Mechanisms of laser induced modification of lead and barium vanadate glasses

Mechanisms of laser induced modification of lead and bariumvanadate glasses

Ruhil Dongol, Landon Tweeton, Cory Faris, Steve Feller, and Mario Affatigatoa�

Department of Physics, Coe College, 1220 First Avenue NE, Cedar Rapids, Iowa 52402, USA

�Received 14 July 2010; accepted 18 October 2010; published online 11 January 2011�

We report on our investigations on the mechanisms for structural and morphological change in leadand barium vanadate glasses modified by 785 nm laser irradiation. The fundamental process isthermal in nature, leading to phase decomposition, phase changes, and mass transport in the centerof the irradiated region, as well as the formation of lead-or barium-rich debris zones. Crystallizationis also a consequence of the exposure to the laser light. We further report on preirradiationexperiments in which low-power exposure above a determined energy threshold results in structuralchanges that bridge the gap between the amorphous state and the ultimate crystalline arrangement.Finally, we discuss the application of the laser irradiation in the making of vanadate microspheres.© 2011 American Institute of Physics. �doi:10.1063/1.3517081�

I. INTRODUCTION

Recently we reported on laser-induced modification oflead vanadate and alkali borate glasses using laserirradiation.1,2 These alterations occurred at relatively lowpowers, ranging from 15 to 200 mW, and they resulted in avariety of changes: crystallization �at powers greater than 38mW for the lead vanadates�, accompanied by ejecta and de-bris deposition; refractive index variation �at powers lessthan 38 mW�. Several questions remain outstanding: the ex-act structural rearrangements, the effects of lower and higherlevels of power in the laser irradiation, and the effect of afirst irradiation on subsequent laser light exposures. In thispaper we discuss a possible mechanism for the structuralchanges, and also report on our analysis of the morphologyof the irradiated areas, mainly for glasses at the metavana-date composition.

The structure of lead metavanadate crystals is quite rel-evant to this work. Previous work has established that thereare three phases, of which only one is stable. The stablephase �I� is an orthorhombic structure,3 with distorted edge-sharing vanadium octahedra. The two metastable phases in-clude a monoclinic variant,4 phase �II�, with double chains ofedge-sharing square pyramids, and an orthorhombicmodification,5 phase �III�, made of zigzag chains of corner-sharing tetrahedra.

In the case of barium metavanadate crystals,6 the struc-ture has been determined to be orthorhombic, consisting ofchains of vertex-shared tetrahedra joined by interstitial Baatoms. It is analogous to the stable lead metavanadate phase�I� as well as the orthorhombic metastable modification �III�.We note that the orthorhombic structures of lead and bariummetavanadates are quite similar6 to the network present inthe glass, also composed of zigzag chains of tetrahedra.

II. EXPERIMENTAL PROCEDURES

A. Glassmaking procedure and thermalcharacterization

Glasses were prepared according to the stoichiometries

xPbO + �1 − x�V2O5 → xPbO �1 − x�V2O5, �1�

xBaCO3 + �1 − x�V2O5 → xBaO �1 − x�V2O5 + xCO2.

�2�

The starting materials for the glasses were lead �II� oxide�Aldrich, 99.9%�, barium carbonate �Aldrich, 99+%�, andvanadium �V� oxide �Aldrich, 99.6%�. Lead vanadate glasseswith stoichiometries ranging from x=0.1 to x=0.6 were pre-pared, while the barium vanadates ranged from x=0.1 to0.66. The powders were thoroughly mixed in a platinum cru-cible and heated to 900–1000 °C in an electric muffle fur-nace, depending on the composition. After 10 min the meltwas removed from the furnace, let cool, and a weight mea-surement taken to ensure composition. All weight losseswere within the expected range ��0.1 g from calculated�.The sample was then placed back in the furnace for a further10 min at the same initial temperature and subsequentlycooled to the glassy state by splat quenching between twostainless steel plates. The cooling rate was approximately10 000 °C /s. Sample thickness was 1.5�0.3 mm.

Glass transition �Tg� and crystallization temperatures�Tx� were measured with a Perkin-Elmer DSC-7. Thermalscans were performed at 40 °C /min and the onset definitionof the Tg and Tx was used. The estimate for experimentalerror in calculating the Tg is �5 °C. To determine the spe-cific heat, we subtracted the baseline heat flow curve fromthat of the sample and encapsulating pan. The value of thedifference �Y at a point in the stable range 60–90 °C wastaken to calculate the specific heat Cp near room temperature.By running a few references such as indium and gold weestimated the specific heat error to be on the order of �2%.

Reference crystals were grown from the glasses by de-vitrification. The 50PbO•50V2O5 glass was held at 420 °Ca�Electronic mail: [email protected].

JOURNAL OF APPLIED PHYSICS 109, 013521 �2011�

0021-8979/2011/109�1�/013521/7/$30.00 © 2011 American Institute of Physics109, 013521-1

Author complimentary copy. Redistribution subject to AIP license or copyright, see http://jap.aip.org/jap/copyright.jsp

http://dx.doi.org/10.1063/1.3517081

http://dx.doi.org/10.1063/1.3517081

http://dx.doi.org/10.1063/1.3517081

for 48 h, and the subsequent crystal was checked with x-raydiffractometry �Bruker Discover D8, Cu K� wavelength�. Wethen used the crystals as references in our Raman measure-ments. We were unable to make pure crystals of the meta-stable phases using faster plate cooling on the melt.

B. Laser irradiation procedure

The Raman system used for the laser irradiation �andanalysis� was a JASCO NRS-3100 microRaman spectropho-tometer. A Torsana Starbright 785 nm SLM diode laser wasused to irradiate the sample at various power settings �16, 41,84, and 160 mW� for durations ranging from 5–20 s. Spectrawere then acquired at 1.7 mW at various sites in the resultingaffected area; this power level was not enough to cause sur-face alterations. The laser beam spot size was circular withapproximately 1 �m radius.

C. Atomic force microscopy

Representative irradiated areas were also imaged using aVeeco Multimode Scanning Probe Microscope. The tappingatomic force mode was used, with a scanner mapping out25�25 �m2 areas with a maximum z-range of 5.1 �m.The tips used were phosphorus-doped Si, with a resonancefrequency in the range of 280 kHz. Other than the standardplane-fitting algorithm, no other filters were applied.

D. Scanning electron microscopy and EDX

Pristine and irradiated areas were studied using a TES-CAN Vega II Scanning Electron microscope running at10–20 keV. Secondary-and backscattered electron detectorconfiguration were both used. The elemental analysis wascarried out using an attached Oxford EDX system. Two di-mensional mapping of the elemental concentrations was alsocarried out with the Oxford software.

To check the concentration results we carried out 15 spotmeasurements of the pristine �unirradiated� region, and theaverage �mean� values were 50.4�2.9 wt % Pb,25.6�1.8 wt % V, and 24�1.2 wt % O. These results areto be compared to the expected values from the stoichiom-etry, 51.15 wt % Pb, 25.15 wt % V, and 23.7 wt % O. Theelemental concentrations appear close to what is expected,but it is not known whether any variations are due to truecompositional changes upon melting, or instrumental error.The comparisons we will be making are relative, and thusthis is not a significant concern. Based on the standard de-viations of the measurements and the instrumental error, weestimate the error in the chemical analysis for each elementto be �3%.

III. RESULTS AND DISCUSSION

We began by looking at the changes in the chemicalcomposition induced by the laser irradiation at sufficientlyhigh powers ��84 mW�. Figure 1�a� shows the electron mi-crograph of the surface morphology of one such modifiedspot on a 50PbO•50V2O5 sample, while Fig. 1�b� illustratesa cross section of a similar spot measured from atomic forcemicroscopy. We note the appearance of three concentric re-

gions: a center depression, a crystalline ridge, and a debriszone. The depression is lower than the original glass surface,while the ridge is higher. The debris zone �marked by thehalf-arc� is at the same level as the original glass, but someejecta has landed on this surface.

A. Modification mechanism

The appearance of the modified spot, as well as thecracks that often accompany it, point to a thermal mecha-nism for the modification. The details of the change in theglass structure are clarified with further examination of thecompositional changes.

EDX analysis of the center depression indicates a drasticdecrease in the concentration of lead. The 50PbO•50V2O5

sample ought to have 51.15 stoichiometric weight percentlead, but the EDX �averaged over the center of 15 irradiationspots� indicates that only 18 wt % is present, as shown inTable I. The surrounding crystalline ridge, on the other hand,remains at the original stoichiometry, and the debris zoneappears enriched in lead. Indeed, the debris zone can attainconcentrations as high as 75 wt % Pb, but more typically inthe range of 60–70 wt % Pb. The depth of the affected regioncan be estimated from Fig. 2, a cross section of an irradiated

(a)

(b)

DebrisZone

Ridge

Center

CenterRidgeRidge

Debris

FIG. 1. �Color online� �a� Scanning electron micrograph of a modified spoton a 50PbO•50V2O5 sample. �b� Cross section of a similar spot measuredwith atomic force microscopy.

013521-2 Dongol et al. J. Appl. Phys. 109, 013521 �2011�


spot. The figure shows both the excavated center of the spot,and a hemispherical modified region approximately 60 �min depth.

The EDX data is confirmed by the use of Raman spec-troscopy. Figure 3 shows a comparison of the 1:1 lead vana-date crystal �PbO•V2O5� with the spectrum from the crystal-line ridge, and a similar comparison between the 2:1 crystal�2PbO•V2O5� and the debris zone.

We believe that the laser heating causes a decompositionof the metavanadate glass into two phases. It is known7 thatat 310 °C the glass transforms according to

a � 50PbO • 50V2O5 ——→310C

c � PbV2O5�III�

+ c � 2PbO • V2O5

+ a � V2O5, �3�

where “c” indicates a crystalline phase and “a” an amor-phous one.

The metastable �III� crystalline phase further transforms7

into the stable �I� phase at temperatures greater than 390 °C.Hence the center depression undergoes a decomposition intoa lead-rich crystal and a lead-poor amorphous liquid. Thelaser heating causes the lead-rich phase to boil off as ejecta,which land closer or farther from the center depending on theexact laser power. Its relative position is also affected by the

TABLE I. The first four columns show the measured chemical compositions, in weight percent from EDX, ofthe centers of 15 different irradiation spots on a 50PbO•V2O5 glass. The last three columns are the results �alsoin weight percent� from 15 spots in a pristine area. The bottom two rows contain the average �mean� values aswell as the standard deviations from the mean. The instrumental error is estimated to be �3%.

Spotnumber

Irradiatedvanadium

Irradiatedlead

Irradiatedoxygen

Pristinevanadium

Pristinelead

Pristineoxygen

1 52.09 6.51 41.4 23.64 53.65 22.712 43.54 20.68 35.78 19.86 59.93 20.223 55.38 1.06 43.56 26.72 48.55 24.734 33.82 36.79 29.39 25.97 49.79 24.245 53.71 3.82 42.47 26.13 49.54 24.346 54.65 2.27 43.08 26.58 48.79 24.637 36.4 32.51 31.09 26.58 48.78 24.648 46.11 16.43 37.47 26.04 49.68 24.289 53.3 4.5 42.2 26.62 48.72 24.6610 27.04 48.02 24.94 26.54 48.85 24.6111 41.61 23.88 34.51 26.45 49.01 24.5512 29.15 44.52 26.33 25.9 49.91 24.1913 43.26 21.14 35.6 25.85 49.99 24.1614 55.19 1.37 43.44 25.78 50.11 24.1115 52.16 6.39 41.45 25.66 50.31 24.03Mean 45 18 37 26 50 24Std. Dev. 10 16 6 2 3 1

FIG. 2. Scanning electron micrograph of a modified spot in a50PbO•50V2O5 sample after a cross-sectional fracture. Note the missingmass from the center region.

FIG. 3. �Color online� Comparison of Raman spectra between crystalline2PbO•V2O5 and the debris region �upper pair�, and between the crystallizedridge and the PbO•V2O5 crystal �lower pair�.



overall size of the affected spot, which also depends on thelaser power. The remaining material in the depression sur-face is amorphous and poor in lead content.

The ridge surrounding the center depression does notattain temperature high enough to cause ablation. Rather, thetransformation is from an amorphous metavanadate phase toa crystalline one. We are not able to identify this phase as thestable one �I�, or as one of the two metastable ones becauseof the small size of the region.

The ablation debris that landed in the outer peripherytook the form of small �approximately 1 �m�, lead-rich par-ticles that carpeted an arc of the region surrounding theridge, and they brought lead enrichment to the region. Thehypothesis that it was the debris causing the enrichment waschecked using an atomic force microscope �AFM�. In Fig.4�a� we show a modified spot with debris after it has beenscanned by the tip from the AFM with settings designed tocause the tip to press down quite firmly on the surface. As aconsequence, some of the debris has been removed by the

sweeping action of the tip. EDX mapping spectra �Fig. 4�b��of those “scrubbed” regions show decreased lead concentra-tions �for spectra 3, 4, 5, and 6 the lead concentrations are28.60 wt %, 32.23 wt %, 30.62 wt %, and 41.06 wt %, re-spectively�. Not many of the ejected particles were detachedby the AFM, pointing to the debris being molten before itlanded, thus leading to a strong attachment to the surface.EDX analysis of the debris �of 15 different irradiation spots�,as previously mentioned, yields excess lead contents of58�5 wt %. We did not establish whether vanadium or va-nadium oxide is also ejected, perhaps in forms that are not soeasily detected or that land much farther out.

The effect of sample composition on crystallization isnot dramatic. All of the samples, whether lead-or barium-modified, could attain some crystallization by laser irradia-tion. Up until the metavanadate composition the ease of crys-tallization increased. Table II shows that the specific heat ofthe glasses decreased with increasing modifier content. Thisdecrease leads to greater increases in the temperature of theirradiated area for the deposited energy. The optical absorp-tion is not a factor as it remains extremely high for all com-positions. The metavanadate composition also crystallizeseasily, but this may well be because it matches a crystalcomposition for both the lead and the barium vanadates. Ingeneral, most vanadate glasses have low thermal stabilities�Tg-Tx�.

B. Effect of preirradiation

A barium metavanadate sample illustrates the alterationthat occurs under a first-stage low power irradiation. Thepreirradiation results in no visible changes in the glass, withthe exception that the extreme edges of the laser light swathdo tend to crystallize. In Fig. 5 we show, in an optical mi-crograph, the differences in the modified spots that were cre-ated in both a pristine glassy region as well as a region thathad been preirradiated. Also shown in the figure is a thirdseries of modified spots made in a region that had been an-nealed at the sample’s crystallization temperature,�Tx,389 °C� for two hours. As the figure shows, the series ofspots made in the preirradiation region are significantly

(a)

(b)

FIG. 4. �Color online� �a� SEM micrograph and �b� EDX map of spectrataken from sample of 50PbO•50V2O5 swept by the AFM tip. Spectra 3, 4,5, and 6 showed missing debris from those locations, and correspondinglylower amounts of lead.

TABLE II. Specific heat measurements for a series of lead- and barium-vanadate glasses.

CompositionCp

�J kg−1 K−1�, �2%

10PbO.90V2O5 69120PbO.80V2O5 63930PbO.70V2O5 55240PbO.60V2O5 50150PbO.50V2O5 44460PbO.40V2O5 410

10BaO.90V2O5 73620BaO.80V2O5 66830BaO.70V2O5 61540BaO.60V2O5 58250BaO.50V2O5 52760BaO.40V2O5 481



larger than those made in the pristine region, a difference thatholds for all power levels. Furthermore, the annealed regionbehaves quite similarly to the pristine region, with little to noimpact on the size of the series of modified spots.

Figure 6 shows that the preirradiation does indeed alterthe structure of the glass 50BaO•50V2O5. The Raman spec-trum shows growth in the band at 860 cm−1, typicallyassigned8,9 to V–O–V “bridge” vibrations, accompanied by adecrease in the 950 cm−1 band, usually associated with V–Ostretching modes. This can be interpreted as a modificationof the amorphous network into a more ordered, crystal-likearrangement. It is most likely that the glass begins to rear-range into longer chains �leading to the increase in V–O–Vbridges�, enhancing its similarity to the orthorhombic BaVphase. Once this is accomplished, the glass is more analo-gous to the crystalline structure, and a second irradiation athigher power results in a faster and more widespread changeto a full crystal region. Other low power measurements withEDX �not reported� indicate that there are no chemical com-position changes in the irradiated areas.

In Fig. 7 the Raman spectra compare the pristine50BaO•50V2O5 glass �a�, the center of the preirradiated re-gion �b�, and the extreme edge of the preirradiated region �c�.The center of the irradiation swath shows the changes wepreviously discussed. The edge, on the other hand, crystal-lized readily. As the swath’s edge is quite close to the physi-cal edge of the glass sample, we believe that this may haveaided the crystallization through the presence of defects nearthe fractured side.

Irradiated swath

Edge Center FIG. 5. �Color online� Optical micrographs of a seriesof modified spots on a pristine swath �left�; a preirradi-ated swath �center�, and a thermally annealed swath�right�. The sample was 50BaO.50V2O5 glass.

FIG. 6. �Color online� Raman spectra of a pristine 50BaO•50V2O5 glass aswell as one that has been preirradiated by a moving laser spot.

FIG. 7. �Color online� Raman spectra of a pristine 50BaO•50V2O5 glass�a�, the center of the preirradiated swath �b�, and the outermost edge of thepreirradiation area �c�. These regions were noted in Fig. 5.



Figure 8 shows that the transformation induced by thepreirradiation has a definite energy threshold. We define theenergy deposited as the net power times the irradiation timeper unit area, which assumes 100% absorption. When theenergy deposited reaches a value of 1�0.05 �J /�m2, thesize of subsequent irradiation spots increase in diameterwhen compared to the same exposure on a pristine glass.This is true regardless of the power level used in thepreirradiation—as long as the net energy deposited attainsthe same levels.

C. Spheroidization

One application of vanadate glasses that arises from theirease of thermal modification is the ability to make micro-spheres. In Fig. 9 we show a variety of microspheres andother shapes made by relatively low power laser irradiation.The spheroidization process is very rapid, and can be carriedout by sweeping the laser beam across a fairly wide swath.Any small glass particles within the swath are spheroidized,though the laser light passes result in only partial spheres,ovoids, or other deformed shapes for larger grains.

The glass fragments that were prepared to be turned intospheres were sieved using a 20 �m sieve. This opens thepossibility of achieving nanosize spheres by simply generat-ing a milled powder that contains smaller precursor grains.

IV. CONCLUSIONS

Our study of the mechanism leading to the modificationof lead and barium vanadate glasses using 785 nm laser ir-radiation has yielded several results:

�a� At the metavanadate composition, the center of the ir-radiation area attains high temperatures and the mate-rial therein is heated to the point where phases sepa-rate. The lead-rich phase �likely chervetite, Pb2V2O7� isthen ejected from the center region.

�b� The mass transport resulting from the irradiation takesthe form of hot ejecta that land in an arc surrounding

the irradiation region. As the debris is high in lead con-tent, this causes a local enrichment of the debris zone.

�c� The concentric zone that surrounds the center spot ap-pears to crystallize into the stoichiometrically-expectedmetavanadate crystal, PbV2O6 �I�. This is the thermo-dynamically stable phase.

�d� Low power preirradiation of the glass resulted in agreater impact by later laser light exposures. The effectof secondary irradiation was the formation of largerand more significant modified spots, building on thechanges induced by the low-power primary irradiation.Raman spectra indicated a modification of the glassnetwork, interpreted as a rearrangement into longerchains which turns the glass into a structure more simi-lar to the orthorhombic BaV phase. This phenomenonhad a clear threshold as a function of the depositedenergy per unit area, approximately 1 �J /�m2.

�e� Evidence for the high temperatures attained by thevanadate glasses also come in the form of our ability tospheroidize small glass fragment by laser irradiation. Italso points to an application in which small spheres ofradii �10 �m can be manufactured by applying awide-swath laser beam.

ACKNOWLEDGMENTS

This work was supported by the National Science Foun-dation under Grant Nos. DMR-RUI-0904615, DMR-MRI-0722682, DMR-MRI-0420539, and DMR-MRI-0320861.Coe College is also thanked for its financial support.

1B. Franta, T. Williams, C. Faris, S. Feller, and M. Affatigato, Phys. Chem.Glasses 48, 357 �2007�.

2B. Franta, L. Tweeton, S. Feller, and M. Affatigato, Phys. Chem. Glasses50, 305 �2009�.

3B. D. Jordan and C. Calvo, Can. J. Chem. 52, 2701 �1974�.4G. Calestani, G. D. Andreetti, A. Montenero, and M. Bettinelli, Acta Crys-tallogr., Sect. C: Cryst. Struct. Commun. 41, 177 �1985�.

5G. Calestani, G. D. Andreetti, A. Montenero, M. Bettinelli, and J. Rebi-

FIG. 9. SEM micrograph of a few grains altered by the laser exposureswath. The glass was the 50BaO•50V2O5 sample.

60

40

20

0

Sp

ot

size

( μμm

)

1.41.21.00.80.60.40.20.0

Energy deposited (μμJ/μμm2)

94 mW

48 mW

18 mW

FIG. 8. �Color online� Dependence of the diameter of the modified spots onthe net energy/unit area deposited during the preirradiation process for a50BaO.50V2O5 glass. Note the threshold at approximately 1 �J /�m2



http://dx.doi.org/10.1139/v74-393

http://dx.doi.org/10.1107/S0108270185003225

http://dx.doi.org/10.1107/S0108270185003225

zant, Acta Crystallogr., Sect. C: Cryst. Struct. Commun. 41, 179 �1985�.6T. Yao, Y. Oka, and N. Yamamoto, Inorg. Chim. Acta 238, 165 �1995�.7S. Hayakawa, T. Yoko, and S. Sakka, Bull. Inst. Chem. Res., Kyoto Univ.71, 411 �1994�.

8D. Maniu, T. Iliescu, and S. Astilean, Rom. Med. Rev. 56, 419 �2004�.9R. Lewandowska, K. Krasowski, R. Bacewicz, and J. E. Garbarczyk, SolidState Ionics 119, 229 �1999�.



http://dx.doi.org/10.1107/S0108270185003237

http://dx.doi.org/10.1016/0020-1693(95)04654-R

http://dx.doi.org/10.1016/S0167-2738(98)00508-6

http://dx.doi.org/10.1016/S0167-2738(98)00508-6

Mechanisms of laser induced modification of lead and barium vanadate glasses

Documents

Transcript of Mechanisms of laser induced modification of lead and barium vanadate glasses