studiu dezinfectie amprente

Embed Size (px)

DESCRIPTION

bla

Citation preview

  • d e n t a l m a t e r i a l s 2 8 ( 2 0 1 2 ) 749755

    Available online at www.sciencedirect.com

    jo u rn al hom epa ge : www.int l .e lsev ierhea l th .com/ journa ls /dema

    The e on the dimensionalstabil on

    Danish MaDepartment stry UThe London School of Medicine and Dentistry, Institute of Dentistry, London, United Kingdom

    a r t i c l e i n f o

    Article histor

    Received 13

    Received in

    20 Novembe

    Accepted 13

    Keywords:

    Cross conta

    Immersion

    Alginates

    Shrinkage

    Expansion

    Diffusion

    Disinfectan

    a b s t r a c t

    1. Int

    Disinfectioavoid crossmicro-orgafrom blood

    CorresponUniversity o

    E-mail a1 Faculty o

    0109-5641/$doi:10.1016/y:

    January 2011

    revised form

    r 2011

    March 2012

    mination

    disinfection

    ts

    Objectives. Dimensional changes occur in set dental alginate impression materials when

    immersed in disinfecting solutions. In this contribution the dimensional changes of two algi-

    nates in two disinfecting solutions, and for two specimen thicknesses, have been studied.

    The results were analyzed theoretically.

    Methods. The dimensional changes of two commercial alginates (Blueprint Cremix and

    Hydrogum), have been measured, in distilled water and two disinfecting solutions (Perform

    ID/sodium hypochlorite), using a traveling microscope, at 5 min intervals over a period of 1 h.

    Samples of simple geometry have been studied, namely rectangular strips with thicknesses

    of 1.5 and 3 mm, respectively.

    Results. In all cases, both alginates continuously shrank with time, in the three immersion

    liquids, over the hour of measurement, indicating transfer of water from the alginate into

    the external water or disinfecting solution. The t1/2 shrinkage plots were generally linear,

    but with an intercept on the t1/2 axis, indicating the possibility of an initial expansion at

    very short times. In most cases, the ratios of slopes for both thicknesses were 1.331.54, in

    contrast to the theoretical value of 2. Perform ID however gave anomalous results for the

    1.5 mm thick samples. At 10 min their shrinkage was 1.341.72%, compared with 0.42% to0.67% in the other two media.

    Signicance. The effects of thickness observed were not in accord with simple Fickian theory

    because of the various ions diffusing into and out of the alginate. Moreover, the water content

    of the alginate decreased consequent on the cross-linking process.

    2012 Academy of Dental Materials. Published by Elsevier Ltd. All rights reserved.

    roduction

    n of dental impression materials is necessary to contamination consequent on the presence ofnisms embedded on/in the impression materials

    and saliva [1]. The actual micro-organisms found

    ding author at: Dental Physical Sciences (Oral Growth and Development), Francis Bancroft Building (2nd Floor), Queen Mary,f London, Mile End Road, London E1 4NS, United Kingdom. Tel.: +44 020 7882 5982; fax: +44 020 7882 7089.ddress: [email protected] (M.P. Patel).f Dentistry, SEGi UniversityCollege, Malaysia.

    have been reported, for example, by Jennings and Sama-ranayake [2], and Look et al. [3].

    There are generally two disinfecting methods available,namely the use of aerosols or immersion in disinfecting solu-tions [4]. Clearly, either method should not compromise thedimensional stability of the impression material. With algi-nate impression materials, dimensional changes are known

    see front matter 2012 Academy of Dental Materials. Published by Elsevier Ltd. All rights reserved.j.dental.2012.03.013ffect of disinfecting solutions ity of dental alginate impressi

    Muzaffar1, Michael Braden, Sandra Parker, of Oral Growth and Development, Dental Physical Sciences in Denti materials

    ngala P. Patel

    nit, Queen Mary, University of London, Barts and

  • 750 d e n t a l m a t e r i a l s 2 8 ( 2 0 1 2 ) 749755

    to occur on immersion in water and articial saliva [5,6]. Oftenthe alginate initially swells, but then shrinks until ultimatelythe processsuggest algin water [1]

    The situnumber of aeral constitdifferencessolutions aalginates aods. Some

    Martin esodium hypas 10% andmaterials, cylindrical intervals, nline for immthe case ofinitial expatraction to

    Abour emade fromimmersed untreated impression24 h beforeno signicacasts betwe

    Bergmandisinfectanchlorhexidbenzalkoni0.5% chlormaterials. Dsteel blocksdimensionmaterials. Tsional chan

    Alginateiodophor, p[10], and thwere exammaster casto simulatestipulated casts. Dimmicroscopecolloid matcompared smoother t

    Peutzfelof disinfecwater wasthe samplhumid envThe resultscontainingage, attribu

    signicant dimensional changes were observed with otherimmersion systems, although it was noted that Blueprint

    ted eanotater

    es w in ainfe

    abotion ds ue msiono invickn

    es duce, t

    inved disity oo difasurture

    inve plottablis

    anald the

    Ma

    Ma

    o corint, enste, Schwdediums adn. Itals, ng pl, Mit was a c

    Me

    Saming w

    10 respknesrs in

    ml/7 intolate equilibrates with a net shrinkage. Other sourcesinates swell continuously with time, on immersion.ation is currently confused because there are alginates commercially available; although the gen-uents will be similar, there are likely to be detailed. Furthermore there are a number of disinfectingvailable. The various studies reported use differentnd disinfecting solutions, and differing test meth-of the studies are reported here.t al. [7] studied three disinfecting solutions, 5.25%ochlorite, Perform ID and Sterilox, the latter both

    concentrated solutions, with a range of impressionincluding one alginate. Changes in dimensions ofsamples (4 mm high) were measured at three timeamely after setting, after the manufactures guide-ersion disinfection, and 24 h after immersion. In

    the alginate, some of the solutions exhibited annsion up to 10 min, and then subsequently con-give an overall net contraction.t al. [8] studied the dimensional stability of casts

    an alginate impression material that had beenin undiluted sodium hypochlorite, for 30 min;impressions were used as controls. Casts of the

    materials were poured after 30 min, and stored for being measured. The results of this study indicatednt differences in the dimensional stability of theen controls and disinfected samples.

    et al. [9] investigated the effects of sixts (2% gluteraldehyde, 24% propanolol, 0.5%ine, 5.0% sodium salt of sulfanochloramide, 1.0%um chloride and a mixture of phenyl phenol andcresol), on four proprietary alginate impressionistilled water was used as the control. Six stainless, indented with three lines, were used to measure

    al stability and surface detail of the set alginatehe results of this study showed signicant dimen-ges, but no signicant changes in surface detail.

    impressions have also been immersed inhenol gluteraldehyde, and glyoxal gluteraldehydee effects on surface detail and dimensional changesined. This study used a mandibular arch for thet. The set impressions were rinsed in water for 10 s,

    clinical practice, immersed in the disinfectant, asby the manufacturers, and then poured to formensional changes were followed by a measuring. It was concluded from the results that hydro-erials showed no signicant changes in accuracywith controls. However, disinfected samples werehan the corresponding controls.dt and Asmussen [11] investigated three typestants with three commercial alginates. Deionized

    used as a control. The immersion time for alles was 1 h, after which they were stored in aironment for 24 h, and then casts were made.

    contra-indicated the use of immersion systems 70% ethanol, due to the resultant excessive shrink-ted to dehydration of the alginates. Otherwise no

    exhibiIn

    sion mchang10 minthe dis

    Theindicamethoto havdimenAlso n(e.g. thchang

    Hen

    (1) Toanbiltwmena

    (2) Tobyest

    (3) Toan

    2.

    2.1.

    The tw(BluepLiechtID (PIDof a po15 g soder wasolutiomateriinfecti(NaOCreagenused a

    2.2.

    2.2.1.Model60 mm3 mm,of thicfacturewas 15ducedglass pxpansion.her study by Jones et al. [12], alginate impres-ials showed clinically unacceptable dimensionalhen impressions were immersed for more than

    2% gluteraldehyde solution. It was also noted thatcted impressions exhibited discoloration.ve list of studies is not exhaustive, but gives anof the differing results obtained with the varioussed. It is noteworthy that no investigators seemedonitored dimensional changes with time to getal change proles (Minagi et al. [13] is an exception).estigation has studied the effect of the dimensionsess) of alginate specimens on their dimensionalring immersion in solutions.he aims of this study were:

    stigate the effects of two disinfecting solutions,tilled water as a control, on the dimensional sta-f two commercial alginates made from samples offerent thicknesses. The dimensional changes wereed as a function of time, up to 1 h, to determine theof dimensional change proles.stigate the effect of specimen size (e.g. thickness)ing the results (% linear shrinkages) against t1/2 toh the role of diffusion in the material/specimen.yze the data in terms of Fickian diffusion theory

    Gibbs modication of this theory.

    terials and methods

    terials

    mmercial alginates studied were Blueprint CremixDentsply, USA) and Hydrogum (Ivoclar Vivadent,in). The two disinfecting solutions were (1) Performulke and Mayr, Germany). This material consistsr containing 20 g potassium peroxomonosulfate,

    benzoate, 10 g tartaric acid. One scoop of pow-ded to 1000 ml of water to form the disinfecting

    is claimed by the manufacturer that impressionincluding alginates, will be stable during the dis-eriod of 10 min; (2) sodium hypochlorite solutionlton Laboratories, Rivadis, France); 500 ml of thiss added to 1000 ml water. Distilled water (DW) wasontrol immersion medium.

    thods

    ple preparationax molds were used, with rectangular cavities

    mm, to make samples of two thicknesses, 1.5 andectively, for the purpose of determining the effects. Each alginate was mixed according to the manu-structions (the water/alginate ratio for Hydrogum

    g, and 34 ml/14.7 g for Blueprint Cremix) and intro- the mold, standing on a glass plate. Then another

    was pressed on to the mold, squeezing out the

  • d e n t a l m a t e r i a l s 2 8 ( 2 0 1 2 ) 749755 751

    Fig. 1 DimHydrogum

    excess matfor Blueprinsample wacut away. Fprocedures

    2.2.2. MeAs soon asbottomed pwas anchoa second psample, takThen the s(30 ml).

    DimensmicroscopeDW and thobtained foon the opppin, and threading. Thto 1 h, on th

    Initiallyage againsagainst theestablish th

    2.3. Sta

    The Mann Blueprint Crite/Performand 3 mm.

    3. Re

    Fig. 1 is an showing tyin 1% NaOC

    Figs. 26samples) vthree immefor the twois to establError bars hacter of the

    Effect of thickness (1.5 and 3.0 mm) on dimensionales as a function of t1/2 Blueprint Cremix in distilled

    1 ogum

    um

    led water S1 NSm hypochlorite S2 NSm ID NS NS

    lueprint > Hydrogum.lueprint > Hydrogum.no signicant difference; probability of the null hypothesis

    nicant difference; probability of the null hypothesis 0.05.S sig

    in DW,is expa

    TabMann Cremixsults

    example of a linear shrinkage (%) versus time plot,pical error bars, for the 1.5 mm samples immersedl for the two alginates studied.

    plot percentage linear shrinkage (average of veersus t1/2 for the two alginates immersed in thersion media, respectively, each graph having plots

    thicknesses (1.5 and 3 mm). Plotting against t1/2

    ish the role of diffusion in the shrinkage process.ave been omitted to emphasize the general char-

    dimensional changes. Note with Fig. 3, Hydrogum

    Fig. 3 Effechanges as(th = regresComparison between Blueprint Cremix and (at 1 h).

    1.5 mm 3.0 mm

    NS NSct of thickness (1.5 and 3.0 mm) on dimensional a function of t1/2 Hydrogum in distilled watersion line for each thickness).

  • 752 d e n t a l m a t e r i a l s 2 8 ( 2 0 1 2 ) 749755

    Fig. 4 Effect of thickness (1.5 and 3.0 mm) on dimensionalchanges as a function of t1/2 Hydrogum in sodiumhypochlorite.

    Table 2 Comparison between sodium hypochlorite andPerform ID at 10 min (and at 1 h in parentheses).

    Material

    BlueprintHydrogum

    NS no si>0.05.S signicS1 HypocS2 HypocS3 Hypoc

    rite/Perform1.5/3 mm. Tages at 10 msion disinfobtain the

    Fig. 5 Effect of thickness (1.5 and 3 mm) on dimensionalchanges as a function of t1/2 Blueprint in sodiumhypochlorite.

    Table 3

    Alginate

    Hydrogum

    Blueprint

    Signies

    Table 4 samples

    Figure

    Fig. 2Fig. 3 Fig. 4 Fig. 5 Fig. 6 1.5 mm 3 mm

    NS (NS) S1 (S1) S2 (NS) S3 (S3)

    gnicant difference; probability of the null hypothesisant difference; probability of the null hypothesis Perform ID.hlorite < Perform ID.hlorite > Perform ID.

    ID at 10 min (and 1 h), for the two thicknessesable 3 summarizes the percentage linear shrink-in (manufacturer recommended time for immer-

    ection), and 1 h (the period of measurement to nature of the shrinkage proles), for the two

    Fig. 6 Effect of thickness (1.5 and 3 mm) on dimensionalchanges as a function of t1/2 Hydrogum in Perform ID.

    Percentage linear shrinkage values at 10 min (and 1 h in brackets).

    Thickness (mm) Distilled water (%) Na hypochlorite (%) Perform ID (%)

    1.5 0.42* (0.2) 0.25 (1.6) 1.72 (3)3.0 0.20 (0.7) 0.42 (1.3) 0.06 (0.5)1.5 0.66 (2.0) 0.67 (2.4) 1.34 (2.3)3.0 0.32 (1.3) 0.42 (1.7) 0.12 (0.6)

    expansion.

    Analysis of graphical data giving the slope, slope ratio and intercept on t1/2 axis, for the 1.5 and 3 mm alginate(derived from % linear shrinkage versus t1/2 plots).

    Material Slope Intercept on t1/2 axis (min1/2)

    1.5 mm 3.0 mm Ratio, S 1.5 mm 3 mm

    Blueprint (DW) 0.33 0.23 1.44 0.90 1.94Hydrogum (DW) 0.14 0.10 1.40 6.40 0.86Hydrogum (NaOCl) 0.20 0.15 1.33 1.70 1.90Blueprint (NaOCl) 0.40 0.26 1.54 1.21 1.21Hydrogum (PID) 0.50 0.10 5.00 0.00 1.92

  • d e n t a l m a t e r i a l s 2 8 ( 2 0 1 2 ) 749755 753

    alginates (of two thicknesses) in the three media. Table 4 sum-marizes the slopes and intercepts of the data in Figs. 25, thatis the gurHydrogum following e

    L

    L(%) = S(

    where L/Lt01/2 = inter

    4. Di

    Fig. 1, a typtime plot, Hydrogum the statisticno signicaof Hydroguples immer1 h. From tences betw(Table 2), anID for bothtant howevinfra).

    The percof time, in and two sanot always in DW (Figthan the 3 min all the o1.5 mm samimum, befothe 1.5 mm3.0 mm samdue to the st1/2 3 minnoted prevNallamuthuis similar inportion, fothere are in

    The ratisamples, atTable 3; fo2.1 for Hydand DW, reswith both ales (see fovalues of thWith the 1.t1/2 axis, suexpansion.much less.ferent at 1for Hydrogbe a distin

    that a constituent of Perform ID was having a powerful effect,possibly on the structure of the set alginate. The constituent

    ossi to coi-viram Fig4) inalginear

    demthickon thage)

    M(

    M iient,as as

    thisq. (2

    (

    D

    , for

    2

    S1.5f theope ier, inenero ploere bD, coasesomaled tor to aorth d alg

    the tain tess oo thodicaass u

    furt con

    duceincorensiocatio:

    c

    T

    xes where there is a major linear region (Fig. 6, forin Perform ID, has markedly different proles). Thequation applies to these linear regions:

    t1/2 t1/20 ) (1)

    (%) = % shrinkage, S = slope of the linear region, andcept of line with the t1/2 axis.

    scussion

    ical example of the percentage shrinkage versusshows evidence that Blueprint shrank more thanover the duration of the experiment (1 h). However,al analysis data in Table 1 indicates that there werent differences between the percentage shrinkagesm and Blueprint, except with the 1.5 mm sam-sed in distilled water and sodium hypochlorite, athe practical viewpoint, there are statistical differ-een thicknesses in disinfecting solutions at 10 mind the effect of thickness is much greater in Perform

    materials (for example see Fig. 6). More impor-er are the differences between release proles (vide

    entage linear shrinkage graphs against square rootFigs. 26, representing the three immersion mediample thicknesses, generally gave linear plots, butgoing through the origin. In the case of Hydrogum. 3), the 1.5 mm samples appeared to shrink less

    m thick samples, the converse of what was seenther Figs. 1, 2, 4, 5 and 6. This was due to theples initially expanding, and going through a max-re reverting to a linear contraction. Consequently,

    samples have a higher shrinkage slope than theples, but actually lower shrinkage values (Table 3),amples initially expanding; shrinkage began after

    1/2, at a value of 0.5%. Such behavior has beeniously for some alginates, immersed in water, by

    [5,6]. It should be noted that the nature of the plots all cases, i.e. there is a usually a transitory initial

    llowed by a continuous steady shrinkage. Clearlydividual differences between the various systems.os of the shrinkage values of the 1.5 mm and 3 mm

    t = 10 min (t1/2 = 3.20) (calculated from the data inr example 0.25/0.42 in NaOCl = 0.6), were 0.6 androgum, and 1.6 and 2.1 for Blueprint, in NaOClpectively. There were however marked differenceslginates in PID, with respect to (i) the shrinkage pro-r example Fig. 6), and (ii) the ratios of the shrinkagee two sample thicknesses, at t = 10 min (t1/2 = 3.20).

    5 mm samples the plot was curved concavely to theggesting there might have been an initial period of

    Clearly, shrinkage of the 3 mm samples was very Most importantly, the shrinkages were very dif-0 min (Table 3), there being a factor of 29-foldum, and 11.2 for Blueprint. This would seem toct disadvantage of Perform ID, and also indicates

    could paddedan ant

    Fro(Table these that linbelow ferent diffusi(shrink

    Mt = 2

    wherecoefc

    It wfollowfrom E

    S = 4M

    Hence

    S1.5S3.0

    =

    whereslope oples slHowevwere gthe twDW, wform Imost cthe antors nein ordeIt is wstudiebut forand agthicknacter tThis inand m

    Onechangewill reneeds an extmodi[16] by

    F = Rbly be potassium peroxomonosulfate, presumablynfer a low pH 3.5. Low pH values are known to havel effect [14].s. 26, the ratios of the two slopes were calculated

    order to see if simple diffusion theory applied toates immersed in the three media (it is assumedshrinkage is proportional to mass loss). The theoryonstrates that the ratio of the slopes of the two dif-ness plots is theoretically 2. It is based on Fickianeory [15]; the linear part of an uptake/desorption

    plot is given by:

    Dt

    L2

    )1/2(2)

    s the equilibrium water uptake/loss, D the diffusion t the time, and 2L the thickness.sumed the dimensional change plots will generally

    equation. The slope (S) of the Mt versus t1/2 plot,), can be written as:

    )1/2 ( 12L

    )(3)

    simple diffusion theory:

    (4)

    is the slope of the 1.5 mm samples and S3.0 the 3 mm samples. Hence the ratio of the 1.5 mm sam-s predicted to be twice that of the 3.0 mm samples.

    this contribution (Table 4), although the t1/2 plotsally linear in Figs. 26, the ratios of the slopes ofts, for the two alginates immersed in NaOCl andetween 1.33 and 1.54 and 5 for Hydrogum in Per-mpared with the theoretical value of 2. Clearly in, simple Fickian Diffusion Theory cannot explainous results seen in Figs. 26. Therefore further fac-

    be taken into consideration, as highlighted below,ttempt to explain the anomalous results obtained.noting in this context, that Nallamuthu et al. [6]inate weight loss in air, which gave linear t1/2 plots,wo thicknesses studied, the slope ratios were 1.2,he thickness ratios were 2:1. Hence, the effects ofn dimensional changes in air were similar in char-se studied here, in DW and disinfecting solutions.tes that the relationship between linear shrinkageptake is not a simple one.her factor that should be considered is the entropysequent on cross-linking as the material sets. This

    the equilibrium water content of the alginate, andporation into the theoretical approach, as follows:n of the Fickian diffusion theory is given by Gibbsn of the diffusion theory, which replaces Ficks Law

    (5)

  • 754 d e n t a l m a t e r i a l s 2 8 ( 2 0 1 2 ) 749755

    where is the chemical potential, F is the ux and c, the con-centration at point x, measured for example, from one of thesurfaces oris measureof the voluThis has belyze the wawater solub

    The Florcurrent sys

    = RT[ln(

    where is stant, T ispolymer (asolvent par2 the denbetween crwithin the for a givencorrespond

    It must known whesystem; furcontain a co

    be a useful omitted, thmer and liqof v2, and tions (i.e. this cross-lin

    The casin that on m(partially) ilater. Whenbecomes pis compatibshrinkage).essarily gov

    It is preobserved fosequent onthis to be demonstranate materon immersdue to the However oreversing tfuses out wshrinkage ofact that wthe entropytion of watis expelled

    The impFickian Thesimple Fick

    observed (i.e. experimental values of slope ratios are 1.331.54,and not 2).

    comlori

    int innd wtial ees (Fted n coith

    n PIDay re

    inteig. 6amalar r

    mmible as lin

    (1.5 kiange d

    solunownsing rthe, thebasiiatio

    l N

    the + + lied

    mo

    chlorm ID

    oth mPID iferen3.5 [2re oiculans.ce,

    with, witherve he

    statsionathun inpendlinic

    intr the center plane; c will equal v1, if concentrationd as the volume of diffusant (water) as a fractionme of the whole polymer system, v2 = 1 v1 = 1 c.en studied by Muniandy and Thomas [17] to ana-ter uptake of hydrophobic elastomers containingle moieties.y Huggins Theory [18,19] is a possible model for thetem, where:

    1 v2) + v2 + v2 + (2V1v21/3)

    Mc

    ](6)

    the chemical potential, R is the universal gas con- the temperature (K), v2 the volume fraction oflginate), v1 the volume fraction of the liquid, theameter, V1 the molar volume of the solvent (water),sity of the polymer, and Mc the molecular weightoss-links. In Eq. (6), considering the four termssquare brackets, terms 1, 2 and 4 give the entropy

    v2 value; the third term (containing ) gives theing enthalpy (H) value.be stressed Eq. (6) is being used as a guide. It is notther it is quantitatively valid for the alginate-waterthermore the alginate formulations studied herensiderable quantity of ller. Nevertheless it should

    guide. If the nal term in Eq. (5) (i.e. [2V1v213]/Mc) is

    e equation is for the interaction between the poly-uid. In this case if < 0.5, is negative for all valuesthe polymer and liquid are miscible in all propor-e polymer is soluble in the liquid). If the polymer

    ked, it will swell until = 0.e of the alginate impression material is interestingixing the powder and water, the alginate dissolves

    n water to give a paste, which then sets (cross-links) cross-linked, Eq. (6) applies, and it is suggested, ositive (i.e. the system now has more water thanle, and it is expelled from the sample, resulting in

    This process is diffusion controlled, but not nec-erned by simple Fickian equations.

    sumed in this work that the dimensional changesr the two commercial alginate materials are con-

    weight changes; Nallamuthu et al. [6] have shownthe case with water and articial saliva, who

    ted an initial expansion in water with some algi-ials, followed by shrinkage. They suggested thation of the alginate in water, there is some uptakepresence of soluble ions in the set alginate matrix.n the entry of water, these ions diffuse out, thushe chemical potential, so that water then dif-ith consequent shrinkage. In this contribution, then immersion in water has been explained by thehen the alginate sets (cross-links), this decreases

    of mixing, resulting in the equilibrium concentra-er in the alginate decreasing, so that excess water

    from the alginate, albeit diffusion controlled.lication for the current work is clear; standardory does not necessarily apply, and explains whyian prediction of thickness dependence of t1/2 is not

    On hypochBlueprto t1/2 athe inisamplattribusolutioture. Wthose iThis mand an

    In Fwas drA simithe 1.5a possplot wslopesple Ficis a larfectingit is kincreabe a futrolyteof the dissoc

    NaOC1 and (1 ) multip

    The

    HypoPerfo

    If bity of the difPID is structuin partsolutio

    Henuouslyperiodto obssion. Tto theimmerNallamafter aThe dehave chas anparing the results of the two alginates in sodiumte (Figs. 4 and 5), both of these were similar to

    water (Fig. 2), in terms of being linear with respectith an intercept on the t1/2 axis. But for Hydrogum,xpansion observed in water, for the 1.5 mm thickig. 3), was not present in NaOCl (Fig. 4). This isto the effect of the osmolarity of the hypochloriteunteracting the effect of ions in the alginate struc-the other samples, both 1.5 and 3 mm, excluding, the straight line had an intercept on the t1/2 axis.ect a transient expansion followed by shrinkage,rcept on the t1/2 axis, before the rst measurement., it is clear that the behavior of Hydrogum in PIDtically different from that of it in DW and NaOCl.esult was observed with Blueprint in PID too. For

    samples, the plot was initially curved, leading intolinear region, whereas for the 3 mm samples theear, with an intercept on t1/2 axis. The ratios of themm/3 mm) were 5 for Hydrogum (cf. 2.0 for sim-

    theory). This raises the question as to why thereifference in characteristics between the two disin-tions. Their osmolarities is one possibility, because

    that water is attracted to solutions, the uptakewith increasing osmolarity, and osmolarity wouldr contribution to shrinkage. In the case of an elec-

    osmolarity of it is inuenced by the dissociationc molecule. NaOCl is an example: if the degree ofn is ,

    a+ + ClO

    total number of moieties in solution is = 1 + , the factor by which the molarity must be

    to give the osmolarity in this case.larities of the disinfecting solutions are:

    rite: 0.044 moles per liter: 0.110 moles per liter

    aterials have the same value, then the Osmolar-s 2.5 times that of NaOCl, and may account force in shrinkage behavior. Furthermore, the pH of0]; this implies the possibility of pH affecting the

    f the alginate. Clearly this area merits further studyr measurement of the osmolarities of disinfecting

    generally all the alginate samples shrank contin- time, in the three immersion media, over the 1 h

    no signs of leveling out. This emphasizes the needthe manufacturers recommended time of immer-ndings in the current work are in marked contrastement of Anusavice [1] that alginates expand on

    in water, but generally agree with the work of [5,6], who showed that alginates generally shrink

    itial period of expansion on immersion in water.ence of shrinkage rates on sample thickness could

    al implications. An impression of a dentate mouthicate structure with some parts of the impression

  • d e n t a l m a t e r i a l s 2 8 ( 2 0 1 2 ) 749755 755

    being thinner than others. Such regions could shrink morerapidly than adjoining thicker regions, indicating the possibil-ity of distortion. This will depend on the ratios of slopes ofshrinkage t1/2 plots. Manufacturers guidelines for the time ofimmersion disinfection should be adhered to.

    5. Conclusions

    Immersion of the two alginates in DW and two disinfectingsolutions generally exhibited shrinkage, albeit sometimeswith an initial expansion.

    The DW and NaOCl solution behaved similarly, exhibitinglinear t1/2 plots.

    The behavior of both alginates in PID was more complex,indicating a different mechanism for the interaction of thedisinfecting solution with the alginate.

    The effect of thickness was less than that predicted by sim-ple diffusion theory in the case of DW and NaOCl.

    Shrinkage readings at 10 min showed much larger differ-ences between 1.5 and 3 mm thickness for PID comparedwith NaO

    The dataTheory fo

    It is suggmay be dA further

    r e f e r e n

    [1] AnusavW.B. Sa

    [2] Jenningmicro-disinfe3827.

    [3] Look JOfrom dProsthe

    [4] Silva SMtechniqmateri2449.

    [5] Nallamuthu NA. Hydrocolloids and cross-infection. PhDthesis. University of London; 2007.

    [6] Nallamuthu N, Braden M, Patel MP. Dimensional changes ofalginate impression materials. J Mater Sci: Mater Med2006;17:120510.

    [7] Martin N, Martin MV, Jedynaliewicz NM. The dimensionalstability of dental impression materials following immersionin disinfecting solutions. J Dent Res 2007;23:7608.

    [8] Abour MAB, ONeilly PJR, Satchel DJ, Pearson GJ. Physicalproperties of casts prepared from disinfected alginate. Eur JProsthodont Rest Dent 1996;4(4):8791.

    [9] Bergman BO, Bergman M, Olsonn S. Alginate impressionmaterials, dimensional stability and surface detailsharpness following treatment with disinfecting solutions.Swedish Dent J 1985;9:25562.

    [10] Johnson GH, Chellis KD, Gordon GE, Lepe X. Dimensionalstability and detail reproduction of irreversible hydrocolloidand elastomeric impressions disinfected by immersions. JProsthet Dent 1998;79:44653.

    [11] Peutzfeldt A, Asmussen E. Effect of disinfecting solutions onthe accuracy of alginate and elastomeric impressions. ScandJ Dent Res 1989;97(5):4705.

    [12] Jones ML, Newcombe RG, Barry G, Bellis H, Bottomley JA. Aex pginatgime88;15inagiagawaterimun86;56mmoh ed.ank Jess; 1bbs Jerm

    unianlymel. 97.gineory Pys 19gginys 19rformrmaCl and DW. was analyzed theoretically using the Flory Hugginsr the interaction of polymers with liquids.ested that the different behavior exhibited by PIDue to its higher osmolarity compared with NaOCl.

    factor may be its low pH (3.5).

    c e s

    ice KJ. Phillips science of dental materials. 11th ed.unders Company; 2003.s KJ, Samaranayake LP. The persistence of

    organisms on impression materials followingction. Int J Prosthodont 1991;4:

    , Clay DJ, Gong K, Messer HH. Preliminary resultsisinfection of irreversible hydrocolloid impressions. Jt Dent 1990;63:7017.LM, Salvador MCG. Effect of the disinfection

    ue on the dimensional stability of dental impressionals. J Appl Oral Sci 2004;12:

    realre19

    [13] MAkmim19

    [14] Di4t

    [15] CrPr

    [16] GiTh

    [17] MPovoEn

    [18] FlPh

    [19] HuPh

    [20] PeGelotter investigation into the dimensional stability ofe impressions following disinfection by varyings employing 2.2% Gluteraldehyde. Br J Orthod(3):18592.

    S, Fuskikima K, Maeda N, Satomu K, Ohkawa S,a Y, et al. Disinfection method for impression

    als; freedom from fear of Hepititus B and acquirede decient syndrome. J Prosthet Dent(4):4514.ck NJ, Primrose SB. Introduction to modern virology.

    Blackwell Science; 1994.. Mathematics of diffusion. 2nd ed. Oxford University979, 48.W. The collected works of J. Willard Gibbs.odynamics, vol. 1. NY: Longmans, Green, & Co; 1928.dy K, Thomas AG. Water absorption in rubbers. In:rs in marine environments. Conference 2, paper 13,

    London: Transactions of the Institute of Marineers; 1984. p. 8794.J. Thermodynamics of high polymer solutions. J Chem42;10:5161.s ML. Solutions of long chain compounds. J Chem41;9:440.-ID materials safety data sheet. Norderstedt,

    ny: Producer/Supplier: Schlke & Mayr GmbH; 2010.

    The effect of disinfecting solutions on the dimensional stability of dental alginate impression materials1 Introduction2 Materials and methods2.1 Materials2.2 Methods2.2.1 Sample preparation2.2.2 Measuring dimensional changes

    2.3 Statistical analysis

    3 Results4 Discussion5 ConclusionsReferences