 ppm), GA (CH2 = 3.88 ppm) and DEG

(HO−CH2− = 3.7 ppm and −O−CH2− = 3.58 ppm). More acidic hydroxyl groups are formed

after hydrolysis and this leads to more hydrogen bonding. Consequently, the hydroxyl group

signal shift towards higher frequency after hydrolysis. The multiple signals for the

methylene groups (number 22) indicate sequential or tacticity effects (starting material is

DL-lactide). Unfortunately, the signal at 1.15 ppm is an isopropanol impurity (an

experimental artifact).

The NMR of PLGA(50:50)1550-diol(HEMA)2 and PLGA(50:50)1550-diol(etLDI-HEMA)2

polymeric biomaterials were not possible because of their lack of solubility in any of the

available NMR solvents. Upon hydrolysis it was possible to dissolve the acidified and non-

acidified degraded backbone (PMAA/PMA-NA) in CD3OD and D2O, respectively.

PLGA(50:50)1550-diol(HEMA)2 is a 3D ladder like polymer network consisting of backbones

of pHEMA interconnected via cross-links of PLGA(50:50)1550-diol. If all the ester bonds are

cleaved, the polymer network should result in polymethacrylate backbone, EG, DEG, LA,

and GA. The 1H NMR of hydrolyzed PLGA(50:50)1550-diol(HEMA)2 (Figure 6C) shows that

the CH3− signal (peak 1) that belongs to the poly methacylate backone is overlapped by the

CH3− signal of LA (peak 8). This was confirmed by 1H-1H gCOSY NMR experiments. The

peak 2 belongs to the −CH2− group of the backbone. Rest of the signals corresponds to the

hydrolyzed PLGA(50:50)1550-diol 1H NMR spectrum (Figure 6B). The peak that appears at

2.72 ppm is unknown.

The PLGA(50:50)1550-diol(etLDI-HEMA)2 is similar in chemical structure to

PLGA(50:50)1550-diol(HEMA)2 except that the backbones of pHEMA and PLGA-links

(PLGA(50:50)1550-diol) are interconnected via ethyl ester of lysine diisocyanate (etLDI)

linkers. So upon complete hydrolytic degradation all the products should be the same as in

case of PLGA1550-diol(HEMA)2 except for lysine and ethanol. The signals 11, 12, 13, 14,


47

and 15 (Figure 6D) belong to lysine (Figure 1). The signal bond connectivity between lysine

signals were established by H-H gCOSY. The peak 16, a triplet, belongs to a −CH2− group

of ethanol. Peak 17 belongs to PEG present as an impurity in the biomaterial. Mn and the

molar ratios of LA and GA in the PLGA(50:50)1550-diol were calculated by integrating the

CH group of LA, −CH2− group of GA and -CH2- groups of DEG (Table 2).

8 25

11

10

6 7

9

-OH

8 2510

6 7

9

12

-OH

-OH

810

6 79

13

14

15 16

-OH CD3OD

CD3OD

CD3OD

CD3OD

17

6 7

A

B

C

D

18 20

19

21

22

23 24

25

?

1

1

8 25

11

10

6 7

9

-OH

8 2510

6 7

9

12

-OH

-OH

810

6 79

13

14

15 16

-OH CD3OD

CD3OD

CD3OD

CD3OD

17

6 7

A

B

C

D

18 20

19

21

22

23 24

25

?

1

1

Figure 6 The 1H NMR spectrum of (A) PLGA(50:50)1550-diol before hydrolysis and (B) PLGA(50:50)1550-diol (C) PLGA(50:50)1550-diol(HEMA)2 (D) PLGA(50:50)1550-diol(etLDI-HEMA)2 after hydrolysis at 120oC for 24 h. The numbering of peaks corresponds to the numbering in Figure 1.

HPLC-ToF-MS is much more sensitive than NMR, but it is an indirect method:

determination of LA and GA occurs after hydrolysis. Still, the Mn of the PLGA-links

measured by HPLC-ToF-MS is considered to be more accurate, because in NMR accurate

integration of the peak area is difficult as the signals of EG and DEG slightly overlap. This

results in apparently higher Mn values by NMR except in case of PLGA(50:50)1550-diol links

(Table 2).

Chapter 2

48

Table 2 Mn of PLGA-links (PLGA(50:50)1550-diol) and molar ratios of LA and GA calculated by 1H NMR and HPLC-ToF-MS.

1H NMR HPLC-ToF-MS Sample

Mn

(Da)

% RSD

(n=3)

Molar ratio

(LA:GA)

Mn

(Da)

% RSD

(n=3)

Molar ratio

(LA:GA)

PLGA(50:50)1550-diol 1317 5 51 : 49 1371 2 49 : 51

PLGA(50:50)1550-diol(HEMA)2 1549 2 50 : 50 1001 3 48 : 52

PLGA(50:50)1550-diol(etLDI-HEMA)2 1791 2 50 : 50 1028 4 49 : 51

3.3 Molar mass characterization and quantification of PMAA

Figure 7 is showing the aqueous SEC separation for the hydrolyzed backbone (in the form of

PMA-Na) for the samples subjected to hydrolysis: pHEMA 20 kDa (a), pHEMA 300 kDa

(b), PLGA(50:50)1550-diol(HEMA)2 (c) and PLGA(50:50)1550-diol(etLDI-HEMA)2 (d). All

the peaks show a non-Gaussian distribution. One potential cause of this bimodal MMD is

that two chains are cross-linked due to the esterification reaction between alcohol in one

chain and carboxylic acid in the other chain. There is no evidence of cross-linking found in 1H NMR spectra, so the extent of cross-linking is very small and the signals may not appear

in the spectra. Zainnuddin et al. suggested that the presence of ions in the solute can promote

the formation of physical crosslinks between two neighboring hydrophilic and hydrophobic

groups of the polymer chains [44]. However, it is unlikely that these physical cross-links

stay intact when subjecting the sample to SEC.

In order to see the effect of hydrolysis on bimodality of the hydrolysate product, PMA-Na

standard with Mp 549 kDa, which showed a uni-modal MMD without hydrolysis, was

hydrolyzed in 1 M KOH with and without EG at 120oC and 3 bar for 24 h. The hydrolysates

were subjected to SEC and no influence of the hydrolysis was observed on the uni-modality

of the MMD. The bimodal distribution of PMA-Na was independent of injection volume.

pHEMA with low a Mw is soluble in water but as the Mw increases, its solubility in water

decreases [45].

Therefore, to further investigate the origin of bimodality in the hydrolyzed backbone, SEC in

DMF of pHEMA (300 and 20 kDa) was performed. The MMD of 300 kDa pHEMA renders

a non-Gaussian behaviour (Figure 8), so the bimodal distribution appears already present in

the starting material.


49

Table 3 SEC-dRI data for the hydrolyzed backbone (PMA-Na)

Sample Mn

(kDa)

% RSD

(n=3)

Mw

(kDa)

% RSD

(n=3)

PDI

% RSD

(n=3)

300 kDa 52 8 108 6 2.1 7

20 kDa - - 16 1 - -

PLGA(50:50)1550-diol(HEMA)2 46 4 223 3 4.7 4

PLGA(50:50)1550-diol(etLDI-HEMA)2 34 8 152 6 4.5 9

The molecular weight data are PMA-Na-relative.

-0.2

0.2

0.6

1

1.4

1.8

0 4 8 12 16 20 24 28 32 36 40

Ret. volume (mL)

dRI

sign

als

c

d

b

-0.2

0.2

0.6

1

1.4

1.8

0 4 8 12 16 20 24 28 32 36 40

Ret. volume (mL)

dRI

sign

als

a

-0.2

0.2

0.6

1

1.4

1.8

0 4 8 12 16 20 24 28 32 36 40

Ret. volume (mL)

dRI

sign

als

c

d

b

-0.2

0.2

0.6

1

1.4

1.8

0 4 8 12 16 20 24 28 32 36 40

Ret. volume (mL)

dRI

sign

als

a

-0.2

0.2

0.6

1

1.4

1.8

0 4 8 12 16 20 24 28 32 36 40

Ret. volume (mL)

dRI

sign

als

a

Figure 7 SEC-dRI chromatograms of (a) pHEMA 20 kDa, (b) pHEMA 300 kDa, (c) PLGA(50:50)1550-diol(HEMA)2, (d) PLGA(50:50)1550-diol(etLDI-HEMA)2 after hydrolysis for 24 h at 120oC and 3 bar pressure.

In 20 kDa pHEMA hydrolysate, the PMAA elutes from 11 mL to 16.2 mL and shows a

small shoulder on the low molecular weight side. Figure 7a indicates that very low molecular

Chapter 2

50

weight (< 1000 Da) PMAA is eluting together with the high concentration of salt. (This

could be solved by application of columns with a resolving range at low molecular weights).

Therefore, the Mn value for 20 kDa pHEMA hydrolysate is not reported in Table 3. The Mn,

Mw and PDI of PMA-Na in the samples are listed in Table 3. The hydrolyzed backbones in

case of PLGA(50:50)1550-diol(HEMA)2 and PLGA(50:50)1550-diol(etLDI-HEMA)2 are more

polydisperse than those of the pHEMA (20 and 300 kDa) standards. The monomeric

products of hydrolysis such as LA, GA, EG, DEG, and lysine elute after the permeation

limit. The quantitative results of hydrolyzed backbone in different biomaterials are tabulated

in Table 5.

-0.05

0.05

0.15

0.25

0.35

0.45

0 5 10 15 20 25

Retentional volume (mL)

dRI

sign

als

a

b

-0.05

0.05

0.15

0.25

0.35

0.45

0 5 10 15 20 25

Retentional volume (mL)

dRI

sign

als

a

b

Figure 8 SEC-dRI chromatograms of starting materials (a) 300 kDa, (b) 20 kDa pHEMA.

3.4 Quantification of monomeric products by HPLC-ToF-MS

The low molecular weight products were analyzed by HPLC-ESI-ToF-MS. The TIC

chromatogram of the PLGA(50:50)1550-diol(etLDI-HEMA)2 hydrolysate is shown in Figure

9. The peaks at 1.59, 2.26, and 4.15 min correspond to lysine, EG, and DEG, respectively.

The peak at 1.68 min is the aggregate of sodium formate clusters. In case of pHEMA (300 or

20 kDa) EG is the major product of hydrolysis as compared to DEG and TEG (triethylene

glycol), present as an impurity in the starting material. TEG elutes at 8.40 min. The peak

intensities of EG are much lower than that of DEG and TEG (Table 4), because alcohols are

not easily charged either in positive or negative mode while DEG can be charged easily due

to the presence of its ether group. The three diols form proton [M+H]+, ammonium

[M+NH4]+, sodium [M+Na]+, and potassium [M+K]+ adducts. It was attempted to promote


51

the formation of protonated molecule or the ammonium and sodium adducts by adding 0.1%

(v/v) formic acid, 0.02% (v/v) ammonia solution (25%) or 0.03% (w/v) NaI to the mobile

phase. The peak intensity increased with NaI addition (cf. Table 4). Therefore, NaI was

selected to make sodium adducts for quantification. Lysine preferably makes [M+H]+ adduct

even in the presence of NaI. Lysine remains un-retained even at very low concentration of

organic modifier. GA and LA in PLGA(50:50)1550-diol, PLGA(50:50)1550-diol(HEMA)2 and

PLGA(50:50)1550-diol(etLDI-HEMA)2 were quantified in negative ESI mode as [M–H] ¯ ion

without NaI in the mobile phase and are eluted at 2.19 and 3.06 min (Figure 10). The

standards prepared in water showed peaks of linear dimer (m/z 161.0432) of LA at 7 and 8

min. To avoid this, all the solutions of standards were first prepared in 1 M KOH to convert

the dimers into monomers and then diluted up to a dilution factor of 20 with deionized water

and pH neutralized with hydrochloric acid.

Table 4 Relative peak intensities of EG, DEG and TEG in pHEMA hydrolysate with different combinations of formic acid (FA), ammonia (NH3) and sodium iodide (NaI) in the aqueous mobile phase for HPLC-ToF-MS containing 1% organic modifier

Intensity (Mcps) Mobile phase

containing

Selected ion

EG DEG TEG

0.1% FA [M+H]+ 0.06 2.60 0.43

0.1% FA + 0.02% NH3 [M+NH4]+ 0.08 1.40 0.40

0.1% FA + 0.03% NaI [M+Na]+ 0.51 4.00 0.56

When quantifying target components in samples one has to take care to avoid matrix

overloading. The undetected co-eluting matrix components may reduce the ionization

efficiency of the analytes and cause poor reproducibility and accuracy [46,47]. As the lysine

elutes at t0, significant signal suppression occurred and the apparent yield (relative to the

assumed structure of the starting materials as in Figure 1) was 25%. To compensate for this

matrix signal suppression, the quantification of components was performed by standard

addition giving a yield of lysine up to 68%. This indicates that either not all of the etLDI is

converted to lysine, may be due to the presence of lysine-diacrylate cross-links (Figure 11B)

or some of the cross-link chains are deficient with etLDI. The summary in Table 5 shows

that all the hydrolyzed samples were recovered quantitatively with respect to the total

amount of sample subjected to hydrolysis, except for pHEMA (20 kDa), which can be

explained by incomplete separation (cf. Figure 7a) and for PLGA(50:50)1550-diol(etLDI-

Chapter 2

52

HEMA)2, because PEG impurities, ethanol and carbon dioxide, which were produced during

the hydrolysis of the last material were not quantified.

In addition to overall recovery, the recovery of separate hydrolysis products when treating a

mixture of the hydrolysis products by the same hydrolysis and analysis procedure was

determined. These recoveries are also indicated in Table 5. Although the recovery of GA is

significantly less than 100%, the yields are acceptable, considering the small sample size.

Figure 9 TIC and XIC chromatogram of PLGA(50:50)1550-diol(etLDI-HEMA)2 hydrolysate. Conditions: positive ESI mode with isocratic elution with water containing 0.1% (v/v) formic acid, 0.03% (w/v) NaI and 1% (v/v) acetonitrile, flow rate 1.5 mL min-1, column C18 Alltech Prevail (250 mm × 4.6 mm i.d., 5 µm). peaks at 1.59, 2.26 and 4.15 min are traces of m/z = 147.12, 85.03, and 129.05 respectively and correspond to [lysine+H]+, [EG+Na]+ and [DEG+Na]+ respectively.


53

Table 6 shows the theoretical and the experimental molar ratios among different components

of pHEMA, PLGA(50:50)1550-diol(HEMA)2, and PLGA(50:50)1550-diol(etLDI-HEMA)2.

The theoretical ratios were estimated from the Figure 1 and the experimental ratios are based

on the quantitative data presented in Table 5. The results indicate that both pHEMA with 300

kDa Mv and PLGA(50:50)1550-diol(HEMA)2 are more deficient in EG than pHEMA with 20

kDa Mv. This suggests either the presence of cross-linking between two neighboring PMA

backbones via esterification and the formation of ethyl diacrylates or the pHEMA as a

starting material is partially hydrolyzed. In case of PLGA(50:50)1550-diol(HEMA)2, the

decrease in the amount of DEG with respect to PMA and EG may either be attributed to the

missing cross-links or to the presence of dangling chains without DEG. For

PLGA(50:50)1550-diol(etLDI-HEMA)2 the molar ratios between different building blocks are

close to those of the ideal structure except the lower amount of lysine compared to DEG.

Figure 10 TIC and XIC chromatogram of PLGA(50:50)1550-diol(HEMA)2 hydrolysate. Conditions: negative ESI mode with isocratic elution with water containing 0.1% (v/v) formic acid and 1% (v/v) acetonitrile, flow rate 1.5 mL min-1, column C18 Alltech Prevail (250 mm × 4.6 mm i.d., 5 µm). Peaks at 2.19 and 3.06 min are traces of m/z = 75.01 and 89.02 respectively and correspond to [GA-H]¯ and [LA-H]¯ respectively.

Chapter 2

54

Based on these experimental ratios between different components, the following structures

could be suggested for PLGA(50:50)1550-diol(HEMA)2 and PLGA(50:50)1550-diol(etLDI-

HEMA)2 (Figure 11).

B

A

BB

AA

Figure 11 Proposed chemical structure of (A) PLGA(50:50)1550-diol(HEMA)2 (B) PLGA(50:50)1550-diol(etLDI-HEMA)2 based on quantitative hydrolysis results (Table 1 and 2). Polymethacrylic acid ( ), ethylene glycol ( ), ethyl ester of lysine diisocyanate ( ), PLGA ( ), and diethylene glycol ( ).

However, it should be stipulated that this aposteriori analysis of the monomers after

hydrolysis only determines averages and cannot discriminate between different distributions,

e.g. of lysines in side chains, which could have implications for degradation of the

material. This stresses that it is imperative to involve analyses in each step of manufacturing,

assessing the starting materials and intermediates as well as the final product. The presented

method was not set up to determine the structure of the biomaterial network, but as a

balancing check accounting for all the resulting degradation products, since response factors

of oligomers are not precisely known, unlike those of the composing monomers.

In degradation studies this method allows quantitative analysis of oligomeric and other

intermediates that constitute the majority of degradation products under physiological

conditions, in the second stage of a two-step procedure: the first step uses physiologically

relevant conditions, while in the second step a fast and complete degradation is executed for

quantification of the final products.


55

Table 5 Sum

mary of the quantitative results of degradation products of biom

aterials performed w

ith HP

LC

-TO

F-MS.

292

4105

595

398

4108

Average %

yield3

20.0

20.05

20.2

20.3

39.8

Average m

ass of sam

ple

41.80

‐‐‐‐

‐‐‐‐

‐‐‐‐

‐‐‐‐

96 ±

4Lysin

e

25.76

47.10

‐‐‐‐

‐‐‐‐

422.43

92 ±

2GA

47.08

49.04

‐‐‐‐

‐‐‐‐

327.18

94 ±

4LA

41.24

61.57

20.13

50.20

43.09

97 ±

2DEG

14

1.39

12.41

77.87

36.27

‐‐‐‐

100 ±3

EG

72.43

25.26

116.81

118.97

‐‐‐‐

99 ±

1PMA‐Na1

% RSD

(n=3)

Amount

(mg)

% RSD

(n=2)

Amount

(mg)

% RSD

(n=2)

Amount

(mg)

% RSD

(n=2)

Amount

(mg)

% RSD

(n=2)

Amount

(mg)

PLG

A(50:50)1550 ‐

diol(etLD

I‐HEM

A)2

PLG

A(50:50)1550 ‐

diol(H

EMA)2

pHEM

A(20 kD

a)pHEM

A(300 kD

a)PLG

A(50:50)1550 ‐d

iol

Reco

very2

+/‐CV (%

)Hydrolysate

samples

292

4105

595

398

4108

Average %

yield3

20.0

20.05

20.2

20.3

39.8

Average m

ass of sam

ple

41.80

‐‐‐‐

‐‐‐‐

‐‐‐‐

‐‐‐‐

96 ±

4Lysin

e

25.76

47.10

‐‐‐‐

‐‐‐‐

422.43

92 ±

2GA

47.08

49.04

‐‐‐‐

‐‐‐‐

327.18

94 ±

4LA

41.24

61.57

20.13

50.20

43.09

97 ±

2DEG

14

1.39

12.41

77.87

36.27

‐‐‐‐

100 ±3

EG

72.43

25.26

116.81

118.97

‐‐‐‐

99 ±

1PMA‐Na1

% RSD

(n=3)

Amount

(mg)

% RSD

(n=2)

Amount

(mg)

% RSD

(n=2)

Amount

(mg)

% RSD

(n=2)

Amount

(mg)

% RSD

(n=2)

Amount

(mg)

PLG

A(50:50)1550 ‐

diol(etLD

I‐HEM

A)2

PLG

A(50:50)1550 ‐

diol(H

EMA)2

pHEM

A(20 kD

a)pHEM

A(300 kD

a)PLG

A(50:50)1550 ‐d

iol

Reco

very2

+/‐CV (%

)Hydrolysate

samples

1Obtained from

SEC

-dRI. 2R

ecoveries of each analytebased on control solution. 3T

he average % yield contains the data corrected w

ith mass of the

buildingblocks in the structure.

Chapter 2

56

Table 6 The theoretical and experimental ratio between different components of biomaterials

Ratio between PMA : EG EG : DEG PMA : DEG Lys: DEG

(Theoretical ratio) (1:1) (2:1) (2:1) (2:1)

PLGA(50:50)1550-diol(HEMA)2 1.0 : 0.74 2.0 : 0.77 2.0 : 0.61 --

PLGA(50:50)1550-diol(etLDI-HEMA)2 1.0 : 1.0 2.0 : 1.05 2.0 : 1.05 2.0 : 1.85

pHEMA (300 kDa) 1.0 : 0.56 - -- --

pHEMA (20 kDa) 1.0 : 0.83 - -- --

4 Conclusions

The current method leads to complete hydrolysis of pHEMA (both high and low Mw),

PLGA(50:50)1550-diol, PLGA(50:50)1550-diol(HEMA)2, and PLGA(50:50))1550-diol(etLDI-

HEMA)2 in 24 h at 120oC. The Teflon lined microwave vial was helpful to avoid contact of

reaction medium with the glass vial. This leads to the complete degradation of biomaterial

without the formation of insoluble residues under harsh conditions (high pH, temperature

and pressure). NMR proved to be a good analytical technique to monitor the cleavage of

bonds in these biomaterials. HPLC-ToF-MS can be utilized to quantify the monomers in the

hydrolysis mixture. The origin of bimodality in the MMD of PMA-Na can be inferred from

the non-Gaussian distribution of the starting material. This study will be helpful to

investigate the hydrolytic degradation and for the compositional analysis of novel polymeric

networks including pHEMA as an intermediate product.


57

5 References

[1] G. Wnek, G.L. Bowlin, Encyclopedia of Biomaterials and Biomedical Engineering, Informa Health Care, 2004.

[2] G. Mabilleau, M.F. Moreau, R. Filmon, M.F. Baslé, D. Chappard, Biomaterials 25 (2004) 5155.

[3] B.D. Ratner, A.S. Hoffman, F.J. Schoen, J.E. Lemons, Biomaterials Science, Second Edition: An Introduction to Materials in Medicine Academic Press, 2004.

[4] J.B. Park, J.D. Bronzino, Biomaterials: Principles and Applications, CRC Press, 2003.

[5] J. Stitzel, J. Liu, S.J. Lee, M. Komura, J. Berry, S. Soker, G. Lim, M. Van Dyke, R. Czerw, J.J. Yoo, A. Atala, Biomaterials 27 (2006) 1088.

[6] S. Suzuki, K. Matsuda, T. Maruguchi, Y. Nishimura, Y. Ikada, British Journal of Plastic Surgery 48 (1995) 222.

[7] S.H. Kim, C. Marmo, G.A. Somorjai, Biomaterials 22 (2001) 3285. [8] M. Schuler, G.R. Owen, D.W. Hamilton, M. de Wild, M. Textor, D.M. Brunette,

S.G.P. Tosatti, Biomaterials 27 (2006) 4003. [9] P. Klement, Y.J. Du, L.R. Berry, P. Tressel, A.K.C. Chan, Biomaterials 27 (2006)

5107. [10] S. Atzet, S. Curtin, P. Trinh, S. Bryant, B. Ratner, Biomacromolecules 9 (2008)

3370. [11] K.D.F. Vlugt-Wensink, X. Jiang, G. Schotman, G. Kruijtzer, A. Vredenberg, J.T.

Chung, Z. Zhang, C. Versluis, D. Ramos, R. Verrijk, W. Jiskoot, D.J.A. Crommelin, W.E. Hennink, Biomacromolecules 7 (2006) 2983.

[12] N.S. Khelfallah, G. Decher, P.J. Mesini, Biointerphases 2 (2007) 131. [13] S.C. Baker, G. Rohman, J. Southgate, N.R. Cameron, Biomaterials 30 (2009) 1321. [14] E.I.d.P.d. Nemours, Ind Eng Chem 28 (1936) 1160. [15] O. Wichterle, D. Lim, Nature 185 (1960) 117. [16] S. Abraham, S. Brahim, K. Ishihara, A. Guiseppi-Elie, Biomaterials 26 (2005) 4767. [17] J.-P. Montheard, M. Chatzopoulos, D. Chappard, Rev Macromol Chem Phys C32

(1992) 1 [18] F.O. Eschbach, S.J. Huang, J.A. Cameron, J Bioact Compatible Polym 9 (1994) 210. [19] J. Gracida, J. Alba, J. Cardoso, F. Perez-Guevara, Polym Degrad Stab 83 (2004) 247. [20] X. Jiang, M.C. Lok, W.E. Hennink, Bioconjugate Chem 18 (2007) 2077. [21] P.Y. Ghi, D.J.T. Hill, A.K. Whittaker, Biomacromolecules 3 (2002) 991. [22] M.Y. Arıca, D. Tuğlu, M.M. Başar, D. Kılıç, G. Bayramoğlu, E. Batislam, J Biomed

Mater Res B Appl Biomater 86B (2008) 18. [23] D. Horák, A. Adamyan, O. Golubeva, N. Skuba, T. Vinokurova, J Mater Sci Mater

Med 17 (2006) 123. [24] J. Lukáš, K. Smetana, P. Petrovický, V. Palečková, J. Vacik, B. Dvořánková, L.

Brož, D. Pospíšilová, Z. Holíková, J. Bartůňková, J Mater Sci Mater Med 12 (2001) 639.

[25] M. Houchin, E. Topp, J Pharm Sci 97 (2008) 2395. [26] L. Sun, S. Zhou, W. Wang, X. Li, J. Wang, J. Weng, Colloids Surf, A 345 (2009)

173. [27] T. Ishihara, T. Kubota, T. Choi, M. Takahashi, E. Ayano, H. Kanazawa, M. Higaki,

Int J Pharm 375 (2009) 148. [28] T. Yoshioka, N. Kawazoe, T. Tateishi, G. Chen, Biomaterials 29 (2008) 3438.

Chapter 2

58

[29] C. Hwang, Y. Park, J. Park, K. Lee, K. Sun, A. Khademhosseini, S. Lee, Biomedical Microdevices 11 (2009) 739.

[30] M. Takahashi, H. Onishi, Y. Machida, J Controlled Release 100 (2004) 63. [31] S.H. Choi, T.G. Park, J Biomater Sci Polym Ed 13 (2002) 1163. [32] Y. Duan, J. Xu, Y. Lin, H. Yu, T. Gong, Y. Li, Z. Zhang, J Biomed Mater Res A 87A

(2008) 515. [33] D.-D. Lu, J.-C. Yuan, L.-G. You, Z.-Q. Lei, J Macromol Sci, Part A: Pure Appl

Chem 45 (2008) 705 [34] S.-W. Jung, Y.-I. Jeong, Y.-H. Kim, K.-C. Choi, S.-H. Kim, J Microencapsulation 22

(2005) 901. [35] K. Sutherland, J.R. Mahoney, A.J. Coury, J.W. Eaton, J Clin Invest 92 (1993) 2360. [36] G.T. Chao, Z.Y. Qian, M.J. Huang, B. Kan, Y.C. Gu, C.Y. Gong, J.L. Yang, K.

Wang, M. Dai, X.Y. Li, M.L. Gou, M.J. Tu, Y.Q. Wei, J Biomed Mater Res A 85A (2008) 36.

[37] J.B. Hutchison, A.S. Lindquist, K.S. Anseth, Macromolecules 37 (2004) 3823. [38] H. Matsubara, S.-i. Hata, Y. Kondo, Y. Ishida, H. Takigawa, H. Ohtani, Anal Sci 22

(2006) 1403. [39] R. Peters, Ph.D thesis, Characterization of polymeric network structures, University

of Amsterdam, 2009. [40] A. A. Dias, A. Petit, in European Patent: EP1996640 (B1), 2009. [41] S.T. Popovici, P.J. Schoenmakers, J Chromatogr A 1099 (2005) 92. [42] S.-T. Popovici, W.T. Kok, P.J. Schoenmakers, J Chromatogr A 1060 (2004) 237. [43] D. Bikiaris, V. Karavelidis, G. Karayannidis, Macromol Rapid Commun 27 (2006)

1199. [44] Zainuddin, D. Hill, A. Whittaker, L. Lambert, T. Chirila, J Mater Sci Mater Med 18

(2007) 1141. [45] J.V.M. Weaver, I. Bannister, K.L. Robinson, X. Bories-Azeau, S.P. Armes, M.

Smallridge, P. McKenna, Macromolecules 37 (2004) 2395. [46] S. Ito, K. Tsukada, J Chromatogr A 943 (2001) 39. [47] M. Villagrasa, M. Guillamón, E. Eljarrat, D. Barceló, J Chromatogr A 1157 (2007)

108.

A. Ghaffar, G. J. J. Draaisma, G. Mihov, A. A. Dias, P.J. Schoenmakers, Sj. van der Wal, Biomacromolecules 12 (2011) 3243-3251.

Chapter 3

3. Monitoring the in vitro enzyme-mediated degradation of

degradable poly(ester amide) for controlled drug delivery by LC-

ToF-MS

To scrutinize materials for specific biomedical applications, we need sensitive and selective

analytical methods that can give more insight into the process of their biodegradation. In the

present study, the enzymatic degradation of multi-block poly(ester amide) based on natural

amino acids, such as lysine and leucine, was performed with serine proteases (α-

chymotrypsin (α-CT) and proteinase K (PK)) in phosphate-buffered saline solution at 37oC

for 4 weeks. Fully and partially degraded water-soluble products were analyzed by liquid

chromatography hyphenated with time-of-flight mass spectrometry using an electrospray

interface (LC-ESI-ToF-MS). Tracking the release of monomeric and oligomeric products

into the enzyme media during the course of enzymatic degradation revealed the preferences

of R-CT and PK toward ester and amide bonds: both α-CT and PK showed esterase and

amidase activity. Although within the experimental time frame up to 30 and 15% weight loss

was observed in case of α-CT and PK, respectively, analysis by size-exclusion

chromatography showed no change in the characteristic molecular-weight averages of the

remaining polymer. This suggests that the enzymatic degradation occurs at the surface of this

biomaterial. A sustained and linear degradation over a period of 4 weeks supports the

potential of this class of poly(ester amide)s for drug delivery applications.

Chapter 3

60

1 Introduction

The demand for design and synthesis of biodegradable polymeric materials for application in

drug-delivery devices, scaffolds for tissue engineering, and surgical implants, such as

sutures, pins, rods, and screws for fixation devices, is increasing [1,2]. From this perspective,

degradable poly(ester amide)s (PEAs), preferably those containing natural amino acids, have

gained much attention. These materials have been studied extensively in terms of their

biocompatibility and degradation and have been tested in vivo as drug-eluting stent coatings

in clinical trials such as the NOBLESSE trial [3,4]. PEAs contain ester bonds that are

susceptible to hydrolytic degradation. The inclusion of α-amino acids results in improved

mechanical and thermal properties through hydrogen bond interactions [5]. In general,

degradation of biomaterials may take place either through bulk degradation or via surface

erosion [6]. It is important to differentiate between these two mechanisms for a given

biomaterial if one is to control the drug release rate by changing the ratios of ester, amide,

and methylene groups. This requires knowledge of the relative rates of hydrolysis at different

sites within the polymeric chains and determination of the structure of degradation products,

which may critically affect the biocompatibility and rate of clearance from the body [7].

The sequence of the monomers in a multi-block polymer affects its crystalline or amorphous

nature, which in turn influences the degradation rate [8,9]. Therefore, it is important to

separate and identify the low-molecular-weight water-soluble products during

biodegradation in order to correlate the structural parameters with rates of degradation [10].

The collected information will be helpful (i) to estimate overall structure and optimize the

synthetic approach towards a functional material, (ii) to evaluate the toxicity of the

degradation products at an early stage, (iii) to determine the rate of hydrolysis at different

sites (e.g., specificity of enzyme for ester or amide bonds) and (iv) for the rational design of

new materials.

Several studies report the use of liquid chromatography for the separation of water-soluble

degradation products of polyesters and their identification with UV-Vis using standards [11],

FAB-MS [12], ESI-MS [13], and so on or their offline analysis with NMR [14]. Rizzarelli et

al. utilized the combination of high-performance liquid chromatography with electrospray-

ionization mass spectrometry (LC-ESI-MS) or tandem mass spectrometry (LC-ESI-MS/MS).

They noticed selective ester hydrolysis in aliphatic copolyesters catalyzed by lipases [15].

Mass spectrometry has emerged as a powerful analytical tool for the characterization of

natural and synthetic macromolecules [7,16], but the technique has limitations for high-

Monitoring the in vitro enzyme-mediated degradation of degradable PEA by LC-ToF-MS

61

molecular-weight synthetic polymers. This is true for both ionization techniques that are

commonly applied for high-molecular-weight analytes, that is, ESI and matrix-assisted laser

desorption/ionization (MALDI). When using ESI, multiply charged ions swamp the

spectrum, amplifying the number of ions arising from the molecular-weight distribution

(MWD). In MALDI, both statistics and charge affinity may cause low-molecular-weight

oligomers to dominate the spectrum.

In the current chapter, stainless steel disks coated with a class of PEAs (Scheme 1)

composed of sebacic acid, 1,6-hexanediol, lysine, and leucine with two different sequences

of repeating units were subjected to in vitro enzymatic degradation with α-chymotrypsin (α-

CT) and proteinase K (PK). The weight-average molecular weight (Mw) and dispersity of the

remaining un-dissolved material were analyzed by SEC before and after degradation. The

water-soluble degradation products (monomers and oligomers) were separated on an LC

column and their structure was characterized by time-of-flight (ToF) MS.

The objectives of the current study were (i) to determine the effectiveness of α-CT and PK as

model enzymatic systems to degrade the present PEA by measuring weight losses, (ii) to

differentiate between surface and bulk degradation by estimating the average molecular

weights and dispersity of the polymer before and after degradation, and (iii) to assess the

specific activity (qualitative assessment of preferences and reaction rates) of the enzymes

with the present PEA by (semi-quantitatively) determining the monomeric and oligomeric

products released during biodegradation.

The implication of this research is that a comparative degradation study of several PEAs by

model enzymes may guide the development of PEA biomaterials towards specific properties,

such as degradation rate, controlled drug-delivery rate, targeted release of active

pharmaceutical molecules, and so on.

Scheme 1 Structure of poly (ester amide)s subjected to enzymatic degradation. Where x = 3 and y = 4. The solid and dashed arrows represent the possible cleavage of ester and amide bonds, respectively.

HN

HN

OO O

n

O

NH

O

O

O

OHN

O

O

y x ym

Chapter 3

62

2 Materials and methods

2.1 Materials

The polymer was synthesized at DSM Ahead following a published procedure [17]. The

chemical structure is shown in Scheme 1. The actual values are m/n = 3:1, x = 3, and y = 4.

Figure 1 shows the 1H NMR spectrum of the PEA recorded in d6-ethanol (Euriso-top, Gif sur

Yvette, France) on a Varian Inova 500 MHz NMR (Varian, Walnut Creek, CA, USA)

equipped with a 5 mm 13C/31P/1H GS Probe 500. A pulse-repetition time of 25 s, a pulse

duration of 3.6 µs, 63 scans and a temperature of 25oC were used to record 1H spectra. The

integration of characteristic protons signals for 1,6-hexanediol (15) and lysine (5) confirmed

the molar ratio of 3:1 between m and n block of the polymer. There is 1:1 ratio between

lysine (5) and benzyl group (18). Lysine (5) and sebacic acid (6) possess 1:1.13 ratio in the n

block of the polymer. The 1,6-hexanediol (15) and sebacic acid (6 + 9) signal ratio also

supports the m/n = 3:1 molar ratio. Accurate integration was difficult in case of overlapping

signals. The HSQC NMR experiments were done to confirm the characteristic signals (see

Figure S1 in the supporting information). Glass transition temperature (Tg) of the analyzed

polymer is 33oC, determined via DSC. The test sample, 5 mg, was vacuum-dried at 65 ± 5º C

and placed in a crucible pan. Next the sample was analyzed on a Mettler Toledo DSC 822e

instrument and Tg was derived from the second heating curve.

2.2 Solubility

The solubility of the biomaterial in water and in a number of common organic solvents was

assessed by combining 2 to 3 mg of the PEA with 1 mL of the respective solvents (Table1)

at room temperature (25oC). The PEA was considered to be soluble when the solutions

became completely transparent.

2.3 Enzyme activity

N-Suc-Ala-Ala-Pro-Phe-pNA (Bachem, Bubenhof, Switzerland) was used as a chromogenic

substrate to determine the activity of α-CT from bovine pancreas (Fluka, Steinheim,

Germany, pr.no. 27270, > 68 units/mg protein). The amount of p-nitrophenyl anilide

released was determined by recording the absorbance at 410 nm and 25°C using a


63

spectrophotometer. The amount of enzyme activity releasing 1 µmol of chromophore per

min is defined as 1 unit [18].

The activity of the PK from Tritirachium album (Sigma Aldrich, Steinheim, Germany, pr.

no. P2308, >30 units/mg protein) was measured by colorimetric analysis according to the

quality-control procedure described by the supplier (Enzymatic Assay of Proteinase K, EC

3.4.21.64). One unit will hydrolyze urea-denatured hemoglobin to produce a colour

equivalent to 1 µmol (181 µg) of tyrosine per min at pH 7.5 at 37 °C [19].

Figure 1 1H NMR spectrum of the PEA in d6-ethanol.

2.4 In vitro enzyme-mediated degradation

Approximately 20 mg of PEA were drop-cast on one side of stainless-steel round disks

(diameter 13 mm, thickness 80 µm, surface area 133 mm2). The coatings were applied in

three layers by pipetting 70 µL of polymer solution prepared in ethanol (0.1 g/mL, filtered

through a 0.45 μm filter). Each layer was allowed to air-dry for at least 2 h at ambient

temperature before the next layer was applied. After the final layer was applied, the coated

disks were air-dried overnight at room temperature, followed by drying at 40oC under

reduced pressure to a constant weight.

Chapter 3

64

The coated dried disks (in triplicate) were immersed in 1.5 mL of phosphate-buffered saline

(PBS) buffer (0.2 g KCl, 0.2 g KH2PO4, 1.15 g Na2HPO4, 8 g NaCl in 1 L, containing

sodium azide (0.5 g/L, to inhibit bacterial growth) with α-CT (17 U/mL; pH 8) or with PK

(5 U/mL; pH 7.4) in 15 mL polypropylene conical tubes (BD, Franklin Lakes, NJ, USA)

and incubated at 37oC and 120 rpm (Innova 44 Incubator Shaker Series, New Brunswick

Scientific, Edison, NJ, USA). The enzyme solutions were refreshed every 48 h and stored at

–20oC for LC-ToF-MS analysis after their pH was monitored. The remaining polymer

samples (on disks, each in triplicate) were collected for gravimetric analysis randomly on

day 7, 14, 21, and 28. The solutions were aspirated and the disks were rinsed three times for

5 min with distilled water. The samples of remaining polymer were then dried under vacuum

at 65oC for 48 h. The samples were dried for additional 24 h at 65oC to enssure constant

weight. The percent weight loss was calculated using the following formula:

100)disk theofweight ()sample of weight initial(

)disk theofweight () sample theofweight (1(%) lossWeight

disk

disk

Enzyme blank and polymer blank samples were also incubated and analyzed to correct the

data. Under the current experimental setup, the enzyme media was replaced with fresh

enzyme after every 48 h. Therefore, to estimate the decrease in the activity of enzymes, we

collected samples from the enzyme media (both α-CT and PK) incubated with coated and

non-coated disks after 8, 24, 32, and 48 h for LC-ToF-MS analysis.

2.5 Size-exclusion chromatography

The SEC experiments were performed on an LC system equipped with a LC-10AD solvent

delivery module, a CTO-6A column oven, a SIL-9A auto injector, an SPD-10AVvp UV-vis

detector, and an RID-10A refractive-index detector (all from Shimadzu, Kyoto, Japan). The

SEC analyses were performed on three PLgel MIXED-B columns (10 µm, 300 × 7.6 mm

i.d.; Polymer Laboratories, Church Stretton, U.K.) connected in series. THF stabilized with

butylated hydroxytoluene (BHT) (BioSolve, Valkenswaard, The Netherlands) was pumped

at a flow rate of 1 mL min-1. The injection volume was 50 µL, and the column oven

temperature was set at 50oC. Polystyrene standards (Polymer Laboratories, Shropshire, U.K.)

were used to calibrate the SEC-UV-dRI system. Data were recorded and chromatographic


65

peaks were treated using Class-VP 7.4 software (Shimadzu). Molar-mass distributions

(MMDs) were calculated from the chromatograms using software written in-house in Excel

2003 (Microsoft).

2.6 LC-ToF-MS study

We injected 5 μL of the enzyme solutions collected after every 48 h without dilution after

filtration through a 0.2 µm pore size PTFE filter (Grace Davison). The stock solutions of

1,6-hexanediol (Aldrich), sebacic acid (Fluka) and benzyl alcohol (Fluka), were prepared in

PBS, and quantification was based on a standard-addition method in positive ESI mode for

1,6-hexanediol and in negative ESI mode for sebacic acid. The concentration of benzyl

alcohol was determined by LC-UV at 254 nm.

The chromatographic separations of soluble degraded products were performed on a Prevail

C18 Column (250 × 4.6 mm i.d., 5 µm particle size) (Grace Davison, Deerfield, IL, USA)

connected to an Agilent 1100-series LC system consisting of a degasser, a gradient pump,

and an auto-sampler (all from Agilent Technologies, Waldbronn, Germany). The column

oven (Temperature Control Module from Waters, Milford, MA, USA) was set at 40oC.

Mobile phase A was 0.1% (v/v) aqueous formic acid (Fluka, Steinheim, Germany) and B

was acetonitrile (Biosolve). The gradient was started at t = 0 min with 5% (v/v) B, reaching

60% (v/v) B in 25 min, held constant for 2 min, and then back to 5% (v/v) at 30 min

(tend = 35 min) at a flow rate of 1.5 mL min-1 and was split between a waste reservoir and the

ESI interface by means of a zero-dead-volume T-piece to enssure a flow of approximately

~0.2 mL min-1 into the electrospray. Organic solvents used for the LC mobile phase were of

LC grade. Highly pure water for mobile-phase preparation was obtained from an Arium 611

Ultrapure (18.2 MΩ*cm) Water System (Sartorius, Goettingen, Germany).

The LC system was hyphenated with a 6210 series time-of-flight mass spectrometer

(Agilent) via an ESI interface. The conditions of the ESI-ToF-MS were as follows. Drying

gas was nitrogen at 8 L min-1, 300oC and 200 kPa. The capillary voltage was 3500 V; the

fragmenter was set at 140 V and the skimmer at 60 V. Octopole dc1 was set at 33 V;

octopole radio frequency at 250 V. The data were acquired in the scan mode from m/z 50 to

3000 with 0.88 scans/s. An Agilent MassHunter Workstation A.02.01 and Analyst QS 1.1

software (Applied Biosystems, Carlsbad, California, USA) were used for data acquisition

and data analysis, respectively.

Chapter 3

66

The LC-ToF-MS separation and identification of the water-soluble degradation products was

optimized by monomers and oligomers, generated by means of chemical degradation. (see

the supporting information).


3.1 Solubility

The solubility of the PEA in a number of organic solvents at room temperature (25oC) is

summarized in Table 1. The PEA is completely soluble in most of the polar (protic and

aprotic) solvents, except in water, acetonitrile, and ethyl acetate.

Table 1 Solubility of the PEA used in the present study in common organic solvents evaluated at 25oC.

solvents Solubility solvents Solubility solvents Solubility

n-Hexane - Dichloromethane + H2O -

Cyclohexane - Tetrahydrofuran + Methanol +

Chloroform + Ethyl acetate - Ethanol +

Toluene - Dimethylformamide + Isopropanol +

N,N`-dimethyl acetamide + Acetonitrile - Acetic acid +

Dimethyl sulfoxide + Acetone +

Soluble +, insoluble -

3.2 Overall effectiveness of in vitro enzyme-mediated degradation

The results in Figure 2 show a greater weight loss (30%) in case of α-CT than when using

PK (15%) after 4 weeks under the current experimental conditions. When incubating with

only PBS buffer, no noticeable weight loss was observed. The biomaterial showed a

sustained and near-linear degradation over a period of 4 weeks (Figure 2). This weight loss

the polymer indicates constant erosion of the surface by the enzymes. No solid particles were

observed in the reaction media. No decrease in the pH of the degradation media was

observed during the course of the enzyme-mediated degradation, implying that the erosion

does not result in an accumulation of acidic degradation products.


67

-5

0

5

10

15

20

25

30

35

0 5 10 15 20 25 30

Days

Gra

vim

etri

c w

eig

ht

loss

(%

)

Figure 2 Percentage of polymer weight loss as a function of the degradation time (days) in α-Chymotrypsin (α-CT, 17 U/mL) and proteinase K (PK, 5 U/mL) in PBS. The “blank” experiment in PBS is shown for comparison. Lines with triangles, squares and diamonds correspond to blank, PK and α-CT experimental data, respectively.

0

1

2

3

4

3 3,5 4 4,5 5 5,5 6

log M

dR

I re

spo

nse

0

7

14

21

28

Figure 3 The molar-mass distribution of the PEA before and after enzymatic degradation with α-CT. The side bar on the right side represents the number of days of treatment.

Chapter 3

68

3.3 Molecular-weight of remaining material

The Mn and dispersity of the starting material were 31 kDa and 2.1, respectively. The

molecular weight and molecular-weight distribution (MWD) of the sample prior to

degradation and of the remaining un-dissolved samples after incubation with α-CT were

determined. The SEC-dRI system was calibrated with polystyrene standards in THF. No

significant differences were observed between the molecular weights of the samples prior to

and after enzymatic degradation (Figure 3). This maintenance of molecular weight is in line

with the enzymatic degradation being a surface-erosion phenomenon. Similar results were

reported by Fan et al. [20,21] and by Guo et al. [22].

3.4 LC-ToF-MS analysis following enzymatic degradation

The chromatographic separation and identification of the monomers and oligomers librated

as a result of enzymatic degradation was monitored using LC-ToF-MS with electro-spray

ionization in the positive and negative modes. Formic acid was used to make proton adducts.

Sodium adducts were also observed because of the presence of sodium ions in the incubating

medium. The presence of multiple charge ions were identified by the differences of m/z

values between the isotopic peaks of respective masses. The negative mode spectra showed a

difference of 2 and 24 Da for proton and sodium adducts, respectively, as compared with

positive mode spectra. Benzyl alcohol showed the [M-OH]+ adduct only in positive mode.

The m/z and the retention time of benzyl alcohol in the enzyme solutions were confirmed by

the benzyl alcohol standard. The ESI-ToF-MS spectra of the identified peaks (Table 1) are

presented in the supporting information. The copolymer consists of five components; leucine

(L), lysine (K), 1,6-hexanediol (C6), sebacic acid (C10), and benzyl alcohol (BnOH). All the

LC10 and KC10 combinations contained an amide bond. In case of LC6, the carboxylic

group of L is connected to C6 via an ester bond. The carboxylic group of the lysine is end-

capped with benzyl alcohol and forms an ester bond. Scheme 1 shows the possible

fragmentation pattern in case of both ester- and amide-bond cleavage. If only the ester bonds

are hydrolyzed, degradation products will be C6 (peak number 4), benzyl alcohol (6), LC10L

(11), LC10LC6 (17), and the oligomers LC10(KC10)nL originating from n block of the

polymer. If amide-bond cleavage is simultaneously taking place along with ester hydrolysis,

then the additional products that may appear are K (peak number 1), L (3), LC6, C10 (8),

LC10 (9), C10K (5), LC10K(7), and LC10KC10 (10).


69

Figure 4 displays the separation of the different degradation products at various intervals of

time during enzymatic degradation of PEA with α-CT. Mass-spectrometric analysis shows

the release of L (peak number 3), C6 (4), C10K (5), benzyl alcohol (6), LC10K (7), C10 (8)

LC10 (9) LC10KC10 (10), LC10L (11). Peaks 12–14 represent oligomeric blocks of

(LC10[KC10]nL) with n number of repeating unites ranging from 1 to 4. The sensitivity of

ESI-ToF-MS decreases significantly with increasing n. Therefore, no direct conclusion can

be drawn from the peak areas in the chromatograms. The products generated at different

stages of the enzymatic degradation (moments in time) indicate that ester hydrolysis has

dominated during enzymatic degradation up to day 18; after that, an unexpected change was

observed and α-CT acted mainly as a protease cleaving predominately the amide bonds.

Figure 4 Selected TIC chromatograms of the degradation products obtained at different time intervals during α-chymotrypsin degradation at 37°C. Val and Phe represent valine and phenylalanine, respectively.

In case of PK, the appearance of L (peak number 3), C6 (4), benzyl alcohol (6), LC10K (7),

LC10 (9), LC10KC10 (12), and so on indicated the cleavage of both ester and amide bonds.

The enzymes in both cases showed preference towards the amide bond between L and C10

compared with the one between K and C10 because no release of K was observed (Figure 5).

Table 2 shows the structural assignments of different peaks separated in Figures 4 and 5

based on the m/z ratios from their ESI-ToF mass spectra.

Chapter 3

70

Figure 5 Selected TIC chromatograms of the degradation products obtained after different time intervals during enzymatic degradation of the PEA with proteinase K at 37oC and 120 rpm.

Figure 6 Cumulative increase (y-axis) in the concentration (µg.mL-1) of 1,6-hexanediol (◊ and ♦ symbols), sebacic acid ( and symbols) and benzyl alcohol (∆ and symbols) during the course of enzymatic degradation of the PEA with α-chymotrypsin (solid lines) and proteinase K (dashed lines). The concentration of benzyl alcohol was determined by LC with UV-absorbance detection at 254 nm.

Because C6 is connected on both sides with L through ester bonds and C10 is bonded via

amide bonds either to two K or to ne L and one K, the emerging C6 and C10 can be used to

monitor the ultimate degradation due to ester hydrolysis or amide bond-cleavage. Figure 6

demonstrates a nearly linear increase in the cumulative concentration of benzyl alcohol, C6


71

and C10 in the enzyme media collected after every 48 h during the enzymatic degradation

with α-CT and PK. It is obvious from the quantitative data that the release of C10 greatly

increased after the 18th day in α-CT degraded hydrolysates, whereas in case of PK, very

little C10 was librated. An interesting observation is that in case of the PK-mediated

degradation the detected C6 in solution corresponds to 100% of the content of this moiety in

the hydrolyzed polymer (estimation is based on the observed weight loss). However, more

than half of the potential C6 appeared in fragments other than C6 itself when using α-CT.

Although it is clear (cf. peaks nos. 3, 7 and 9) that PK exhibits amidase activity, the amount

of C10 was negligible in the medium. This stresses the need for quantitative analysis of the

intermediate degradation products.

Unfortunately, we can only analyze the oligomers in a semi-quantitative manner so far. The

contribution of K, L, valine (Val) and phenylalanine (Phe) from the enzymes to the reaction

media was confirmed by LC-ToF-MS analysis of the sample blank used in the study and also

by the hydrolytic degradation of the enzymes.

3.5 Factors affecting enzyme activities

The effects of equilibrium or saturation of the degradation products on the activity of the

enzymes were estimated by analyzing the reaction media incubated at 37oC and 120 rpm

after 8, 24, 32, and 48 h by LC-ToF-MS. Figure 7 shows the peak areas for peaks from 11 to

15. In case of α-CT, the heights of these peaks are almost approaching almost constant

levels. However, the peaks heights at 48 h are decreasing gradually as the molecular weights

of the oligomers are increasing. This suggests further degradation of the water-soluble

oligomers in the enzyme media at these prolonged incubating times. In case of PK, this

situation has not yet been reached after 48 h.

The decrease in enzyme activity could be caused either by self-digestion or by denaturation

of the enzyme. A gel-electrophoresis experiment was performed to compare a freshly

prepared enzyme solution with a solution which had been stored at 37ºC for 3 days [23].

This experiment clearly showed several small enzyme fragments in the stored sample.

The change in the activity of the α-CT during the enzymatic degradation was measured after

6, 18, 24, 30, and 48 h. At 0 h, the activity measurement involved only enzyme solution and

the standard substrate (N-Suc-Ala-Ala-Pro-Phe-pNA). But after 6, 18, 24, 30, and 48 h, the

enzyme media collected from the incubated coated disks also contained water-soluble

degradation products from PEA. These products in the presence of standard substrate do not

Chapter 3

72

appear to inhibit the activity of enzyme. Nevertheless, we found a decrease of >90% in the

activity of enzyme after 48 h under the current experimental conditions.

Figure 7 TIC LC-ToF-MS chromatograms of the reaction media injected after 8 (∆), 24 (), 32 (), and 48 (◊) hrs of enzymatic degradation. (a) α-chymotrypsin (b) proteinase K.

In summary, the PEA degraded primarily via ester hydrolysis during in vitro enzyme-

mediated degradation (with α-CT and PK at 37oC). The survival of KC10-containing

fragments under the applied enzymatic conditions suggested that the amide bonds between K

and C10 are more stable than those between L and C10. The analysis shows only fragments,

which theoretically could be derived directly from the anticipated polymer structure. This

suggests a polymerization process without side reactions. A significant acceleration of the

degradation may possibly be obtained if the enzyme solution is more frequently or even

continuously refreshed. An on-stream-analysis system to study and explore these synergistic


73

effects is currently being developed in our laboratory. The rate-determining factor should

still be the accessible surface area of the biomaterial, but the conditions will be more

favourable for degradation than those used here.

Table 2 Structural assignments for the fragments arising from different monomers and oligomers, separated as shown in Figures 4 and 5 following the enzyme-mediated degradation of the PEA in α-Chymotrypsin and Proteinase K. The symbols K, L, C6 and C10 represent lysine, leucine, 1,6-hexanediol, and sebacic acid, respectively.

Peak no.

Symbol Structure Number of amide bond broken

m/z [M+H]+

1 K

2 147.1129

2 formic acid

3 L

1 132.1033

4 C6

0 119.11099

5 C10K

2 331.2211

6 benzyl alcohol

0 91.0554

[M-OH]+

7 LC10K

1 444.3106

8 C10

2 201.1154

[M-H]-

9 LC10

1 316.2143

Chapter 3

74

10 LC10KC10

1 628.4277

11 LC10L

0 429.3072

12 LC10(KC10)1L

0 741.5221

13 LC10(KC10)2L

0 1053.7141

14 LC10(KC10)3L

HO

(CH3)2HC

O

NH

O

O

HN

HN

4 2

HO OO

O

NH

OH

CH(CH3)2

O4

3

0 1365.9289

15 LC10(KC10)4L

0 1678.1283

16 C10LC6

1 416.2977

17 LC10LC6

HO

(CH3)2HC

O

NH

O

O

HN

4

CH(CH3)2

O

OOH3

0 529.3803


75

4 Conclusions

Poly(ester amide) polymer was subjected to enzymatic degradation conditions. Hyphenation

of liquid chromatography with electrospray – time-of-flight – mass spectrometry proved to

be a powerful analytical tool for the chromatographic separation and identification of the

water-soluble degradation products after enzymatic degradation. The technique allowed the

identification of fully and partially degraded polymer fragments, thus providing information

on the polymerization process and on the intrinsic polymer structure. The polymer was found

to degrade at a steady rate with both enzymes during this study. A lack of significant

changes in the average molecular weight of the remaining polymer strongly suggests that

surface erosion occurred during the enzyme-mediated degradation. Furthermore, no

accumulation of acidic byproducts was observed during the course of the experiment. In this

respect, the polymer performs better than conventional polyesters. The experiments

confirmed that this class of (polyester amide)s shows a remarkable hydrolytic stability in the

absence of enzymes.

Moreover, to avoid the further degradation of the water-soluble degradation products in the

enzyme media due to long incubation time and to enhance the degradation rate by

continuously refreshing the enzyme solution, the development of a system for on-stream

analysis of degradation products is in progress.

Chapter 3

76

5 References

[1] in 't Veld, P. J. A.; Dijkstra, P. J.; Feijen, J. Clin. Mater. 1993, 13, (1-4), 143-147. [2] Mihov, G.; Draaisma, G.; Dias, A.; Turnell, B.; Gomurashvili, Z. Abstract book: 11th

European Symposium on Controlled Drug Delivery 2010, 340-342. [3] Sarkar, D.; Lopina, S. T. Polym. Degrad. Stab. 2007, 92, (11), 1994-2004. [4] M.DeFife, K.; Grako, K.; Cruz-Aranda, G.; Price, S.; Chantung, R.; Macpherson, K.;

Khoshabeh, R.; Gopalan, S.; Turnell, W. G. J. Biomater. Sci., Polym. Ed. 2009, 20, 1495-1511.

[5] Paredes, N.; Rodriguez-Galán, A.; Puiggalí, J.; Peraire, C. J. Appl. Polym. Sci. 1998, 69, (8), 1537-1549.

[6] Burkersroda, F. v.; Schedl, L.; Göpferich, A. Biomaterials 2002, 23, (21), 4221-4231.

[7] Adamus, G.; Hakkarainen, M.; Höglund, A.; Kowalczuk, M.; Albertsson, A.-C. Biomacromolecules 2009, 10, (6), 1540-1546.

[8] Montaudo, G.; Rizzarelli, P. Polym. Degrad. Stab. 2000, 70, (2), 305-314. [9] Gan, Z.; Abe, H.; Doi, Y. Biomacromolecules 2001, 2, (1), 313-321. [10] Rizzarelli, P.; Puglisi, C.; Montaudo, G. Rapid Commun. Mass Spectrom. 2005, 19,

(17), 2407-2418. [11] Scherer, T. M.; Fuller, R. C.; Lenz, R. W.; Goodwin, S. Polym. Degrad. Stab. 1999,

64, (2), 267-275. [12] Kitakuni, E.; Yoshikawa, K.; Nakano, K.; Sasuga, J.; Nobiki, M.; Naoi, H.; Yokota,

Y.; Ishioka, R.; Yakabe, Y. Environ. Toxicol. Chem. 2001, 20, (5), 941-946. [13] Scandola, M.; Focarete, M. L.; Adamus, G.; Sikorska, W.; Baranowska, I.;

Świerczek, S.; Gnatowski, M.; Kowalczuk, M.; Jedliński, Z. Macromolecules 1997, 30, (9), 2568-2574.

[14] Abe, H.; Doi, Y.; Aoki, H.; Akehata, T.; Hori, Y.; Yamaguchi, A. Macromolecules 1995, 28, (23), 7630-7637.

[15] Rizzarelli, P.; Impallomeni, G.; Montaudo, G. Biomacromolecules 2004, 5, (2), 433-444.

[16] Wang, N.; Li, L. J. Am. Soc. Mass Spectrom. 2010, 21, (9), 1573-1587. [17] Chu, C.-C.; Katsarava, R. US Patent # 7304122 B2, 2007. [18] Graham, L. D.; Haggett, K. D.; Jennings, P. A.; Le Brocque, D. S.; Whittaker, R. G.;

Schober, P. A. Biochemistry 1993, 32, (24), 6250-6258. [19] Folin, O.; Ciocalteu, V. J. Biol. Chem. 1927, 73, (2), 627-650. [20] Fan, Y.; Kobayashi, M.; Kise, H. J. Polym. Sci., Part A: Polym Chem. 2001, 39, (9),

1318-1328. [21] Fan, Y.; Kobayashi, M.; Kise, H. J. Polym. Sci., Part A: Polym Chem. 2002, 40, (3),

385-392. [22] Guo, K.; Chu, C. C. Biomacromolecules 2007, 8, (9), 2851-2861. [23] Castro, G. R. Enzyme Microb. Technol. 2000, 27, (1-2), 143-150.


77

6 Supporting information

Two-dimensional H,C-correlated NMR spectrum of the starting PEA. Mass spectra of the

peaks identified during enzymatic degradation of PEA with α-CT and PK. Details of LC-

ToF-MS method optimization for the separation and identification of the water soluble

degradation products generated by means of chemical degradation of PEA. This material is

available free of charge via the Internet at http://pubs.acs.org/.

6.1 Two-dimensional H,C-correlated spectrum (HSQC) of PEA

Figure S1 Two-dimensional H,C-correlated spectrum (HSQC) of the starting PEA. in d6-ethanol recorded on a Bruker Avance 400 MHz NMR spectrometer. symbol ‡ are assigned to NH groups of amide bonds. The numbering to each signal corresponds to the numbering of protons and carbon present in Figure 1 in the Chapter 3.

6.2 ESI-ToF-MS spectra of the identified peaks – enzymatic degradation

The chromatographic peaks for the respective monomers and oligomers in Figure 4 and 5

were identified by their ESI-ToF-MS spectra (Table 2). Figure S2-S5 show the m/z values.

Chapter 3

78

All the spectra showed protonated adducts. Sodium adducts were also observed due to the

presence of sodium ions in the incubating buffer.

Figure S2 The LC-ToF-MS spectra of peaks 1, 3, 4, 5, and 7 in positive mode. The spectrum for peak 8 was recorded in the negative mode.


79

Figure S3 The LC-ToF-MS spectra of peaks nos. 9-13 in positive mode.

Chapter 3

80

Figure S4 The LC-ToF-MS spectra of peaks nos. 14-17 in positive mode.


81

Figure S5 The LC-ToF-MS spectra of peak 6 (Table 2) in positive mode. (A) Benzyl alcohol spectra in the hydrolysate (B) XIC-chromatogram of benzyl alcohol in the hydrolysate (C) spectrum of Benzyl alcohol standard (D) XIC chromatogram of Benzyl alcohol standard.

6.3 Chemical Degradation – optimization of the LC-ToF-MS method

6.3.1 Chemical degradation

Approximately 20 mg of PEA were hydrolyzed at 120oC and 300 kPa pressure for 24 h

while stirring in a 10 mL glass vessels with 2 mL of a mixture of 1 M NaOH and ethanol

(75:25) in a microwave instrument (CEM Corporation, Methews, NC, USA)33. Similarly,

approximately 20 mg of PEA were mixed with 2 mL of a mixture of 3 M HCl and ethanol

(75:25) and hydrolyzed at 90oC and 300 kPa for 24 h in the same apparatus as described

above. Both hydrolysates were twofold diluted with LC-grade water and neutralized (pH ≈

7) with HCl or NaOH prior to their injection into the LC-ToF-MS. Similarly, PEA in both

mixtures were also hydrolyzed at room temperature.

Chapter 3

82

6.3.2 Chemical degradation: LC-ToF-MS analysis

Figure S6a displays the TIC chromatograms for the partially hydrolyzed products in 1 M

aqueous NaOH solution at room temperature. The peaks are labeled according the structural

assignments based on their ESI-ToF mass spectra (Table S1). The formation of leucine (L)

and 1,6-hexanediol (C6) indicates that the polymer is degraded primarily via ester

hydrolysis. The amide bond between L and sebacic acid (C10) appears to be more

susceptible to cleavage than the amide bonds between lysine (K) and C10. Peaks from 20.2

to 22.5 min represent LC10L and oligomeric blocks of (LC10(KC10)nL) with the number of

repeating units (n) ranging from 1 to 5 in order of increasing retention. Because ethanol was

added to enhance the solubility of the polymer in 1 M NaOH, ethylation may occur. This

was observed for various peaks. For example, peak EtLC10K is the ethylated product of

peak LC10K. Peaks from 22.5 to 26 min indicate the presence of ethylated products of peaks

LC10, LC10KC10,LC10L, and oligomeric series LC10(KC10)1-3 (eluting in reverse order,

as hydrophobicity increases upon ethylation). The ESI-ToF-MS spectra of peaks from 22.5

to 26 min show the co-elution of several singly ethylated products originating from LC10,

LC10KC10, and LC10(KC10)1-3L (Table S1).

The polymer degraded nearly completely to its monomers in 1 M NaOH at 120oC in the

microwave instrument (peaks 1, 3, 4 and 9) except for four fragments, namely C10K,

LC10K, LC10, and LC10L. No ethylated products were observed (Figure S6b).

In case of acid hydrolysis (3 M HCl, at room temperature) the oligomers were separated

according to the degree of ethylation i.e. without ethylation, once ethylated and twice

ethylated oligomers, and once ethylated oligomers containing C6 at the other side (Figure

S7a). Peaks EtLC10KEt, EtLC10(KC10)2Let, and EtLC10(KC10)1LEt reflect twice

ethylated oligomers originating from the compounds associated with peaks LC10K,

LC10(KC10)2L, and LC10(KC10)1L , respectively. Peaks EtLC10(KC10)2LC6,

EtLC10(KC10)1LC6, and EtLC10LC6 contain additionally one ethyl group and C6 in

comparison with the analytes of peaks LC10(KC10)2L, LC10(KC10)1L, and LC10L,

respectively. Thus, the degradation products form a complex mixture of oligomers

possessing different end-group functionalities (amine, alcohol, and carboxylic acid or ethyl

esters). Based on hydrophobicity each class is eluted in a retention order with specific

retention increments.


83

K

L

C6C6L

C10

LC10

LC10L

L

Ile

C6BnOH

LC10K

EtLC10K

LC10

LC10KC10

LC10KC10K

LC10L

LC10(KC10)1LLC10(KC10)2L


LC10(KC10)5L

(7) EtLC10(KC10)1-3L

(8) EtLC10(KC10)1-3L(9) EtLC10L

a

b

(6) EtLC10, EtLC10KC10

67 8

9

K

K

L

C6C6L

C10

LC10

LC10L

L

Ile

C6BnOH

LC10K

EtLC10K

LC10

LC10KC10

LC10KC10K

LC10L



LC10(KC10)5L

(7) EtLC10(KC10)1-3L

(8) EtLC10(KC10)1-3L(9) EtLC10L

a

b

(6) EtLC10, EtLC10KC10

67 8

9

K

Figure S6 TIC chromatograms of the PEA degraded in 1 M NaOH aqueous solution at (a) room temperature and (b) 120oC and 3 bar in a Microwave instrument.

Figure S7b shows the TIC chromatogram of the polymer degradation products after nearly

complete degradation in 3M HCl:ethanol (75:25) at 90°C. At high temperature and under

acidic conditions, esterification of the carboxylic-acid groups in L and C10 was observed

(peaks EtL and EtC10). However, only one carboxylic acid was esterified in case of C10.

Similarly, only one hydroxyl group of the C6 was etherified with ethanol to generate an ether

group (peak EtC6). It has been reported in literature that the esterification of carboxylic acid

with alcohols to produce esters is very slow at room temperature, but that direct microwave

heating of the materials greatly speeds up the reaction [Pipus, G.; Plazl, I.; Koloini, T.

Chemical Engineering Journal 2000, 76, (3), 239-245]. Lysine remained un-retained in all

the chromatographic separations under the applied conditions and no esterification of its

carboxylic-acid group was observed. The analysis shows only fragments which theoretically

could be derived from the polymer structure, which suggests a polymerization process

without side-reactions.

Chapter 3

84

L

L

C6EtL

C6L LC10KEtLC10K

LuC10KC10K

EtLC8KEt

LC8

LC10KC10

C6 EtC6

EtL

C10 EtC10

LC8LC6

K

LC10L

LC10(KC10)1-5L

1

23

4

5

EtLC10(KC10)2LC6

EtLC10(KC10)2LEt

EtLC10(KC10)1LC6

EtLC10(KC10)1LEt

EtLC10LC6

a

b

67 8

9

K

L

L

C6EtL

C6L LC10KEtLC10K

LuC10KC10K

EtLC8KEt

LC8

LC10KC10

C6 EtC6

EtL

C10 EtC10

LC8LC6

K

LC10L

LC10(KC10)1-5L

1

23

4

5

EtLC10(KC10)2LC6

EtLC10(KC10)2LEt

EtLC10(KC10)1LC6

EtLC10(KC10)1LEt

EtLC10LC6

a

b

67 8

9

K

Figure S7 TIC chromatograms of the PEA degraded in 3 M HCl aqueous solution at (a) room temperature and (b) 90oC and 3 bar in a Microwave instrument.

A. Ghaffar, G. J. J. Draaisma, G. Mihov, P.J. Schoenmakers, Sj. van der Wal, to be submitted.

Chapter 4

4. A versatile system for studying the enzymatic degradation of

multi-block poly(ester amide)s*

The suitability of biomaterials for specific biomedical applications can be investigated

through their in-vitro biodegradability with selected enzymes and through their degradation

kinetics. A system was developed for studying the enzymatic degradation of poly(ester

amide) (PEA) coatings under sink conditions, with on-stream analysis of degradation

products by liquid chromatography coupled to time-of-flight mass spectrometry (LC-ToF-

MS). A coated capillary was treated by an enzyme solution in pulses (pulse-feed mode) or

continuously (continuous-feed mode) with different flow rates. The water-soluble products

resulting from the interaction of enzyme with the PEA coating were deposited on-line on a

reversed-phase LC column, separated by gradient-elution LC, ionized by electrospray-

ionization (ESI), and identified based on ToF-MS data.

The experiments underline the benefits of the experimental set-up, which requires only small

amounts of coating and enzyme and produces detailed results rapidly. The system was

investigated using different injection volumes (pulses) of an α-Chymotrypsin solution in

varying concentrations, different flow rates, and different lengths of coated capillary. The

versatility of the system makes it easy to follow the course of degradation and to

differentiate between primary and secondary degradation products. The system was applied

to study the degradation of a di-block and a tri-block PEA. Specific degradation products

showed different time profiles than the (more gradual) overall weight loss. Continuous-feed

mode analysis allowed the convenient determination of highly stable amide-bond-containing

fragments, while pulse-feed mode analysis revealed benzyl ester-containing products as

primary degradation products.

Chapter 4

86

1 Introduction

Biodegradable polymeric implants that degrade gradually and that yield degradation

products that are excreted benignly by the body prevent surgical re-interventions to remove

them after their role (e.g. as drug-delivery carrier) has subsided [1, 2]. However, the

suitability of such materials for biomedical applications requires an extensive evaluation in

term of their biocompatibility (tissues response), mechanical properties and, most

importantly, their degradation behaviour [3]. Chemical and enzymatic hydrolysis are the

primary biodegradation mechanisms for such materials. The highly reactive species

produced during the foreign-body response may degrade the polymer chain and contribute to

the overall degradation of biomaterials [2, 4]. The hydrophilic or hydrophobic nature of the

polymeric chains can affect the degradation rate and mechanism. Bulk erosion and surface

erosion are typically distinguished [5]. Poly(ester amide)s (PEAs) have attracted the

attention of the biomedical field for temporary implants due to their good physical

properties, biocompatibility and controlled degradability [6, 7]. PEAs contain ester bonds

susceptible to both hydrolytic and enzymatic degradation, while the amide bonds are prone

to degradation by enzymes.

To optimize the use of such biomaterials for specific biomedical applications and to assess

the potential risks of intermediate and final degradation products, it is crucial to understand

their degradation kinetics. The amounts and toxicological nature of the degradation products

define the acceptability of a biodegradable device by the biological environment [8]. The

kinetics of the enzymatic degradation of biomaterials are conventionally studied first under

in vitro conditions by incubating the specimen in a medium containing enzyme at 37oC with

or without agitation. Such a batch-mode study does not closely mimic the in vivo conditions.

Therefore, it does not allow rigorous modelling of the kinetics. However, the selection of

appropriate enzymes, incubation media, surface-to-volume ratios and duration of the

experiment may help to approach the physiological conditions where the biomaterial is

implanted, [3]. The degradation of degradable synthetic polymeric devices is also influenced

by the static or dynamic conditions of the media [9]. Agrawal et al. proposed the schematics

of an apparatus, which they used to study the degradation of biodegradable scaffolds based

on copolymers of polylactic acid (PLA) and polyglycolic acid (PGA) under dynamic (fluid-

flow) conditions. They reported a decrease in degradation rate under dynamic as compared

to static conditions [9]. Gorman et. al. studied the encrustation of urinary-tract devices based

on polyurethanes, Percuflex® and silicone under static and dynamic conditions with

A versatile system for studying the enzymatic degradation of multi-block poly(ester amide)s

87

artificial urine in a reaction vessel. Significantly higher levels of calcium and magnesium

were observed in the static mode, but the level of encrustation was the same in both cases

[10]. In another study, a dynamic simulated system was developed to study the effect of fluid

flow on the degradation of poly(lactide-co-glycolide acid) (PLGA) in Hank’s simulated body

fluid (SBF) for 30 days. Slower degradation was observed in case of dynamic conditions

[11]. On the other hand, porous scaffolds prepared from poly(L-lactic acid)/β-tricalcium

phosphate (PLLA/ β-TCP) composite, when degraded with phosphate-buffered-saline (PBS)

solution in a dynamic loading set-up, showed significantly faster degradation than in static

incubation [12].

During static studies high concentrations of degradation products may affect the activity of

an enzyme and oligomeric products released as a result of surface erosion are prone to

further degradation in the enzyme-containing media due to long incubation times [7].

Therefore, we decided to develop an on-stream-analysis system to study and explore these

synergistic effects. We constructed a small, flexible experimental set-up to (i) quickly

establish the effectiveness of a particular enzyme towards biodegradable polymer coatings

using small amounts of enzyme, (ii) assess the preference of a selected enzyme for ester or

amide bonds, (iii) allow on-line mass-spectrometric analysis of the intermediate and final

degradation products, with the aim of estimating their toxicological nature.

In the developed on-line system the enzyme solution passes through a capillary coated with

PEA either in continuous-feed mode or in pulse-feed mode. The introduced enzyme

degrades the surface of the polymer under flow conditions and the degradation products are

collected along with enzyme on a reversed-phase octadecyl-silane (C18) column. Attaching

this column to a gradient pump by switching a valve allows the chromatographic separation

of the degradation products and their identification with ESI-ToF-MS. A class of multi-block

PEAs, composed of sebacic acid, 1,6-hexanediol, lysine-benzyl ester and leucine with two

different sequences of repeat units (Scheme 1), was used in this study. The experimental

system was optimized by varying the concentration of enzyme, injection (pulse) volume,

switching time, flow rate of enzyme through the coated capillary, and length of the coated

capillary. The generated data were used to assess the effects of these parameters on the rate

of degradation of PEA and the generation of specific degradation products. The applicability

of the system was ultimately demonstrated for the degradation of a tri-block PEA containing

additionally isosorbide (1,4-dianhydrosorbitol (DAS)) under both continuous-feed and pulse-

feed conditions.

Chapter 4

88

Scheme 1 Structures of (A) di-block poly (ester amide) based on sebacic acid (i.e. y=4), 1,6-hexanediol (x=3), lysine-benzyl ester and leucine. For di-block PEA, m ≈ 3×n. (B) tri-block poly(ester amide), additionally containing isosorbide (1,4-dianhydrosorbitol (DAS)). The solid and dashed arrows represent the possible cleavage of ester and amide bonds upon enzymatic degradation with α-CT, respectively. In case of tri-block PEA, m:n:o = 6:5:9.

The developed system provides a convenient approach and a miniaturized system to study

the kinetics under sink conditions. The low enzyme activities required to use this on-line

system widen the scope of degradation studies, allowing a broader range of appropriate (but

possibly expensive or with a limited availability) enzymes to be tested.

2 Experimental

2.1 Materials

Both the di- and tri-block PEAs (Scheme 1) were synthesized at DSM following an open

literature procedure [13]. The chemical structure is shown in Scheme 1, where x = 3, y = 4.

The structure and the molar ratios between different structural components were confirmed

by 1H NMR and HSQC (heteronuclear single quantum coherence) NMR experiments [7] (for

details see the supporting information). For di-block PEA, m 3n; for tri-block PEA, m:n:o

6:5:9. Solubility in a range of common solvents was assessed (see the supporting

information). The molar-mass distributions were determined by size-exclusion

chromatography with differential-refractive-index detection (SEC-dRI) in tetrahydofuran

(THF). The molecular weight and dispersity of the tri-block PEA are 56 kDa and 2,

respectively (see the supporting information). The enzymatic degradation of coated

capillaries was carried out by protease α-Chymotrypsin (α-CT) from bovine pancreas (Fluka,

A

B

HN

HN

OO O

n

O

NH

O

OO

OHN

O

O

y x ym


89

Steinheim, Germany, pr. C4129, > 40 units/mg protein). Different concentrations of α-CT

were prepared in PBS buffer (0.2 g KCl, 0.2 g KH2PO4, 1.15 g Na2HPO4, 8 g NaCl in 1 L

demineralized water, containing 0.5 g/L sodium azide to inhibit bacterial growth) of pH ≈ 8.

2.2 Dynamic coating of stainless-steel capillaries

The PEA coatings were applied to pre-weighed, dried and cleaned stainless-steel capillaries

(i.d. = 1.15 or 0.98 mm, length = 40, 60, 80 or 100 mm) by drawing the polymer

formulations (0.1 g/mL solution in ethanol) through the capillary at a flow rate of 5 μl/min

with a syringe pump using withdrawal/injection mode. After passing four capillary volumes

back and forth the coated capillaries were placed on a rolling bar mixer (Stuart SRT9D,

Keison Products, Chelmsford, Essex, U.K.) at 40 rmp for 1 h, to homogenize the thickness

of the coating. Subsequently, the coated capillaries were dried at 40oC while flushing with a

very gentle stream of nitrogen gas (approximately one bubble/s, tested in water) for 24 h.

The coating procedure was optimized by cutting the coated capillary (with known amount of

coating) in four pieces of equal lengths. The amounts of coating were calculated in each part

by weighing before and after the coatings were removed. Approximately the same amount of

coating was obtained from each part with a relative standard deviation (r.s.d) of 10%.

Katsarava [14, 15] reported that polymers containing amino acids with fatty lateral

substituents (e.g. those possessing –CH2-CH(CH3)2 groups, such as PEAs based on leucine)

showed contraction of the film upon exposure to PBS during biodegradation. Therefore,

film-thickness measurements could be misleading and these were not taken into account.

The results obtained by coating capillaries of different lengths are given in Table 1. The

coating weight was always about 20 µg/mm2, except for the shortest capillary.

Table 1 Data obtained upon coating capillaries of different length and diameter with PEA.

Number L of capillary

(mm)

ID of capillary

(mm)

Surface area of

capillary (mm2)*

PEA coating

(mg)

coating

(g/mm2)

n RSD

(%)

1 40 0.98 125.7 1.61 13 6 13.7

2 60 0.98 188.6 3.53 19 4 11.4

3 80 0.98 251.4 5.03 20 4 8.2

4 100 0.98 314.3 6.45 21 4 11.1

5 107 1.15 386.7 8.09 21 4 7.7

*Calculated.

Chapter 4

90

2.3 On-line LC-ToF-MS analysis

An Agilent 1100 series HPLC system consisting of a degasser, an isocratic pump, a gradient

pump, an auto-sampler (all from Agilent Technologies, Waldbronn, Germany) and a column

oven (Waters Temperature Control Module, Milford, MA, USA) set at 37oC was used for the

on-line analyses. The system was configured by connecting the isocratic pump and a PEAs

coated capillary in series to port 6 of a 2/6 micro-switching valve (Agilent, Waldbronn,

Germany). The loop of the valve, connected at port 1 and 4, contained a Zorbax XDB C18

column (150 × 4.6 mm i.d., 5 µm particle size) (Agilent Technologies, Wilmington, DE,

USA). Port 2 and 3 were connected to the gradient pump and to a 6210 series ESI-ToF-MS

(Agilent Technologies, Waldbronn, Germany), respectively. Figure 1 describes the system

for the two positions of the valve. Port 5 was used for waste in either position.

The system thus configured was used for an on-line study of the enzymatic degradation of

PEA coatings in two different modes. (i) Continuous-feed mode involves continuous

pumping of enzyme solution through the coated capillary at different flow rates. The release

of degradation products is monitored by collecting these through switching the valve at

given time intervals during the course of the degradation. (ii) Pulse-feed mode, in which the

isocratic pump is pumping water at different flow rates through the coated capillary. Enzyme

solutions are injected using sandwich injection [16], comprising of 10 µl PBS plugs before

and after the enzyme plug in valve position A. The contact time of the enzyme in the coated

capillary (at 37oC) was varied by changing the flow rate of the isocratic pump from 10 to

100 µL min-1. The enzymatic degradation products were collected at the top of the C18

column. Subsequently, the valve was switched from position A to B for gradient elution of

the degradation products.

The separations of the adsorbed degradation products on the C18 column were carried out by

running a gradient from 5% (v/v) B at t = 0 min to 60% (v/v) B at t = 25 min, held until t =

27 min and then back to 5% (v/v) B at t = 30 min (tend = 35 min after a 5-min final hold). The

flow rate was 1.5 mL min-1 and the flow was split post-column between the waste reservoir

(approximately 1.3 mL min-1) and the electrospray-ionization (ESI) interface (approximately

0.2 mL min-1) by means of a zero-dead-volume T-piece. 0.1% (v/v) aqueous formic acid

(Fluka) (mobile phase A) and acetonitrile (HPLC grade, Biosolve) (mobile phase B) were

used to constitute the gradient. Highly pure water for mobile-phase preparation was obtained

by means of an Arium® 611 Ultrapure (18.2 MΩ*cm) Water System (Sartorius AG,

Goettingen, Germany). The sensitivity of ESI-ToF-MS proved different from day to day.


91

Therefore, for each set of experiments a leucine standard was injected during the run to

adjust for differences in sensitivity. Blank experiments without coating were performed to

determine the enzyme-related background.

1

6

5 4

3

2

37oC

Waste

A B

Internally coated stainless steel tube

TOF-MS

C18 Column 37oC

Grad-Pump

Water

Iso-Pump

α-CT

Auto-sampler

A

1

6

5 4

3

2

37oC

Waste

A B


TOF-MS

C18 Column 37oC

Grad-Pump

Water

Iso-Pump

α-CT

Auto-sampler

A

1

6

5 4

3

2

37oC

Waste

A B


Iso-Pump

TOF-MS

Grad-Pump

α-CT

Auto-sampler

B

C18 Column 37oC

Water

1

6

5 4

3

2

37oC

Waste

A B


Iso-Pump

TOF-MS

Grad-Pump

α-CT

Auto-sampler

B

C18 Column 37oCC18 Column 37oC

Water

Figure 1 Schematic diagram of the on-line testing for the enzymatic degradation of PEAs coatings. (A) Position A; injected enzyme (in PBS) comes in contact with the coatings in the capillary; the degradation products are adsorbed on the C18 column (present in the loop); (B) Position B; Gradient is run to separate the degradation products and to identify them by ESI-ToF-MS. Iso-pump and Grad-pump represents isocratic and gradient-elution pumps, respectively. The auto-sampler is used to inject pulses of enzyme-containing solution in case of pulse-feed mode. In case of continuous-feed mode the cooled enzyme solution in PBS was pumped by the iso-pump.

The HPLC system was hyphenated with a 6210 series Time-of-Flight Mass Spectrometer

(Agilent Technologies, Waldbronn, Germany) via an ESI interface. The conditions of the

ESI-ToF-MS were as follows: drying gas was nitrogen (N2) at 8 L min-1, 300oC and 200kPa.

The capillary voltage, fragmenter and skimmer, were set at 3500 V, 140 V and 60 V,

respectively. The octopole dc1 and octopole radio frequency were set at 33 V and 250 V,

respectively. The data were acquired in the scan mode from m/z 50 to 3000 with 0.88

Chapter 4

92

scans/sec. An Agilent MassHunter Workstation A.02.01 and AnalystTM QS 1.1 software

(Applied Biosystems) were used for data acquisition and data analysis, respectively.

Figure 2 Selected TIC chromatograms the degradation products of (A) di-block PEA and (B) tri-block PEA obtained in continuous-feed mode with 2 min switching time (20 μL) with α-CT (0.025 mg/mL, 1 U/mL in PBS); iso-pump flow rate 10 μL/min. The peak numbers correspond to the structural assignments based on m/z values in Table 2 and Schemes 2 and 3. Peaks with * represent the α-CT peak in each chromatogram (see the supporting information for ESI-ToF-MS spectrum). Symbol § represents the switching time from position A to B (Figure 1).


3.1 Continuous-feed mode

Figure 2a shows the degradation products of di-block PEA (Scheme 1) generated as a result

of continuous-feed enzymatic degradation. Table 2 lists the m/z values and the structures

assigned to each peak. All combinations of LC10 and KC10 contained an amide bond


93

between respective amino acid and C10. In case of K(Bn), the carboxylic acid group of the K

is end-capped with benzyl alcohol and forms a lysine-benzyl ester bond. The carboxylic acid

group of L is connected to C6 via ester bond in all LC6 combinations. For the optimization

of the continuous-feed mode peaks 8, 9, 10, 14, and 18 (Scheme 2) were used. In this mode

the enzyme was dissolved in PBS and the iso-pump (Figure 1) was continuously pumping

the enzyme solution through the coated capillary. After running for an hour at the constant

flow-rate used, the valve was switched from position B to A and then back to position B

(Figure 1) to direct a specified volume of the effluent (containing water-soluble degradation

products as a result of enzymatic degradation in the coated capillary) to the C18 column. By

changing the switching time different volumes of the solution of degradation products were

loaded on the C18 column before analysis by gradient-elution LC (Figure 3a).

0

5000000

10000000

15000000

20000000

25000000

30000000

35000000

40000000

45000000

50000000

0 5 10 15 20 25 30 35 40 45

Inj. Vol. (μL)

Pea

l are

a (c

ps

)

0

200000

400000

600000

800000

1000000

1200000

1400000

1600000

Pea

k are

a (c

ps

)

0

2000000

4000000

6000000

8000000

10000000

12000000

0.0 0.4 0.8 1.2 1.6 2.0 2.4 2.8 3.2 3.6 4.0 4.4

Reaction time (min)

Pe

al a

rea

(cp

s)

0

50000

100000

150000

200000

250000

Pea

k are

a (c

ps)

A

B

0

5000000

10000000

15000000

20000000

25000000

30000000

35000000

40000000

45000000

50000000

0 5 10 15 20 25 30 35 40 45

Inj. Vol. (μL)

Pea

l are

a (c

ps

)

0

200000

400000

600000

800000

1000000

1200000

1400000

1600000

Pea

k are

a (c

ps

)

0

2000000

4000000

6000000

8000000

10000000

12000000

0.0 0.4 0.8 1.2 1.6 2.0 2.4 2.8 3.2 3.6 4.0 4.4

Reaction time (min)

Pe

al a

rea

(cp

s)

0

50000

100000

150000

200000

250000

Pea

k are

a (c

ps)

A

B

Figure 3 (A) Peak areas of peak 8 (♦) and peak 9 () (left-hand-axis) and peak 14 () and peak 18 () (righ-hand axis) vs. injection volume (calculated from switching times) of α-CT solution (0.025 mg/mL, 1 U/mL) passing at a flow rate of 10 μL/min through the coated capillary. (B) Peak areas of peak 8 (◊), peak 9 (∆), peak 14 (), and peak 18 () as a function of flow rate for passing 10 µL of α-CT solution (0.025 mg/mL, 1 U/mL) through the coated capillary. Capillary dimensions are L = 40 mm, i.d. = 1 mm.

Chapter 4

94

The results indicated that the linearity of the ESI-MS detection was acceptable. Similarly,

identical injection volumes (10 µL) were loaded with different flow rates of α-CT solution

(0.025 mg/mL, 1 U/mL) (Figure 3b). The contact (direct reaction) time decreased with

increasing flow rate. Contact times up to 4 min led to a proportional increase of the

concentration of peaks 8 and 9 at the selected enzyme concentration. Peaks 14 and 18

showed opposite release profiles as the contact time increased. This indicates further

degradation at longer contact times of C6LC10L (ester bonds between L and C6 in peak 14)

and LC10[K(Bn)C10]1L (ester bond between Bn and K in peak 18), respectively. Previous

results showed that the ester bond in the structure that gives rise to peak 14 is more stable

than that of peak 18, since no benzyl ester peak was observed in case of batch-mode analysis

[7]. An injection volume of 10 µL and a flow rate of 10 µL/min were selected for further

experiments.

To study the effect of enzyme concentration, the degradation of the PEA coating was

monitored with α-CT solutions with concentrations of 0.5, 1, 2, and 5 U/mL for 24 h at a

flow rate of 10 µL/min. After every 30 min, 10 µL of the effluent of the PEA coated

capillary were loaded on the C18 column and characterized by gradient-elution LC-MS.

Figure 4a shows a decrease in the peak area of peak 8 (Figure 2) over a period of 24 h of

32% at 5 U/mL and 22% at 0.5 U/mL. The decrease in peak area is due to a decrease in

enzyme activity, the faster decrease at higher enzyme concentration likely indicates that the

binding sites of the enzyme saturate the surface area of the coating [17]. Figure 4b and 4c

show the rate of formation of different degradation products. Again, peak 14 and 18 showed

a non-monotonous, contrasting profile due to further hydrolysis.

After 24 h degradation each capillary was dried in an oven at 40oC while flushing with

nitrogen, to calculate the percentage weight loss for different enzyme concentrations. A non-

linear increase in the percentage weight loss was observed with an increasing enzyme

concentration. The on-line system in its current mode of operation appears to be very

efficient for determining normalized relative degradation rates (cf. Figure 4d). A very small

amount of enzyme (here 0.2 U α-CT) is required, because of the low mass detection limits of

LC-ESI-ToF-MS.


95

0100000200000300000400000500000600000700000800000900000

1000000

0 1 2 3 4 5 6

[Eo] U/mL

Ra

teo

ffo

rma

tion

(cp

s/h

ou

rs)

0

5000000

10000000

15000000

20000000

0.5 3.0 5.5 8.0 10.5 13.0 15.5 18.0 20.5 23.0

Time (hours)

Pe

ak

are

a (

cps)

0

5000000

10000000

15000000

20000000

25000000

30000000

35000000

40000000

45000000

0 1 2 3 4 5 6

[Eo] U/mL

Rat

e of

for

mat

ion

(cps

/hou

rs)

0

100

200

300

400

500

600

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

[E] (U/ml)

loss

(ng/

day/

U/m

m2 ) flow

batch

A

BD

C

0100000200000300000400000500000600000700000800000900000

1000000

0 1 2 3 4 5 6

[Eo] U/mL

Ra

teo

ffo

rma

tion

(cp

s/h

ou

rs)

0

5000000

10000000

15000000

20000000

0.5 3.0 5.5 8.0 10.5 13.0 15.5 18.0 20.5 23.0

Time (hours)

Pe

ak

are

a (

cps)

0

5000000

10000000

15000000

20000000

25000000

30000000

35000000

40000000

45000000

0 1 2 3 4 5 6

[Eo] U/mL

Rat

e of

for

mat

ion

(cps

/hou

rs)

0

100

200

300

400

500

600

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

[E] (U/ml)

loss

(ng/

day/

U/m

m2 ) flow

batch

A

BD

C

Figure 4 (A) Decrease in the peak area of peak 8 for 10 μL α-CT injected at different intervals of time for a coated capillary (L = 40 mm, ID = 1 mm, surface-to-volume ratio 4.1 mm2/μL) flushed at a flow rate of 10 µL/min with four different α-CT concentrations 0.5 (), 1 (♦), 2 (), and 5 () U/mL. (B and C) Rate of formation with respect to peak 8 (◊), peak 9 (), peak 10 (), peak 14 (∆) and peak 18 (). (D) normalized weight loss during 24 h as a function of enzyme concentration with continuous-feed mode (♦) and batch-mode analysis () [7].

It is clear that the specific degradation (Figure 4b and 4c, peaks 8, 9, 10, 14, 18) and the

overall degradation (figure 4d, weight loss) are rendering different information. At high

enzyme concentrations the normalized weight loss per unit enzyme slows down twice as fast

as the formation of degradation product LC10L (peak 8) (and at different contact times).

Thus, these two effects are not directly correlated. The results may be compared with the

approximately 1% weight loss of PEA coating with 17 U/mL of enzyme in 24 h observed in

a batch-mode study [7].

The effect of the surface area on the rate of degradation was studied by delivering α-CT (1

U/mL) at a flow rate of 10 µL/min to capillaries with different lengths (L = 40 mm, 60 mm,

80 mm, 100 mm) but constant internal diameter (1 mm) for 24 h. After every 30 min, 10 µL

of the effluent was loaded on the C18 column for gradient-elution LC.

Chapter 4

96

0

5000000

10000000

15000000

20000000

25000000

30000000

35000000

40000000

0.0 4.0 8.0 12.0 16.0 20.0 24.0

Time (hours)

Pe

ak

are

a(c

ps)

0

500000

1000000

1500000

2000000

2500000

3 4 5 6 7 8 9 10 11

Capillary Lenght (cm)

Rat

eof

form

atio

n(c

ps/

hou

rs)

490

500

510

520

530

540

550

560

570

580

100 120 140 160 180 200 220 240 260 280 300 320 340

Surface area (mm²)

Loss

(n

g/m

m²/U

/24hr)

0

10000000

20000000

30000000

40000000

50000000

60000000

70000000

3 4 5 6 7 8 9 10 11


Rat

e of

for

mat

ion

(cps

/hou

rs)

A

BD

C

0

5000000

10000000

15000000

20000000

25000000

30000000

35000000

40000000

0.0 4.0 8.0 12.0 16.0 20.0 24.0

Time (hours)

Pe

ak

are

a(c

ps)

0

500000

1000000

1500000

2000000

2500000

3 4 5 6 7 8 9 10 11


Rat

eof

form

atio

n(c

ps/

hou

rs)

490

500

510

520

530

540

550

560

570

580

100 120 140 160 180 200 220 240 260 280 300 320 340

Surface area (mm²)

Loss

(n

g/m

m²/U

/24hr)

0

10000000

20000000

30000000

40000000

50000000

60000000

70000000

3 4 5 6 7 8 9 10 11


Rat

e of

for

mat

ion

(cps

/hou

rs)

A

BD

C

Figure 5 (A) Decrease in the peak area of peak 8 for 10 μL α-CT injected at different intervals of time for different lengths of coated capillary 40 (♦), 60 (), 80 (), and 100 () mm (each i.d. = 1 mm) flushed at flow rate of 10 µL/min with α-CT solution (1 U/mL). (B, C) The rate of formation with respect to peak 8 (◊), peak 9 (), peak 10 (), peak 14 (∆) and peak 18 (). (D) The normalized hydrolysis rate during 24 h as a function of capillary area.

Figure 5a shows a gradual decrease in the peak area of peak 8 (Figure 2) for each coated

capillary with time. The rate of formation of peak 8, 9 and 10 increases almost linearly with

an increase in capillary length (Figure 5b). The weight loss of PEA coating as a function of

surface area (varied through the capillary length) is shown in Figure 5d. There is a slight

gradual decrease in the weight-loss per unit area as the length of the capillary increases.

However, not all peak areas increase regularly with capillary length. This is most obvious

from the release profiles of peaks 14 (C6LC10L) and 18 (LC10[K(Bn)C10]1L) in Figures

3b, 4c, and 5c. It can be concluded that these instable products showed deviant behaviour

when varying the contact time, enzyme concentration, and capillary length, respectively.

Therefore, the rate of formation of different degradation products cannot be equated with the

overall weight loss of the polymer.

The coatings were also flushed with PBS for 24 h to study its contribution to the degradation

of the PEA coating. No water-soluble products were observed in these latter experiments.


97

O

(CH3)2HC

ONH

O

O

HN

4HO 3OH

CH(H3C)2

O

HO

(CH3)2HC

ONH

O

O

HN

4 OH

CH(H3C)2

O

H2N NH

HO O

2

O

O

OH4

H2NOH

CH(H3C)2

O

OH

H2NHN

2

HO OO

O

NH

OH

CH(CH3)2

O4 HO

O

O

NH

OH

CH(CH3)2

O4

O

(CH3)2HC

ONH

O

O

OH4HO 3

OHHO 3

H2N NH2

HO O

2

O

(CH3)2HC

ONH

O

O

HN

HN

4 2

O O

*

m

O 3

HN

CH(CH3)2

O

*

O

O4 n

3 24

56713

8

14

9 10

11

12

15

16

17

18 18

19

20

21

22

23 24

24

24

n = 1n = 2 n = 3

n = 4

O

(CH3)2HC

ONH

O

O

HN

4HO 3OH

CH(H3C)2

O

HO

(CH3)2HC

ONH

O

O

HN

4 OH

CH(H3C)2

O

H2N NH

HO O

2

O

O

OH4

H2NOH

CH(H3C)2

O

OH

H2NHN

2

HO OO

O

NH

OH

CH(CH3)2

O4 HO

O

O

NH

OH

CH(CH3)2

O4

O

(CH3)2HC

ONH

O

O

OH4HO 3

OHHO 3

H2N NH2

HO O

2

O

(CH3)2HC

ONH

O

O

HN

HN

4 2

O O

*

m

O 3

HN

CH(CH3)2

O

*

O

O4 n

3 24

56713

8

14

9 10

11

12

15

16

17

18 18

19

20

21

22

23 24

24

24

n = 1n = 2 n = 3

n = 4

Scheme 2 Degradation pattern of di-block poly(ester amide) based on sebacic acid, 1,6-hexanediol, lysine-benzyl ester and leucine, subjected to enzymatic degradation (see text for experimental conditions). The structures proposed are numbered according to their separation in Figures 2a and 6a. The m/z values are listed in Table 2.

Chapter 4

98

3.2 Pulse-feed mode

Figure 6a shows the separation of water-soluble products (mainly monomers and oligomers)

released as a result of enzymatic degradation of di-block PEA (Scheme 1A) during pulse-

feed analysis. The numbering of each peak corresponds to the list of structures deduced from

the m/z values (Table 2).

The switching times of the valve were optimized by recording the area of the α-CT peak (1

mg/mL) at different switching times using no capillary, an un-coated capillary (L = 107 mm,

i.d. = 1.15 mm) or a coated capillary of the same size, with a 13 µl capillary loop instead of

an LC column installed.

Figure 6 TIC chromatograms of the degradation products of (A) di-block PEA and (B) tri-block PEA obtained in pulse-feed mode with 20 μL injection volume of α-CT (0.125 mg/mL, 5 U/mL in PBS) at 100 μL/min iso-pump flow rate. The peak numbers correspond to the structural assignments based on m/z values in Table 2 and structures in Schemes 2 and 3. Peak marked * represents the α-CT peak in each chromatogram.


99

Figure 7a and 7b show the resulting profiles as a function of flow rate and configuration. The

40 µl injection broadens into a much wider band. This is monitored by sampling via the 13

µl capillary loop at different switching times. The (coated) capillary causes most of the band

broadening, while the system adds approximately 0.1 mL band width to the 40 µL plug of α-

CT injected. This increased band broadening (and accompanying dilution) is almost constant

across the range of flow rates studied and it is not affected by the coating on the capillary

wall. Despite the variation in concentrations, mutual comparisons can be made based on the

total enzyme activity, because the entire resulting plug is analyzed by LC. The peak areas of

oligomers were observed to increase linearly with the injection volume of α-CT.

0

500000

1000000

1500000

2000000

0 1 2 3 4 5 6 7 8 9 10

Switching time (min)

CT

-Pe

ak

are

a

05000000

10000000150000002000000025000000300000003500000040000000

0 5 10 15 20 25 30 35 40 45 50 55 60Time (hours)

Pe

ak

Are

a (

cps)

0

500000

1000000

1500000

2000000

0 2 4 6 8 10 12 14 16 18 20 22


CT

-Pe

ak

are

a

A

B

C

0

500000

1000000

1500000

2000000

0 1 2 3 4 5 6 7 8 9 10


CT

-Pe

ak

are

a

05000000

10000000150000002000000025000000300000003500000040000000

0 5 10 15 20 25 30 35 40 45 50 55 60Time (hours)

Pe

ak

Are

a (

cps)

0

500000

1000000

1500000

2000000

0 2 4 6 8 10 12 14 16 18 20 22


CT

-Pe

ak

are

a

A

B

C

Figure 7 (A) Peak areas of α-CT (40 μL, 1 mg/mL, 40 U/mL, injected with no (un)coated capillary installed) as a function of the switching time. Flow rates 100(),50() and 25() µL/min. (B) Band profile without capillary (), with uncoated () and coated () capillary. Flow rate 25 µl/min (C) areas of peaks 8 and 9 for 10 μL α-CT injected as a function of time. Capillary dimensions L = 107 mm, ID = 1.15 mm.

When 10 μL of α-CT was injected on the coated capillary after different time periods, a

decrease in the peak areas of water-soluble oligomers was observed. Figure 7c shows a

decrease in the areas of peaks 8 (LC10L) and 9 (LC10[KC10]1L) with increasing time. This

can be attributed to a decrease in enzyme activity in the supply vessel at room temperature.

The flow rate of the iso-pump to transport the enzyme plug and the length of the plug

determine the contact time of the enzyme with the coating in the capillary. Figure 8 shows

the peak areas of selected oligomers generated at different reaction times, when 20 µL of α-

Chapter 4

100

CT solutions of different activities were injected at different flow rates. As expected, the

formation rate for different types of degradation products is clearly different and strongly

dependent on enzyme concentration and reaction time. It appears that ester-containing

reaction products require a lower optimal enzyme concentration than other products (e.g.

diacids). Two effects are thought to contribute to this observation, viz. the preference of α-

CT for ester bonds and subsequent hydrolysis of the primary degradation products.

0

10000000

20000000

30000000

40000000

50000000

60000000

0 3 6 9 12 15

Reaction time (min)

Pea

k ar

ea (

cps)

0

2000000

4000000

6000000

8000000

10000000

0 3 6 9 12 15

Reaction time (min)

Pea

k ar

ea (

cps)

0

5000000

10000000

15000000

20000000

25000000

0 3 6 9 12 15

Reaction (min)

Pea

k ar

ea (

cps)

0

10000000

20000000

30000000

40000000

50000000

0 3 6 9 12 15

Reaction time (min)

Pea

k ar

ea (

cps)

B

C D

A

0

10000000

20000000

30000000

40000000

50000000

60000000

0 3 6 9 12 15

Reaction time (min)

Pea

k ar

ea (

cps)

0

2000000

4000000

6000000

8000000

10000000

0 3 6 9 12 15

Reaction time (min)

Pea

k ar

ea (

cps)

0

5000000

10000000

15000000

20000000

25000000

0 3 6 9 12 15

Reaction (min)

Pea

k ar

ea (

cps)

0

10000000

20000000

30000000

40000000

50000000

0 3 6 9 12 15

Reaction time (min)

Pea

k ar

ea (

cps)

B

C D

A

Figure 8 Change in the areas of peaks 8 (A) 14 (B) 9 (C) 21 (D) with concentration of α-CT enzyme. 1 U/mL (♦), 5 U/mL (), 10 U/mL (). Reaction time was varied by changing the flow rate of iso-pump from 10 to 100 µL/min.

The effect of PBS on the degradation of PEA coating was also tested in pulse mode by

injecting 40 µL of PBS solution after every 30 min for 24 h. No water-soluble products were

observed. When different α-CT concentrations (1U, 5U, and 10 U/mL) were injected at

different flow rates (10, 25, 50, 75, and 100 μL/min), the peak areas of amide-bond-

containing fragments (peaks 8 to 12) increased with a decrease in flow rate. However, the

peak areas of the diol- and benzyl ester-containing fragments (peaks 14 and 21, respectively)

showed an optimum reaction time (flow rate). This suggests that at lower flow rates (longer

contact times) products such as those reflected in peaks 14 and 21 convert to products such

as those of peaks 8 and 9 due to further ester hydrolysis (Figure 8).


101

3.3 Comparison of pulse-feed mode and continuous-feed mode

The on-line system was tested in both the pulse-feed and continuous-feed modes for the

enzymatic degradation of PEA, coated in capillaries, with α-CT as the enzymatic model. In

continuous-feed mode, the coated capillary was continuously flushed with enzyme solution

and a certain volume of post-capillary degradation solution was injected in the LC column

by switching from position B to A and then back to position B (Figure 1). Under the

dynamic sink conditions used, the benzyl ester containing degradation products released

from the surface of the coating to the flowing solution were degraded further. Peaks 8 to 12,

the oligomers containing amide bonds originating from the n block of the PEA (Scheme 1),

were stable under these dynamic conditions. This continuous-feed mode of analysis can be

used to study the weight loss of PEA as a result of enzymatic degradation under flow

conditions with on-line detection of water-soluble degradation products. The results obtained

in this mode showed a faster normalized degradation rate of PEA as compared to the batch-

mode degradation experiments performed before [7].

In pulse-feed mode only a band of α-CT passes through the coated capillary. The moment

this band is injected, it starts diffusing in the surrounding PBS. Therefore its reaction time

and concentration can only be approximated and semi-quantitative results (for mutual

comparison) are obtained. A 40 μL plug is diluted about five-fold during passage through the

coated capillary. The separation of degradation products formed in this mode shows benzyl-

containing products (Figure 2a). Such products were neither observed in case of batch-mode

analysis of this PEA [7], nor in case of continuous-feed mode analysis, except for peak 18.

This shows that pulse-feed analysis makes it possible to determine the primary degradation

products. With this mode of analysis a number of enzymes can be tested in very small

amounts and in a short time for their ability to degrade a certain PEA coating and their

selectivity towards ester and/or amide bonds. Delivery of microliter volumes of active

enzyme can be realized from the cooled tray of an auto-sampler. In addition, the effect of

different peroxide-containing solutions (2 < pH < 9) and simulated body fluids on the

degradation of coatings can be studied. The information provided by the pulse-feed analysis

is ideally suitable to design long-term degradation experiments with selected enzymes.

Chapter 4

102

3.4 Application to tri-block PEA coatings

The system was applied to test the degradation of tri-block PEA coatings with α-CT in both

the continuous-feed and pulse-feed modes. Figure 2b shows the water-soluble products

released during continuous-feed mode analysis with α-CT (0.025 mg/mL, 1U/mL in PBS) at

a flow rate of 10 µL/min with 2 min switching time (20 µL loaded on C18 column). Peaks 8

to 12 represent the oligomeric series from the n block of the polymer. Benzyl ester and DAS-

containing peaks were not observed. The chromatogram differs from that obtained from the

di-block PEA (Figure 2a) in that peak 18 is absent in Figure 2b. The degradation of tri-block

PEA in the pulse-feed mode (Figure 8b) exhibits benzyl ester (peaks 18, 22, and 23), diol

(peaks 14, 15, and 21), and DAS (peaks 25, 26, and 27) containing fragments in addition to

stable amide-containing (peaks 8, 9, 10, 11) n-block oligomers. The ester bond associated

with DAS (o-block) also exhibits degradability under the current experimental conditions

with α-CT.

3 2

25

8

14

910

11

12

18

18

2122

23

n = 1 n = 2n = 3

n = 4

4

26

27

o = 1

3 2

25

8

14

910

11

12

18

18

2122

23

n = 1 n = 2n = 3

n = 4

4

26

27

o = 1

Scheme 3 Degradation pattern of tri-block poly(ester amide) based on sebacic acid, 1,6-hexanediol, lysine-benzyl ester, leucine and isosorbide (1,4-dianhydrosorbitol (DAS)) subjected to enzymatic degradation (experimental conditions see text). The structures proposed are numbered according to their separation in Figures 2b and 6b. The m/z values are listed in Table 2.


103

The major degradation patterns of the di- and tri-block PEA are given in Schemes 2 and 3,

respectively. The oligomers in Scheme 2 that are not mentioned in Scheme 3 are present in

low concentrations. The schemes differ through the presence of isosorbide-containing

oligomers in Scheme 3, originating from the third block of the tri-block PEA. The m:n block

ratio is clearly reflected in the concentration ratios of the respective oligomers in Figures 6a

and 6b.

Table 2 Structural assignments for the water soluble fragments (monomers and oligomers), separated as shown in Figure 2a and 2b as a results of enzymatic degradation during pulse feed mode and continuous feed mode analysis, respectively. K, K(Bn), L, C6 and C10 represent lysine, lysine-benzyl ester, leucine, 1,6-hexanediol and sebacic acid, respectively.

Peak No.

Symbol No. of Amide bond broken m/z [M+H]+

1 formic acid

2 L 1 132.11

3 C6 0 119.11

4 Benzyl alcohol 0 91.055 [M-OH]+

5 KC10 1 331.23

6 LC10 1 316.21

7 C10LC6 1 416.31

8 LC10L 0 429.31

9 LC10[KC10]1L 0 741.51

10 LC10[KC10]2L 0 1053.71

11 LC10[KC10]3L 0 1365.92

12 LC10[KC10]4L 0 1679.12

13 KC10L 1 443.36

14 LC10LC6 1 529.39

15 C6LC10[KC10]1L 0 841.59

16 LC10[K(Bn)C10]1 1 718.46

17 C6LC10K(Bn)C10K 1 946.58

18 LC10[K(Bn)C10]1L 0 831.54

18 LC10[K(Bn)C10]-[KC10]L 0 1143.78

19 C6LC10[KC10]2 0 1039.77

20 LC10[K(Bn)C10]-[KC10]LC6 0 1243.86

Chapter 4

104

21 LC10[K(Bn)C10]1LC6 0 931.65

22 LC10[K(Bn)C10]2-[KC10]L 0 1546.03

23 LC10[K(Bn)C10]2L 0 1233.81

24 LC10[KC10]3 1 1251.87

24 LC10[KC10]2-[K(Bn)C10] 1 1341.86

24 LC10[K(Bn)C10]2-[KC10]LC6 0 1646.10

25 (DAS)LC10L 0 557.34

26 LC10K(Bn)C10L(DAS) 0 959.59

27 LC10L(DAS)LC10L 0 967.62

4 Conclusions

On-line analysis by LC-ToF-MS is useful to separate and identify the enzymatic degradation

products of PEAs coatings generated under dynamic sink conditions. Low flow rates of the

enzyme solution and longer coated capillaries enhance the extent of degradation. The

percentage weight loss increases gradually with an increase in α-CT activity, but it is hardly

affected by the capillary length. Normalized formation rates of degradation products depend

on enzyme activity and contact time and are highly compound specific, so that specific-

product profiles may be quite different from the overall weight-loss trend. Pulse-feed

analysis generated benzyl ester-containing degradation products, which were not observed in

continuous-feed analysis. The developed system allows testing of highly expensive enzymes

for their ability to degrade a particular novel PEA-based biomaterial. The biodegradation of

two types of PEA with a selected enzyme was investigated.

The use of very small amounts of enzymes, no sample preparation steps, conditions closer to

those of physiological degradation, and shorter analysis times are significant advantages of

the proposed system to study the degradation of biomedical materials and to identify the

nature of their degradation products. This is expected to result in better insight in the

underlying mechanisms.


105

5 References

[1] Aylvin A. Dias, M. Hendriks, Drug Delivery Technology 10 (2010) 20. [2] A. Ghaffar, P.G. Verschuren, J.A.J. Geenevasen, T. Handels, J. Berard, B. Plum,

A.A. Dias, P.J. Schoenmakers, S. van der Wal, Journal of Chromatography A 1218 (2011) 449.

[3] R.L. Reis, J.S. Román, Biodegradable Systems in Tissue Engineering and Regenerative Medicine, CRC Press, 2004.

[4] R.S. Labow, Y. Tang, C.B. McCloskey, J.P. Santerre, Journal of Biomaterials Science, Polymer Edition 13 (2002) 651.

[5] A.K. Burkoth, J. Burdick, K.S. Anseth, Journal of Biomedical Materials Research 51 (2000) 352.

[6] L. Castaldo, P. Corbo, G. Maglio, R. Palumbo, Polymer Bulletin 28 (1992) 301. [7] A. Ghaffar, G.J.J. Draaisma, G. Mihov, A.A. Dias, P.J. Schoenmakers, S. van der

Wal, Biomacromolecules 12 (2011) 3243. [8] A.-C. Albertsson, M. Hakkarainen, in Chromatography for Sustainable Polymeric

Materials -Renewable, Degradable and Recyclable ; Advances in Polymer Science 211, Springer, 2008, p. 85.

[9] C.M. Agrawal, J.S. McKinney, D. Lanctot, K.A. Athanasiou, Biomaterials 21 (2000) 2443.

[10] S.P. Gorman, C.P. Garvin, F. Quigley, D.S. Jones, Journal of Pharmacy and Pharmacology 55 (2003) 461.

[11] Y.-y. Huang, M. Qi, M. Zhang, H.-z. Liu, D.-z. Yang, Transactions of Nonferrous Metals Society of China 16 (2006) s293.

[12] Y. Yang, Y. Zhao, G. Tang, H. Li, X. Yuan, Y. Fan, Polymer Degradation and Stability 93 (2008) 1838.

[13] C.-C. Chu, R. Katsarava, in U.P. B2 (Editor), US Patent 7304122 B2, 2007. [14] G. Tsitlanadze, M. Machaidze, T. Kviria, N. Djavakhishvili, C.C. Chu, R. Katsarava,

Journal of Biomaterials Science, Polymer Edition 15 (2004) 1. [15] R. Katsarava, V. Beridze, N. Arabuli, D. Kharadze, C.C. Chu, C.Y. Won, Journal of

Polymer Science Part A: Polymer Chemistry 37 (1999) 391. [16] Y. Mengerink, R. Peters, M. Kerkhoff, J. Hellenbrand, H. Omloo, J. Andrien, M.

Vestjens, S. van der Wal, Journal of Chromatography A 876 (2000) 37. [17] K. Mukai, K. Yamada, Y. Doi, International Journal of Biological Macromolecules

15 (1993) 361.

Chapter 4

106

6 Supporting information

6.1 Solubility

The solubility of the tri-block PEA in water and in a number of common organic solvents

was assessed by combining 10 to 30 mg of the polymer with 0.5 mL of the respective

solvents (Table S1) at room temperature (20oC). The samples were vortexed and allowed to

dissolve overnight. The PEA was separated on a 250 x 4.6 mm Zorbax Eclipse XDB-C18

column from the solvent, using THF as a mobile phase and sandwich injection in the solvent

to prevent precipitation in the injection system.

Table S1 Solubility of the tri-block PEA used in the present study in common organic solvents evaluated at 25oC

solvents Solubility solvents Solubility solvents Solubility

Toluene - Tetrahydrofuran + Methanol +

Dichloromethane + Ethyl acetate - Acetic acid +

Isopropanol + Dimethylformamide + Formic acid +

Chloroform + Acetone -

N,N`-dimethyl acetamide

+ Acetonitrile - water -

Dimethyl sulfoxide + Ethanol +

Soluble +, insoluble -

6.2 Molecular weight (Mw) and dispersity of tri-block PEA

The SEC experiments were performed on an LC system equipped with a Waters 2590

Alliance separation module (Waters, Milford, MA, USA) and an RID-10A refractive-index

detector (Shimadzu, Kyoto, Japan). The SEC analyses were performed on three PLgel

MIXED-B columns (10 µm, 300 × 7.6 mm i.d.) (Polymer Laboratories, Church Stretton,

U.K.) connected in series. THF stabilized with butylated hydroxytoluene (BHT), (BioSolve,

Valkenswaard, The Netherlands) was pumped at a flow rate of 1 mL min-1. The injection

volume was 50 µL and the column-oven temperature was set at 50oC. Polystyrene standards

(Polymer Laboratories) were used to calibrate the SEC-UV-dRI system. Data were recorded

and chromatographic peaks were processed using Class-VP 7.4 software (Shimadzu). Molar-


107

mass distributions (MMD) were calculated from the chromatograms using software written

in-house in Excel 2003 (Microsoft).

Table S2 Peak molecular weight (Mp), weight-average molecular weight (Mw), number-average molecular weight (Mn), and dispersity (PDI) of the polystyrene standards. Data specified by the supplier (Polymer Laboratories).

Polystyrene Standards

Mp Mw Mn (PDI)

PS1 580 640 555 1.16 PS2 1310 1300 1220 1.07 PS3 2100 2100 2010 1.05 PS4 4920 4890 4740 1.03 PS5 9920 9910 9700 1.02 PS6 19880 19680 19220 1.02 PS7 30230 30500 29600 1.01 PS8 52400 51950 51050 1.02 PS9 70950 69200 67350 1.03

PS10 96000 94650 92350 1.03 PS11 126700 124100 121200 1.03 PS12 197300 201500 196800 1.02

PS13 299400 297100 292200 1.02

y = -0.00248x3 + 0.15813x2 - 3.71023x + 35.55551

R2 = 0.99985

0

5

10

15

20

25

30

35

40

45

50

17 18.5 20 21.5 23 24.5 26

Ret. Vol. (mL)

dR

I R

esp

on

se

0.000

1.000

2.000

3.000

4.000

5.000

6.000

Lo

g M

PS-1 PS-2 PS-3 PS-4 PS-5 PS-6 PS-7 PS-8

PS-9 PS-10 PS-11 PS-12 PS-13 PEA3

Figure S1 Size-exclusion separation of tri-block PEA (PEA3) and polystyrene standards (overlayed chromatograms).

Chapter 4

108

Table S3 Mw, Mn, and dispersity of tri-block PEA (Scheme1B)

Tri-block PEA Mw Mn PDI

I 55445 27100 2.05

II 56233 28360 1.98

III 56707 28128 2.02

Average 56128 27863 2

% RSD (n=3) 1 3 2

6.3 NMR experiments

Aproximately 20 mg of the sample was dissolved in 1 mL of d6-ethanol and 1H, 13C, 1H-1H

gCOSY (two-dimenssional homonuclear H, H gradient-correlated spectroscopy) and HSQC

(heteronuclear single quantum coherence) NMR experiments were recorded on a Varian

Inova 500-MHz NMR (Varian, Palo Alto, CA, USA) equipped with a 500 5mm 13C/31P/1H

GS probe.

2.2113.43

2.00

1.61

4.62

1.54

5.51

5.02

13,14

*

8

17

34

2

16, 712, 11

6

9

5

15

1, 1022, 23

*

*

21

18,20,23,

19, 24

25

6

7

7

9 8

8

8

89 8

8

8

87

7

9

13

1412

10

11

1

2

3

4

5

15

1516

17

17

16

19

20

21

22

23

24

18

25

2.2113.43

2.00

1.61

4.62

1.54

5.51

5.02

13,14

*

8

17

34

2

16, 712, 11

6

9

5

15

1, 1022, 23

*

*

21

18,20,23,

19, 24

25

6

7

7

9 8

8

8

89 8

8

8

87

7

9

13

1412

10

11

1

2

3

4

5

15

1516

17

17

16

19

20

21

22

23

24

18

25

Figure S2 1H NMR spectrum of tri-block PEA in d6-ethanol.


109

The integration of characteristic protons signals for 1,6-hexanediol (15), lysine (5),

isosorbide (21) confirmed the molar composition of the tri-block polymer (m:n:o =

0.31:0.27:0.42; see Scheme 1), which is very close to the intended ratio 0.3:0.25:0.45.

Lysine (5) and benzyl (25) groups show a 1:1 ratio. Lysine (5) and sebacic acid (6) exhibit a

1:1.1 ratio in the n-block of the polymer. Accurate integration was difficult in the case of

overlapping signals. The HSQC experiment was performed to positively identify the

characteristic signals.

13,14

8

17

3426

9

515

21

18,20,23

* **

*Solvent

16, 712, 11

19, 241, 1022, 23

13,14

8

17

3426

9

515

21

18,20,23

* **

*Solvent

16, 712, 11

19, 241, 1022, 23

Figure S3 HSQC spectrum of tri-block PEA in d6-ethanol.

Chapter 4

110

6.4 ESI-ToF-MS spectrum of α-chymotrypsin

In Figure 2 and 6, the peaks labelled with the symbol (*) are identified as α-CT enzyme

based on the ESI-ToF-MS spectrum of the peak (Figure S4). The molecular weight

calculated from this spectrum is 25 kDa, which is in agreement with the α-CT supplier’s data

(Sigma, C4129, 25 kDa).

+10+11

+12+13+14

+15

+16

+17

+18

+19

+20+21

+22

+23

+24

+25

+26

+27+28+9

+10+11

+12+13+14

+15

+16

+17

+18

+19

+20+21

+22

+23

+24

+25

+26

+27+28+9

Figure S4 ESI-ToF-MS spectrum of α-CT for the peaks labeled with symbol (*) in Figure 2 and 6.

111

Summary

In recent years, the demand for synthetic degradable polymeric biomaterials has been

growing continuously for temporary applications in the field of drug-delivery devices, for

tissue engineering, in scaffolds, or in surgical implants. The information obtained from their

chemical analysis can be used to modify existing biomaterials and for rational design of new

materials. The suitability of such materials for specific applications strongly depends on their

rate of degradation and their biocompatibility. Therefore, the objective of the present study

was to develop methods for the analysis of biomaterials by degrading them under non-

physiological and physiological conditions, followed by the analysis of their degradation

products.

The thesis starts with a review of different degradation methods and analytical strategies,

which may be applied to gain in-depth knowledge on the chemical structure of degradable

polymers and the toxicological nature of their degradation products (chapter 1). Methods for

the degradation and analysis of degradable materials using chromatographic separations and

spectroscopic or spectrometric detection methods provide useful information. Each

analytical technique has its limitations, but combination of techniques may provide powerful

tools to aid in optimizing the performance of degradable polymers.

In chapter 2 an approach is described to analyse the structure of different polyesters and

networks based on polyester urethane acrylates, based on completely degrading the polymers

at harsh conditions, such as elevated temperature and extreme pH. Degradation was

performed in alkaline media in a microwave instrument. Residues in the glass vessels were

prevented by incorporating an internal PTFE (“Teflon”) liner. Hydrolysis until completion

was monitored using NMR spectroscopy. The amount and the kinetic chain length of

poly(methacrylic acid) backbone were determined by size-exclusion chromatography (SEC).

The monomeric products were separated and quantified by liquid chromatography

hyphenated to mass spectrometry (LC-MS). The study provided useful insights in the

composition of these novel polymeric networks.

LC-MS was exploited to analyse the products of multi-block poly(ester amide) (PEA),

partially degraded at extreme pH or high temperature (chapter 3). This analysis provided the

basis to monitor the release of different monomers and oligomers generated by the enzyme-

mediated degradation of PEA with α-chymotrypsin and proteinase K under physiological

conditions. The compositional analysis by LC-MS yielded useful information on the

Summary

112

polymerization process. Quantitative analysis of several degradation products revealed the

different esterase and amidase activities of both enzymes. The potential of the studied

polymer for drug-delivery applications was supported by the steady degradation of the

polymer under physiological conditions, as was evident from the observed weight loss and

from the average molecular weight of the remaining polymer the degradation, which was

preserved according to SEC measurements. This confirmed a surface-degradation

mechanism. The molar ratios between blocks of the copolymer and different components of

the polymer were ascertained by 1D and 2D NMR measurements.

In chapter 4 a miniaturized and automated system is described to study the kinetics of

degradation and the products formed in an efficient way. Further degradation of the

generated products in the enzyme media could be avoided. The system allowed tracking the

degradation products of PEA, coated internally in stainless-steel tubing, under dynamic sink

and static conditions. The system was investigated using different reaction times, enzyme

concentrations, and capillary lengths. LC-MS proved to be good analytical technique for the

on-stream separation and identification of the released degradation products. Analysis in

continuous-feed and pulse-feed modes allowed to differentiate between primary and

secondary degradation products. Weight loss curves and concentration profiles of specific

degradation products often did not coincide. Shorter analysis times, no sample preparation

and more favourable conditions make the developed system a versatile analytical tool to

study the degradation of polymers. The minimal requirements on sample quantities and

enzyme activities highlight the efficient application of this system.

This thesis resulted in the quantitative analysis of complex polyacrylate networks and an

innovative system for fast assessment of degradable polymeric coatings.

113

Samenvatting

De laatste jaren laten een gestage groei zien van de vraag naar degradeerbare biomaterialen

op basis van synthetische polymeren. Deze worden onder andere toegepast op de gebieden

van gecontroleerde medicijnafgifte, in weefselkweek, als botschraag en als implantaat.

De analyse van de chemische structuur van biomaterialen levert informatie die kan worden

gebruikt voor verbetering van de huidige en het rationeel ontwerpen van nieuwe materialen.

De geschiktheid van zulke materialen voor specifieke doeleinden is sterk afhankelijk van

hun afbraaksnelheid en hun biocompatibiliteit. De doelstelling van de onderhavige studie is

dan ook om methoden te ontwikkelen voor de analyse van biomaterialen door ze (deels) af

te breken onder al dan niet fysiologische omstandigheden, gevolgd door de analyse van hun

afbraakproducten.

Het proefschrift begint met een overzicht van verschillende afbraakmethoden en analytische

strategieën die toegepast zouden kunnen worden om diepgaand inzicht te verkrijgen in de

structuur van degradeerbare polymeren en de toxicologische aard van hun afbraakproducten

(hoofdstuk 1). Methoden voor afbraak en analyse van afbreekbare materialen met behulp

van chromatografische scheidingen en spectroscopische of spectrometrische

detectiemethoden leveren nuttige informatie omtrent de structuur van de materialen en hun

afbraakproducten. Elke afzonderlijke analytische techniek heeft zijn beperkingen, maar

combinaties van technieken kunnen krachtige hulpmiddelen zijn ter ondersteuning van de

optimalisering van de eigenschappen van afbreekbare polymeren.

In hoofdstuk 2 wordt de analyse beschreven van de structuren van verschillende polyesters

en netwerken op basis van polyester urethaan acrylaten door de polymeren geheel af te

breken bij extreme omstandigheden zoals hoge temperatuur en pH. Een PTFE (“teflon”)

voering was nodig om neerslag te voorkomen in de monsterbuis van de magnetron. Met

NMR werd de hydrolyse tot het eindpunt gevolgd. De hoeveelheid en ketenlengte van het

gevormde poly(methacrylzuur) werden bepaald met size exclusion chromatografie (SEC).

De overige (monomere) producten werden gescheiden en gekwantificeerd met behulp van

vloeistofchromatografie gekoppeld met massaspectrometrie (LC-MS). Dit onderzoek

verschafte nuttige inzichten met betrekking tot de samenstelling van deze nieuwe polymere

netwerken.

LC-MS werd gebruikt om de producten van gedeeltelijk bij extreme pH of hoge

temperatuur afgebroken multi-blok poly(esteramide) (PEA) te analyseren (hoofdstuk 3).

Samenvatting

114

Deze analyse maakte het mogelijk om het vrijkomen van verschillende monomeren en

oligomeren te volgen, die worden gevormd door de enzymatische afbraak van PEA onder

invloed van α-chymotrypsine of proteinase K onder fysiologische omstandigheden. Analyse

van de samenstelling met LC-MS leverde informatie over het polymerisatieproces.

Kwantitatieve analyse van verscheidene afbraakproducten bracht ook de verschillende

esterase en amidase activiteiten van beide enzymen aan het licht. De goede vooruitzichten

van dit polymeer in toepassingen voor medicijnafgifte werden ondersteund door de

geleidelijke afbraak onder fysiologische omstandigheden, zoals bleek uit het

gewichtsverlies en het gelijk blijven van het gemiddeld molucuulgewicht van het

overblijvend polymeer, dat met SEC tijdens de afbraak werd onderzocht. Dit bevestigde dat

deze afbraak een oppervlakteverschijnsel betrof. Tenslotte werden met 1D- en 2D-NMR

technieken de molaire verhoudingen tussen de copolymeerblokken en tussen de

verschillende delen van het polymeer vastgesteld.

In hoofdstuk 4 wordt een geminiaturiseerd en geautomatiseerd systeem beschreven om de

kinetiek van de afbraak op een snelle en efficiënte manier te bestuderen en de verdere

afbraak van de geproduceerde verbindingen in de enzymoplossing te voorkomen. De

afbraakproducten van PEA, dat is aangebracht op de binnenkant van een roestvrijstalen

buisje, kunnen met dit systeem worden gevolgd onder statische, maar ook onder

dynamische (niet-evenwichts) omstandigheden. Het systeem is onderzocht met

verschillende reactietijden, enzymconcentraties en lengtes van het capillair. De vrijkomende

afbraakproducten werden on-line gescheiden en geïdentificeerd met LC-MS. Onderscheid

tussen primaire en secundaire afbraakproducten bleek mogelijk door analyse met continue

of gepulseerde toevoer van enzymoplossing. Het gewichtsverlies en het concentratieprofiel

van specifieke afbraakproducten vielen vaak niet samen. De kortere analysetijden,

ontbreken van de noodzaak voor monstervoorbereiding en gunstige reactieomstandigheden

maken het systeem tot een veelzijdig analytisch hulpmiddel voor het bestuderen van de

afbraak van polymeren. Andere prominente voordelen van dit systeem zijn de geringe

hoeveelheden polymeer en lage enzymactiviteit die vereist zijn.

Dit promotieonderzoek heeft geresulteerd in de kwantitatieve analyse van de samenstelling

van complexe polyacrylaatnetwerken en een handig, innovatief systeem voor de snelle

analyse van afbreekbare polymere coatings.

Acknowledgements

115

Acknowledgments

With the name of ALLAH, the Beneficent, the Merciful. Countless and humble thanks to

Almighty God, who blessed and bestowed me the courage and patience to fulfil this

assignment.

Finally the time has arrived. I can appreciate, and acknowledge the support and help of those

special people I met and worked with. First of all, I owe my deepest gratitude to my

promotor, Sjoerd van der Wal, for his kind support, effort and worthy advices during the

accompanied period. This thesis would not have been possible without his brilliant guidance

and the highly constructive critical discussions I had with him during the research work.

Despite the distance, he always responded to my queries by e-mail, even on weekends. I

sincerely appreciate the very kind, friendly and motivating attitude of my co-promotor, The

Big Peter, who provided me with an opportunity to work in his very dynamic and

international research group. Indeed, highly useful and informative discussions with him

always added value to my research work. Sjoerd and Peter, thank you very much for your

kind help and valuable contributions to my education and research. It was an honour for me

to work under your supervision.

I am very grateful to Prof. W. E. Hennink, Prof. D. W. Grijpma, Prof. J. G. M. Janssen, Prof.

C. G. de Koster, Dr. W. Th. Kok, and Dr. A. A. Dias for being members of my promotion

committee. The feedback and comments from Hans-Gerd Janssen were very much

appreciated.

Definitely, I cannot forget the wonderful time I spent at the University of Amsterdam and in

the cosmopolitan Amsterdam. My special thanks go to Wim, a true Amsterdam native, for

his expertise, cooperation, and precious suggestions throughout my PhD. Your warm

reception in the very early morning of 1st November, 2007 at Schiphol is highly regarded.

I am truly indebted and thankful to my wonderful colleagues of the Analytical-Chemistry

Group (former Polymer-Analysis Group) for making the work environment very friendly. To

Gabriel, for helping me out with data analysis and understanding MATLAB, whenever I

visited your office. To Erwin, for guidance in understanding the art of connecting and

disconnecting different HPLC systems and detectors to make suitable combinations. To

Peter Pruim, for providing me the opportunity to use the nano-pump, which was very fruitful

for me. To Rob, for providing useful tips on SEC/GPC. To Linda, for her hospitality, shukria

olivia. To Elena and Dominique, for useful discussions. To Rudy, for proofreading my thesis

Acknowledgements

116

and for being very supportive. To Peter verschuren, for pleasant technical discussions to help

tackle the difficulties during my research work. To Tom, for his generosity and tremendous

amount of cooperation that definitely facilitated my research. To Petra Aarnoutse, for her

kind help. My warmest and special thanks go to Aleksandra and Daniela, for their utmost

help and for being very caring throughout my stay. Importantly, I cordially thank Rashid for

the wonderful time we spent together in Gein and at the University of Amsterdam. Playing

tennis, cooking, and roaming around the Gasperplas with you will be a part of my memories.

Thanks for helping me in designing the cover of my thesis in photoshop. Your absence was

very much felt during the last couple of months of my stay in Amsterdam. I also enjoyed the

time I spent with to Stella, Francisco, Filippo, Sonja, Arend, Dini, Xulin Jiang, Hanneke,

Ngoc A, Katya, Maria, Jana, Chuchu, and Henrik (the viking). I also thank my students

Friso, Rolf, Roy, Merel, and Paul for working with me at different levels of my PhD and for

their contributions to this dissertation. I am grateful to Azmatullah Khan Baloch for inviting

me frequently for tasty pakoras and tea. I am privileged to have had Hannan Tahir – a very

kind, humble and sober person – as my housemate during the last year of my stay. To Sajjad

Zahir, thanks for bringing delicious food several times.

I owe sincere and earnest thankfulness to Marianne (from international office), Gerda, Petra

Hagen, Maureen, Renate (from HIMS office), and Marijke Duyvendak (from UvA Library)

for their indispensable assistance. Great discussions with the skilled Jan van Maarseveen and

benign/kind support from Hans Bieraugel (from the synthetic-organic-chemistry group), Jan

Geenevasen and Jan Meine (from the NMR department), Louis Hartog (biocatalysis and bio-

organic chemistry group), and Marjo Mittelmeijer-Hazeleger (from heterogeneous catalysis

and sustainable chemistry) were very much appreciated. I would like to mention Marco

Scholten (from Agilent Technologies) for his help in understanding different aspects of the

time-of-flight mass spectrometer.

It was a pleasure to collaborate with Janine Jansen and Dirk Grijpma; I always enjoyed my

visits to the University of Twente and to have useful discussions with you.

I am thankful for the support of DSM during my research. In particular, I would like to take

the opportunity to thank Aylvin for his critical remarks and highly useful suggestions; Ron

Peters, Bart Plum, and Tristan Handels for their kind assistance with the polyester-urethane-

acrylates project. The knowledge, expertise, and the exceptional support of George Mihov

and Guy Draaisma helped me a lot during the analysis of multi-block poly(ester amide)s. I

also would like to convey my thanks to Marc Hendriks.

Acknowledgements

117

I sincerely acknowledge the financial support for my PhD from the Higher Education

Commission (HEC) of Pakistan and Nuffic, The Netherlands. Certainly, the kind and caring

responses from Rana Shafiq Ahmad, Muhammad Ashfaq (from HEC), Charlene and Loes

(from Nuffic) are highly appreciated.

I am also grateful to my employer “University of Engineering and Technology Lahore” for

granting me the study leave to pursue my PhD at the University of Amsterdam. I would like

to express my gratitude to my colleagues at the Chemistry Department of UET for their best

wishes.

I am thankful to Prof. Robert Langer and Dr. Bozhi Tian (Massachusetts Institute of

Technology) for their kindness to provide me the high resolution picture and for the

generosity in granting me the permission to use it on the cover of my thesis.

I also acknowledge the support of my friends Shahzad Ahmed, Adnan Tahir, Mian Tariq,

Ijaz-ul-Mohsin, Khurram, Muhammad Abu Naeem (Lionel Gomes), and many more, who

helped me from the beginning and who made my stay comfortable.

If I forgot anybody, I humbly apologize. Thanks anyway.

Finally, I would like to thank my parents and family members for their prayers,

encouragement, and support…..

شايد .مجهے اپنی دعاۇں ميں ياد رکها وقت ہر نے گزار ہوں جنہوں شکر کا والدين پيارے ميں، ميں اپنے آخر..…

اهللا تعالى آپ کو صحت اور تندرستی .ميرے پاس وہ الفاظ نہيں جن سے ميں انکے احسانات کا شکريہ ادا کر سکوں

حيات اور اپنی پياری بيٹی اور بيٹے کو شريک ميں اپنی .دير ہمارے سروں پر قائم رکهے دے اور آپ کا سايہ تا

ميں .ہوں کرتا پيار بہت سے سب آپ ميں. ہوں سراہتا کو ہمت اور حوصلے کے ان اور ميں کبهی نہيں بهول سکتا

گزار ہوں جنہوں ئيوں سے شکراور اپنی بہنوں کا دل کی گہرا....) عمير حاکم، عامر، آصف،( بهائيوں اپنے پيارے

احمد باسم اور بهانجوں پيارے سے ہی بہت ميں اپنے. نے ميری غير موجودگی ميں والدين کا خيال رکها

ہوں اٹهتا مسکرا کر ميں ہميشہ سن سکتا، جن کو بهول کو کبهی نہيں باتوں لطف، پر ميٹهی، ميٹهی کی

.ہوں جاتا بهول پريشانياں تمام اپنی اور

Abdul Ghaffar

Amsterdam, September 2011

Bibliography

Printed by Ipskamp Drukkers, Enschede, The Netherlands 118

Bibliography

[1] A. Ghaffar, P.G. Verschuren, J.A.J. Geenevasen, T. Handels, J. Berard, B. Plum, A.A. Dias, P.J. Schoenmakers, S. van der Wal, Fast in vitro hydrolytic degradation of polyester urethane acrylate biomaterials: structure elucidation, separation and quantification of degradation products, J. Chromatogr. A, 1218 (2011) 449-458.

[2] A. Ghaffar, G. J. J. Draaisma, G. Mihov, A. A. Dias, P.J. Schoenmakers, Sj. van der Wal, Monitoring the in vitro enzyme-mediated degradation of degradable poly(ester amide) for controlled drug delivery by LC-ToF-MS, Biomacromolecules 12 (2011) 3243-3251.

[3] A. Chojnacka, A. Ghaffar, A. Feilden, K. Treacher, H.-G. Janssen, P. Schoenmakers, Pyrolysis–gas chromatography–mass spectrometry for studying N-vinyl-2-pyrrolidone-co-vinyl acetate copolymers and their dissolution behaviour, Analytica Chimica Acta (2011) doi:10.1016/j.aca.2011.05.052

[4] A. Ghaffar, G. J. J. Draaisma, G. Mihov, P.J. Schoenmakers, Sj. van der Wal, A versatile system for studying the enzymatic degradation of multi-block poly(ester amide)s, to be submitted.

[5] A. Ghaffar, P.J. Schoenmakers, Sj. van der Wal, Methods for the chemical analysis of degradable synthetic polymeric biomaterials, to be submitted.

[6] Janine Jansen, Abdul Ghaffar, Thomas N.S. van der Horst, George Mihov, Sjoerd van der Wal, Jan Feijen, Dirk W. Grijpma, Controlling the kinetic chain length of the crosslinkages in photo-polymerized biodegradable networks, to be submitted.

Course Attended

[1] Molecular Spectroscopy, 2009, organized by AIO Network Analytical Chemistry,

The Netherlands.

[2] Chemometrics, 2010, organized by AIO Network Analytical Chemistry, The Netherlands.

[3] Polymer Chemistry, RPK-A 2010, organized by National Dutch Graduate School of Polymer Science and Technology (ptn).

Documents

UvA-DARE (Digital Academic Repository) Degradation and ... · Proefschrift – Degradation and Analysis of Synthetic Polymeric Materials for Biomedical Applications by Abdul Ghaffar