Biomaterials 25 (2004) 1019–1028

ARTICLE IN PRESS

*Correspondin

E-mail addres

0142-9612/$ - see

doi:10.1016/S014

Nanoscopic behavior of polyvinylpyrrolidone particles onpolysulfone/polyvinylpyrrolidone film

Masayo Hayamaa, Ken-ichiro Yamamotoa, Fukashi Kohoria, Tsutomu Uesakab,Yoshiyuki Uenob, Hiroyuki Sugayab, Ichiro Itagakib, Kiyotaka Sakaia,*

aDepartment of Chemical Engineering, Waseda University, 3-4-1 Okubo, Shinjuku-ku, Tokyo 169-8555, JapanbToray Industries, Inc., Shiga Plant 2-1, Sonoyama 3-chome, Otsu, Siga 520-0842, Japan

Received 2 July 2003; accepted 29 July 2003

Abstract

We revealed morphology and physicochemical behavior of a widely used powerful hydrophilizing agent, polyvinylpyrrolidone

(PVP), present on polysulfone (PS)/PVP films by atomic force microscopy (AFM). This is the first time such clear PS/PVP phase-

separated morphology was observed by nanoscopic technique. The film surfaces were observed by the identical observation mode,

probe and scanning conditions to reveal the change of PVP morphology and behavior between dry and wet conditions. Morphology

was related to biocompatibility by combining AFM data with results of surface element composition, contact angle, adhesion

amount of rabbit platelet and relative amount of adsorbed fibrinogen. PVP nano-particles of one or several molecules were formed

on the dry PS/PVP film surfaces. Amount of PVP present on the surfaces increased with the molecular weight of PVP. At a mixed

amount of 1–5 wt%, PVP K90 formed crowded particles on the dry surface. When wet, they swelled, followed by their union to

produce a smooth surface leading to improved biocompatibility. The highest biocompatibility with excellent mechanical strength is

achieved by blending the highest molecular weight PVP K90 at 1–5 wt%.

r 2003 Elsevier Ltd. All rights reserved.

Keywords: PVP nanoparticles; Polysulfone/PVP films; Atomic force microscopy; Biocompatibility; Surface analysis; Surface roughness

1. Introduction

Polyvinylpyrrolidone (PVP) is widely used for hydro-philizing biomaterials of artificial organs [1] andindustrial membrane materials [2], and also for phar-maceutical purposes, e.g. a filling agent in tablets.Physicochemical properties of PVP [3–5] and PVP-blended materials [6–8] have been studied in detail bysuch techniques as X-ray photoelectron spectroscopy(XPS), differential scanning calorimetry (DSC), Fouriertransform infrared spectroscopy (FTIR), dynamic lightscattering (DLS), electron spin resonance imaging(ESRI).

In the field of hemodialysis membranes, PVP is usedfor hydrophilizing comparatively hydrophobic materialssuch as polysulfone (PS), poly(acrylonitrile) (PAN) andpolyester polymer alloy (PEPA). The mechanism howPVP improves biocompatibility of these materials and

g author.+81-3-5286-3216; fax: +81-3-3209-7957.

s: kisakai@waseda.jp (K. Sakai).

front matter r 2003 Elsevier Ltd. All rights reserved.

2-9612(03)00629-X

why the same material containing PVP often showsdifferent biocompatibility by the polymer blendingprocess and by the sterilization process have not beenclarified yet.

The present study is focused on PVP-blended PSbiomaterials for developing a patient-friendly dialysismembrane with excellent biocompatibility, and to aim atquantitatively evaluating PVP morphology and itsbehavior in wet and dry conditions by atomic forcemicroscopy (AFM) and also at revealing the mechanismhow PVP contributes to excellent biocompatibility ofmembrane materials. However, it is difficult to observePVP morphology directly within membranes havingcomplicated structures composed of tiny pores. Hence,films without any pores were prepared from a mixture ofPS and PVP. They were evaluated by AFM to reveal themechanism described above.

PS material hardly swells when wet. Its waterabsorption rate has been reported to be as low as0.30 wt% per 24 h based on the method of ASTM D570[9]. The interaction between PVP and water will cause

ARTICLE IN PRESS

Table 2

AFM observation conditions

Equipment SPM-9500J3 (Shimadzu, Kyoto, Japan)

Observation mode Contact mode

Probe NP-S (120 mm, wide) (Nihon Veeco KK,

Tokyo, Japan)

Scanner Standard (30mm� 30 mm� 5mm)

Scanning area 2 mm� 2mm (512 pixels� 512 pixels)

Scanning speed 0.2 Hz

Operating point 0.3 V

M. Hayama et al. / Biomaterials 25 (2004) 1019–10281020

formation of some varied structures of PS blended withPVP.

Structure parameters are obtained by AFM quanti-tative analysis of the observed surface images [10–14].Surface roughness, surface morphology, and polymerparticle structure are determined as root mean square(RMS) roughness and particle diameter. Amount ofPVP present on the film surfaces is quantitativelymeasurable by XPS [14]. Hydrophilicity is determinedby contact angle measurements. Biocompatibility wasevaluated by measuring such parameters as adhesionamount of rabbit platelet and relative amount ofadsorbed fibrinogen [14–16].

2. Experimental

2.1. Materials

All of the PS (Udel-P3500; SOLVAY S.A., Brussels,Belgium)/PVP (ISP Technologies Inc., NJ, USA) filmsevaluated in the present study (Table 1) were preparedfrom a mixture of PS and PVP using a solvent of N;N 0-dimethylacetamide (DMAc). PS/PVP mixing ratio wascontrolled by varying PVP concentration maintainingPS concentration to be constant, 10 wt%. The mixturewas cast to form a film of a thickness of 203 mm on aglass plate thermostated at 373 K with a hot plate. Aftercasting, the film was dried on a hot plate for 5min sothat solvent was evaporated, and then put in a waterbath. To eliminate residual DMAc, the films wereautoclaved at 383 K for 30 min three times.

2.2. Morphology by AFM [17]

Three-dimensional structures of the film surfaces wereobserved with SPM-9500J3 (Shimadzu, Kyoto, Japan).

Table 1

Technical data on the PS/PVP films tested

Sample PS/PVP mixing ratio (%) PVP grade Molecular weight

PS PVP

A 100 0 K90 1200,000

B 99.9 0.1

C 99.5 0.5

D 99 1

E 95 5

F 67 33

G 50 50

H 33 67

I 95 5 K30 50,000

J 33 67

K 95 5 K15 10,000

L 67 33

M 33 67

N 95 2.5/2.5 K90/K15 1200,000/10,000

Films were fixed on a steel disk with a transparentmanicure for dry samples, which were then placed inreverse osmosis (RO) water (conductivity o2 ms/cm) for24 h for preparing wet samples. AFM observationconditions are summarized in Table 2. To quantitativelyevaluate a change in surface structure from dry to wetconditions, the identical observation mode (contactmode) and probe (NP-S) were used in both air andwater. This is the key to succeeding in observing PVPbehavior. Scanning area was set as widely as possible toobserve nano-scale structure and also not to detectpeculiar morphology. Furthermore, scanning speed andoperating point were controlled as low as possible not todamage film surface that is remarkably softer than metaland inorganic material surfaces. RMS was measuredusing an image analysis software (n ¼ 3). PVP particlediameter was measured on a cross-sectional image(n ¼ 30).

2.3. Amount of PVP

Amount of PVP present on the surface of the filmswas measured by XPS (ESCALAB220iXL; Thermo VGScientific, East Grinstead, UK) equipped with a mono-chromatized Al Ka source (1486.6 eV) of 1 mm diameteroperating at a voltage of 10.0 kV, a current of 20 mAand a photoelectron escape angle of 90�. Nitrogen (N) isan element present only in PVP, and sulfur (S) is thatpresent only in PS. Comparing the ratio of N to S,amount of PVP present on the surface is determinable.

2.4. Hydrophilicity

Contact angle of the film was measured with a contactangle meter CA-D (Kyowa surface Science Co., Ltd.,Saitama, Japan) at 298 K to evaluate hydrophilicity ofthe films.

2.5. Biocompatibility

Adhesion amount of rabbit platelet and relativeamount of adsorbed fibrinogen, assuming that amountof adsorbed fibrinogen of 100% PS film was 100, weremeasured to evaluate biocompatibility of the films.

ARTICLE IN PRESS

wet

wet

(A)

PVP K90, 0 %

(B)

PVP K90, 0.1 %

dry

dry

10.68[m]

0.00

0.00

2.00 × 2.00 [µm] Z-Max 10.68[nm]

2.00 × 2.00 [µm] Z-Max 9.50[nm]

0.50

0.50

1.001.00

1.50

1.50

9.50[m]

0.00

0.00

0.50

1.00 1.00

1.50

1.50

10.06[m]

0.00

0.00

2.00 × 2.00 [µm] Z-Max 10.06[nm] 2.00 × 2.00 [µm] Z-Max 8.31[nm]

0.50

0.50

1.001.00

1.50

1.50

8.31 [m]

0.00

0.00

0.50

0.50

1.001.00

1.50

1.50

0.50

wet

wet

PVP K90, 0.5 %

PVP K90, 1 %

dry

dry

(C)

(D)

9.46[m]

0.00

0.00

2.00 × 2.00 [µm] Z-Max 9.46[nm]

0.50

0.50

1.00 1.00

1.50

1.50

10.15 [m]

0.00

0.00

2.00 × 2.00 [µm] Z-Max 10.15[nm]

0.50

0.50

1.00 1.00

1.50

1.50

9.87 [m]

0.00

0.00

2.00 × 2.00 [µm] Z-Max 9.87[nm]

0.50

0.50

1.00 1.00

1.50

1.50

10.58 [m]

0.00

0.00

2.00 × 2.00 [µm] Z-Max 10.58[nm]

0.50

0.50

1.00 1.00

1.50

1.50

Fig. 1. AFM surface images of dry and wet PS/PVP films (K90, K30, K15, and K90/K15 mixed) (scale: 2mm� 2 mm�E10 nm).

M. Hayama et al. / Biomaterials 25 (2004) 1019–1028 1021

ARTICLE IN PRESS

wet

wet

PVP K90, 5 %

PVP K90, 33 %

dry

dry

(E)

(F)

9.98[m]

0.00

0.00

2.00 × 2.00 [µm] Z-Max 9.98[nm]

0.50

0.50

1.00 1.00

1.50

1.50

10.00[m]

0.00

0.00

2.00 × 2.00 [µm] Z-Max 10.00[nm]

0.50

0.50

1.00 1.00

1.50

1.50

9.63 [m]

0.00

0.00

2.00 × 2.00 [µm] Z-Max 9.63[nm]

0.50

0.50

1.00 1.00

1.50

1.50

10.06 [m]

0.00

0.00

2.00 × 2.00 [µm] Z-Max 10.06[nm]

0.50

0.50

1.00 1.00

1.50

1.50

wet

wet

PVP K90, 50 %

unobservable

PVP K90, 67 %

dry

dry

(G)

(H)

10.54 [m]

0.00

0.00

2.00 × 2.00 [µm] Z-Max 10.54[nm]

0.50

0.50

1.00 1.00

1.50

1.50

10.13[m]

0.00

0.00

2.00 × 2.00 [µm] Z-Max 10.13[nm]

0.50

0.50

1.001.00

1.50

1.50

10.04 [m]

0.00

0.00

2.00 × 2.00 [µm] Z-Max 10.04[nm]

0.50

0.50

1.00 1.00

1.50

1.50

Fig. 1 (continued).

M. Hayama et al. / Biomaterials 25 (2004) 1019–10281022

ARTICLE IN PRESS

wet

wet

PVP K30, 5 %

PVP K30, 67 %

dry

dry

(I)

(J)

10.09[m]

0.00

0.00

2.00 × 2.00 [µm] Z-Max 10.09[nm]

0.50

0.50

1.00 1.00

1.50

1.50

10.02[m]

0.00

0.00

2.00 × 2.00 [µm] Z-Max 10.02[nm]

0.50

0.50

1.00 1.00

1.50

1.50

10.05 [m]

0.00

0.00

2.00 × 2.00 [µm] Z-Max 10.05[nm]

0.50

0.50

1.001.00

1.50

1.50

10.05 [m]

0.00

0.00

2.00 × 2.00 [µm] Z-Max 10.05[nm]

0.50

0.50

1.00 1.00

1.50

1.50

wet

wet

PVP K15, 5 %

PVP K15, 33 %

dry

dry

(K)

(L)

9.87[m]

0.00

0.00

2.00 × 2.00 [µm] Z-Max 9.87[nm]

0.50

0.50

1.00 1.00

1.50

1.50

10.00[m]

0.00

0.00

2.00 × 2.00 [µm] Z-Max 10.00[nm]

0.50

0.50

1.00 1.00

1.50

1.50

10.02[m]

0.00

0.00

2.00 × 2.00 [µm] Z-Max 10.02[nm]

0.50

0.50

1.001.00

1.50

1.50

9.77 [m]

0.00

0.00

2.00 × 2.00 [µm] Z-Max 9.77[nm]

0.50

0.50

1.00 1.00

1.50

1.50

Fig. 1 (continued).

M. Hayama et al. / Biomaterials 25 (2004) 1019–1028 1023

ARTICLE IN PRESS

wet

wet

PVP K15, 67 %

PVP K90/15,2.5 %/2.5 %

dry

dry

(M)

(N)

10.35[m]

0.00

0.00

2.00 × 2.00 [µm] Z-Max 10.35[nm]

0.50

0.50

1.00 1.00

1.50

1.50

10.11 [m]

0.00

0.00

2.00 × 2.00 [µm] Z-Max 10.11[nm]

0.50

0.50

1.001.00

1.50

1.50

9.28 [m]

0.00

0.00

2.00 × 2.00 [µm] Z-Max 9.28[nm]

0.50

0.50

1.001.00

1.50

1.50

10.02 [m]

0.00

0.00

2.00 × 2.00 [µm] Z-Max 10.02[nm]

0.50

0.50

1.001.00

1.50

1.50

Fig. 1 (continued).

M. Hayama et al. / Biomaterials 25 (2004) 1019–10281024

2.5.1. Platelet adhesion

Film was set at the bottom of an 18 mm+polystyrene cylindrical tube, which was filled with saline.Mixture of aqueous 3.2% citric acid three-sodiumdihydrate and fresh rabbit blood at a volume ratio of1:9 was centrifuged at 1000 rpm for 10 min, and itssupernatant was collected (Plasma 1). The residual wascentrifuged at 3000 rpm for 10 min, and its supernatantwas also collected (Plasma 2). Plasma 1 was diluted withPlasma 2 to obtain platelet-rich plasma (PRP) of20� 106 number/ml. After removing saline from thetube, 1.0 ml of PRP was added and shaken at 310 Kfor 1 h, followed by rinsing with saline three times.Blood component was fixed with 3% glutaraldehyde.The film was rinsed with distilled water and dried at areduced pressure below 0.1 mmHg over 5 h, and subse-quently coated with Pt–Pd to observe it by scanningelectron microscopy (SEM) (S800; Hitachi, Tokyo,Japan) at a magnification of 3000. The number ofadhesion platelets was counted on an area of1.12� 104 mm2 (n ¼ 10).

2.5.2. Fibrinogen adsorption

Fifteen pieces of 5mm2 film were put in a tube andrinsed with sterile PBS. Elbow venous blood fromvolunteers infused with 5U/ml heparin was collected,iced and further centrifuged at 281 K. One milliliter ofplasma was added in the tube and shaken at 310 K for1 h. Successively, the samples were put in 2% PFA/PBSfor fixing, rinsed with PBS, blocked with � 1/4 BlockAce, rinsed in � 1/10 Block Ace-0.05% Tween20,allowed to contact with anti-fibrinogen antibody (� 1/5000) for 2 h, rinsed in � 1/10 Block Ace-0.05%Tween20, rinsed in � 1/10 Block Ace, and afterwardagitated in TMB one solution at a room temperature for7min. The reaction was stopped by adding 6n HCl.Absorbance was measured at 450 nm.

3. Results and discussion

AFM surface images of dry and wet PS/PVP films areshown in Fig. 1. RMS and RMS wet/dry ratio

ARTICLE IN PRESS

PVP ratio [%] 0 0.1 51 33 67500.5

RM

S [n

m]

RM

S w

et/d

ry r

atio

[-]dry

we twet/dry ratio

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

A B C D E F G H0

0.5

1

1.5

2

2.5

3

Fig. 2. RMS and RMS wet/dry ratio from the AFM surface images

(Samples A, B, C, D, E, F, G, H: K90).

0

10

20

30

40

50

60

70

80

0 4020 60 80

PVP K90

PVP K30

PVP K15

PVP K90/K15

Am

ount

of

PV

P o

n fi

lm s

urfa

ce [

wt%

]

Mixed amount of PVP [wt%]

Fig. 3. Relationship between amount of PVP present on film surface

and mixed amount of PVP (K90, K30, K15, and K90/K15 mixed).

0102030405060708090

0 20 40 60 80

Con

tact

ang

le [

°]

Amount of PVP on film surface [wt%]

PVP K90PVP K30PVP K15PVP K90/K15

Fig. 4. Relationship between contact angle and amount of PVP

present on film surface (K90, K30, K15, and K90/K15 mixed).

0

10

20

30

40

50

60

7080

0 20 40 60 80

Adh

esio

n am

ount

of

rabb

it p

late

let

[num

ber/

103 �

m2 ]

Amount of PVP on film surface [wt%]

PVP K90PVP K30PVP K15PVP K90/K15

Fig. 5. Relationship between adhesion amount of rabbit platelet and

amount of PVP present on film surface (K90, K30, K15, and K90/K15

mixed).

0

20

40

60

80

100

120

0 20 40 60 80

Amount of PVP on film surface [%]Rel

ativ

e am

ount

of

adso

rbed

fib

rino

gen

[%]

PVP K90

PVP K30

PVP K15

PVP K90/K15

Fig. 6. Relationship between relative amount of absorbed fibrinogen

and amount of PVP present on film surface (K90, K30, K15, and K90/

K15 mixed).

M. Hayama et al. / Biomaterials 25 (2004) 1019–1028 1025

determined from the AFM surface images (K90) areshown in Fig. 2. A relationship between amount of PVPpresent on the film surfaces by XPS analysis and mixedamount of PVP is shown in Fig. 3. The amount of PVPK90 on the film surfaces significantly increased at mixedamounts of PVP below 1–5 wt%mixed and gentlyincreased above 1–5 wt%mixed which is equal to20–30 wt% of amount of PVP present on the filmsurface. PVP was concentrated at the film surface below1–5 wt%mixed and tended to be distributed homoge-neously inside the film and on the surface at higheramounts of PVP. The surface properties reached theirbest performance at 1–5 wt%mixed with PVP K90. PVPof lower molecular weights was hard to appear on thesurface than PVP of higher molecular weights. Contactangle, adhesion amount of rabbit platelet and relativeamount of adsorbed fibrinogen against the amounts ofPVP present on the film surface are shown in Figs. 4–6,respectively. The reason why platelet adhesion andfibrinogen adsorption profiles were different and whyplatelet adhesion profile overlapped with contact angleprofile while fibrinogen adsorption profile did not areexplicable by differences in physicochemical properties

such as molecular weight and size (fibrinogen MW:340,000, platelet size: 2–3 mm) and adhesion/adsorptionmechanism. Further, based on our results and the factthat PVP improves biocompatibility of materials by the

ARTICLE IN PRESS

rino

gen

atel

et

100 1.2

Contact angleRabbit plateletFibrinogenRMS wet/dry ratio

M. Hayama et al. / Biomaterials 25 (2004) 1019–10281026

so-called ‘‘wiping effect’’ as well as hydrophilicity,fibrinogen adsorption is affected by both the wipingeffect and hydrophilicity while platelet adhesion isaffected mainly by hydrophilicity alone.

3.1. PVP morphology

PVP tiny particles of nanometer size were observed onthe dry film surfaces (Fig. 1). A phase-separation is two-dimensionally observable by AFM in the phase mode.The PVP/PS phase-separated morphology of oursamples was three-dimensionally observed by AFM inthe contact mode based on PVP swelling property.

Particle diameters of K90, K30 and K15 wereobtained to be 33.6711.3, 9.172.6, 10.873.6 nm fromthe images of samples H, J and M with higher amount ofPVP, respectively. Stokes diameters of PVP K90, K30and K15 were calculated from their molecular weights,assuming that molecules are perfectly spherical, usingthe following empirical equation [18] to be 49.6, 11.7,5.6 nm, respectively:

Stokes diameter ¼ 6:13 � 10�9MW0:456: ð1Þ

R

elat

ive

amou

nt o

f ad

sorb

ed f

ib[%

]

Adh

esio

n am

ount

of

rabb

it p

l[n

umbe

r/10

3 �

m2 ]

0102030405060708090

K90 K30 K15 K90/K150.0

0.2

0.4

0.6

0.8

1.0

RM

S w

et/d

ry r

atio

[-]

Con

tact

ang

le [

°]

Fig. 8. Comparison of films containing the same amount of PVP of

various molecular weights (Samples E: K90, I:K30, K:K15, N:K90/

K15 mixed, at 1–5wt%mixed).

These results indicate that only one molecule of PVPforms a particle with PVP K90 and K30, while severalmolecules gather and form a particle with PVP K15.With the film consisting of two kinds of PVP of differingmolecular weights (sample N; K90: 2.5 wt%mixed, K15:2.5 wt%mixed), a particle with a diameter of 26.077.8 nmwas formed. PVP of differing molecular weights formedwidely size-distributed particles.

This is the first time we revealed that PVP formednano-scale particles consisting of only one molecule withPVP K90, K30 and several molecules with PVP K15 onthe PS/PVP film surfaces.

Low

Dry Wet Dry

PVP particle

PS

Water

1-5 wt%

PVP

Mixed amount o

Fig. 7. PVP K90 behavior image belo

3.2. PVP behavior

PVP particles swelled in wetting (Fig. 1). PVPparticles of lower molecular weights swelled a little,while PVP particles of higher molecular weightssignificantly swelled, which united at higher amountsof PVP. Surface morphology of dry and wet PS filmshydrophilized with K90 was dramatically changed at1–5 wt%mixed (Fig. 7). Below 1–5 wt%mixed, PVP parti-cles observed on the dry surfaces swelled to formsmooth surfaces with some holes in the wet condition.On the other hand, above 1–5wt%mixed, PVP particlescovered the whole surface in the dry condition. PVPpresent on and inside the film swelled to heave up in thewet condition. Fig. 2 shows that RMS and RMS wet/dryratio have the same tendency as AFM images. RMSof the film surfaces decreased in drying above1–5 wt%mixed, while RMS increased in drying below1–5 wt%mixed. At 1–5 wt%mixed, PVP K90 formed

High

Wet Dry Wet

mixed

f PVP

w, at and above 1–5wt%mixed.

ARTICLE IN PRESS

Low High

Dry WetDry Wet Dry Wet

Molecular weight of PVP

K90 K30 K15

PVP particle

PVP

PS

At a mixed amount of 5wt%

Water

Fig. 9. Behavior image of PVP of various molecular weights at 1–5wt%mixed (K90, K30, K15).

M. Hayama et al. / Biomaterials 25 (2004) 1019–1028 1027

crowded particles on the dry surface. When wet, theyswelled, followed by their union to produce a smoothsurface leading to improved biocompatibility. Adheredwater was observed on the dry surface of the sample A.Gelled surface of the sample H in wetting wasunobservable because it was unstable on the fixingmaterial. AFM observation results agreed with the dataon contact angle (Fig. 4), adhesion amount of rabbitplatelet (Fig. 5) and relative amount of adsorbedfibrinogen (Fig. 6), which decreased remarkably at 1–5wt% mixed (PVP K90) (contact angle: 72� at1wt%mixed-35� at 5 wt%mixed, adhesion amount ofrabbit platelet: 20� 103 mm2 at 1 wt%mixed-12� 103 mm2 at 5wt%mixed, relative amount of adsorbedfibrinogen: 83 at 1wt%mixed-58 at 5wt%mixed).

Comparing films containing the same amount of PVPof different molecular weights (samples E, I, K, N;5wt%mixed) as shown in Fig. 8, the higher molecularweights PVP had, the more easily PVP particlesappeared on the film surfaces, achieving excellentbiocompatibility. Biocompatibility, hydrophilicity andRMS wet/dry ratio of a PVP K90/K15 blended filmwere located between those of PVP K90 and PVP K15,and almost the same as PVP K30. Fig. 9 illustratesimages of the PS/PVP K90, K30 and K15 film surfacesat 1–5 wt%mixed. PVP K90 formed crowded particles onthe dry surface. They swelled and united to form asmooth surface in the wet condition. PVP K30 and K15formed isolated particles on the dry surfaces, hence theywere incapable of completely covering the surface toform small holes on the wet surfaces.

Holes of sub-micron size were observed on both dryand wet surfaces of the sample K. These holes are notrelated to PVP behavior. They were formed during thepreparation of film samples because of low viscosity ofthe PS/PVP solution.

AFM images and RMS obtained from the AFMimages are useful for characterizing surfaces of mixedpolymers.

3.3. Future study

In the future study, we will reveal the difference ofPVP distribution between on and inside the film byAFM and transmission electron microscopy (TEM) andthe relationship between surface softness and biocom-patibility by AFM in the force mode. Furthermore, wewill try to reveal the mechanism how platelet adheresand fibrinogen is adsorbed on the PS/PVP surface byrelating surface morphology to their adhesion/adsorp-tion force using the force mode.

4. Conclusions

This is the first time such clear PS/PVP phase-separated morphology was observed by nanoscopictechnique. PVP nano-particles of one or severalmolecules are formed on dry PS/PVP film surfaces.Amount of PVP present on the surfaces increased withthe molecular weight of PVP. At 1–5 wt% of mixedamount of PVP, PVP K90 formed crowded particles onthe dry surface. When wet, they swelled, followed bytheir union to produce a smooth surface leading toimproved biocompatibility. The highest biocompatibil-ity with excellent mechanical strength is achievedby blending the highest molecular weight PVP K90 at1–5 wt%.

Acknowledgements

This work was supported in part by Grant-in-Aid for21COE ‘‘Practical Nano-Chemistry’’ from MEXT,Japan. The authors thank Professor Hiroyuki Nishidefor helpful discussion.

ARTICLE IN PRESSM. Hayama et al. / Biomaterials 25 (2004) 1019–10281028

References

[1] Bowry SK, Ronco C. Surface topolography and surface elemental

composition analysis of Helixones, a new high-flux polysulfone

dialysis membrane. Int J Artif Org 2001;24:757–64.

[2] Ochoa NA, Pr!adanos P, Palacio L, Pagliero C, Marchese J,

Hern!andez A. Pore size distributions based on AFM imaging and

retention of multidisperse polymer solutes characterisation of

polyethrsulfone UF membranes with dopes containing different

PVP. J Membr Sci 2001;187:227–37.

[3] Pila&r J, Labsk!y J, Marek A, Ko &n!ak &C. Translation diffusion of

paramagnetic tracers in poly(1-vinylpyrrolidone) hydrogel and

concentrated aqueous solutions by 1D electron spin resonance

imaging. Macromolecules 1999;32:8230–3.

[4] Maeda Y, Nakamura T, Ikeda I. Hydration and phase behavior

of poly(N-vinylcaprolactam) and poly(N-vinylpyrrolidone) in

water. Macromolecules 2002;35:217–22.

[5] S .uvegh K, Zelk !o R. Physical aging of poly(vinylpyrrolidone)

under different humidity conditions. Macromolecules

2002;35:795–800.

[6] Kuo SW, Chang FC. Studies of miscibility behavior and

hydrogen bonding in blends of poly(vinylphenol) and poly(vinyl-

pyrrolidone). Macromolecules 2001;34:5224–8.

[7] Motzer HR, Painter PC, Coleman MM. Interactions in miscible

blends of poly(styrene-co-methacrylic acid) with copolymers

containing vinylpyrrolidone and vinylpyridine groups. Macro-

molecules 2001;34:8390–3.

[8] Cowie JMG, McEwen IJ, Pethrick RA. Investigation of hydro-

gen-bonding structure in blends of poly(N-vinylpyrrolidone) with

poly(vinyl acetate-co-vinyl alcohol) using positron annihilation.

Macromolecules 2001;34:7071–5.

[9] Udels polysulfone catalog. Solvay Advanced Polymer, Co., Ltd.,

Tokyo, Japan; 2002.

[10] Kwak S, Yeom M, Roh I, Kim D, Kim J. Correlations of

chemical structure, atomic force microscopy (AFM) morphology,

and reverse osmosis (RO) characteristics in aromatic polyester

high-flux RO membranes. J Membr Sci 1997;132:183–91.

[11] Kowal A, Nowak S, Sulowicz W, Pietrzyk JA, Krawentek L,

Drozdz M, Nowogrodzka-Zagorska M, Bal W. Use of modern

microscopic techniques for examining dialysis membrane proper-

ties. Przegl Lek 2000;57:702–6.

[12] Khulbe KC, Matsuura T. Characterization of synthetic mem-

branes by Raman spectroscopy electron spin resonance and

atomic force microscopy: a review. Polymer 2000;41:1917–35.

[13] Bowen WR, Doneva TA, Yin HB. Polysulfone-sulfonated

poly(ether ether) ketone blend membranes: systematic synthesis

and characterization. J Membr Sci 2001;181:253–63.

[14] Kannno M, Kawakami H, Nagaoka S, Kubota S. Biocompatibility

of fluorinated polyimide. J Biomed Mater Res 2002;60:53–60.

[15] Marsh LH, Coke M, Dettmar PW, Ewen RJ, Havler M, Nevell

TG, Smart JD, Smith JR, Timmins B, Tsibouklis J, Alexander C.

Adsorbed poly(ethyleneoxide)-poly(propyleneoxide) copolymers

on synthetic surfaces: spectroscopy and microscopy of polymer

structures and effects on adhension of skin-borne bacteria.

J Biomed Mater Res 2002;61:641–52.

[16] Fushimi F, Nakayama M, Nishimura K, Hiyoshi T. Platelet

adhesion, contact phase coagulation activation, and C5a genera-

tion of polyethylene glycol acid-grafted high flux cellulosic

membrane with varieties of grafting amounts. Artif Org

1998;22:821–6.

[17] Hayama M, Kohori F, Sakai K. AFM observation of small

surface pores of hollow-fiber dialysis membrane using highly

sharpened probe. J Membr Sci 2002;197:243–9.

[18] Takesawa S, Ozawa K, Mimura R, Sakai K. Solute transport

mechanism through membrane in a dialyzer. Jpn J Artif Org

1984;13:1460–7 [in Japanese].