Saudi Journal of Kidney Diseases and Transplantation

: 2014  |  Volume : 25  |  Issue : 1  |  Page : 156--160

Polyflux® 210h hemodialysis membrane targets to improve filtration

A Hedayat1, A Shoker2,  
1 College of Dentistry, University of Saskatchewan, Saskatoon, Canada
2 Saskatchewan Transplant Program, St. Paul's Hospital, Saskatoon, Canada

Correspondence Address:
A Shoker
Saskatchewan Transplant Program, St. Paul«SQ»s Hospital, Saskatoon

How to cite this article:
Hedayat A, Shoker A. Polyflux® 210h hemodialysis membrane targets to improve filtration.Saudi J Kidney Dis Transpl 2014;25:156-160

How to cite this URL:
Hedayat A, Shoker A. Polyflux® 210h hemodialysis membrane targets to improve filtration. Saudi J Kidney Dis Transpl [serial online] 2014 [cited 2021 Jul 29 ];25:156-160
Available from:

Full Text

To the Editor,

The ideal hemodialysis filter should have pores large enough to clear all uremic toxins, including large molecules. However, this poses the threat of allowing back-diffusion of small toxins and loss of albumin, which forfeits the goal of toxin removal. Thus, the real challenge is the design of pores that strike an optimal ba­lance between clearing as many uremic toxins while at the same time minimizing back-diffusion and retaining useful molecules.

It is fairly understood how the pore size in ultrafiltration membranes affects its perfor­mance characteristics. [1],[2] Solute clearance de­pends on both molecule characteristics [3] and membrane properties. [4] Understanding the con­tribution of both factors is essential to improve the design of better membranes. Thus far, the membrane properties have been mainly studied by their manufacturers, and there are only very independent research on this subject. Effective diffusivity (D eff ) is a significant parameter in hemodialysis, as it quantitatively describes how effectively a molecule diffuses through a medium as compared with its free diffusivity (D o ) in solution. It is dependent on the hin­drance of the capillary wall that the molecules are passing through. As the diffusion hin­drance increases, the clearance of uremic to­xins decreases.

Polyflux® 210H, a popular capillary hemodialysis filter produced by Gambro Dialysatoren GmbH, Hechingen, Germany, is made up of 12,000 Polyamix TM fibers with an overall surface area of 2.1 m 2 . Recently, we have determined the inner pore density as 5.45 ± 1.41% and the capillary's wall thickness as 46.88 μm ± 1.65 for this dialyser. [5] Each fiber consists of three layers: Innermost skin layer, outermost finger layer and a large middle spongy layer for structural support. [6] Industrial tools are available to modify these structures. With this knowledge, we aimed this study to examine how changes in inner pore number and fiber diameter affect clearances of known molecules such as urea, glucose, Endothelin, β2-microglobulin and complement factor D through Gambro's Polyflux® 210H.

Initially, the characteristics of the fenestra­tions on the surfaces of the Polyflux® 210H capillaries were ascertained using scanning electron microscopy (SEM), as has been done by our team earlier in a previous study using the SEM . Next, the tension in the capillary fibers as a result of applied pressure was esti­mated using LaPlace's law as follows: T = (P × R) / t, where, T is the tension in the fiber wall, P is the pressure across the wall, R is the outside radius of the fiber and t is the fiber thickness.

Following this, the hindrance factors affec­ting molecular flux were calculated. There are two types of hindrance factors: The hindrance due to diffusion (K diff ) and the hindrance due to convection (Kc onv ). Both are attributed to solute-wall hydrodynamic interactions. [7] Basi­cally, free diffusivity (Do) of toxins is a function of the toxin's molecular weight (MW), and can be calculated as follows: D o = 1.76 ×10 -4 (MW) (-0 .552 ) . [8]

The effective diffusivity (D e ff) of toxins through hemodialysis hollow fiber membranes can be calculated as follows: D eff = D o [ε / (2 - ε)] = D o Kdiff, [9] where ε is the fraction of open space on the inner surface and [ε / (2 - ε)] 2 is K diff , the hindrance factor due to diffusion. K conv is the hindrance factor due to convection, and is a function of the pressure gradient across the membrane wall.

SEM was performed to ascertain the charac­teristics of the fenestrations on the surfaces of the Polyflux® 210H capillaries. [Figure 1] shows an SEM photomicrograph of the cross-section of a Polyflux® 210H capillary.{Figure 1}

Capillary wall pressure was measured assu­ming that the inlet pressure at the proximal end of the dialyzer is 300 mm Hg applied pressure; the tension exerted on each capillary's wall was estimated to be 0.066 mm Hg/fiber (1.3 × 10 psi/fiber). T he calculation was made using LaPlace's law.

[Table 1] summarizes the free diffusivity, D o , and the effective diffusivity (D eff ) of albumin, glucose and selected uremic toxins. The D e ff is the product of D o and the hindrance due to dif­fusion, K diff . The diffusion hindrance that exists at the surface in contact with the blood is equivalent to (ε/2- ε) [2] , where ε= 0.054, and is the mean of the fenestrated inner surface of the capillary's surface.{Table 1}

The theoretical potential of modifying pore density and capillary diameters to improve clearances of known molecules was examined. As the pore density increases, the diffusion hindrance drops. For example, if, theoretically, the pore density is doubled, the calculated ε increases to 10.8% and the effective diffusivities of the molecules will multiply by a factor of 4.23.

Similarly, if the pressure at the inlet of the dialyzer is maintained at 300 mm Hg and the wall thickness of each capillary remains the same, but the capillary's width was doubled, the capillary wall pressure estimated from Laplace law will then double to 0.132 mm Hg/fiber. Note that the pressure inside the ca­pillary is different from the trans-membrane pressure. We measured the intra-capillary wall pressure from LaPlace's Law, T = (P × R) /t defined earlier.

There are different mathematical models to measure diffusion through asymmetrical struc­tures such as tortuous dialysis membranes. Numerous studies have addressed the pore size dispersion in membranes. [10],[11],[12] While the pore size dispersion was modeled in one study, [13] another study focused on the effect of the dialyzer design on solid clearance. [14] In another approach, researchers used the perception of fractional rejection of solutes to determine the pore size distribution of ceramic and polymeric membranes. Also, mathematical mode­ling of mass transfer in an artificial kidney yielded results that coincided within 10% of the clinical ones. [16] Utilizing mini-modules proved very advantageous to determine the solutes' transit through hollow fiber membranes. [17] Also, the pore dispersion in cellulose membranes was found to affect the passage of large molecules. [18] Image examination is a no­vel technique that proved competent in esti­mating the practical pore size in permeable membranes. [10] The newly developed Al2O3 membranes that have nanopores have also been investigated with respect to their sieving characteristic [19] and the improvement of hemodialysis. [14] Different methods were also fol­lowed to characterize the pore sizes and their densities in those Al 2 O 3 membranes. For example, the pore count was estimated using area fraction in aluminum oxide tubular mem­branes. In this study, NIH Image J software was used to calculate the pore density. [20]

We used the equation that correlates the effective diffusivity (D eff ) through a fenestrated surface to the free diffusivity (D o ). [9] As illustrated in [Table 1], the effective diffusivity increased by at least 4.1-times when the pore density on the inner walls of the capillaries was only doubled. This is significant for future designs of capillaries.

From LaPlace's law, our theoretical calculation of the stress on the capillary walls prove that doubling the diameter of the capillaries while maintaining the wall thickness will double the wall tension in the membrane. This will lead to better clearance of middle mole­cules that are mainly cleared by convection. In addition, assuming that the inside diameter of the capillaries is doubled, pore density is doubled, but the fiber number is halved to maintain the final dialyzer size, K diff decreased by a factor of 6.61 from 7.7 × 10 -4 to 5.09 × 10 -3 . Thus, these modifications can have posi­tive results on the total clearance, particularly the middle molecules, as illustrated through our calculations in [Table 1].

To summarize, our results demonstrate that (a) reducing K diff will increase the clearance of toxins, (b) doubling the pore density will in­crease D eff , (c) doubling the capillaries' width will increase D eff and (d) doubling the pore density together with doubling the capillary width will increase D eff further than each sce­nario by itself.

However, there are several limitations to this study. The calculations used in this study are based on theoretical concepts and require experimental support in vivo. The contribution of molecular characteristics to hindrance factor was not addressed. In addition, applied fields such as the electrical fields on the surface of the membrane can add to the complexity of filtration. For example, some molecules can change shape in the presence of electrical fields. [21] Further, the calculation of convective hindrance requires further experiments and therefore was not addressed. Convection plays a major role in the clearance of lager mole­cules. Nevertheless, we feel that the data is robust enough to put the pore structure into the picture for future studies.


This research was supported by a grant from the Western Economic Diversification and Interprice Saskatchewan, Canada.

 Conflict of Interest Statement

The authors had no involvements that might raise the question of bias in the work reported or in the conclusions, implications or opinions stated. We declare that we have no conflict of interest pursuing this study. We also certify that neither this manuscript nor one with subs­tantially similar content under our authorship has been published or is being considered for publication elsewhere. We certify that all the data calculated during this study is presented in this manuscript, and no data from the study has been or will be published separately.


1Kanani DM, Fissel WH, Roy S, Dubnisheva A, Fleishman A, Zydney A. Permeability - selec­tivity analysis for ultrafiltration: Effect of pore geometry. J Mem Sci 2010;349:405-10.
2Clark WR, Ronco C. Determinants of hemodialyser performance and the potential effect on clinical outcome. Nephrol Dial Transplant 2001;15(Suppl 5):56-60.
3Vanholder R, De Smet R, Vogeleere P, Hsu C, Ringoir S. The Uremic Syndrom. In: Jacobs C, Kjellstrand CM, Koch KM, Winchester JF, eds. Replacement of Renal Function by Dia­lysis. 4 th ed. Netherlands: Kluwer Academic Publishers; 1996. p. 1-33.
4Vigano SM, Di Filippo S, Manzoni C, et al. Membrane Characteristics. In: Ronco C, Cruz DN, eds. Hemodialysis From Basic Research to Clinical Trials. Vol. 161. Contrib Nephrol, Basel, Karger 2008. p. 162-7.
5Hedayat A, Szpunar J, Kiran Kumar NAP, et al. Morphological Characterization of the Polyflux 210H Hemodialysis Filter Pores. International Journal of Nephrology 2012; doi: 10.1155/2012/304135.
6Ambalavanan S, Rabetoy G, Cheung AK. High-efficiency and high-flux hemodialysis; In: Schrier RW, ed. Atlas of Diseases of the Kidney. Current Medicine, Philadelphia, 1999, Vol. 5, p. 1-10.
7Dechadilok P, Deen WM. Hindrance factors for diffusion and convection in pores. Ind Engr Chem Res 2006;45:6953-9.
8Peterson PA, Ervin PE, Berggard I. Differen­tiation of glomerular, tubular, and normal proteinuria: Determinations of urinary excretion of beta2-microglobulin, albumin, and total protein. J Clin Invest 1969;48:1189-98.
9Keurentjes JT, Janssen AE, Broek AP, Van Der Padt A, Wesselingh JA, Van't Riet K. Multicomponent Diffusion in dialysis mem­branes. Chem Eng Sci 1992;47:1963-71.
10She FH, Tung KL, Kong LX. Calculation of effective pore diameters in porous filtration membranes with image analysis. Robitics Comput Integr Manufacturing 2008;24:427-34.
11Mey-Marom A, Katz MG. Measurment of active pore size distribution of microporous membranes - a new approach. J Mem Sci 1986;27:119-30.
12Ho CC, Zydney AL. Measerment of membrane pore interconnectivity. J Mem Sci 2000;170:101-12.
13Kassotis J, Shmidt J, Hodgins LT, Gregor HP. Modelling of the pore size distribution of ultrafiltration membranes. J Mem Sci 1985; 22:61-76.
14Huang Z, Zhang W, Yu J, Gao D. Nanoporous alumina membranes for enhancing hemodialysis. J Med Dev 2007;1:79-83.
15Lee S, Park G, Amy G, et al. Determination of membrane pore size distribution using the fractional rejection of non-ionic and charged macromolecules. J Mem Sci 2012;201:191-201.
16Liao Z, Poh C, Huang Z, Hardy P, Clark W, Gao D. A numerical and experimental study of mass transfer in the artificial kidney. J Biomech Eng 2003;125:472-80.
17Liao Z, Klein E, Poh CK, et al. Measurement of hollow fiber membrane transport properties in hemodialyzers. J Mem Sci 2005;256:176-83.
18Hoenich NA, Stamp S. Clinical investigation of the role of membrane structure on blood contact and solute transport characteristics of a cellulose membrane. Biomaterials 2000;21: 317-24.
19Attaluri AC, Huang Z, Belwalker A, Van Geertruyden W, Gao D, Misolek W. Eva­luation of nano-porous alumina membranes for hemodialysis application. ASAIO J 2009;55: 217-23.
20Belwalker A, Grasing E, Van Geertruyden W, Huang Z, Misiolek WZ. Effect of processing parameters on pore structure and thickness of anodic aluminum oxide (AAO) tubular mem­branes. J Mem Sc 2008;319:192-8.
21Eijkel JC, Van den Berg A. Nanotechnology for membranes, filters, and sieves. A series of mini - reviews covering new trends in funda­mental and applied research, and potential applications of miniaturized technologies. Lab Chip 2006;6:19-23.