The phase behavior in supercritical carbon dioxide of hyperbranched copolymers with architectural variations

Accepted Manuscript Title: The Phase Behavior in Supercritical Carbon Dioxide of Hyperbranched Copolymers with Architectural Variations Author: Paweł G. Parzuchowski Jacek Gregorowicz Edyta P. Wawrzy´nska Dominik Wi˛acek Gabriel Rokicki PII: DOI: Reference:

S0896-8446(15)30081-4 http://dx.doi.org/doi:10.1016/j.supflu.2015.07.028 SUPFLU 3404

To appear in:

J. of Supercritical Fluids

Received date: Revised date: Accepted date:

20-4-2015 22-7-2015 23-7-2015

Please cite this article as: P.G. Parzuchowski, J. Gregorowicz, E.P. Wawrzy´nska, D. Wi˛acek, G. Rokicki, The Phase Behavior in Supercritical Carbon Dioxide of Hyperbranched Copolymers with Architectural Variations, The Journal of Supercritical Fluids (2015), http://dx.doi.org/10.1016/j.supflu.2015.07.028 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

The Phase Behavior in Supercritical Carbon Dioxide of Hyperbranched Copolymers with Architectural Variations

ip t

Paweł G. Parzuchowski,a*Jacek Gregorowicz,b Edyta P. Wawrzyńska, a Dominik Wiącek,a Gabriel

cr

Rokickia

Warsaw University of Technology, Faculty of Chemistry, Noakowskiego 3, 00-664 Warsaw, Poland

b

Institute of Physical Chemistry of the Polish Academy of Sciences, Kasprzaka 44/52, 01-224 Warsaw,

us

a

e-mail:

[email protected]*,

an

Poland

[email protected],

M

[email protected], [email protected]

[email protected],

Abstract

d

Hyperbranched polymers (HBPs) have been known and extensively investigated for over two decades. However, there are still areas that need to be explored. Recently, much effort has been placed

Ac ce pt e

in developing drug delivery systems based on macromolecules of three-dimensional structure. This paper describes the synthesis of hyperbranched copolyesters of 2,2-bis(hydroxymethyl)propionic acid (bis-MPA) containing small amounts (5 or 10%) of chain extending units of various lengths and their phase behavior in supercritical carbon dioxide (scCO2) after modification with trifluoroacetic anhydride. The structure of the copolyesters was confirmed with 1H and

13

C NMR, FTIR spectroscopies and

MALDI-TOF mass spectrometry. The phase behavior of the polymers in supercritical carbon dioxide was explored as a function of concentration and temperature. It was shown that polymers containing 5% (in respect to bis-MPA) of chain extending units of the moderate length (5-hydroxypentanoate or 6hydroxyhexanoate) exhibited the lowest phase transition parameters in supercritical carbon dioxide.

Keywords: hyperbranched polymer, polyester, phase behavior, supercritical carbon dioxide

Introduction 1

Page 1 of 30

Recently, much effort has been placed in developing drug delivery systems based on macromolecules of three-dimensional structure.[1] One of the investigated topics is introduction of drug molecules into the polymeric matrix. For this purpose an easily available, cheap and biocompatible solvent – supercritical carbon dioxide (scCO2) can be applied. Over the past thirty years there has been an intense interest in the use of scCO2 for laboratory and industrial applications. Supercritical carbon dioxide is believed to be a good choice for polymer synthesis [2,3] and processing.[4] Unfortunately, carbon

ip t

dioxide’s solvent power is low, especially for polar and high molecular weight polymers. Design and synthesis of CO2-soluble surfactants, ligands and phase transfer agents demand a broad knowledge on

cr

solubility of polymers in dense carbon dioxide. Identification of highly CO2-soluble polymers has been a subject of intense research for the last twenty years.[5-10] In our previous paper [11] a more detailed

us

description of results on phase behavior of polymers in CO2 solutions published in the open literature up to 2009 was presented. We refer a reader to this publication for more information on the subject. In the recent years a number of research projects have been undertaken to explore the possibility to synthesize

an

CO2-philic material based on poly(vinyl acetate) (PVAc). PVAc has been recognized as one of the most CO2-philic hydrocarbon material and it was expected that their modifications would make it possible to

M

obtain inexpensive hydrocarbon material as CO2-philic as fluoro and silicon based polymers. Tan et al.[12] have examined solubility of PVAc in CO2 by a high throughput gravimetric extraction (HTGE) screening method and determined cloud-point pressures for this system using a variable volume view

d

cell (VVVC). They have shown that the solubility of PVAc strongly depends on the molecular weight

Ac ce pt e

and end groups functionalities. Recently a random copolymers of vinyl acetate and vinyl butyrate (PVAc-PVB) and a copolymer of vinyl acetate and dibutylmaleate (PVAc-PDBM) have been shown to exhibit higher solubility than PVAc.[13-14] Girard et al. have also examined the structure – property relationship between PVAc-based copolymers and their solubility in supercritical CO2.[15,16] Hyperbranched polymers (HBPs) have been known and extensively investigated for over two decades.[17-21] However there are still areas, such as their physicochemical properties, that need to be explored. In contrast to dendrimers, HBPs can be easily prepared in a one-pot procedure. Their properties are affected by the nature of the backbone and the chain-end functional groups, degree of branching, chain length between branching points, and the molecular weight distribution. Hyperbranched polymers can be easily modified to tailor their properties for a specific purpose. The first extensive experimental studies on the solubility of hyperbranched polyesters and polyethers in scCO2 showed that the nature of the end groups is an important factor that influences phase transition parameters.[11] The results have also shown that the nature of the interior of the hyperbranched macromolecules is also significant. The general observation was that hyperbranched polyesters dissolve at lower pressures than hyperbranched polyethers.[11] Other polyesters - aliphatic hyperbranched 2

Page 2 of 30

polycarbonates modified with fluorinated end-groups showed only moderate solubility in scCO2. It seemed that enhancement in solubility of the hyperbranched polymer in comparison to the linear polycarbonates was caused mainly by the difference in the number of the fluorinated end groups.[22] The first studies conducted on the phase behavior of hyperbranched polymers with architectural variation showed that it is possible to adjust the phase transition pressure in scCO2 by incorporation of linear homologous oligo(6-hydroxyhexanoate)s of the different lengths into the structure of

ip t

hyperbranched poly(2,2-bis(hydroxymethyl)propionic acid).[23] It was shown that incorporation of linear oligo(6-HH) segments increased the distances between branching points and made the structure of

cr

the HBP less dense and better accessible for the solvent. Actually, the lowest phase transition pressures were obtained for the polyesters containing 5 and 10mol% of oligo(6-HH) units in respect to bis-MPA

us

units.

In this work we propose to expand previous studies by investigation of the influence of various oligo(alkylcarboxylate) structures as chain extender units on the solubility of the hyperbranched

Ac ce pt e

d

M

an

polyester in supercritical carbon dioxide. This idea is illustrated in Figure 1.

Figure 1. Illustration of the idea of incorporation of linear units increasing distances between branching points of 2,2-bis(hydroxymethyl)propionic acid based polyesters (Boltorn®). To date, the most popular and commercially available hyperbranched polymers are hydroxyl terminated polyethers (based on glycidol – Polyglycerol®)[24] and polyesters (based on 2,2bis(hydroxymethyl)propionic acid – Boltorn®).[25-28] They have already been modified in numerous 3

Page 3 of 30

ways by end-capping with various functional groups [29-30] or even outer or inner spheres differentiation.[31] Due to the growing interest in biodegradable materials, new HBPs were synthesized utilizing 2,2-bis(hydroxymethyl)propionic acid (bis-MPA) as a branching molecule. Trollsas et al. synthesized a series of ε-caprolactone-based AB2 macromonomers through a living ring opening polymerization, using aluminum benzyloxide as the initiator. The AB2 macromonomers were condensed into polymers

through

a

room-temperature

esterification

using

DCC

and

ip t

hyperbranched

4-

(dimethylamino)pyridinium 4-toluenesulfonate (DPTS).[32] Very similar HBPs were synthesized from

cr

AB2 macromonomers containing two oligo(ε-caprolactone) chains terminated with hydroxyl groups grown from a bis-MPA initiator by ROP by Choi and Kwak.[33] The polymers were designed to

us

incorporate different lengths of linear oligomeric segments consisting of 5, 10, and 20 ε-caprolactone monomer units on the branched backbone chains. The dynamic viscoelastic relaxation behavior and the molecular mobility for these polymers were investigated showing that the molecular mobility of three

an

hyperbranched PCLs was higher than that of linear one, and was observed to enhance with decreasing lengths of oligo(ε-caprolactone) segments and increasing relative degree of branching (DB).[34]

M

Copolymers of bis-MPA with another cyclic ester - δ-valerolactone were obtained via lipase-mediated

Materials

Ac ce pt e

Experimental

d

syntheses using compressed fluids: supercritical carbon dioxide or liquid 1,1,1,2-tetrafluoroethane.[35]

2,2-Bis(hydroxymethyl)propionic acid, glycolide, lactide, δ-valerolactone, ε-caprolactone, ωhydroxydodecanoic acid, trimethylolpropane and trifluoroacetic anhydride were purchased from Aldrich Chemical (Poznan, Poland) and used as received. Solvents were purchased from POCh (Gliwice, Poland) and dried prior to use. Instrumentation

FTIR spectra were recorded on a Bruker ALPHA FTIR spectrometer equipped with Platinum ATR single reflection diamond ATR module. 1H NMR and 13C NMR spectra were recorded on a Varian VXR 400 MHz spectrometer using tetramethylsilane as an internal standard and deuterated solvents (CDCl3, DMSO-d6) and analyzed with MestReNova v.6.2.0-7238 (Mestrelab Research S.L) software. MALDI-TOF measurements were performed on a Bruker UltraFlex MALDI TOF/TOF spectrometer (Bremen, Germany) in a linear or reflectron mode using DHB (2,5-dihydroxybenzoic acid) or HABA (2-(4'-hydroxybenzeneazo)benzoic acid) matrix and Bruker Peptide Calibration Standard (1047.193149.57 Da) as a calibrant and analyzed with flexAnalysis v.3.3 (Bruker Daltonik GmbH) and 4

Page 4 of 30

Polymerix v. 2.0 (Sierra Analytics Inc.) software. The molecular weight and molecular weight distribution were determined by a GPC on a Viscotek system comprising GPCmax and TDA 305 unit equipped with one guard and two DVB Jordi gel columns (102-107, linear, mix bed) in CH2Cl2 as an eluent at 35 °C at a flow rate of 1.0 mL/min using the RI detector and PS calibration. The detailed description of the apparatus and the experimental procedure for phase behavior

ip t

measurements is given in our previous work.[11]

cr

Syntheses Preparation of hyperbranched polyesters (HBP)

us

In a 100 mL reactor equipped with a mechanical stirrer Heidolph RZR 2020, a thermometer, an argon inlet and a distillation condenser 1.12g (8.33 mmol) of trimethylolpropane (TMP) was placed followed by 33.53g (0.25 mol) of 2,2-bis(hydroxymethyl)propionic acid, an appropriate amount of chain extender

an

precursor (see Table 1) and 70.5mg (38.3µL, 0.68 mmol) of 95% H2SO4. The reaction mixture was virgously stirred and heated at 120°C for 10h at atmospheric pressure and then for an additional 3h

M

under reduced pressure (0.01 mmHg). The polymers were then precipitated from acetone into hexane. Polymers were then dried under vacuum for a week in a thin layer and stored under neutral atmosphere.

d

Exact amounts of reagents are given in Table 1.

Sample

Ac ce pt e

Table 1. Amounts of reagents used for synthesis of HBP copolyesters bis-MPA

TMP

g (mmol)

g (mmol)

Chain extender precursor g (mmol)

Yield g

33.53 (250)

1.12 (8.33)

[none]

[23]

0.00 (0.00)

29.2

33.53 (250)

1.12 (8.33)

glycolide

0.78 (6.72)

29.1

HBP-G (10)

33.53 (250)

1.12 (8.33)

glycolide

1.64 (14.1)

30.3

HBP-L (5)

33.53 (250)

1.12 (8.33)

L-lactide

0.88 (6.10)

30.3

HBP-L (10)

33.53 (250)

1.12 (8.33)

L-lactide

1.83 (12.7)

31.0

HBP-V (5)

33.53 (250)

1.12 (8.33)

δ-valerolactone

1.30 (13.0)

30.1

HBP-V (10)

33.53 (250)

1.12 (8.33)

δ-valerolactone

HBP-0 HBP-G (5)

*

2.83 (28.3)

32.6

1.52 (13.3)

31.5

HBP-C (5)

33.53 (250)

1.12 (8.33)

ε-caprolactone

[23]

HBP-C (10)

33.53 (250)

1.12 (8.33)

ε-caprolactone

[23]

3.23 (28.3)

32.4

HBP-D (5)

16.77 (125)

0.56 (4.16)

12-hydroxydodecanoic acid

1.45 (6.7)

15.4

5

Page 5 of 30

HBP-D (10) *

16.77 (125)

0.56 (4.16)

12-hydroxydodecanoic acid

3.09 (14.3)

16.7

theoretical molar pre-cent of chain extending units

HBP-0 see ref. [23] HBP-G (5) Yield 29.1g (94%), 1H NMR (DMSO-d6, 400 MHz); δ (ppm) = 4.94 (bs, 6.6H, OHL), 4.62 (bs, 6.0H, OHT), 4.26-3.85 (m, 23.5H, CH2OC(O)), 3.56-3.22 (m, 31.8H, CH2OH), 1.30-0.95 (m, 34.6H,

ip t

CH3 and CH2 TMP), 0.86-0.75 (m, 3H, CH3 TMP); 13C NMR (DMSO-d6, 100 MHz) = spectrum analogous to HBP-G (10) (lower intensity of glycolic unit peaks); FTIR (ATR): ν (cm-1) = 3400, 2955, 2870, 1723,

HBP-G (10)

cr

1469, 1370, 1304, 1209, 1119, 1039.

Yield 30.3g (95%), 1H NMR (DMSO-d6, 400 MHz); δ (ppm) = 4.24-3.81 (m, 48.3H,

us

CH2OC(O)), 3.60-3.20 (m, 37.1H, CH2OH), 1.33-0.95 (m, 44.8H, CH3 and CH2 TMP), 0.86-0.76 (m, 3H, CH3 TMP); 13C NMR (DMSO-d6, 100 MHz); δ (ppm) =174.6-171.1 (m, C=O), 65.7-64.4 (m, CH2O(CO),

an

64.1-63.1 (m, CH2OH), 59.6-59.2 (m, CH2OGLYCOLIC), 50.3 (s, Cterm.), 48.5-47.8 (m, Clin.), 46.3 (s, Cbran.), 22.2 (bs, CH2 TMP), 17.2-16.2 (m, CH3), 7.5 (bs, CH3 TMP); FTIR (ATR): ν (cm-1) = 3400, 2945, 2870,

M

1722, 1469, 1370, 1304, 1209, 1120, 1039.

HBP-L (5) Yield 30.3g (98%), 1H NMR (DMSO-d6, 400 MHz); δ (ppm) = 4.95 (bs, 6.0H, OH), 4.233.85 (m, 21.6H, CH2OC(O) and CHOC(O)), 3.60-3.21 (m, 22.7H, CH2OH and CHOH), 1.30-0.95 (m, TMP

), 0.88-0.73 (m, 3H, CH3

d

29.6H, CH3 and CH2

TMP

); 13C NMR (DMSO-d6, 100 MHz) = spectrum

Ac ce pt e

analogous to HBP-L (10) (lower intensity of lactic unit peaks); FTIR (ATR): ν (cm-1) = 3381, 2930, 2870, 1722, 1468, 1371, 1304, 1207, 1120, 1038. HBP-L (10)

Yield 31.0g (97%), 1H NMR (DMSO-d6, 400 MHz); δ (ppm) = 4.98 (bs, 12.5H, OH),

4.22-3.88 (m, 36.5H, CH2OC(O) and CHOC(O)), 3.27-3.27 (m, 41.3H, CH2OH and CHOH), 1.28-0.97 (m, 48.5H, CH3 and CH2 TMP), 0.876-0.76 (m, 3H, CH3 TMP); 13C NMR (DMSO-d6, 100 MHz); δ (ppm) =174.4-172.5 (m, C=O), 66.1-64.5 (m, CH2O(CO) and CHOC(O)), 64.3-63.0 (m, CH2OH), 50.3 (s, Cterm.), 48.4-48.1 (m, Clin.), 46.4 (s, Cbran.), 22.6 (bs, CH2 TMP), 20.7-20.4 (m, CH(CH3)), (17.5-16.5 (m, CH3), 7.5-7.3 (m, CH3 TMP); FTIR (ATR): ν (cm-1) = 3394, 2929, 2870, 1722, 1467, 13980, 1304, 1209, 1121, 1038.

HBP-V (5) Yield 30.1g (96%), 1H NMR (DMSO-d6, 400 MHz); δ (ppm) = 4.99 (bs, 7.2H, OH), 4.253.89 (m, 35.3H, CH2OC(O)), 3.63-3.27 (m, 41.3H, CH2OH), 2.24 (bs, 1.3H, CH2 VL), 1.60 (bs, 2.6H, CH2 ), 1.28-0.96 (m, 48H, CH3 and CH2 TMP), 0.91-0.75 (m, 3H, CH3 TMP); 13C NMR (DMSO-d6, 100 MHz)

VL

spectrum analogous to HBP-V (10) (lower intensity of pentanoate peaks); FTIR (ATR): ν (cm-1) = 3377, 2947, 2885, 1720, 1467, 1398, 1302, 1209, 1120, 1037. 6

Page 6 of 30

HBP-V (10) Yield 32.6g (99%), 1H NMR (DMSO-d6, 400 MHz); δ (ppm) = 4.97 (bs, 10.5H, OH), 4.24-3.85 (m, 38.6H, CH2OC(O)), 3.60-3.20 (m, 51H, CH2OH), 2.25 (bs, 2.9H, CH2 VL), 1.62 (bs, 5.8H, CH2 VL), 1.29-0.97 (m, 48H, CH3 and CH2 TMP), 0.91-0.78 (m, 3H, CH3 TMP); ); 13C NMR (DMSO-d6, 100 MHz); δ (ppm) =174.63-171.4 (m, C=O), 65.9-64.6 (m, CH2O(CO), 64.1-63.4 (m, CH2OH), 50.3 (s, Cterm.), 48.5-47.9 (m, Clin.), 46.5-45.6 (m, Cbran.), 33.3-32.6 (m, CH2(CO)VL), 28.3-27.2 (m, CH2CH2OVL), (cm-1) = 3397, 2945, 2885, 1721, 1469, 1398, 1301, 1211, 1120, 1038.

cr

HBP-C (5) see ref. [23]

ip t

22.3 (bs, CH2 TMP), 21.4-20.6 (m, CH2CH2(CO)VL), 17.6-16.6 (m, CH3), 7.5 (bs, CH3 TMP); FTIR (ATR): ν

HBP-C (10) see ref. [23]

us

HBP-D (5) Yield 15.4g (94%), 1H NMR (DMSO-d6, 400 MHz); δ (ppm) = 4.6 (bs, 9H, OH), 4.26-3.88 (m, 28H, CH2OC(O)), 3.66-3.24 (m, 28H, CH2OH), 2.24 (bs, 1.2H, CH2 DD), 1.46 (bs, 2.4H, CH2 DD), 1.24

an

(bs, 8.4H, CH2 DD), 1.17-0.98 (m, 32H, CH3 and CH2 TMP), 0.88-0.71 (m, 3H, CH3 TMP); 13C NMR (DMSOd6, 100 MHz) spectrum analogous to HBP-D (10) (lower intensity of dodecanoate peaks); FTIR (ATR):

M

ν (cm-1) = 33810, 2930, 1722, 1468, 1371, 1304, 1208, 1120, 1038.

HBP-D (10) Yield 16.7g (93%), 1H NMR (DMSO-d6, 400 MHz); δ (ppm) = 4.95 (bs, 6H, OHL), 4.64 (bs, 6H, OHT), 4.28-3.85 (m, 24H, CH2OC(O)), 3.61-3.25 (m, 33H, CH2OH), 2.25 (bs, 2.2H, CH2 DD),

d

1.46 (bs, 4.4H, CH2 DD), 1.22 (bs, 15.4H, CH2 DD), 1.19-0.95 (m, 32H, CH3 and CH2 TMP), 0.83-0.78 (m, 3H,

Ac ce pt e

CH3 TMP); 13C NMR (DMSO-d6, 100 MHz); δ (ppm) =174.6-172.5 (m, C=O), 66.2-64.4 (m, CH2O), 64.263.4 (m, CH2OH), 50.3 (s, Cterm.), 48.5-48.0 (m, Clin.), 46.4-46.0 (m, Cbran.), 33.8-33.1 (m, CH2(CO)DD), 29.3-27.9 (m, CH2DD), 25.7-24.2 (m, CH2DD), 22.3 (bs, CH2 TMP), 17.4-16.5 (m, CH3), 7.7-7.3 (m, CH3 TMP); FTIR (ATR): ν (cm-1) = 3394, 2929, 1722, 1466, 1398, 1304, 1210, 1121, 1038.

Modification of hyperbranched polyesters with trifluoroacetic anhydride HBP-(G,L,V,C,D)f polymers 5g of a hyperbranched polyester was placed in a 25 mL three-neck flask equipped with a magnetic stirrer, a thermometer, a rubber septum and a reflux condenser end-capped with a calcium chloride tube. The flask was flushed with argon. Then a 50% molar excess (in respect to the theoretical amount of hydroxyl groups in the polymer) of an anhydride was slowly added with a syringe. The temperature of the reaction mixture was kept below 0 °C. When the whole polymer dissolved the temperature was raised up and the reaction mixture stirred overnight at room temperature. The excess of anhydride and acid was removed on a rotary evaporator. The polymer was then dissolved in dichloromethane, washed with saturated sodium bicarbonate solution, then several times with water and dried with magnesium 7

Page 7 of 30

sulfate. Inorganic salt was filtered off, solvent removed in vacuo and the polymer kept under high vacuum (0.001 mmHg) for 8h. Polymers were obtained as a light yellow oils. Exact amounts of reagents are given in Table 2.

Table 2. The amounts of reagents used for esterification of HBP with TFA anhydride

mmol**

g

% ***

g

HBP-0f [23]

3610

5

68.5

14.4

92

7.8

HBP-Gf (5)

3710

5

66.7

14.0

92

8.0

HBP-Gf (10)

3820

5

64.8

13.6

91

7.8

HBP-Lf (5)

3720

5

66.5

14.0

91

7.9

HBP-Lf (10)

3840

5

64.5

13.5

93

7.9

HBP-Vf (5)

3780

5

65.5

13.8

92

8.0

3960

5

62.5

13.1

92

7.8

3800

5

65.2

13.7

90

7.7

HBP-Cf (10) [23]

4000

5

61.8

13.0

94

7.6

HBP-Df (5)

3940

5

62.8

13.2

90

7.7

HBP-Df (10)

4290

5

12.1

90

7.5

an

M

HBP-Cf (5)

[23]

57.7

theoretical values for polymers: HBP-0 – HBP-D;

Ac ce pt e

cr

g

us

ip t

Yield

Mn*

HBP-Vf (10)

*

TFA anhydride

d

Polymer

**

for theoretical number of OH groups;

***

calculated for actual degree of substitution (approx.80%). HBP-0f see ref. [23]

HBP-Gf (5) Yield 8.0g (92%), 1H NMR (CDCl3, 400 MHz); δ (ppm) = 5.05 (bs, 3.9H, OH), 4.69-4.40 (m, 48.3H, CH2OC(O)CF3), 4.40-3.98 (m, 52.3H, CH2OC(O)), 3.83-3.60 (m, 7.7H, CH2OH), 1.44-1.06 (m, 77H, CH3 and CH2 TMP), 0.99-0.87 (m, 3H, CH3 TMP); FTIR (ATR): ν (cm-1) = 2985, 1787, 1737, 1471, 1346, 1218, 1122, 999, 773, 731.

HBP-Gf (10) Yield 7.8g (91%), 1H NMR (CDCl3, 400 MHz); δ (ppm) = 5.07 (bs, 4.7H, OH), 4.674.40 (m, 52.7H, CH2OC(O)CF3), 4.40-3.98 (m, 62.3H, CH2OC(O)), 3.83-3.60 (m, 9.5H, CH2OH), 1.441.00 (m, 86.3H, CH3 and CH2 TMP), 1.00-0.87 (m, 3H, CH3 TMP); FTIR (ATR): ν (cm-1) = 2985, 1788, 1737, 1472, 1386, 1347, 1219, 1122, 994, 773, 731.

8

Page 8 of 30

HBP-Lf (5) Yield 7.9g (91%), 1H NMR (CDCl3, 400 MHz); δ (ppm) = 5.20 (bs, 4.8H, OH), 4.69-4.39 (m, 53.1H, CH(CH3)OC(O)CF3 and CH2OC(O)CF3), 4.39-4.02 (m, 58H, CH(CH3)OC(O) and CH2OC(O)), 3.82-3.61 (m, 9.6H, CH2OH), 1.68-1.00 (m, 91H, CH3, CH3 lactic and CH2 TMP), 1.00-0.85 (m, 3H, CH3 TMP); FTIR (ATR): ν (cm-1) = 2987, 1787, 1737, 1471, 1386, 1347, 1219, 993, 773, 731. HBP-Lf (10) Yield 7.9g (93%), 1H NMR (CDCl3, 400 MHz); δ (ppm) = 5.09 (bs, 5.0H, OH), 4.61-

ip t

4.40 (m, 50.8H, CH(CH3)OC(O)CF3 and CH2OC(O)CF3), 4.40-4.00 (m, 57.3H, CH(CH3)OC(O) and CH2OC(O)), 3.84-3.60 (m, 10H, CH2OH), 1.66-1.11 (m, 91.6H, CH3, CH3 lactic and CH2 TMP), 1.02-0.82 (m,

cr

3H, CH3 TMP); FTIR (ATR): ν (cm-1) = 2987, 1788, 1737, 1472, 1346, 1219, 1121, 993, 773, 732.

HBP-Vf (5) Yield 8.0g (92%), 1H NMR (CDCl3, 400 MHz); δ (ppm) = 4.86 (bs, 4.2H, OH), 4.64-4.39

us

(m, 52.6H, CH2OC(O)CF3), 4.39-3.98 (m, 57.6H, CH2OC(O)), 3.81-3.63 (m, 8.5H, CH2OH), 2.36 (bs, 2.5H, CH2 VL), 1.66 (bs, 5H, CH2 VL), 1.45-1.02 (m, 84.7H, CH3 and CH2 TMP), 1.00-0.86 (m, 3H, CH3 TMP);

an

FTIR (ATR): ν (cm-1) = 2985, 1787, 1736, 1472, 1386, 1346, 1219, 1121, 993, 773, 731. HBP-Vf (10) Yield 7.8g (92%), 1H NMR (CDCl3, 400 MHz); δ (ppm) = 4.93 (bs, 4.7H, OH), 4.61-4.40

M

(m, 53.1H, CH2OC(O)CF3), 4.40-4.02 (m, 61.7H, CH2OC(O)), 3.85-3.59 (m, 9.3H, CH2OH), 2.36 (bs, 5.4H, CH2 VL), 1.65 (bs, 10.8H, CH2 VL), 1.46-1.05 (m, 92H, CH3 and CH2 TMP), 0.98-0.85 (m, 3H, CH3 TMP);

Ac ce pt e

HBP-Cf (5) see ref. [23]

d

FTIR (ATR): ν (cm-1) = 2981, 1787, 1736, 1471, 1385, 1346, 1219, 1122, 1001, 773, 731.

HBP-Cf (10) see ref. [23]

HBP-Df (5) Yield 7.7g (90%), 1H NMR (CDCl3, 400 MHz); δ (ppm) = 5.23 (bs, 4.3H, OH), 4.63-4.40 (m, 52.3H, CH2OC(O)CF3), 4.39-4.01 (m, 57.6H, CH2OC(O)), 3.82-3.61 (m, 8.7H, CH2OH), 2.30 (bs, 2.6H, CH2 DD), 1.66-1.02 (m, 108H, CH3, CH2 DD and CH2 TMP), 0.99-0.89 (m, 3H, CH3 TMP); FTIR (ATR): ν (cm-1) = 2945, 1788, 1737, 1472, 1346, 1219, 1122, 1001, 773, 731. HBP-Df (10) Yield 7.5g (90%), 1H NMR (CDCl3, 400 MHz); δ (ppm) = 5.26 (bs, 4H, OH), 4.65-4.41 (m, 49.5H, CH2OC(O)CF3), 4.41-4.02 (m, 56.9H, CH2OC(O)), 3.83-3.60 (m, 8H, CH2OH), 2.29 (bs, 5.4H, CH2 DD), 1.73 -1.05 (m, 128.2H, CH3, CH2 DD and CH2 TMP), 0.98-0.85 (m, 3H, CH3 TMP); FTIR (ATR): ν (cm-1) = 2934, 1788, 1737, 1470, 1347, 1219, 1225, 1003, 773, 731. Results and Discussion

Syntheses 9

Page 9 of 30

The synthetic pathway towards hyperbranched copolyesters of bis-MPA with various ωhydroxycarboxylic acids (or their derivatives) is given in Scheme 1. For the synthesis we applied a direct one-pot procedure of polymerization/polycondensation, assuming that at applied reaction conditions (120 °C, H+) all: a ring opening polymerization (ROP), polycondensation and transesterification reactions would proceed leading to polymers of random monomers distribution. The polymers were designed to maintain a constant bis-MPA/TMP molar ratio (the same number of

ip t

hydroxyl groups per average macromolecule) and to contain 5 or 10 mol% of various spacer units. O

+

OH

HO

Spacer Precursor (G-D)

OH

+ OH

O

(i)

O

O

O

OH

O

O

R

OH

O O O

O O

O

R

O

O

O

O

O

O

O

O

O

O

O

R

O

O

O

O

HBP-V

R = (CH2)4

HBP-C

R = (CH2)5

HBP-D

R = (CH2)11

CF3

O

HBP-Gf

O

HBP-Lf

CF3

O

HBP-Vf

O

R

O

O

R = CH(CH3)

(ii)

O

O

O

HBP-L

O O

O

Ac ce pt e O

R = CH2

OH

d

O

CF3

R

M

O

HBP-G

an

O

OH

cr

HO

us

OH

O

O O

HBP-Cf O

R O

HBP-Df

CF3

Scheme 1. The synthesis of hyperbranched copolyesters poly((2,2-bis(hydroxymethyl)propionate)-co(ω-hydroxyalkylate)); (i) polycondensation: TMP, H2SO4, 120 °C, 0.01 mmHg, 96-99%; G – glycolide, L – lactide, V - δ-valerolactone, C - ε-caprolactone, D – 12-hydroxydodecanoic acid (ii) modification: trifluoroacetic anhydride (TFA), dichloromethane, r.t., 85-94%. A series of copolyesters consisting of poly(4,4-bis(hydroxymethyl)propionic acid) (bis-MPA) units containing ω-hydroxyalkylate units was synthesized in a two stage process. Trimethylolpropane (TMP) was used as a core and sulfuric acid as a catalyst of polycondensation reaction. Application of a 10

Page 10 of 30

core molecule (TMP) made it possible to control the molecular weight and helped to reduce the molarmass dispersity of hyperbranched polymers.[36],[37] Presence of an ethyl group in a core molecule helped in determination of an average molecular weight by 1H NMR spectroscopy. Actual amounts of reagents are given in Table 1. In the first stage a mixture of TMP, bis-MPA, chain extender precursor and a catalyst (H2SO4) were mixed together in one portion and heated at 120 °C under argon atmosphere. Temperature was

ip t

raised slowly to avoid distillation of low molecular weight reactants out of the mixture. The viscosity of the polymer gradually increased during the reaction. After approx. 10 hours, the pressure in the flask

cr

was lowered to 0.01 mmHg and the reaction continued for an additional 3 hours. To remove any unreacted monomers the polymers were dissolved in acetone or acetone/methanol solvent and

us

precipitated into hexane. Polymers were then carefully dried under reduced pressure. The yields varied from 93 to 99% (Table 1).

Since polyols of the Boltorn® type (Figure 1,A) show very limited solubility in supercritical esters, known to increase the polymer solubility.

an

carbon dioxide, OH groups of the polymers HBP-G – HBP-D were transformed into trifluoroacetic acid

M

Briefly, all the parent copolymers HBP-G – HBP-D were reacted with an excess of TFA anhydride below 0 °C under argon atmosphere. The polymers dissolved gradually in the reaction mixture and were left to stir overnight. Then the excess of TFA anhydride and the TFA acid formed

d

during reaction were removed on a rotary evaporator protected with a trap containing KOH pellets. Any

Ac ce pt e

residues of the acid were removed by washing with bicarbonate solution and water.

Characterization of the polymers Parent copolyesters

The chemical structure of the synthesized co-polymers is quite complex. There are 11 possible substructures which can be divided into five different groups of units: starting, dendritic, linear, terminal and chain extending.[23]

Figure 2 shows 1H NMR spectra of all synthesized co-polyesters containing 10mol% of chain extender units. The amount of the additive is low but in each case it is possible to find and assign signals coming from the chain extender units.

11

Page 11 of 30

ip t cr us an M d Ac ce pt e

Figure 2. 1H NMR (400 MHz, DMSO-d6) spectra of hyperbranched copolyesters: HBP-0, HBP-G(10), HBP-L(10), HBP-V(10), HBP-C(10) and HBP-D(10). The degrees of branching (DB) of obtained copolyesters were estimated from the integrals of the signals of methyl groups (e, Figure 2) and within the margin of error, were close to the theoretical 0.5 value. The 13C NMR data confirmed incorporation of chain extender units into the structure of the polymer in each case.

12

Page 12 of 30

ip t cr us an M d Ac ce pt e Figure 3. MALDI-TOF (DHB or HABA matrix, linear or reflectron mode) spectra of unsubstituted hyperbranched copolyesters. 13

Page 13 of 30

In the MALDI-TOF mass spectra the majority of signals were observed as sodium adducts, however, for the most intensive series, potassium adducts were seen as well. Since the molar mass of trimethylolpropane (molecular peak 134.094) is very close to the molar mass of bis-MPA (molecular peak 134.058) it was not possible to judge from the MALDI-TOF which of those molecules worked as a starting unit for the polymerization process. We assumed that on average, there was one TMP structure per macromolecule.

ip t

Figure 3 shows the MALDI-TOF mass spectra of all synthesized co-polyesters containing 10mol% of chain extender units and a base polymer without an additive. Each spectrum consists of the series related to the poly(bis-MPA) and copolymers containing one or more chain extender units.

cr

Spectrum of the base poly(bis-MPA) HBP-0 contains four sets of signals: [C6H14O3 + (C5H8O3)n](Na+

us

and K+), [C6H12O2 + (C5H8O3)n]Na+ and [C5H10O + (C5H8O3)n]Na+ that can be assigned to a TMP core poly(bis-MPA), TMP core poly(bis-MPA) with one water molecule loss (eg. cyclic structures) and TMP core poly(bis-MPA) with the loss of one water and one formaldehyde molecule respectively. The last

an

two series are most probably a result of fragmentation that occurred during the MALDI-TOF ionization process. The mentioned above series are present in the spectra of all of the investigated polymers. Other

M

signals are related to the molecules containing chain extender used for copolymerization. They are collected in Table 3

Chain

Ac ce pt e

Polymer

d

Table 3. Series of signals observed in the MALDI-TOF spectra of synthesized co-polyesters. extender

HBP-0

none

HBP-G

glycolic

HBP-L

lactic

HBP-V

pentanoic

HBP-C

HBP-D

hexanoic dodecanoic

Series

[C6H14O3 + (C5H8O3)n]Na+, [C6H14O3 + (C5H8O3)n]K+, [C6H12O2

+ (C5H8O3)n]Na+, [C5H10O + (C5H8O3)n]Na+

HBP-0 series and [C8H16O5 + (C5H8O3)n]Na+, [C8H14O4 + (C5H8O3)n]Na+, [C7H12O3 + (C5H8O3)n]Na+

HBP-0 series and [C9H18O5 + (C5H8O3)n]Na+, [C9H16O4 + (C5H8O3)n]Na+, [C8H14O3 + (C5H8O3)n]Na+

HBP-0 series and [C11H22O5 + (C5H8O3)n]Na+, [C11H20O4 + (C5H8O3)n]Na+, [C10H18O3 + (C5H8O3)n]Na+ HBP-0 series and [C12H24O5 + (C5H8O3)n]Na+, [C12H22O4 + (C5H8O3)n]Na+, [C11H20O3 + (C5H8O3)n]Na+

HBP-0 series and [C18H36O5 + (C5H8O3)n]Na+, [C18H34O4 + (C5H8O3)n]Na+, [C17H32O3 + (C5H8O3)n]Na+

14

Page 14 of 30

The signals collected in Table 3 refer to macromolecules containing incorporated one chain extender unit. In this case signals of molecules with the loss of one water molecule or water and formaldehyde molecules are present as well. Signals of the macromolecules containing two or more chain extender units are on the threshold level and were not listed. This is in agreement with our previous results.[23] It must be also stated, that the results obtained by MALDI-TOF refer only to a fraction of the polymer that was ionized during the MALDI-TOF experiment, not the whole polymer

ip t

sample.

cr

Modified copolyesters

The parent polymers HBP-0 – HBP-D were modified with TFA residues. The presence of the

us

TFA terminal groups in the polymer structure can be observed by FTIR spectroscopy. Figure 4 shows a spectrum of polymer HBP-Df (10). This picture is representative for all the synthesized materials since the chain extender in each sample is quite low. Introduction of fluorinated residues is represented by an

Ac ce pt e

d

M

an

appearance of a new trifluoroacetic ester absorption band at 1787 cm-1.

Figure 4. FTIR spectrum (ATR) of the HBP-Df (10) TFA modified polymer. The FTIR spectra taken directly after synthesis revealed no absorption bands at 3600 cm-1 related to unsubstituted hydroxyl groups of the copolyester. However, 1H NMR spectra showed approx. 10% unreacted hydroxymethyl groups (3.8-3.6 ppm). This suggest that the polymers show some propensity to hydrolysis.

15

Page 15 of 30

ip t cr

us

Figure 5. 1H NMR (400 MHz, CDCl3) spectrum of derivative HBP-Gf (10).

Figure 5 shows a sample proton NMR spectrum of a fluorinated polymer HBP-Gf (10). Substitution of hydroxyl groups led to appearance of new methylene group peaks in the range of 4.75 to

an

4.3 ppm. There are also mentioned above signals of methylene groups next to unreacted hydroxyl groups. The ratio of integrals of those peaks allowed to calculate the degree of substitution of hydroxyl

M

groups. (see Table 4) Determination of molecular weight

d

The knowledge of molecular weights of the obtained copolyesters and their derivatives were crucial

Ac ce pt e

for the interpretation of phase equilibria. For this reason a thorough characterization of the synthesized polymers was performed. Our previous studies showed that NMR spectroscopy is a good enough method for the characterization of the Boltorn® type hyperbranched polymers. For comparison, we have also performed SEC-RI measurements with polystyrene as a calibration standard. Despite the fact that as expected the molecular masses obtained were lower than those obtained from NMR spectra, the SEC-RI measurements made it possible to estimate the molecular mass polydispersities of the polymers. Based on NMR spectra of fluorinated polymers it was possible to calculate the molecular weight of the polymers, the degree of substitution of hydroxyl groups, as well as calculate back the MW of parent co-polyesters. Relative numbers of all kinds of repeating units were obtained directly or indirectly by comparing the integrals of all groups of signals to the integral of well separated signal of methyl group of TMP (-CH3 TMP , Figure 5). The results are given in Table 4. Table 4. Composition of copolymers substituted with TFA chain extender

c.e %

M n1

M n2

Mn3

% subs t.

Mn4

Mn

Mw

PDI

(GPC)

(GPC)

(GPC)

16

Page 16 of 30

0

3614

3428

6782

80

5840

-

-

-

HBP-Gf (5)

glycolic

4.4

3707

3101

6875

91

5544

1544

2181

1.41

HBP-Gf (10)

glycolic

9.9

3811

3571

6979

91

6295

1246

1711

1.37

HBP-Lf (5)

lactic

4.5

3729

3517

6897

89

6201

1204

1559

1.30

HBP-Lf (10)

lactic

8.3

3859

3489

7027

90

6118

1184

1528

1.30

HBP-Vf (5)

pentanoic

4.2

3774

3531

6942

92

6275

1478

2198

1.49

HBP-Vf (10)

pentanoic

8.5

3954

3749

92

-

-

hexanoic

4.4

3797

3658

81

6557 6151

-

HBP-Cf (5)

7122 6965

980

1590

1.62

HBP-Cf (10)

hexanoic

9.6

4001

3791

7169

78

6154

-

-

-

HBP-Df (5)

dodecanoic

4.4

3931

3658

7099

92

6405

1262

1797

1.42

HBP-Df (10)

dodecanoic

9.3

4287

3737

7455

90

6277

5605*

7700*

1.37

us

cr

ip t

none

HBP-0f

Mn1: theoretical value for OH terminated polymer; Mn2: NMR determined value for OH terminated polymer; Mn3: theoretical value for 100% substituted polymer, Mn4: NMR determined value for

an

substituted polymer. * the refractive indexes of the solution and the pure solvent showed positive and

M

negative values (see Ref. [23]). Phase equilibria

In this study the copolymers of bis-MPA and various ω-hydroxycarboxylic acids (or their

d

derivatives) with TMP as a central unit were investigated. All of the polymers contained TFA terminal

Ac ce pt e

groups. In contrary to the previous work[23] the amount of ω-hydroxycarboxylate units in the structure was kept constant (5 and 10%) and the variable was the chemical structure of the chain extender hydroxycarboxylate unit. The systematic changes of the interior structure of the investigated hyperbranched polyesters make it possible to assess their influence on the phase behavior of the polymers in supercritical carbon dioxide.

The phase behavior for a series of the systems composed of hyperbranched polyesters HBP-Gf (5)– HBP-Gf (10) and carbon dioxide was investigated in a temperature range from 300 to 355 K. For most of the systems the LCST phase behavior was observed in the temperature window in which measurements were performed. Phase diagrams for polymer + CO2 systems are described below in more details.

17

Page 17 of 30

ip t cr us an M

Figure 6. Phase behavior for the system HBP-Gf + CO2. (A) Pressure – temperature cloud point curves

d

at constant composition for the polymer HBP-Gf (5); (B) pressure – weight fraction phase diagram at

Ac ce pt e

constant temperature for the polymer HBP-Gf (5); (C) Pressure – temperature cloud point curves at constant composition for the polymer HBP-Gf (10); (D) pressure – weight fraction phase diagram at constant temperature for the polymer HBP-Gf (10). The results of the cloud pressure measurements for the system HBP-Gf (5) + CO2 and HBP-Gf (10) + CO2 are presented in Figure 6. The polymer HBP-Gf (5) was fairly well soluble in supercritical carbon dioxide. The cloud point pressures spanned from about 550 bar at 355 K to about 370 bar at 303 K. The maximum cloud point pressure was observed at about 0.05 weight fraction of the polymer. This is clearly seen in Figure 6B where miscibility gaps in the pressure – weight fraction coordinates at four temperatures are presented. These results also clearly showed that temperature had significant impact on the phase transition: increase of temperature by 50 K increased the cloud point pressure by about 200 bar. At higher polymer concentrations the cloud point pressure was less sensitive to the polymer concentration in the system. In Figure 6C and 6D results of the cloud pressure measurements for the system HBP-Gf (10) + CO2 are presented. The increase of the amount of chain extender units from 4.4 to 9.9 mol% (Table 2) for polymer HBP-Gf (10) resulted in a slight reduction of the homogenization pressure. The cloud point pressures spanned from about 550 bar at 355 K to about 350 bar at 303 K. The 18

Page 18 of 30

maximum on the cloud point curve was observed again at about 0.05 weight fraction of the polymer (Figure 6D). However, as can be seen the cloud point pressures were less sensitive to the polymer concentration than for the system HBP-Gf (5) + CO2. The temperature had significant impact on the phase transition: increase of temperature by 50 K increased the cloud point pressure by about 250 bar. The results of the cloud pressure measurements for the system HBP-Lf (5) + CO2 and HBP-Lf (10)

Ac ce pt e

d

M

an

us

cr

ip t

+ CO2 are presented in Figure 7.

Figure 7. Phase behavior for the system HBP-Lf + CO2. (A) Pressure – temperature cloud point curves at constant composition for the polymer HBP-Lf (5); (B) pressure – weight fraction phase diagram at constant temperature for the polymer HBP-Lf (5); (C) Pressure – temperature cloud point curves at constant composition for the polymer HBP-Lf (10); (D) pressure – weight fraction phase diagram at constant temperature for the polymer HBP-Lf (10).

The HBP-Lf copolymers containing lactate units dissolved in scCO2 at higher pressures than their glycolate analogues. The cloud point pressures varied from 650 bar at 355 K to approx. 350 bar at 303 K for the HBP-Lf (5) polymer and from 750 bar at 355 K to approx. 500 bar at 303 K for the HBPLf (10) polymer. The maxima on the cloud point curves for both systems were observed at about 0.06 19

Page 19 of 30

weight fraction of the polymer. This is clearly seen in Figure 7B and 7D where miscibility gaps in the pressure – weight fraction coordinates are presented. These results also showed that the influence of the temperature on the phase transition depended on the polymer concentration and the amount of the lactate chain extender units in the polymer. The HBP-Lf (5) polymer showed typical LCST phase behavior. Increasing temperature by 50 K raised the pressure required to generate single-phase solution by 10 to 100 bar depending on the polymer weight fraction. In case of HBP-Lf (10) polymer containing

ip t

twice as much lactate units both LCST and UCST were observed depending on the polymer weight fraction. From the above results it may be concluded that replacement of the glycolate chain extending

cr

units with lactate ones in the structure of the hyperbranched poly(bis-MPA) increases the pressures needed to dissolve the polymer in scCO2. Most probably the steric effect of the side methyl group in

us

case of hyperbranched structure is not as efficient as in case of linear polymers. Moreover, introduction of additional methyl group to the chain extender repeating unit does not result in spreading of the

an

polymer chains but in making the structure more dense and hydrophobic.

The results of the cloud pressure measurements for the system HBP-Vf (5) + CO2 are presented in

Ac ce pt e

d

M

Figure 8.

Figure 8. Phase behavior for the system HBP-Vf (5) + CO2. (A) Pressure – temperature cloud point curves at constant composition; (B) pressure – weight fraction phase diagram at constant temperature. The polymer HBP-Vf (5) containing 5-hydroxypentanoate chain extender units was more soluble in supercritical carbon dioxide than its analogue HBP-Gf (5). For the HBP-Vf (5) the cloud point pressures spanned from about 520 bar at 355 K to about 320 bar at 303 K. The maximum cloud point pressure was observed for the system HBP-Vf (5) + CO2 at about 0.02 weight fraction of the polymer. This is clearly seen in Figure 8B where miscibility gaps in the pressure – weight fraction coordinates at four temperatures are presented. The results also clearly showed that temperature had significant impact 20

Page 20 of 30

on the position of the boundary of the phase transition: increase of temperature by 50 K increased the cloud point pressure by about 200 bar. The phase behavior of the polymers HBP-Cf (5) i HBP-Cf (10) was discussed before.[23] Just to compare, the most soluble polymer HBP-Cf (5) dissolved in scCO2 at lowest pressures, spanned from about 400 bar at 355 K to about 200 bar at 303 K. The maximum on the cloud point curve was observed at about 0.05 weight fraction of the polymer. The increase of temperature by 50 K increased the cloud

ip t

point pressure by about 200 bar.

The results of the cloud pressure measurements for the system HBP-Df (5) + CO2 and HBP-Df

Ac ce pt e

d

M

an

us

cr

(10) + CO2 are presented in Figure 9.

Figure 9. Phase behavior for the system HBP-Df + CO2. (A) Pressure – temperature cloud point curves at constant composition for the polymer HBP-Df (5); (B) pressure – weight fraction phase diagram at constant temperature for the polymer HBP-Df (5); (C) Pressure – temperature cloud point curves at constant composition for the polymer HBP-Df (10); (D) pressure – weight fraction phase diagram at constant temperature for the polymer HBP-Df (10).

21

Page 21 of 30

These polymers showed the lowest solubility in scCO2 among investigated materials. The polymer HBP-Df(5) was reasonably soluble in supercritical carbon dioxide at low concentration (w = 0.01) only. The cloud point pressures for this polymer spanned from about 850 bar at 340 K to about 330 bar at 303 K. The maximum on the cloud point curve was observed at about 0.06 weight fraction of the polymer. However, rising the amount of the polymer up to weight fraction 0.18 caused increase of the pressure needed to dissolve it up to about 800 bar. This is clearly seen in Figure 9B where miscibility gaps in the

ip t

pressure – weight fraction coordinates at three temperatures are presented. For the polymer HBP-Df (5) temperature had medium impact on the phase transition: increase of temperature by 50 K increased the

cr

cloud point pressure by about 50 to 100 bar. At the highest polymer concentration increase of temperature by 50 K increased the cloud point pressure by about 250 bar. In Figure 9C and 9D results

us

of the cloud pressure measurements for the system HBP-Df (10) + CO2 are presented. The increase of the amount of chain extender units from 4.4 to 9.3 mol% (Table 2) resulted in a significant increase of the homogenization pressure. The cloud point pressures in case of the polymer HBP-Df (10) spanned

an

from about 1000 bar at 355 K to about 600 bar at 303 K. The results also showed that temperature had limited impact on the position of the boundary of the phase transition: increase of temperature by 50 K

M

changed the cloud point pressure only by up to 50 bar. However, the influence of the polymer weight fraction had significant impact on the solubility of the polymer. The maximum on the cloud point curve was observed at about 0.06 weight fraction of the polymer and the increase of the weight fraction from

Ac ce pt e

d

0.01 to 0.06 resulted in the increase of the cloud point pressure by approx. 350 bar.

22

Page 22 of 30

Figure 10. Comparison of the cloud point pressure curves for solutions of polymers HBP-Gf – HBP-Df in carbon dioxide (w≈0.05); chain extenders: G – glycolic, L – lactic, V – hydroxypentanoic, Chydroxyhexanoic, D - hydroxydodecanoic.

Figure 10 presents a comparison of the cloud point pressure curves for solutions of polymers HBPGf – HBP-Df in carbon dioxide. It is clear that hyperbranched polyesters containing 5mol% of chain

ip t

extending units showed better solubility in scCO2 than their analogues containing 10mol% of chain extending units. This behavior was not dependent on the type of chain extender. Increase of its amount

cr

by 5% caused increase of the cloud point pressure by approx. 100 bar. The change of the pressure was

Ac ce pt e

d

M

an

us

lower only for the glycolic chain extender.

Figure 11. The dependence of the cloud point pressure of the polymer solution in scCO2 (5% of chain extender units, T=320K, w≈0.05) on the number of the carbon atoms in a chain extender. The comparison of the influence of the size of the chain extender on the solubility of the copolyesters is given in Figure 11. The lowest phase transition pressures were observed for a moderate size of the chain extending units. Polymers HBP-Vf (5) and HBP-Cf (5) containing 5 or 6 carbon atoms showed better solubility in scCO2 than polymer HBP-Gf(5) consisting of two carbon atoms. However, further increasing of the chain extender unit size to 12 carbon atoms yielded an increase of the phase transition pressure (polymer HBP-Df (5)). As it was mentioned before, application of lactic acid derivatives (HBP-Lf(5)) instead of glycolic acid (HBP-Gf (5)) ones also caused increase of the phase transition pressure. 23

Page 23 of 30

It can be concluded that to a certain point increasing the size of the chain extending units makes the hyperbranched structure better accessible for carbon dioxide. However introduction of methyl side groups or extending the length of the hydrocarbon chains leads to increased hydrophobicity of the material and decrease of the solubility in supercritical carbon dioxide. Summary and conclusion

ip t

In this work we described the synthesis of hyperbranched copolyesters of 2,2bis(hydroxymethyl)propionic acid containing small amounts (5 or 10%) of chain extending units of various lengths and their phase behavior in supercritical carbon dioxide (scCO2) after modification with

cr

trifluoroacetic anhydride. The structure of the copolyesters was confirmed with 1H and 13C NMR, FTIR spectroscopies and MALDI-TOF mass spectrometry. The phase behavior of the polymers in

us

supercritical carbon dioxide was explored as a function of concentration and temperature. It was shown that polymers containing the chain extending units of the moderate length (5-hydroxypentanoate or 6-

Corresponding Author

Notes

Ac ce pt e

E-mail: [email protected].

d

AUTHOR INFORMATION

M

transition parameters in supercritical carbon dioxide.

an

hydroxyhexanoate) in the amount of 5 molar per-cent in respect to bis-MPA exhibited the lowest phase

The authors declare no competing financial interest. ACKNOWLEDGMENTS

Mr. Zbigniew Fraś is kindly acknowledged for performing of the sc(CO2) solubility experiments. This paper is based upon work supported by the Polish National Science Centre research grant (N N209 028440) and the Polish Foundation of Science International PhD program (MPD/2010/4). REFERENCES

[1] S. Suttiruengwong, J. Rolker, I. Smirnova, W. Arlt, M. Seiler, L. Luderitz, Y.P. de Diego, P.J. Jansens, Hyperbranched polymers as drug carriers: Microencapsulation and release kinetics, Pharmaceutical Development and Technology, 11 (2006) 55-70. [2] D.A. Canelas, J.M. DeSimone, Polymerizations in liquid and supercritical carbon dioxide, in: Metal Complex Catalysts Supercritical Fluid Polymerization Supramolecular Architecture, 1997, pp. 103-140. [3] J.L. Kendall, D.A. Canelas, J.L. Young, J.M. DeSimone, Polymerizations in supercritical carbon dioxide, Chemical Reviews, 99 (1999) 543-563. [4] R.B. Yoganathan, R. Mammucari, N.R. Foster, Dense Gas Processing of Polymers, Polymer Reviews, 50 (2010) 144-177. 24

Page 24 of 30

[5] Y. Wu, Hyperbranched Polymers in a Supercritical Fluid: Recent Progress

Ac ce pt e

d

M

an

us

cr

ip t

on Phase Behavior and Modeling, Journal of Applied Solution Chemistry and Modeling, 2 (2013) 3346. [6] K. Langenbach, S. Enders, C. Browarzik, D. Browarzik, Calculation of the high pressure phase equilibrium in hyperbranched polymer systems with the lattice-cluster theory, Journal of Chemical Thermodynamics, 59 (2013) 107-113. [7] N.A. Sheremetyeva, N.V. Voronina, A.V. Bystrova, V.D. Miakushev, M.I. Buzin, A.M. Muzafarov, Fluorine-Containing Organosilicon Polymers of Different Architectures. Synthesis and Properties Study, in: S.J. Clarson, M.J. Owen, S.D. Smith, M.E. VanDyke (Eds.) Advances in Silicones and Silicone-Modified Materials, 2010, pp. 111-134. [8] M.K. Kozlowska, B.F. Jurgens, C.S. Schacht, J. Gross, T.W. de Loos, Phase Behavior of Hyperbranched Polymer Systems: Experiments and Application of the Perturbed-Chain Polar SAFT Equation of State, Journal of Physical Chemistry B, 113 (2009) 1022-1029. [9] V. Martinez, S. Mecking, T. Tassaing, M. Besnard, S. Moisan, F. Cansell, C. Aymonier, Dendritic core-shell macromolecules soluble in supercritical carbon dioxide, Macromolecules, 39 (2006) 39783979. [10] A. Garcia-Bernabe, M. Kramer, B. Olah, R. Haag, Syntheses and phase-transfer properties of dendritic nanocarriers that contain perfluorinated shell structures, Chemistry-a European Journal, 10 (2004) 2822-2830. [11] J. Gregorowicz, Z. Fras, P. Parzuchowski, G. Rokicki, M. Kusznerczuk, S. Dziewulski, Phase behaviour of hyperbranched polyesters and polyethers with modified terminal OH groups in supercritical solvents, Journal of Supercritical Fluids, 55 (2010) 786-796. [12] B. Tan, C.L. Bray, A.I. Cooper, Fractionation of Poly(vinyl acetate) and the Phase Behavior of End-Group Modified Oligo(vinyl acetate)s in CO2, Macromolecules, 42 (2009) 7945-7952. [13] H. Lee, J.W. Pack, W.X. Wang, K.J. Thurecht, S.M. Howdle, Synthesis and Phase Behavior of CO2-Soluble Hydrocarbon Copolymer: Poly(vinyl acetate-alt-dibutyl maleate), Macromolecules, 43 (2010) 2276-2282. [14] H. Lee, E. Terry, M. Zong, N. Arrowsmith, S. Perrier, K.J. Thurecht, S.M. Howdle, Successful dispersion polymerization in supercritical CO2 using polyvinylalkylate hydrocarbon surfactants synthesized and anchored via RAFT, Journal of the American Chemical Society, 130 (2008) 1224212243. [15] E. Girard, T. Tassaing, J.D. Marty, M. Destarac, Influence of macromolecular characteristics of RAFT/MADIX poly(vinyl acetate)-based (co)polymers on their solubility in supercritical carbon dioxide, Polymer Chemistry, 2 (2011) 2222-2230. [16] E. Girard, T. Tassaing, S. Camy, J.S. Condoret, J.D. Marty, M. Destarac, Enhancement of Poly(vinyl ester) Solubility in Supercritical CO2 by Partial Fluorination: The Key Role of PolymerPolymer Interactions, Journal of the American Chemical Society, 134 (2012) 11920-11923. [17] C. Gao, D. Yan, Hyperbranched polymers: from synthesis to applications, Progress in Polymer Science, 29 (2004) 183-275. [18] B. Voit, Hyperbranched polymers - All problems solved after 15 years of research?, Journal of Polymer Science Part a-Polymer Chemistry, 43 (2005) 2679-2699. [19] M.G. McKee, S. Unal, G.L. Wilkes, T.E. Long, Branched polyesters: recent advances in synthesis and performance, Progress in Polymer Science, 30 (2005) 507-539. [20] C.R. Yates, W. Hayes, Synthesis and applications of hyperbranched polymers, European Polymer Journal, 40 (2004) 1257-1281. [21] M. Jikei, M. Kakimoto, Hyperbranched polymers: a promising new class of materials, Progress in Polymer Science, 26 (2001) 1233-1285. [22] M. Tryznowski, K. Tomczyk, Z. Fras, J. Gregorowicz, G. Rokicki, E. Wawrzynska, P.G. Parzuchowski, Aliphatic Hyperbranched Polycarbonates: Synthesis, Characterization, and Solubility in Supercritical Carbon Dioxide, Macromolecules, 45 (2012) 6819-6829. [23] J. Gregorowicz, E.P. Wawrzy ska, P.G. Parzuchowski, Z. Fra , G. Rokicki, K. Wojciechowski, S.A. Wieczorek, A. Wi niewska, A. Plichta, K. D browski, M. Tryznowski, Synthesis, Characterization, 25

Page 25 of 30

Ac ce pt e

d

M

an

us

cr

ip t

and Solubility in Supercritical Carbon Dioxide of Hyperbranched Copolyesters, Macromolecules, 46 (2013) 7180–7195. [24] A. Sunder, R. Hanselmann, H. Frey, R. Mulhaupt, Controlled synthesis of hyperbranched polyglycerols by ring-opening multibranching polymerization, Macromolecules, 32 (1999) 4240-4246. [25] E. Malmstrom, M. Johansson, A. Hult, Hyperbranched aliphatic polyesters, Macromolecules, 28 (1995) 1698-1703. [26] H. Magnusson, E. Malmstrom, A. Hult, Structure buildup in hyperbranched polymers from 2,2bis(hydroxymethyl)propionic acid, Macromolecules, 33 (2000) 3099-3104. [27] E. Zagar, M. Zigon, S. Podzimek, Characterization of commercial aliphatic hyperbranched polyesters, Polymer, 47 (2006) 166-175. [28] E. Zagar, M. Zigon, Aliphatic hyperbranched polyesters based on 2,2-bis(methylol)propionic acidDetermination of structure, solution and bulk properties, Progress in Polymer Science, 36 (2011) 53-88. [29] E. Malmstrom, A. Hult, Hyperbranched polymers: A review, Journal of Macromolecular ScienceReviews in Macromolecular Chemistry and Physics, C37 (1997) 555-579. [30] A. Hult, M. Johansson, E. Malmstrom, Hyperbranched polymers, Branched Polymers Ii, 143 (1999) 1-34. [31] R. Haag, J.F. Stumbe, A. Sunder, H. Frey, A. Hebel, An approach to core-shell-type architectures in hyperbranched polyglycerols by selective chemical differentiation, Macromolecules, 33 (2000) 81588166. [32] M. Trollsas, B. Atthoff, H. Claesson, J.L. Hedrick, Hyperbranched poly(epsilon-caprolactone)s, Macromolecules, 31 (1998) 3439-3445. [33] J. Choi, S.Y. Kwak, Synthesis and characterization of hyperbranched poly(epsilon-caprolactone)s having different lengths of homologous backbone segments, Macromolecules, 36 (2003) 8630-8637. [34] S.Y. Kwak, J. Choi, H.J. Song, Viscoelastic relaxation and molecular mobility of hyperbranched Poly(epsilon-caprolactone)s in their melt state, Chemistry of Materials, 17 (2005) 1148-1156. [35] A. Lopez-Luna, J.L. Gallegos, M. Gimeno, E. Vivaldo-Lima, E. Barzana, Lipase-catalyzed syntheses of linear and hyperbranched polyesters using compressed fluids as solvent media, Journal of Molecular Catalysis B-Enzymatic, 67 (2010) 143-149. [36] H. Galina, J.B. Lechowicz, M. Walczak, Methods of narrowing the molecular size distribution in hyperbranched polymerization involving AB(2) and B-2 monomers, Journal of Macromolecular Science-Physics, B44 (2005) 925-940. [37] H. Galina, M. Walczak, A theoretical model of hyperbranched polymerization involving an AB(f) monomer - Part II. The average polymerization degree and dispersity index, Polimery, 50 (2005) 713717.

Table 1. Amounts of reagents used for synthesis of HBP copolyesters Sample

bis-MPA

TMP

g (mmol)

g (mmol)

Chain extender precursor g (mmol)

Yield g

33.53 (250)

1.12 (8.33)

[none]

[23]

0.00 (0.00)

29.2

33.53 (250)

1.12 (8.33)

glycolide

0.78 (6.72)

29.1

HBP-G (10)

33.53 (250)

1.12 (8.33)

glycolide

1.64 (14.1)

30.3

HBP-L (5)

33.53 (250)

1.12 (8.33)

L-lactide

0.88 (6.10)

30.3

HBP-0 HBP-G (5)

*

26

Page 26 of 30

33.53 (250)

1.12 (8.33)

L-lactide

1.83 (12.7)

31.0

HBP-V (5)

33.53 (250)

1.12 (8.33)

δ-valerolactone

1.30 (13.0)

30.1

HBP-V (10)

33.53 (250)

1.12 (8.33)

δ-valerolactone

2.83 (28.3)

32.6

HBP-C (5)

33.53 (250)

1.12 (8.33)

ε-caprolactone

[23]

1.52 (13.3)

31.5

HBP-C (10)

33.53 (250)

1.12 (8.33)

ε-caprolactone

[23]

3.23 (28.3)

32.4

HBP-D (5)

16.77 (125)

0.56 (4.16)

12-hydroxydodecanoic acid

1.45 (6.7)

15.4

HBP-D (10)

16.77 (125)

0.56 (4.16)

12-hydroxydodecanoic acid

3.09 (14.3)

16.7

theoretical molar pre-cent of chain extending units

Ac ce pt e

d

M

an

us

cr

*

ip t

HBP-L (10)

Table 2. The amounts of reagents used for esterification of HBP with TFA anhydride Polymer

TFA anhydride

Yield

Mn*

g

mmol**

g

% ***

g

HBP-0f [23]

3610

5

68.5

14.4

92

7.8

HBP-Gf (5)

3710

5

66.7

14.0

92

8.0

HBP-Gf (10)

3820

5

64.8

13.6

91

7.8

HBP-Lf (5)

3720

5

66.5

14.0

91

7.9

HBP-Lf (10)

3840

5

64.5

13.5

93

7.9

27

Page 27 of 30

3780

5

65.5

13.8

92

8.0

3960

5

62.5

13.1

92

7.8

3800

5

65.2

13.7

90

7.7

4000

5

61.8

13.0

94

7.6

HBP-Df (5)

3940

5

62.8

13.2

90

7.7

HBP-Df (10)

4290

5

57.7

12.1

90

7.5

HBP-Vf (10) HBP-Cf (5)

[23]

HBP-Cf (10)

*

[23]

theoretical values for polymers: HBP-0 – HBP-D;

**

for theoretical number of OH groups;

***

ip t

HBP-Vf (5)

Ac ce pt e

d

M

an

us

cr

calculated for actual degree of substitution (approx.80%).

Table 3. Series of signals observed in the MALDI-TOF spectra of synthesized co-polyesters. Polymer

HBP-0

HBP-G

Chain

extender none

glycolic

HBP-L

lactic

HBP-V

pentanoic

Series

[C6H14O3 + (C5H8O3)n]Na+, [C6H14O3 + (C5H8O3)n]K+, [C6H12O2

+ (C5H8O3)n]Na+, [C5H10O + (C5H8O3)n]Na+ HBP-0 series and [C8H16O5 + (C5H8O3)n]Na+, [C8H14O4 + (C5H8O3)n]Na+, [C7H12O3 + (C5H8O3)n]Na+ HBP-0 series and [C9H18O5 + (C5H8O3)n]Na+, [C9H16O4 + (C5H8O3)n]Na+, [C8H14O3 + (C5H8O3)n]Na+ HBP-0 series and [C11H22O5 + (C5H8O3)n]Na+, [C11H20O4 + 28

Page 28 of 30

(C5H8O3)n]Na+, [C10H18O3 + (C5H8O3)n]Na+ HBP-C

hexanoic

(C5H8O3)n]Na+, [C11H20O3 + (C5H8O3)n]Na+

dodecanoic

HBP-0 series and [C18H36O5 + (C5H8O3)n]Na+, [C18H34O4 + (C5H8O3)n]Na+, [C17H32O3 + (C5H8O3)n]Na+

Ac ce pt e

d

M

an

us

cr

ip t

HBP-D

HBP-0 series and [C12H24O5 + (C5H8O3)n]Na+, [C12H22O4 +

Table 4. Composition of copolymers substituted with TFA

6782

% subs t. 80

3101

6875

3811

3571

4.5

3729

lactic

8.3

HBP-Vf (5)

pentanoic

HBP-Vf (10) HBP-Cf (5)

Mn

Mw

PDI

(GPC)

(GPC)

(GPC)

5840

-

-

-

91

5544

1544

2181

1.41

6979

91

6295

1246

1711

1.37

3517

6897

89

6201

1204

1559

1.30

3859

3489

7027

90

6118

1184

1528

1.30

4.2

3774

3531

6942

92

6275

1478

2198

1.49

pentanoic

8.5

3954

3749

92

-

-

-

hexanoic

4.4

3797

3658

7122 6965

980

1590

1.62

c.e %

M n1

none

0

3614

3428

HBP-Gf (5)

glycolic

4.4

3707

HBP-Gf (10)

glycolic

9.9

HBP-Lf (5)

lactic

HBP-Lf (10)

chain extender HBP-0f

M n2

Mn3

81

Mn4

6557 6151

29

Page 29 of 30

HBP-Cf (10)

hexanoic

9.6

4001

3791

7169

78

6154

-

-

-

HBP-Df (5)

dodecanoic

4.4

3931

3658

7099

92

6405

1262

1797

1.42

HBP-Df (10)

dodecanoic

9.3

4287

3737

7455

90

6277

5605*

7700*

1.37

Mn1: theoretical value for OH terminated polymer; Mn2: NMR determined value for OH terminated polymer; Mn3: theoretical value for 100% substituted polymer, Mn4: NMR determined value for substituted polymer. * the refractive indexes of the solution and the pure solvent showed positive and

us

cr

ip t

negative values (see Ref. [23]).

an

Highlights

A simple method for preparation of hyperbranched CO2-philic co-polyesters was developed.

•

The co-polyesters showed very good solubility in supercritical carbon dioxide,

•

We showed, that polymers containing 5% of chain extending monomer of the moderate length

d

M

•

Ac ce pt e

(5-hydroxypentanoate or 6-hydroxyheksanoate) exhibited the lowest phase transition parameters in supercritical carbon dioxide.

30

Page 30 of 30