Product Specials


Inside Dentistry

September 2007, Volume 3, Issue 8
Published by AEGIS Communications

Polymerization Shrinkage—A Clinical Review

Deniz Cakir, DDS, MS; Robert Sergent, DMD; and John O. Burgess, DDS, MS

The greatest limitation in the use of composite resin as a posterior restorative material seems to be shrinkage during polymerization, which leads to poor marginal seal, marginal staining, and recurrent caries. Because no method guarantees a perfectly sealed restoration for adhesive restorative materials, clinicians must address problems of polymerization shrinkage and resulting destructive shrinkage stress. Only a thorough understanding of the mechanisms that cause shrinkage stress and the techniques that may reduce its effect will allow clinicians to gain a better use of resin composites. Because recurrent caries is one of the leading causes of restoration replacement, it is imperative that low-shrinkage composite resins be developed. The objectives of this article are to review the origin of polymerization shrinkage, the clinical factors affecting polymerization stress, and methods advocated to reduce shrinkage stress and the effectiveness of these methods.


Composite resins have four primary components: an organic matrix, inorganic fillers, a coupling agent that binds the filler to the matrix, and the initiator/accelerator system. In most composites, the organic matrix is a dimethacrylate, generally bisphenol-A glycidil dimethacrylate (bis-GMA) or urethane dimethacrylate (UDMA) blended with triethylene glycol dimethacrylate (TEGDMA). The matrix contains reactive carbon-carbon double bonds, which crosslink to form a polymer network. Composite resin polymerizes by free radical polymerization generated when a photo-initiator, such as camphoroquinone, absorbs light energy (photons) emitted from the curing light and initiates polymerization by reacting with a photoreducer, a tertiary amine forming free radicals and initiating crosslinking.1 Camphoroquinone has a maximum absorption at 468 nm and can be polymerized with LED curing lights. Some composite resins use other photoinitiators, such as 1-phenyl-1,2-propanedione (with a peak absorption of 410 nm), bisacylphosphine oxide, or triacylphosphine oxide (with peak absorptions of 320 nm to 390 nm), which fall outside the curing range of most LED curing lights.1 These photoinitiators are used to reduce the strong yellow color produced by camphoroquinone.

The exothermic reaction created when the monomer converts to the polymer produces a volume reduction in the polymer with a resulting decrease in molecular vibration and intermolecular distances.2 As the polymer is formed, the resin matrix changes from a paste or pregel state to a viscous solid3 and the composite resin contracts by about 1.5% to 5%. The gel point is the point at which the resin changes from a viscous paste to an elastic solid. When the gel point is reached, stress is transmitted from the composite resin to the surrounding tooth structures. When composite resin is a paste, or pregel state, no stress is conducted to surrounding tooth structure. As curing begins, the material flows from unbound surfaces to accommodate for shrinkage. As the composite resin becomes more rigid because of the increasing modulus of the composite, flow stops and the bonded composite resin transmits shrinkage stresses generated to the surrounding tooth. This point is called the gel point and the stress generated may exceed the adhesive bond or the cohesive strength of the tooth or the composite, producing a marginal defect.

When composite resin is bonded on all surfaces, shrinkage must be compensated by strain (flow) of the composite, tooth, or adhesive.4 If this stress is greater than the cohesive strength of the composite, damage occurs within the composite.5 If the stress exceeds the tensile strength of enamel, the enamel fractures. If the adhesive was placed improperly, then it will fail. These failures can be seen as a white line that appears during restoration finishing because the finishing debris collects in the defect and changes the index of refraction of light.6 Cracks or fractures are seen in teeth with bucco-lingually wide restorations7 because the cavity walls are primarily enamel, which is brittle and too thin to withstand the forces generated by polymerization shrinkage.


Polymerization shrinkage, stress, and modulus development are dependent on the conversion of the monomer to the polymer. Stress development is affected by the preparation geometry and substrate compliance7-9 (whether the composite resin and or tooth structure can flex). The C-factor is the ratio of bonded and unbonded surfaces in the restored tooth. Feilzer and colleagues8 used C-factor to describe the stress generated during polymerization shrinkage of composite resin. When the ratio of bonded to unbonded surfaces increases, the stress placed on the tooth increases because the composite resin cannot flow to relieve the shrinkage stresses.

The greatest stress occurs when composite is bonded to five walls of a prepared cavity (C = 5) as in Class 1 or Class 5 restorations. The composite attempts to shrink toward the bonded surface but will be restrained by the bonded areas on the opposing surface. The lowest C-factor values are obtained with class IV cavities because the material has enough unbonded surfaces to flow, providing stress relief. A high C-factor creates a risk for debonding of the restoration. Shallow and large designs reduce the C-factor; therefore, it is important to have a lower configuration cavity. If a light-curing technique could produce a difference in marginal adaptation, it could be most easily demonstrated in the Class 1 or Class 5 restoration.

The authors could not find any clinical study that demonstrated that any curing method produced an improvement in marginal adaptation or marginal discoloration compared with a standard incremental placement and curing technique. The ability of polymer to flow, and thereby relieve some portion of the stress, is documented. Most recent studies measuring the shrinkage of composite resins have reported about 2% to 3% polymerization shrinkage by volume in highly filled composite resin restorative material.10,11 Flowable composite resins have greater shrinkage, ranging from 4% to 5% per volume.12 The polymerization shrinkage produced in a given composite resin is related to the shade, opacity, and composition of the composite resin, the irradiance levels or exposure times of the curing light used, any incompatibility between a photo-initiator system and the spectral output of the curing light, cavity preparation geometry, and composite layer thickness.13,14

When composite resins are cured, light passes through the composite attenuates, which means that deeper layers of composite resin are less cured. Any factor that decreases the light intensity passing through the composite will lower the conversion rates of the composite resin. If inadequate levels of conversion are achieved during polymerization, mechanical properties and wear resistance are reduced. With incomplete curing, leachable residual monomers and initiators become greater biocompatibility issues, and color instability can also become a problem. The following is a list of factors that affect polymerization shrinkage stress:15,16
• curing-light guide placement (how far away from the surface it is);
• intensity and wavelength of the curing light;
• curing mode of the composite resin (light-cure or chemical-cure);
• flow of the composite—early compensation before development of significant modulus of material;
• water sorption of the composite—a mechanism for compensating for shrinkage and giving improved marginal adaptation;
• composition composites with low filler increases shrinkage (flowables vs restorative composites);
• shade and opacity of the composite resin; and
• type of composite resin—flowable vs highly filled.

Different resins cure differently. Light-cured composite materials may be undercured through reduced irradiance levels, inadequate exposure times, or incompatibility between a photo-initiator and the spectral output of the curing light. A curing unit may register 700 mW/cm2, but as light passes through composite it attenuates, diminishing rapidly from 248 mW/cm2 at 0.5-mm thickness to 25 mW/cm2 at 3-mm thickness of composite resin. Curing composite in 2-mm increments is recommended. 

The depth of cure of the composite resin varies not only with the power density of the curing light and the amount of photo-initiator in the composite, but also with the type and shade of the composite, with darker, more opaque shades and microfills requiring longer curing times. The further the light guide is from the surface being cured, the lower the energy received by the composite and the longer that particular material must be cured. Light attenuates when passing through tooth structure as well as the composite. Therefore, curing through the tooth reduces light intensity and it is not recommended as an effective method for poly-merizing composite resin restorations.


Several articles have suggested that modifying curing light output may reduce polymerization shrinkage and improve marginal integrity. Three modes are available. A continuous cure is used when the output is constant for a specified period of time. The step- or ramp-cure begins at low intensity and switches to higher intensity. The pulse-delay cure is a discontinuous curing procedure. With this technique, increments of composite resin are inserted and cured using the continuous cure. The final occlusal increment receives a brief low-intensity cure and, after a delay during which the resin is finished, the material is then fully polymerized to gain final mechanical properties. Although it has been suggested that the curing mode and composite resin placement technique may affect the marginal integrity of a composite resin restoration, in vitro results to date have been mixed, with some investigators showing improved resistance to leakage while others have been unable to demonstrate efficacy with these techniques.

Originally, a vertical placement incremental technique would reduce the total shrinkage in a composite resin restoration. Since then, many variations of the incremental placement technique have been advocated. A gingivo-occlusal layering (horizontal)17,18 and wedge-shape layering (oblique) method is to place and polymerize wedge-shaped composite increments from the occlusal surface;19 the successive cusp build-up technique20 is to apply the first composite increment to a single dentin surface without contacting the opposing cavity walls, and to build up the restoration by placing a series of wedge-shaped composite increments; with this technique each cusp is built up separately.

One early publication21 advocated a three-sided light-curing technique and incremental placement of the composite to decrease the polymerization stress generated in the tooth structure during composite resin curing. In this technique, a transparent matrix is used and a light-reflecting wedge is placed interproximally at the gingival margin of the preparation. Light-curing progressed by curing through the wedge to cure the initial gingival increment of composite, then the buccal, and then the facial increment was placed and cured. Even though this technique has been used by many clinicians, Losche22 reported that little light reaches the center of the preparation. The three-sided curing technique’s success is not due to the three-sided method, but to decreased light transmission and poor composite resin polymerization. This brings into question all techniques where increments greater than 2 mm are used or the composite resin is cured through the tooth. In the centripetal build-up technique,28 developed for class II cavity restorations, an initial vertical composite increment is applied on the cervical margin against the metal matrix. Cavity filling is then completed by horizontally layering. This technique allows transformation of class II cavities into class I cavities.

Bulk placement and curing has been recommended to reduce stress at the cavosurface margins.23 Using transenamel polymerization, advocated by Belvedere,24 the adhesive, a flowable composite, and a composite resin are placed into the preparation in bulk and then polymerized by curing through the tooth from the buccal and lingual. Polymerization is completed by curing from the occlusal. This method of curing composite resin was tested by measuring the leakage in Class I restorations. In two separate studies,25,26 composite resin was used to restore premolars using four different placement and curing techniques. No difference in leakage was found from the bulk-filling technique vs any of the incremental curing techniques, even the pulse-delay curing technique. In these studies, the composite resin was placed incrementally using different types of increments—horizontal, diagonal, and with a slit in the center of the composite that was filled with a final incremental of composite. There was no statistical difference among any group and the bulk fill had the same leakage as the other placement methods. After the microleakage was measured, the hardness of the sectioned teeth was measured, beginning from the occlusal and proceeding toward the pulp. The hardness of the bulk-filled restorations was significantly less than the incrementally cured restorations, which again demonstrates the limited depth of cure of composite resin.

Composite placed incrementally ensures more complete curing. Incompletely cured composite resins may release components into the oral cavity that may be detrimental.27 The reason for the reduced shrinkage with the bulk-curing technique is obvious—uncured composite resin does not shrink as much as completely cured resin. A significant factor in the reduction of curing effectiveness with a bulk-filling technique is that, as previously discussed, light attenuates while penetrating through the tooth structure. As light passes through the tooth structure, it drops dramatically from 500 mW/cm2 to 80 mW/cm2 when curing through 2.5 mm of tooth.

Is incremental curing the answer to curing completely without stress? Many investigators have compared incremental curing with bulk polymerization of composite resin. Eakle and Ito28 compared four incremental insertion methods and noted that diagonal insertion was best. Crim and Chapman29 reported that incremental placement of composite resin was no more effective than bulk placement in reducing leakage. Coli and Brannstrom30 reported that in composite resin restorations with bulk insertion, the number of restorations with gaps was similar to a two-stage insertion. Versluis and colleagues31 reported that incremental filling techniques reduce cusp movement in teeth with a well-established bond. This brief and incomplete survey of the investigations evaluating the bulk and incremental insertion of composite resin reveals that neither method consistently produces superior results. Some report less leakage with the incremental technique; others less with bulk placement.

In 1992, Goracci and co-workers32 slowly polymerized composite resin over a 4-minute period, while controlling the output of the curing unit with a rheostat. They showed fewer gaps and marginal defects with this technique. The slow polymerization technique was verified, but required so much time to polymerize composite resin that it was clinically ineffective. They did show that slow polymerization methods have merit.

Highlight by 3M ESPE (St. Paul, MN) was the first curing light with a step- or soft-cure, but was discontinued and replaced with the 3M ESPE Elipar® TriLight, which has an exponential output mode in which the output slides from low to high. In two studies,33,34 no significant difference could be found between using this technique and bulk-curing. Another technique is the pulse-delay, or the pulse-cure technique.35 This requires placing increments of composite resin and curing for 20 seconds. The final enamel replacement increment is cured with a brief burst of energy for 2 to 3 seconds. A 3-minute delay is then allowed to enable the composite time to flow and shrink while the restoration is finished and polished. After finishing, the restoration is cured at high intensity to totally polymerize the material. Mechanical properties are maintained when these techniques are applied.

In several studies,36-38 no clear benefit to the soft-curing or the pulse-delay technique could be seen. The effectiveness of the soft-cure or ramp-cure techniques in decreasing leakage and stress at the margins of Class 2 restorations is not clear and it has not been reproduced in clinical trials. Either the evaluation methods in clinical trials are not refined enough to detect these differences or the amount of shrinkage is compensated for by other factors (water sorption39 or compliance of the tooth). Composites are able to compensate for volumetric shrinkage by flow before the resin reaches a solid state, although this compensation is limited;40 in this study the authors found only 20% of the shrinkage completed at the maximum flow. To accomplish improved marginal integrity, the composite resin must flow during its change from a viscous paste to an elastic solid to accommodate the resin shrinkage and to yield sealed margins. A composite resin that does not shrink is necessary to consistently improve marginal integrity, and continued work in this area is essential.


To summarize the main points of this article, the authors reiterate that among the factors affecting composite-resin shrinkage, cavity preparation size and configuration are significant. Light intensity decreases rapidly as light passes through tooth or composite resin. Curing lights have large differences in light output, but light attenuation reduces the ability of curing lights to polymerize thick increments of resin; therefore, composite resin is most completely polymerized in 2-mm increments. Even though different curing units have different curing modes, the composite selected affects shrinkage more than the method of curing. The clinical effectiveness of the soft-, ramp-, or pulse-delay cure is questionable. Continued development of composite resins with reduced shrinkage is critically needed.


Dr. Burgess has received grant/research support from 3M ESPE.



1. Stansbury JW. Curing dental resins and composites by photopolymerization. J Esthet Dent. 2000;12(6):300-308.


2. Ferracane JL. Developing a more complete understanding of stresses produced in dental composites during polymerization. Dent Mater. 2005;21(1):36-42.

3. Charton C, Colon P, Pla F. Shrinkage stress in light-cured composite resins: Influence of material and photoactivation mode. Dent Mater. 2006 Oct 5; [E-pub ahead of print].

4. Davidson CL, Feilzer AJ. Polymerization shrinkage and polymerization shrinkage stress in polymer-based restoratives. J Dent. 1997;25(6): 435-440.

5. Stansbury JW, Trujillo-Lemon M, Lu H, et al. Conversion-dependent shrinkage stress and strain in dental resins and composites. Dent Mater. 2005;21(1): 56-67.

6. Pensak T. Clinical showcase. J Can Dent Assoc. 2004;70(2): 118-119.

7. Sarrett DC. Clinical challenges and the relevance of materials testing for posterior composite restorations. Dent Mater. 2005;(1):9-20.

8. Feilzer AJ, De Gee AJ, Davidson CL. Setting stress in composite resin in relation to configuration of the restoration. J Dent Res. 1987;66(11): 1636-1639.

9. Rees JS, Jacobson PH. The polymerization shrinkage of composite resins. Dent Mater. 1989;5(1):41-44.

10. Puckett AD, Smith R. Method to measure the polymerization shrinkage of light-cured composites. J Prosthet Dent. 1992;68(1):56-58. 

11. Norris C, Burgess JO. Polymerization shrinkage of seventeen composite resins [abstract]. J Dent Res. 2002;81:A-424.

12. Labella R, Lambrechts P, Van Meerbeek B, et al. Polymerization shrinkage and elasticity of flowable composites and filled adhesives. Dent Mater. 1999;15(2): 128-137.

13. Rueggeberg FA, Caughman WF, Curtis JW Jr. Effect of light intensity and exposure duration on cure of resin composite. Oper Dent. 1994;19: 26-32.

14. Alster D, Venhoven BA, Feilzer AJ, et al. Infuence of compliance of the substrate materials on polymerization contraction stress in thin resin composite layers. Biomaterials. 1997;18(4):337-341.

15. Giachetti L, Scaminaci Russo D, Bambi C, et al. A review of polymerization shrinkage stress: current techniques for posterior direct resin restorations. J Contemp Dent Pract. 2006;7(4):79-88.

16. Braga RR, Ferracane JL. Alternatives in polymerization contraction stress management. Crit Rev Oral Biol Med. 2004;15(3): 176-184. 

17. Spreafico RC, Gagliani M. Composite resin restorations on posterior teeth. In: Roulet JF, Degrange M. Adhesion: The Silent Revolution in Dentistry. Chicago: Quintessence; 2000:253-276. 

18. Tjan AH, Bergh BH, Lidner C. Effect of various incremental techniques on the marginal adaptation of class II composite resin restorations. J Prosthet Dent. 1992;67(1):62-66. 

19. Weaver WS, Blank LW, Pelleu GB Jr. A visible light activated resin cured through tooth structure. Gen Dent. 1988;36(3):236-237.

20. Liebenberg WH. Successive cusp build-up: an improved placement technique for posterior direct resin restorations. J Can Dent Assoc. 1996;62(6):501-507.

21. Lutz F, Krejci I, Luescher B, et al. Improved proximal margin adaptation of Class II composite resin restorations by use of light-reflecting wedges. Quintessence Int. 1986;17(10):659-664.

22. Losche GM. Marginal adaptation of Class II composite resin fillings: guided polymerization vs reduced light intensity. J Adhes Dent. 1999;1(1):31-39.

23. Bichacho N. The centripetal build-up for composite resin posterior restorations. Pract Periodontics Aesthet Dent. 1994; 6(3):17-23.

24. Belvedere PC. Contemporary posterior direct composites using state-of-the-art techniques. Dent Clin North Am. 2001;45(1):49-70.

25. Smith M, Burgess JO, Xu X. Microleakage of Class I composite resin restorations cured with three methods [abstract]. J Dent Res. 2001;80:197.

26. Manuel P, Ripps A, Burgess JO, et al. Microleakage of Class I composite resin restorations cured with four methods [abstract]. J Dent Res. 2001;80:197.

27. Schmalz G, Preiss A, Arenholt-Bindslev D. Bisphenol-A content of resin monomers and related degradation products. Clin Oral Investig. 1999;3(3): 114-119.

28. Eakle WS, Ito RK. Effect of insertion technique on microleakage in mesio-occlusodistal composite resin restorations. Quintessence Int. 1990;21(5): 369-374.

29. Crim GA, Chapman KW. Effect of placement techniques on microleakage of a dentin-bonded composite resin. Quintessence Int. 1986;17(1):21-24.

30. Coli P, Brannstrom M. The marginal adaptation of four different bonding agents in Class II composite resin restorations applied in bulk or in two increments. Quintessence Int. 1993;24(8):583-591.

31. Versluis A, Douglas WH, Cross M, et al. Does an incremental filling technique reduce polymerization shrinkage stresses? J Dent Res. 1996;75(3):871-878.

32. Goracci G, Mori G, de Martinis L. Curing light intensity and marginal leakage of resin composite restorations. Quintessence Int. 1996;27:355-362.

33. Walker RS, Burgess JO. Microleakage of Class V restorations with different visible-light curing methods. J Dent Res. 1999;78:155.

34. Walker R, Burgess JO. Microleakage of Class V restorations with different curing methods. J Dent Res. 2000;79:45.

35. Kanca J 3rd, Suh BI. Pulse activation: reducing resin-based composite contraction stresses at the enamel cavosurface margins. Am J Dent. 1999;12(3):107-112.

36. Bernardo MF, Martin MD, Johnson GH, et al. Clinical evaluation of composite restorations polymerized by two different methods. Two year results [abstract]. J Dent Res. 2002;81: A-81.

37. Brackett WW, Covey DA, St. Germain HA Jr. One-year clinical performance of a self-etching adhesive in class V resin composites cured by two methods. Oper Dent. 2002;27(3):218-222.

38. Oberlander H, Friedl KH, Schmalz G, et al. Clinical performance of polyacid-modified resin restorations using “softstart-polymerization.” Clin Oral Investig. 1999;3(2): 55-61.

39. Feilzer AJ, de Gee AJ, Davidson CL. Relaxation of polymerization contraction shear stress by hygroscopic expansion. J Dent Res. 1990;69(1):36-39.

40. Feilzer AJ, Dauvillier BS. Effect of TEGDMA/BisGMA ratio on stress development and viscoelastic properties of experimental two-paste composites. J Dent Res. 2003;82(10):824-828.

Deniz Cakir, DDS, MS
Department of Prosthodontics
University of Alabama at Birmingham
School of Dentistry
Birmingham, Alabama

Robert Sergent, DMD
Comprehensive Dentistry Department
Louisiana State University
School of Dentistry
New Orleans, Louisiana

John O. Burgess, DDS, MS
Assistant Dean for Clinical Research
Department of Prosthodontics
University of Alabama at Birmingham
School of Dentistry
Birmingham, Alabama

Share this: