J. BIOMED. MATER. HES.
VOL. 9, PP. 27-34 (1975)
Degree of Polymerization of Acrylic Bone Cement S. C. BAYXE, E . P. LAUTENSCHLAGER, C. L. COMPERE, and R. WILDES, Northwestern University, Chicago, Illinios 60611
Summary Self-curing powder-liquid admixed acrylic systems are used for internal fixation of total hip and total knee prostheses. Gel permeation chromatography revealed that the polymer chain length distributions of set cements were basically unaffected by their curing pressures. However, a decrease of approximately 11% in porosity coupled with a measured increase in mechanical strengths could be induced through the use of high curing pressures well beyond those attainable by the surgeon in the current arthroplasties. The conclusion of the investigation was that, to i-mprove such cements, attention should be focused on elimination of porosity rather than attempting to produce higher degrees of polymerization.
INTRODUCTION Set bone cement normally appears as an opaque, weak, porous polymer mass (Fig. l a ) . However, under carefully controlled laboratory conditions the cement may be set while the dough is under high pressure, producing a clear, strong, nonporous mass (Fig. lb). This latter curing condition is typical of various commercial processes used t o manufacture acrylic plastics; however, the necessary high pressures are completely unattainable by the surgeon working with bone cement, due to the nature of arthroplastic surgery. I n searching for more effective bone cements, the task can be simplified by determining which variables are capable of causing significant changes in current materials. I n chain polymers of the methyl methacrylate type, stengthening can often be effected by increasing the degree of polymerization (DP). Moreover, the removal of structural defects such as pores improves the mechanical 27
@ 1975 by John Wiley
&L
Sons, Inc.
UAYNE E T AL.
28
(a)
(b)
Fig. 1. Two specimens of the same weight and diameter made from a single batch of acrylic bone cement dough. (a) was allowed to set a t room pressure while (b) was cured a t 27.6 MPa.
properties. Therefore, the purpose of the present investigation will bc to ascertain if a strengthening of acrylic cements, brought about in this case by elevated curing prrssures, is caused by an increase in DP and/or by elimination of porosity.
MATERIALS AND METHODS Thc acrylic bone cement ckmployed in this investigation was Simplex-P* without radiopacifying BaS04. The cement is supplied as a two-component system consisting of a powder and a liquid, to be mixed just prior to use. Each component is supplied in prcweighed and presterilized packages. The recommended 2 to 1 ratio of components for mixing was employed throughout this investigation ( 2 parts powder b y weight to 1 part liquid by volume; e.g., 40 g of powder to 20 ml of liquid). The components w-erc mixed a t 20 f 1°C by adding the liquid t o the powd6.r component in a glass bcakcr. The mixture was immediately stirred for 60 see a t 150 rpm. The mixture was then allowcd to sit in the beaker for an additional minute, which concurred approximately with the onset of the dough stage. Portions of the mixturc. were then quickly transferred to * North Hill Plastics, Ltd., 60 Gmyland Road, London, N16, England.
POLYiMERIZATION OF ACRYLIC BONE CEMENT
29
type 304 stainless steel dies each containing a 4 mm diameter hole. In one series, the dough, after being shaped in the die, was removed and allowed to cure a t ambient pressure. I n a second series, 27.6 MPa pressure was applied to the dough in the die a t 1 min after the onset of the dough stage, and this pressure was maintained for 3 hr, which was well past the setting time. Pressure was applied by inserting a close fitting plunger into the die on top of the dough, and adding a n appropriate load ( - 3 5 kgf) to maintain the desired pressure during curing. After setting, the specimen ends were then squared off on a wet grinder to a length of approximately 8 mm, thereby producing a specimen with a 2 : 1 ratio of length to diameter. Each cement specimen was then appropriately tested for density, mechanical strength, and degree of polymerization. The apparent density of each cement specimen was calculated via a geometric method. The weight of each cylindrically shaped mass was recorded on a n Ainsworth type 10 analytical balance capable of measuring f0.0003 g. The volume of each cylinder was determined from measurements of external length and diameter using a micrometer capable of measurements to f0.0005 cm. The apparent density (g/cm3) for each cement specimen was appropriately calculated by dividing the specimen weight by the external specimen volume. The apparent density, papp,can be viewed as Papp
=
wt of polymer vol of polymer
+ wt of pores + vol of pores
If the weight of pores is assumed to be zero and the volume of pores is taken t o be zero in the specimens cured a t the high pressure, then in the ambient pressure cured specimens
% porosity
=
vol of pores vol of polymer vol of pores
+
=
) 100
(1 -
2)
100
(a)
where papp and p H P are the apparent densities of specimens cured a t ambient and high pressure, respectively. Five specimens from each of the two curing conditions were tested in compression along the longitudinal axis in a n Instron testing machine deforming the specimens at a crosshead speed of 0.05 cmi
30
BAYNE ET AL.
min. The compressive yield stress of t h r material, lated as uzI =
P/A,
u ~ \\as ,
calcu(3)
where P = the load a t which the mode changes from dastic to predominately plastic or permanent deformation, and A = thc original cross-sectional area (0.1256 cmz). A second group of 5 specimens from each curing condition were tested for diametral tension yield strength, uD. Her(. the right circular cylinder specimens were placed on their side and comprcssed diametrically between two flat platens a t a crosshead speed of 0.05 cm/min. The resulting calculated diametral trnsion yield strength is as follows UD
=
2L/?rdl
(4)
where L = the load a t which either brittle failure occurs or the mode changes from elastic to predominately plastic deformation, d = the specimen diameter, and 1 = the specimen length. The degree of polymerization of the bone cement for diff cwnt curing conditions was determined via gel permeation Chromatography (GPC). GPC is a relatively new and efficient method of dctrrmining the molecular size distribution in a polymer sample. This separation method involves column chromatography wherein a dissolved polymer sample passes through columns containing gel (usually a polystyrene gel) which is a porous, solvent-swollen polymer network available with different permeabilities varying over many orders of magnitude. The sample is dissolved in an elutrient which is made t o flow through the columns. The smaller molecules of the dissolved polymer sample penetrate into the gel more easily and arc, therefore, retained in the particular column for longer times. The larger molecules in the sample permeate less and pass through the column more quickly. Samples of cement from the two curing conditions were submitted for GPC analysis.* The cement was dissolved in spectrograde tetrahydrofuran solvent to a polymer concentration of 0.5 w/o. The dissolved sample was injected into the GPC over 120 sec a t ambient temperature while the equipment was operating at a continuous flow rate of 1ml/min. Using four sequential columns with gels of 5 X lo6,
* Arro Laboratories, P.O. Box 686, Joliet, Ill.
POLYMERIZATION OF ACRYLIC BONE CEMENT
31
7 X lo5,2 X lo4, and 1 X lo3 A pore sizes, the emerging elutrient was separated into fractions of decreasing average molecular size. Refractometry was used t o characterize each emerging fraction via comparison t o standards of known molecular length. The amount and distribution of fractions were appropriately reported.
RESULTS AND DISCUSSIONS The mechanical properties and apparent densities for the two curing conditions are reported in Table I. It should be noted that the cement specimens cured at 27.6 RiZPa are stronger and denser than those cured a t ambient temperature. I n addition the diametral tension strength was influenced to a much greater degree than was the compressive strength. Such behavior is typical of porous bodies whose pores, being squeezed shut during compression, do not degrade the mechanical strength nearly as much as pores being pulled open and acting as stress concentrators during tension. It must be remembered that Table I lists the properties of samples made from one batch of Simplex. While the trends of changing strength and density with pressure will probably be the same with any batch of Simplex or other bone cement of the same type, there may be some variation in exact values depending on batch, mixing conditions, and curing temperature as well as differences resulting from changing specimen size and rate of mechanical testing. TABLE I Set Cement Properties * as Influenced by Curing Pressure Ambient-Pressure Cured Mechanical properties compressive yield strength, ug diametral tension yield strength, U D Apparent density
27.6 M P a b Curing Pressure
7 6 . 5 f 2 . 5 MPa
8 3 . 9 f 2 . 3 MPa
1 3 . 2 f 3.6 MPa
23.7 f 1 . 7 MPa
1.044 f 0.040 gm/cm3 1.175 f 0.023 gm/cm3
Data reported as average value plus or minus one standard deviation. In accordance with Systeme International d’Unites one pound per square inch (1 psi) = 0.00689475 MPa (mega Pascals). a
b
BAYNE ET AL.
32
Figure 2 is a plot of the GPC data of the measured polymer chain lengths vs. their cumulative weight percentage. Attention should be focused on the fact that therc is virtually no difference between the resulting curves for the two curing ronditions. On the other hand, as seen in Table I, the difference in curing pressures effects a significant change in mechanical strength and density. Assuming the high pressure cured material t o be porous free, the porosity of the ambient pressure material can be estimated from eq. 2 to be some 11%. With such a large amount of porosity i t would appear that such standard remedial methods for improving polymer strength, namely, crosslinking, irradiation and branching, would probably be ineffective in light of the gross structural defects. Rather, attention should be dirccted toward the elimination of porosity. It is of academic interest to convert the chain length data of Figure 2 into a number representing the degree of polymerization. The DP is often times more physically significant to the researcher than the chain length, because it ropresents the number of monomer units which have added to each other to form the final polymer. This conversion may be performed by dividing the average polymer
Ambient Cured At P
a.
A,
e ,
VOL. 9, PP. 27-34 (1975)
Degree of Polymerization of Acrylic Bone Cement S. C. BAYXE, E . P. LAUTENSCHLAGER, C. L. COMPERE, and R. WILDES, Northwestern University, Chicago, Illinios 60611
Summary Self-curing powder-liquid admixed acrylic systems are used for internal fixation of total hip and total knee prostheses. Gel permeation chromatography revealed that the polymer chain length distributions of set cements were basically unaffected by their curing pressures. However, a decrease of approximately 11% in porosity coupled with a measured increase in mechanical strengths could be induced through the use of high curing pressures well beyond those attainable by the surgeon in the current arthroplasties. The conclusion of the investigation was that, to i-mprove such cements, attention should be focused on elimination of porosity rather than attempting to produce higher degrees of polymerization.
INTRODUCTION Set bone cement normally appears as an opaque, weak, porous polymer mass (Fig. l a ) . However, under carefully controlled laboratory conditions the cement may be set while the dough is under high pressure, producing a clear, strong, nonporous mass (Fig. lb). This latter curing condition is typical of various commercial processes used t o manufacture acrylic plastics; however, the necessary high pressures are completely unattainable by the surgeon working with bone cement, due to the nature of arthroplastic surgery. I n searching for more effective bone cements, the task can be simplified by determining which variables are capable of causing significant changes in current materials. I n chain polymers of the methyl methacrylate type, stengthening can often be effected by increasing the degree of polymerization (DP). Moreover, the removal of structural defects such as pores improves the mechanical 27
@ 1975 by John Wiley
&L
Sons, Inc.
UAYNE E T AL.
28
(a)
(b)
Fig. 1. Two specimens of the same weight and diameter made from a single batch of acrylic bone cement dough. (a) was allowed to set a t room pressure while (b) was cured a t 27.6 MPa.
properties. Therefore, the purpose of the present investigation will bc to ascertain if a strengthening of acrylic cements, brought about in this case by elevated curing prrssures, is caused by an increase in DP and/or by elimination of porosity.
MATERIALS AND METHODS Thc acrylic bone cement ckmployed in this investigation was Simplex-P* without radiopacifying BaS04. The cement is supplied as a two-component system consisting of a powder and a liquid, to be mixed just prior to use. Each component is supplied in prcweighed and presterilized packages. The recommended 2 to 1 ratio of components for mixing was employed throughout this investigation ( 2 parts powder b y weight to 1 part liquid by volume; e.g., 40 g of powder to 20 ml of liquid). The components w-erc mixed a t 20 f 1°C by adding the liquid t o the powd6.r component in a glass bcakcr. The mixture was immediately stirred for 60 see a t 150 rpm. The mixture was then allowcd to sit in the beaker for an additional minute, which concurred approximately with the onset of the dough stage. Portions of the mixturc. were then quickly transferred to * North Hill Plastics, Ltd., 60 Gmyland Road, London, N16, England.
POLYiMERIZATION OF ACRYLIC BONE CEMENT
29
type 304 stainless steel dies each containing a 4 mm diameter hole. In one series, the dough, after being shaped in the die, was removed and allowed to cure a t ambient pressure. I n a second series, 27.6 MPa pressure was applied to the dough in the die a t 1 min after the onset of the dough stage, and this pressure was maintained for 3 hr, which was well past the setting time. Pressure was applied by inserting a close fitting plunger into the die on top of the dough, and adding a n appropriate load ( - 3 5 kgf) to maintain the desired pressure during curing. After setting, the specimen ends were then squared off on a wet grinder to a length of approximately 8 mm, thereby producing a specimen with a 2 : 1 ratio of length to diameter. Each cement specimen was then appropriately tested for density, mechanical strength, and degree of polymerization. The apparent density of each cement specimen was calculated via a geometric method. The weight of each cylindrically shaped mass was recorded on a n Ainsworth type 10 analytical balance capable of measuring f0.0003 g. The volume of each cylinder was determined from measurements of external length and diameter using a micrometer capable of measurements to f0.0005 cm. The apparent density (g/cm3) for each cement specimen was appropriately calculated by dividing the specimen weight by the external specimen volume. The apparent density, papp,can be viewed as Papp
=
wt of polymer vol of polymer
+ wt of pores + vol of pores
If the weight of pores is assumed to be zero and the volume of pores is taken t o be zero in the specimens cured a t the high pressure, then in the ambient pressure cured specimens
% porosity
=
vol of pores vol of polymer vol of pores
+
=
) 100
(1 -
2)
100
(a)
where papp and p H P are the apparent densities of specimens cured a t ambient and high pressure, respectively. Five specimens from each of the two curing conditions were tested in compression along the longitudinal axis in a n Instron testing machine deforming the specimens at a crosshead speed of 0.05 cmi
30
BAYNE ET AL.
min. The compressive yield stress of t h r material, lated as uzI =
P/A,
u ~ \\as ,
calcu(3)
where P = the load a t which the mode changes from dastic to predominately plastic or permanent deformation, and A = thc original cross-sectional area (0.1256 cmz). A second group of 5 specimens from each curing condition were tested for diametral tension yield strength, uD. Her(. the right circular cylinder specimens were placed on their side and comprcssed diametrically between two flat platens a t a crosshead speed of 0.05 cm/min. The resulting calculated diametral trnsion yield strength is as follows UD
=
2L/?rdl
(4)
where L = the load a t which either brittle failure occurs or the mode changes from elastic to predominately plastic deformation, d = the specimen diameter, and 1 = the specimen length. The degree of polymerization of the bone cement for diff cwnt curing conditions was determined via gel permeation Chromatography (GPC). GPC is a relatively new and efficient method of dctrrmining the molecular size distribution in a polymer sample. This separation method involves column chromatography wherein a dissolved polymer sample passes through columns containing gel (usually a polystyrene gel) which is a porous, solvent-swollen polymer network available with different permeabilities varying over many orders of magnitude. The sample is dissolved in an elutrient which is made t o flow through the columns. The smaller molecules of the dissolved polymer sample penetrate into the gel more easily and arc, therefore, retained in the particular column for longer times. The larger molecules in the sample permeate less and pass through the column more quickly. Samples of cement from the two curing conditions were submitted for GPC analysis.* The cement was dissolved in spectrograde tetrahydrofuran solvent to a polymer concentration of 0.5 w/o. The dissolved sample was injected into the GPC over 120 sec a t ambient temperature while the equipment was operating at a continuous flow rate of 1ml/min. Using four sequential columns with gels of 5 X lo6,
* Arro Laboratories, P.O. Box 686, Joliet, Ill.
POLYMERIZATION OF ACRYLIC BONE CEMENT
31
7 X lo5,2 X lo4, and 1 X lo3 A pore sizes, the emerging elutrient was separated into fractions of decreasing average molecular size. Refractometry was used t o characterize each emerging fraction via comparison t o standards of known molecular length. The amount and distribution of fractions were appropriately reported.
RESULTS AND DISCUSSIONS The mechanical properties and apparent densities for the two curing conditions are reported in Table I. It should be noted that the cement specimens cured at 27.6 RiZPa are stronger and denser than those cured a t ambient temperature. I n addition the diametral tension strength was influenced to a much greater degree than was the compressive strength. Such behavior is typical of porous bodies whose pores, being squeezed shut during compression, do not degrade the mechanical strength nearly as much as pores being pulled open and acting as stress concentrators during tension. It must be remembered that Table I lists the properties of samples made from one batch of Simplex. While the trends of changing strength and density with pressure will probably be the same with any batch of Simplex or other bone cement of the same type, there may be some variation in exact values depending on batch, mixing conditions, and curing temperature as well as differences resulting from changing specimen size and rate of mechanical testing. TABLE I Set Cement Properties * as Influenced by Curing Pressure Ambient-Pressure Cured Mechanical properties compressive yield strength, ug diametral tension yield strength, U D Apparent density
27.6 M P a b Curing Pressure
7 6 . 5 f 2 . 5 MPa
8 3 . 9 f 2 . 3 MPa
1 3 . 2 f 3.6 MPa
23.7 f 1 . 7 MPa
1.044 f 0.040 gm/cm3 1.175 f 0.023 gm/cm3
Data reported as average value plus or minus one standard deviation. In accordance with Systeme International d’Unites one pound per square inch (1 psi) = 0.00689475 MPa (mega Pascals). a
b
BAYNE ET AL.
32
Figure 2 is a plot of the GPC data of the measured polymer chain lengths vs. their cumulative weight percentage. Attention should be focused on the fact that therc is virtually no difference between the resulting curves for the two curing ronditions. On the other hand, as seen in Table I, the difference in curing pressures effects a significant change in mechanical strength and density. Assuming the high pressure cured material t o be porous free, the porosity of the ambient pressure material can be estimated from eq. 2 to be some 11%. With such a large amount of porosity i t would appear that such standard remedial methods for improving polymer strength, namely, crosslinking, irradiation and branching, would probably be ineffective in light of the gross structural defects. Rather, attention should be dirccted toward the elimination of porosity. It is of academic interest to convert the chain length data of Figure 2 into a number representing the degree of polymerization. The DP is often times more physically significant to the researcher than the chain length, because it ropresents the number of monomer units which have added to each other to form the final polymer. This conversion may be performed by dividing the average polymer
Ambient Cured At P
a.
A,
e ,
