冻干过程中管制西林瓶破损现象

冻干工艺是将液体产品在容器内进行冷冻，然后在低压环境下，通过升华形式进行干燥。而冻干制剂生产过程中可能会遇到的一个问题，就是作为容器包材的玻璃西林瓶偶尔出现破裂或破损，虽然这种现象相对罕见，但一旦发生，就可能是一个严重的问题，因为它会导致产品损失、甚至带来溢出产品和破碎玻璃渣对设备内部造成的污染。由于整个冻干过程会处于一定温差范围内进行，因此一些观点认为，这种破损现象与包材热应力有关，可以通过改变西林瓶的热性能来减少发生概率。 Downloaded fromjournal.pda.orgon February 25, 2019 Failure of Glass Tubing Vials during Lyophilization David R. Machak and Gary L. Smay PDA J Pharm Sci and Tech2019,7330-38 Access the most recent version at doi:10.5731/pdajpst.2017.008276 RESEARCH Failure of Glass Tubing Vials during Lyophilization DAVID R. MACHAK* and GARY L. SMAY American Glass Research, a Division of AGR International, Inc., 603 Evans City Road, Butler, PA, USA. ©PDA, Inc.2019 ABSTRACT Lyophilization is a commonly used and often preferred method for preparing certain drug products. In this process, the liquid pharmaceutical product is packaged in glass vials, frozen, and then dried via sublimation at low pressures. One problem that can be encountered during lyophilization is the occasional failure of the glass vial, a condition that will be referred to in this paper as “lyo-breakage.” Lyo-breakage, while relatively rare, can be a serious problem, as it results in lost product, additional costs to remediate any spillage, and inspection time to ensure that all broken vials are discarded. Some companies have suggested that lyo-breakage is related to thermal stress and, subsequently, can be reduced through changes to the thermal properties of the vials. In this paper, we will show that when the most common form of lyo-breakage occurs, the stresses in the glass are caused by an internal force from product expansion during freezing and not due to thermal stress from processing temperatures. KEYWORDS: Borosilicate glass, Fracture diagnosis, Fracture patterns,Lyophilization,Thermal shock, Internal force. LAY ABSTRACT:Lyophilization, or freeze drying, is often the preferred method for preparing certain drug products following manufacture. In this process, the liquid pharmaceutical product is packaged in a small glass cylindrical containercalledavial,frozen,andthendriedatlow pressures. One problem thatcanbeencounteredduring lyophilization is the occasional failure of the glass vial. While relatively rare, this failure can result in lost drug product, additional costs to clean up any spillage, and increased inspection time to ensure that all broken vials are discarded.The data presented in this paper demonstrate thatwhen themost common form of lyophilization-associated breakage occurs, the stresses in the glass are caused by an internal force from drug product expansion during freezing and not due to thermal stress on the glass from processing temperatures. Introduction Lyophilization consists of changing a liquid pharma-ceutical product into a dry solid “cake” by means of a freeze/dry process. This involves reducing the temper-ature of the liquid content in a glass vial over a period of several hours while holding the vial at atmospheric pressure. Once the liquid product is frozen, the pres-sure surrounding the vial is reduced to a relatively low value and a slight amount of heat is added to sublime thefrozen water. The temperatureandpressureare then returned to normal atmospheric values to com-plete the drying process. Occasionally, there can be a problem with failureofthevial(eithercrackingor *Corresponding Author: American Glass Research,a Divisionof AGR International,Inc.,603 Evans City Road, Butler, PA 16001; E-mail: dmachak@agrintl.com doi: 10.5731/pdajpst.2017.008276 complete breakage) during this process, and that fail-ure will be termed “lyo-breakage” in this paper. The process during which this type of failure occurs is the same as what others in the literature have termed “freeze–thaw” breakage (1). It is possible that lyo-breakage canoccuronboth molded and tubing vials, although the preponderance of problems that we have encountered is with broken tubing vials. We anticipate that this is not necessarily related to the physical characteristics of tubing vials compared to molded vials but rather to the predomi-nant use of tubing vials for pharmaceuticals that un-dergo the lyophilization process. Properly diagnosing the cause of vial failures is com-plicated,astherecanbeseveraldistinctlydifferent types of breakage occurring during lyophilization. Thesebreakagetypeshavedifferentcausesandre-quire different corrective actions. This paper will fo-cus on the more common type of lyo-breakageof Figure 1 Lyo-breakage: typical fracture patterns from commercial practice. tubing vials, which, in our experience with fracture diagnosis of vial breakage during lyophilization, man-ifests in a majority of instances by the fracture pattern shown in Figure 1. This pattern is characterized by a vertical fracture in the lower sidewall region some-times with forking above and/or below the origin site, which occurs on the outside glass surface. Lyo-breakage has been investigated to some extent in previous studies. In 1993, this type of breakage was attributed to differences in the coefficient of expansion of the frozen product in comparison to the glass vial (2). More recently, this type of breakage was attrib-uted to the number of degree-hours of subzero expo-sure of the vial during lyophilization (3). A thorough study involving detailed testing was undertaken using a strain gage that was mounted to a vial and subjected to a simulated lyophilization cycle (1,4).These studies quantified the existence of a significant strain in the glass during both freezing and subsequent thawing of the product that was created by a notable expansion of the frozen pharmaceutical product. While previous published papers and our laboratory experience have provided information relative to the failure of glass vials during lyophilization, some mis-conceptions in the industry exist about the manner in which lyo-breakage occurs. For example, lyo-break- age is sometimes ascribed to outward-directed forces related to the expansion of the frozen product against the inside surface of the glass vials. Other times, break-age is attributed to thermal differentials that are assumed to occur between the inside and outside glass surfaces during lyophilization. The purpose of this paper is to offer adefinitive explanation of the forces that are acting on the glass vials when the more common typeoflyo-breakage occurs. This explanation will be based on studies of the fracture patterns that are observed on vials after breakage.Fractographyisa well-known and valued means of understanding the forces that are involved in any glass article at the time of failure. When forces act on a glass object, the glass elastically deforms (strain), which in turn results in the creation of both compres-sive and tensile stresses. These stresses are uniquely distributed in the glass depending on design factors, glassthicknessdistribution,andthetypeofforce being applied to the object. Glass only fails under the influence of tensile stresses, and cracks will propagate in directions normal to the distribution of the tensile stresses.Thus, the crack patternwill be unique to the type of force that was acting on the glass object at the time of failure and can be used to identify the force after the fracture event. Figure 2 Examples of vials that failed due to an internal pressure force. Examples of different crack patterns of broken glass vials are shown in Figures 2 and 3. The vials in Figure 2 were broken by an internal pressure force that was created by filling the vials to overflowing with water and subjecting the filled vials to hydraulic pressure. The pressure was initially low and was increased until the vial failed. The fracture pattern consisted of a vertical crack that could exhibit branching above and below the precise location where the fracture origi-nated. The vial in Figure 2A exhibited extensive frac- turing, which is typical of relatively high pressures. The vial in Figure 2B failed at a much lower pressure and exhibited a relatively simple pattern consisting of only a straight vertical crack with looping in the lower end. The vials in Figure 3 were broken by a thermal shock force that was created by vials that had been heated in an oven and then immersed in a cold-water bath. The fracture pattern consisted of many meandering cracks Figure 3 Examples of vials that failed due to a thermal shock force. Glass thickness distribution. throughout the sidewall and base regions. The vial in Figure 3A exhibited extensive cracking in the sidewall indicative of a relatively high temperature differential at the time of failure. The vial in Figure 3B failed at a much lower temperature differential and exhibited a relatively simple pattern consisting of only a single circumferential crack around the base of the vial. Discussion Based onthefracturediagnosistechniquessumma-rized in the published literature (5–8) and as shown by the examples in Figures 2 and 3, it was concluded that the fracture patterns shown in Figure 1 are uniquely characteristic of breakage caused by a force applied to the inside surface of a vial causing it to expand out-ward. To confirm this conclusion,afiniteelement computer stress analysis (FEA) of a tubing vial that had been produced under normal commercial opera-tions was undertaken. The profile and glass thickness distribution of the vial used in these analyses is shown in Figure 4. In these analyses, a 3D symmetricalmodel was created using Solidworks, and the model was then imported into Autodesk® Simulation for the purpose of performing the FEA.The horizontal force simulated the expansion of water when it freezes into ice. The results of the FEA in Figure 5 show that the outward expandingforcegeneratestensilestressesofnearly equal magnitude on both the inside and outside glass Frozen Product Figure 5 Finite element analysis of tensile stress pattern due to expansion of frozen product. surfaces consistentwith the expansion of a thin-walled cylinder in which the thickness is much less than the cylinder diameter. Fracture origins will occur in this region on the outside surface owing to a greater like-lihood of having flaws of sufficient severity on this surface compared to the inside surface, consistentwith previous findings (2). Fracture diagnosis of origins from numerous 51 ex-pansion borosilicate tubing vials that broke during typical commercial lyophilization has shown that the tensile breaking stresses range from 27.6 to 69.0 MPa, as determined from measurements of the physical di-mensions of fracture origin mirrors (5) and as shown in Figure 6. The relationship between the dimensions of a fracture mirror and the breaking stress, o in MPa, is given in Shand’s book (9): where r inmeters is themirror radius (1/2 of themirror diameter) as shown in Figure 6. The proportionality constant, 1.87 MPa/m1/2, is for borosilicate glass, which is being considered in this example. The breaking stresses that were calculated from frac-ture analysis of vials that were broken in normal Figure 6 Fracture origin mirror from a lyo-breakage vial. lyophilization are on the same order of magnitude as the stresses calculated from the data of Milton (4) and Jiang (1). Based on their measured strain levels and assuming theuseofborosilicateglass,stresslevels ranging from 20.0 to 29.0 MPawere calculated, which they attributed to the expansion of the frozen product. Thus,thefracturepatternandthebreakingstresses indicate that the most likely cause for the most com-mon form of lyo-breakage is from the outward expan-sion of the frozen product. Even with this information, it is necessary to consider the assertion that lyo-breakage is caused by stresses generated by temperature gradients that are assumed to be created during the lyophilization process. If a thermal gradient werethe majorfactorthatcaused lyo-breakage,thefracturepattern would consistof meandering cracks in the sidewall and bottom regions with origins that would most likely be located on the outside glass surface in the bottom or heel areas as discussedintheliterature(5,7,8) and asshown in Figure 3. This is indirectcontrast tothefracture patternthatisobservedforvialsthatbreakduring commercial operations as shown in Figure 1. Breakingstressesfrom thermal gradients werealso considered relative to the failure of glass vials during lyophilization. For a rapid change in the surface tem- perature of a thin-walled cylinder (10), the magnitude of stress that is generated in the glass, b,y is given by: where E is Young’s modulus oftheglass, a isthe coefficientofthermal expansionoftheglass, i´sis Poisson’sratio,and TAT is thetemperaturediffer-ence between the outside and inside glass surfaces. Rearrangingthisequationtocreateastressindex value (stress generated per unit temperature differ-ence) gives: For a typical 51 expansion borosilicate composition that is used for pharmaceutical tubular vials, E is 69.0GPa, n´ gis 0.22, and a is 51× 10n7g cm/cm/°C. Using thesevaluesineq3givesastressindexvalueof 0.23MPa/°C. This stress index value can be used to calculatethetemperaturedifferentialthat wouldbe required to cause failure if the glass strength is known or it can be used to calculate the magnitude of stress that would be generated in the glass if the temperature differential is known. During a normal lyophilization process, vials filled with a pharmaceutical product are placed on shelves inside a lyophilization chamber. Refrigerant, such as liquid nitrogen, passes through cavities in the shelves, slowly cools the bearing surface region of the vials via conduction,andcoolstheenvironmentsurrounding the filled vials via convection. Because the total cool-ingtime ofafilledvialfrom room temperatureto approximatelyr4s0°C typically requires a few hours to complete, it is assumed that any temperature gradi-ents that might be created in the glass between the inside and outside surfaces of the vials would be very small. To test this hypothesis, eq 3 was used to estimate the temperature gradient that would be required to gener-ate the stress magnitudes that have been observed for numerous commercial breakage incidents. To achieve a total breaking stress of27.6 MPa, a temperature differential of 125°C between the inside and outside surfacesoftheglassvial wouldberequired.Fora breaking stress of 69.0 MPa, a temperature differential of 314°C would be required. It is unlikely that such high temperature gradients would be generated in the glass due to themanner inwhich filled vials are slowly cooled in normal commercial lyophilization processes. Thus, the expected fracture pattern and the overall magnitudeofstressforthermal gradientbreakage are inconsistent with the observations of commer-cial practice. Laboratory Tests As ameans of further investigating the load types that can lead to lyo-breakage, three laboratory tests were performed. In all tests, 51 expansion type borosilicate tubingvialsof70 mL capacity wereused. These sampleshadbeenselectedfrom commerciallypro-duced tubing vials exhibiting the design shown in Figure1and with normal thicknessprofilesforthe sidewall and bottom regions as shown in Figure 4. To normalize and control the glass surface strength during these tests,the entire outside sidewall and bottom surfaces of the vials were first manually abraded with emery paper consisting of 150 grit silicon carbide particles. This abrasion results in a glass surface strength of approximately 28 MPa for the load dura-tions that were encountered in these studies (11). Freezer Test Abraded vials were divided into two groups of 24samples each. In the first group, 40 mL deionized water was introduced into the vials, and the vials were left unclosed. In the second group, the vials were empty and unclosed. Both groups were placed into a freezer (a1t8°C) for 6 h, a time and temperature that were sufficient to solidly freeze the water in the filled vials as noted by visual observations. Underthesetestconditions,thefilledvialsexperi-enced an outward-directed force on the inside glass surface that was created by the expansion of the water asitfroze.It was assumed thatthermal gradients would be miniscule since the temperature change of the vials would be very gradual, and the inside and outside glass temperatures would remain essentially equal during the entire time period. The unfilled vials experienced no physical or significant thermal stresses. Afterstorage,thevials wereremovedandvisually inspectedforthepresenceofcrackingorcomplete failure. Twenty-three of the filled vials broke during this test, and all broken vials were examined to doc-ument the extent and nature of the fracture pattern. None of the unfilled vials failed. Liquid Nitrogen Immersion Test Abraded vials were divided into two groups for these tests. For the first group, 40 mL deionized water was introduced into five vials, which were left unclosed. The second group consisted of eight vials that were empty and unclosed.Both groupswere physically held by the finish and were individually immersed up to the lower neck region in liquid nitrogen ( 1396°C) for 3min. During immersion, care was used to assure that none of the liquid nitrogen was allowed to enter the unclosed finish of the test vials. It was visually ob-served that the water in the filled vials froze in about 2 min of this time interval. Underthesetestconditions,thefilledvialsexperi-enced an outward-directed force on the inside glass surface thatwas created by the expansion of the frozen water and a substantial thermal shock proportional to a temperature differential of 217°C (room temperature vials, at 21°C, immersed into liquid nitrogen, at 196°C). The unfilled vials experienced only a sub-stantial thermal shock. After testing, the vials were removed and visually inspected for the presence of cracking or complete failure. Two of the filled vials broke during this test, and both of them were examined to document the extent and nature of the fracture pattern. None of the unfilled vials failed. Oven to Water Bath Test Because this test initially involved heating the vials to elevated temperatures, all of these tests were under-taken withempty,unclosedvials. Thevials were placed into an oven at 218°C.After 30 min, the vials were individuallyremovedand, within3s,trans-ferred by the finish and physically held in a room temperature water bath (21°C) for 30 s with the water level at the lower portion of the neck. Care was taken to assure that water did not enter the open portion of the vials during testing. Four vials total were used in this test. Under these test conditions, the vials experienced a substantial thermal gradient of 197°C (heated vials, at 218°C, immersed in room temperature water, at 21°C) but no outward-directed mechanical force, as the vials were empty. After testing, the vials were removed and visually inspectedforthepresenceofcrackingorcomplete Figure 7 Typical fracture pattern from freezer test breakage. failure. All four vials broke during this test, and they were examined to document the extent and nature of the fracture pattern. Results of Laboratory Tests Freezer Test (Outward-Directed Force Only) Of the 24 filled vials, 23 failed, and a representative example of the fracture pattern is shown in Figure 7. These failures occurred only after the vials had been in the freezer for sufficient time for the water to freeze. None of the vials failed within the first fewminutes of insertion into the freezer. In addition, there were no failures of the 24 empty vials, indicating that the thermal stresses experienced by the vials were minis-cule (lessthanthesurface strengthoftheabraded glass). The breakage pattern shown in Figure 7 consisted of asingleverticalsidewallfracture withafracture origin on the outside glass surface in the lower sidewallareaconsistent withthesuperpositionof tensile stress created by the expanding ice with the surfacedamagecreatedbytheemerypaperabra-sion. This fracture pattern was identical to the pat-tern that is typically observed on the most common type of breakage that occurs during lyophilization as shown in Figure 1. Liquid Nitrogen Immersion (Combination of an Outward-Directed Force Plus a Significant Thermal Gradient) None of the eight empty vials failed, indicating that the thermal stresses did not exceed the surface strength of the abraded vials and therefore were relatively low in magnitude. Two of the five filled vials failed, and a representative example of the fracture pattern of the vials that broke in this test is shown in Figure 8. This fracture pattern was the same as was observed for the freezer test (Figure 7). It was noted during this test that the failures occurred at about 2 min of immersion when the water had frozen, thereby exerting an out-wardforce on thevial. Duringthefirst minuteof immersionwhen the thermal shockwould be the great-est and the water had not yet frozen, no failures were noted. Thus, based on the visual observations and on the similarity of the fracture pattern to the freezer test, it was concluded that the outward expansion force of the frozen water was the sole cause of failure in this test,andthermalstressesdidnotcontributetothe breakage. Oven to Cold Bath Thermal Shock Test (Thermal Gradient Only) A representative example of the fracture pattern of the four vials that broke in this test is shown in Figure 9. The fracture pattern was very complex and consisted of numerous meandering fractures in the sidewall Figure 8 Typical fracture pattern from liquid nitrogen immersion breakage. region plus a circumferential crack around the bearing surface.This fracture pattern substantially differs from the pattern that was observed in the previous two laboratory tests (see Figures 7 and 8) and from the fracture pattern that is typically observed in the most common form of lyo-breakage that occurs during com-mercial practice, as shown in Figure 1. The fracture pattern observed in this test was consistent with the unique pattern that is expected from failures due to the creation of a thermal gradient between the inside and outside glass surfaces as noted in the literature (5) and as observed from the exemplar fracture pattern in Figure 3. Conclusions Based on the nature of the fracture patterns, on the measuredbreakingstresses of tubingvialsthatfail during commercial lyophilization, and on the calcu-lated stress values from thermal differentials, it was concluded that the common type of lyo-breakage dis-cussed in this paper is due to the outward expansion force generated by the frozen pharmaceutical product and not due to thermal gradients. Thus, changes to the thermal properties of the glass vials (design changes to the vials or the use of glass having a lower coefficient of thermal expansion) are unlikely to make any significant difference in the frequency of breakage that may be experienced in typical lyoph-ilization processes. Solutions to lyo-breakage can be best realized by performing detailed fracture analyses. Such analyses will clearly differentiate the cause of breakage as either due to excessively high forces due to the ex-panding product or due to low glass strength caused by problems during vial production, transportation, or filling. Conflict of Interest Declaration The authors declare that they have no competing in-terests. References 1.Jiang, G. E.; Akers, M. Mechanistic Studies of Glass Vial Breakage for Frozen Formulations I. PDA J. Pharm. Sci. Technol. 2007, 61 (6),441– 451. 2. Williams, N. A.; Guglielmo, J. Thermal Mechan-ical Analysis of Frozen Solutions of Mannitol and Some Related Stereoisomers.PDA J. Pharm. Sci. Technol. 1993, 47 (3), 119–123. 3. Chow, E. J.; Kitaguchi, B.; Trier, M. Effects of Subzero Temperature Exposure and Supercooling onGlassVialBreakage.PDAJ.Pharm. Sci. Tech-nol. 2012, 66 (1), 55–62. 4. Milton, N.; Gopalrathnam, G.; Craig, G. D.; Mishra, D. S.; Roy, M. L.; Yu, L. Vial Breakage during Freeze-Drying: Crystallization of Sodium Chloride in Sodium Chloride-Sucrose Frozen Aqueous Solutions. J. Pharm. Sci. 2007, 96 (7),1848–1853. 5. Quinn, G. D. Fractography of Ceramics and Glasses; Special Publication 960-16e2; National Institute of Standards and Technology: 2007; Chapters 4, 5, 7. 6. ASTM C1678–10: Standard Practice for Fracto-graphic Analysis of Fracture Mirror Sizes in Ce-ramics and Glasses, March 2010. 7. Bradt, R. C.; Tessler, R. E. Fractography of Glass; Plenum Press: 1994; pp 207–252. 8. Lomax, J. Fracture Analysis as an Aid to Quality Improvement.Glass,March, 1972, 1–6. 9. Mecholsky,J.J.; Rice, R. W.; Freiman,S. W. Predictions of Fracture Energy and Flaw Size in Glasses from Measurements of Mirror Size. J. Am. Cer. Soc. 1974, 57 (10), 440. 10. Shand, E. B. Glass Engineering Handbook; McGraw-Hill Book Co.: NewYork, 1958; p 112. 11. Mould, R. E.; Southwick, R. D. Strength and Static Fatigue of Abraded Glass Under Controlled Ambient Conditions: II Effect of Various Abra-sionsandthe UniversalFatigue Curve.J. Am. Cer. Soc. 1958, 42 (12), 582. PDA Journal PDA Parenteral Drug Association of Pharmaceutical Science and Technology An Authorized User of the electronic PDA Journal of Pharmaceutical Science and Technology (the PDA Journal) is a PDA Member in good standing. Authorized Users are permitted to do the following: ·Search and view the content of the PDA Journal ·Download a single article for the individual use of an Authorized User ·Assemble and distribute links that point to the PDA Journal ·Print individual articles from the PDA Journal for the individual use of an Authorized User ·Make a reasonable number of photocopies of a printed article for the individual use of an Authorized User or for the use by or distribution to other Authorized Users Authorized Users are not permitted to do the following: ·Except as mentioned above, allow anyone other than an Authorized User to use or access the PDA Journal · Display or otherwise make any information from the PDA Journal available to anyone other than an Authorized User ·Post articles from the PDA Journal on Web sites, either available on the Internet or an Intranet, or in any form of online publications ·Transmit electronically, via e-mail or any other file transfer protocols, any portion of the PDA Journal ·Create a searchable archive of any portion of the PDA Journal ·Use robots or intelligent agents to access, search and/or systematically download any portion of the PDA Journal ·Sell, re-sell, rent, lease, license, sublicense, assign or otherwise transfer the use of the PDA Journal or its content ·Use or copy the PDA Journal for document delivery, fee-for-service use, or bulk reproduction or distribution of materials in any form, or any substantially similar commercial purpose ·Alter, modify, repackage or adapt any portion of the PDA Journal ·Make any edits or derivative works with respect to any portion of the PDA Journal including any text or graphics ·Delete or remove in any form or format, including on a printed article or photocopy, any copyright information or notice contained in the PDA Journal