Commercialization of laser-induced breakdownspectroscopy for lead-in-paint inspection
Richard A. Myers,* Noah J. Kolodziejski, and Michael R. SquillanteRadiation Monitoring Devices, Inc., 44 Hunt Street, Watertown, Massachusetts 02472-4699, USA
Since 1977, the United States Environment Protec-tion Agency (EPA) and department of Housing andUrban Development (HUD) have prohibited themanufacture and use of commercial paint containing>0:06 dry weight percent (wt. %) of lead (Pb) (equiva-lent to 600ppm) [1,2]. However, prior to that time,some paints contained up to 50 dry wt. % of Pb[3,4]. Older homes with Pb paint are now consideredenvironmental hazards, especially if the paint is ac-cessible to the young. In 2004, the National SafetyCouncil estimated that there are 38 million U.S.homes still containing Pb-based paints above theabatement levels [5]. Although the exact values de-fining this level vary based on regional laws, HUDdefines Pb-based paint hazards to be dry paint witha Pb content greater or equal to 1:0mg=cm2 or0:5wt: % [6]. Therefore, special procedures areneeded when remodeling or rebuilding a structurethat has Pb content above these levels. This impliesthat the measurement accuracy at the hazard level is
a critical factor, especially to the owner who has topay for the extra precautions required when a posi-tive reading is found. Although this work focuses onthe U.S., which was one of the last industrializedcountries to take action to eliminate Pb-based paints,other countries are also dealing with this problemand are instituting more rigorous policies towardspaint inspections [7].
The most common method for Pb-in-paint inspec-tion relies upon removing paint chips and sendingthem to the laboratory for analysis, where methodssuch as inductively coupled plasma spectroscopyare employed. While the accuracy of these methodscan be very high, they are destructive, require severaldays of waiting for results, and greatly depend on thesample collectionmethod. The primary alternative tolaboratory analysis is x-ray fluorescence (XRF) usinginstruments such as the LPA-1 manufactured byRadiation Monitoring Devices (Watertown, Mass.).These instruments are considered the gold standardfor real-time Pb-based paint inspection, providing re-liable sensitivity to Pb in paint at levels below0:3mg=cm2 with an accuracy of �0:05mg=cm2 [6,8].Because XRFmonitors the fluorescence return signalfrom a 57Co or 109Cd gamma ray source, the most at-tractive aspect of the measurement is that it does not
mark or damage the painted surface. However, one ofthedisadvantages of these instruments is the reducedaccuracy (�0:1mg=cm2) when analyzing paint onme-tal or concrete substrates. In addition, proper dispo-sal, costly training, and extensive record keeping arerequired when working with both the real and per-ceived hazards of radioactive sources. Eliminatingthe use of the source and its additional costs wouldbe a favorable market factor for any new instrumenttrying to replace those using XRF.
B. LIBS Technology
Matured over the past decade, laser-induced break-down spectroscopy (LIBS) has been used as an ele-gant method for material analysis [9–13]. In brief,a basic LIBS system consists of a high power pulsedlaser, collection optics, and a spectrometer. When thelaser is discharged it ablates a very small amount ofmaterial, which instantaneously superheats, gener-ating a plasma plume that dissociates the ablatedmaterial into excited ionic and atomic species. Initi-ally, the plasma emits a continuum of radiation, butas the plasma expands and cools (0.1 to 10 μs), thecharacteristic atomic emission lines of the elementscan be observed. These spectral signatures can be re-corded and rapidly processed for material analysis.This is particularly useful for on-site inspections,high volume analyses, or on-line industrial monitor-ing. However, despite the compelling performance inthe laboratory, few commercially successful LIBSsystems have appeared on the market that can com-pete with established analytical methods.Complex matrix interactions between atomic spe-
cies as well as the plasma and the sample surfacemakes the use of LIBS for simultaneous multiele-ment analysis difficult. However, portable LIBS ispossible if the sample set is controlled and the ana-lysis is limited to one or two elements. Consequently,because of its strong and isolated emission line, thedetection of Pb-in-paint was one of the early targetapplications for LIBS analysis [14,15]. Early re-search established the ability to measure very lowconcentrations of Pb in paint with a custom probewith a mainframe laser, but no additional workwas preformed to develop a portable instrument orfully assess its market potential [14]. In anothereffort, a portable system was developed using theKigre MK-367 (Kigre, Inc., Hilton Head Island,S.C.) compact neodymium-doped yttrium aluminumgarnet (Nd:YAG) laser and S2000 (Ocean Optics,Dunedin, Fla.) compact spectrometer [15]. While thissystem showed encouraging results, only prelimin-ary work was reported, and, as far as we are aware,further studies to assess the performance of the por-table system as a commercial tool for Pb-in-paint in-spection were not fully explored. Despite the furtherdevelopment of LIBS instrumentation over the pastdecade, no federally approved LIBS instruments forPb-in-paint inspection of which we are aware haveappeared on the market.
2. Experimental Studies
A. Experimental Setup
To assess the technical performance of LIBS for Pbpaint inspection, we used aQ-switched Nd:YAG solidstate laser operating at 1064nm. The laser had apulse width of 10ns, a maximum repetition rate of10Hz, and a maximum output energy of ∼200mJper pulse. Although a portable laser was not avail-able during this study, the mainframe laser provideda high degree of control over its repetition rate andpulse energy and provided both internal and exter-nal triggering capabilities. This control allowed usto simulate the laser output for a range of portablelasers and synchronize the timing of the laser plasmawith the spectrometer. Beam splitters and a spatialfilter were used to reduce the laser energy to <15mJper pulse, consistent with the laser pulse energies ofcompact Nd:YAG lasers viable for portable LIBS sys-tems. Off-the-shelf optics were used in a collinearbackscatter collection geometry. The principal opticsincluded a 35mm focal length lens for both focusingthe laser excitation and collecting the backscatteredemission. The 35mm focal length maintained (1) anappropriate separation from the target, preventingthe ejected sample from coating the lens, (2) a focalspot size as small as 100 μm, and (3) a depth-of-focusover 1mm, which reduced the sampling uncertaintydue to surface imperfections and provided consonantintensities as the laser “drilled” through the sample.
Using a lens, the backscattered plasma emissionwas imaged into an optical fiber with a 600nm dia-meter whose opposite end was coupled to a 1=4mmonochromator. After confirming that the primaryPb emission line at 405:78nm could be used to moni-tor the level of Pb in paint, we set the resolution ofour spectrometer to 0:01nm and used a cooled 256 ×1024 element CCD from Princeton Instruments(Trenton, N.J.). Our cooled CCD did not allow tem-poral separation of the initial plasma signal fromthe emission signal. We also utilized an Ocean OpticsHR2000 spectrometer with a grating centered at∼440nm and a 120nm bandwidth that provided aspectral resolution of ∼0:1nm. While this unit oper-ated at room temperature, it used a CCD that has ahigh-speed gating feature for reducing the initialplasma signal. Software from Ocean Optics allowedus to externally trigger the laser and adjust the delaybetween the plasma formation and the start of thesignal collection. The delay following the laser trig-ger pulse was typically set to 2:5 μs.
We compensated for the shot-to-shot signal varia-tion (as much as 20%) by subtracting the backgroundsignal within 0:5nm of the Pb emission signal. Theonly element with a strong spectral emission linewithin 0:1nm of the Pb emission is from Gd, but itwas never observed during our studies [16]. Figure 1shows our setup and spectra of a latex-based paintwith and without added Pb. Because the Nd:YAGlaser radiation is not visible, a 632nm He–Ne laserwas used for system alignment.
We monitored the LIBS signal from solid Pb, alumi-num, and stainless steel. We also tested NIST-traceable Pb paint calibration samples designedfor use with XRF instruments along with variousPb-containing paint samples from old homes. For aset of samples with a known range of Pb concentra-tions, we produced paints by mixing commercialwater- or oil-based paints with either lead oxide(PbO) or lead carbonate (PbCO3) [14]. We used sev-eral different paint bases with varying color tints aswell as amounts of Ti. The dry weights of the paintswere determined by consulting themanufacture datasheet and confirmed in the laboratory. Paint sampleswere applied to a substrate with a brush, resulting ina single layer thickness ranging from 20 to 40 μm.Multiple layers were created following appropriatedrying times. To simulate years of drying, some sam-ples were placed in a laboratory oven for 24 to 48 h atan elevated temperature of ∼90 °C. Plastic, metal,wood, and glass substrates were used.Paint sample thickness was measured using a
thickness gauge with ∼3 μm accuracy. To determinethe laser removal rate, paint was applied to glassslides and the thickness was measured. By illumi-nating the opposite side of the glass slide and view-ing the sample through amicroscope, we were able todetermine the number of laser shots required to pe-netrate through the paint as a function of laser en-ergy and focal spot size.
C. Experimental Results
Laboratory-prepared paint samples containingequivalent Pb concentrations displayed Pb emissionstrengths that were within experimental error ofeach other, independent of the color tint, the amountof Ti, or the paint base. This was important since itindicates that a reliable set of calibration standardscan be created.To assess LIBS as an alternative to XRF for rapid,
on-site paint inspection, we optimized the amount ofmaterial removed during each laser pulse. As seen in
Fig. 2, while the signal-to-background ratio in-creased with laser energy, it began to saturate whenthe pulse energy exceeded 15mJ and the penetrationdepth of each laser shot decreased. This phenomenonwas observed in all of the paints tested, and the de-crease in penetration depth with laser energy was at-tributed plasma screening [17,18]. This occurs whenthe intense plasma spark prevents the tail end ofthe laser pulse from interacting with the target. Inaddition, the plasma ablation formed craters withdiameters from <100 μm to over 500 μm as the laserpulse energy was increased across the tested range.Using these data, we determined that a pulse energybetween 6 and 8mJ provided the best trade-off be-tween the strength of signal and removal rate.
Figure 3 shows the measured Pb emission signalarising from LIBS experiment using paint sampleswith PbCO3 concentrations ranging from 0:1wt:%to 10wt:% The strength of the emission signal fromPb was recorded by averaging the signal generatedwith six laser shots and laser pulse energy of 7mJ.Several random sampling locations were selectedfor each measurement to determine the signal uni-formity and measurement error across each sample.The figure indicates that, at the HUD federal abate-ment level of 0:5wt:% Pb, at least a 10 : 1 signal-to-background ratio can be achieved. The response waslinear at this federal action level, which is importantfor achieving accurate measurements. We measuredthese samples with RMD’s XRF instrument, but acomprehensive comparison was not realized becausethe relatively thin paint layers resulted in readingsbelow the maximum reliable sensitivity of the instru-ment (0:3mg=cm2). A separate measurement of thePb sample using a more sensitive XRF instrument
Fig. 1. Diagram showing the LIBS experimental setup used dur-ing our studies. The He–Ne laser was used for alignment. Typicallaser energies ranged from 5 to 7mJ and both a bench top and acompact spectrometer were used during our studies. The insetshows the spectrum from latex paint with and without Pb. ThePb emission at 405:8nm is noted with an arrow.
Fig. 2. Pb emission signal-to-background ratio of paint increasesas a function of laser energy, but the penetration depth per laserpulse decreases with increasing energy. This decrease is likely dueto plasma screening. Data were collected using the cooled CCD and1=4m spectrometer.
showed that our estimates were accurate for the pur-pose of our study with error towards overestimatingthe concentrations.Once the laser penetrated through the paint, the
spectrum also included elements from the substrate.The plasma created at the substrate surface encom-passed the surrounding paint layers, resulting in aconvolution of the emission signals from both mate-rials. During our test, if the substrate wasmetal suchas Al, which has strong emission lines within ourspectral range, the location of the substrate wasreadily resolved. Although we did not test for it inthis study, it may be possible to monitor calcium’semission at 396:8nm when paint is on drywall orplaster. However, painted wood lacks a dominant re-ference emission line, making it difficult to deter-mine when the substrate was reached. Alternatively,there are low-cost instruments that can measurepaint thickness on metal, drywall, wood, or concrete(e.g., the PosiTector gauges from DeFelsko Corpora-tion of Ogdensburg, N.Y.). Typically utilizing ultraso-nic methods, these instruments are hand-held,nondestructive, rapid (<2 s=reading), and can poten-tially be integrated into a LIBS analysis inspec-tion tool.To demonstrate portability, we made measure-
ments with the compact Ocean Optics HR2000 spec-trometer. The same set of calibration samples thatwere used to test the bench-top system were usedto calibrate the signal from the HR2000 unit. Wefound that the signal-to-background ratio reduced
by 25% relative to the cooled CCD spectrometer,but it was still easy to detect Pb concentrations belowthe 0:5wt:% abatement level. A spectrometer unitwith higher resolution and efficiency at the Pb emis-sion line would further improve the performance.
Paint boundaries between layers of Pb paint andPb-free paint were difficult to distinguish withoutsignificant drying times between layer applications,especially if the Pb paint was applied over the Pb-free paint. To circumvent this problem we character-ized a sample of painted wood trim from a home builtin 1880 with both the LIBS setup and portable XRFinstrument. Multiple paint layers (more than six dif-ferent colors) were readily visible. The total thick-ness ranged from 400 to 600 μm depending on thespot and wood surface contour.
We compared the XRF and the LIBS measure-ments of our painted trim to show that the LIBSmethod had an advantage for detecting the Pb asa function of depth and providing historic informa-tion about the home. RMD’s LPA-1 took several sec-onds to give a reading of>10mg=cm2, indicating thatthere was a high concentration of Pb paint presentwithin the sample. The LPA-1 is not required to pro-vide a more exact number when the Pb concentra-tions reach these high levels. The comparable LIBSdata were recorded using our experimental setupwith an 8mJ laser pulse and the HR2000 spectro-meter. Three laser pulses were averaged per datapoint through ∼550 μm of paint and ∼300 μm ofsubstrate. At a 10Hz rate, the data collection took∼10 s. Figure 4 gives the LIBS emission data from
Fig. 3. Signal-to-noise ratio as a function of the dry wt. % ofPbCO3 added to a commercial latex paint. The measurementswere performed at 7mJ with a cooled CCD detector. Data were re-corded following averaging of six laser shots. The federal guidelineof 0.05 dry wt. % of Pb (5000ppm) is noted with a dashed line. Con-centration errors result from sample preparation and achievinghomogeneous blending. Data were collected using the cooledCCD and 1=4m spectrometer.
Fig. 4. Pb signal with background subtracted as a function ofdepth for a sample that came from a house built in∼1880 and con-tained several layers of paint. The dashed line indicates the ap-proximate boundary between the paint and the wood substrate.The Pb emission signal beyond the boundary layer is likely dueto absorption of the primer into the wood substrate as well asthe continued sampling of the Pb paint. Data were recorded withthe HR2000 spectrometer from Ocean Optics.
a representative paint chip as a function of the laserpenetration depth. The y axis represents the rawemission signal above background as well as the cor-responding Pb concentration. The calibration of theHR2000 was determined from a fit of emission sig-nals obtained from the same series Pb paint samplesused to generate Fig. 3.It is readily apparent from Fig. 4 that the Pb con-
centrations above federal abatement levels were de-tected. While the first few paint layers (up to 50 μm)of the paint chip are free from appreciable Pb, therewere intermediate layers at approximately 100 μmthat had a Pb-emission signal consistent with a∼1wt:% Pb concentration. Paint with this Pb con-centration was commonly used between 1940 and1970 [4]. Although the Pb emission signal decreasedagain, there was a dramatic increase in signal con-sistent with Pb concentrations above 10% at a depthof 300 μm. This is not surprising, especially if the Pbwas in the primer, which was common between 1900and 1940. The signal variation at depths beyond300 μm could be due to further paint layers with var-ious Pb concentrations or the error introduced as theplasma interacts with the already exposed layers.Several locations on different parts of the wood trimwere tested and showed a similar profile.The lack of distinct emission lines from wood made
it difficult to determine when the end of the paintwas reached. Although there was still an emissionsignal from Pb at locations deeper than the esti-mated paint thickness of 500 μm, it would be ex-pected that the paint and primer would haveabsorbed into the wood. This theory was supportedwhen Pb-emission signals consistent with those gi-ven in Fig. 4 were observed following the removalof paint from an area of wood much larger thanthe laser spot size. In other words, only paint thathad soaked into the wood was sampled. The backsideof the wood did not show any measurable Pb.
3. Comparison of LIBS and XRF for Pb PaintInspection
A. Methodology
The primary advantage of the XRF instrument is itsability to determine the Pb concentration in a single,nondestructive measurement, independent of thepaint thickness. These instruments are designed toaccurately determine if a property contains Pb inpaint above or below the required action level. How-ever, because of this integrated measurement, thesubstrate material affects the accuracy, particularlyif it is metal. While the LIBS measurement makes iteasier to isolate the sample from the substrate, it re-quires “drilling” through the paint and, although themajority old paint chips have a thickness of no morethan 0:5 μm, the measurement accuracy and sam-pling time must be investigated.A quantitative comparison between XRFand LIBS
is difficult. While LIBS measures the Pb concentra-tion per unit volume, a XRF measurement is weight
per unit area (or loading). The HUD guidelinespermit either type of measurement (1mg=cm2 or0:5wt:%), but the measurement of one cannot be re-lated to the other without knowledge of the paintdensity and thickness. This difference can be illu-strated by considering a NIST-traceable sample usedfor calibration of portable XRF instruments. Thesecalibration samples consist of a ∼40 μm layer ofPb-doped paint on a polyester sheet covered with alaminate. While the commercial XRF instrument isnear its detection limits when measuring a samplewith 0:5mg=cm2 Pb loading, the LIBS method pro-duces a signal that is nearly 30 times above the back-ground. For a 40 μm layer of household paint,between 10% and 15% of the total dry weight mustbe Pb to achieve 0:5mg=cm2. This makes LIBS po-tentially useful for postabatement inspection, wherelimited quantities of material are collected. On theother hand, for samples with a thickness of 1mmand a XRF reading of 0:5mg=cm2, the average con-centration of Pb will be <0:5 dry wt. %, significantlyreducing the LIBS signal per sample depth.
B. Sensitivity
Using a LIBS arrangement consistent with what canbe achieve with a portable system, it is possible todetermine if paint is above or below the abatementlevel with an accuracy better than 5%, comparableto current XRF instruments. With additional calibra-tion effort, an integrated measurement could bemade for a sample of known thickness.
C. Sampling Time
The time it takes to make a measurement is criticalfor the inspector who wants to perform as many mea-surements as possible in the shortest time possible.The average two to three bedroom home inspectionrequires up to 350 interior and exterior sampling lo-cations. If the inspection method reduces the totalnumber of inspections that the inspector can performin one day, then the interest in the alternative meth-od will diminish.
Most XRF measurements are performed in 2 or 3 s,although for more accurate readings or if the sourceis old, integration times of 30 s are not uncommon.The speed of the LIBS method is determined by timethat it takes to “drill” through the paint. Most mod-ern paints have a thickness of 30 to 50 μm, while old-er paints tended to be thicker but require fewer coats.Assuming an average removal rate of 10 μm=pulseand a laser operating at a repetition rate of 10Hz,it would require ∼5 s of sampling time to interrogatea paint sample with a thickness of 500 μm. For theextreme case of inspecting paint with a thicknessof 1:5mm, or if the thickness is unknown, the LIBSmethod would take three to four times longer thanthe XRF method.
Our research indicates that there are very fewinstances where the paint would be prohibitivelythick across an entire surface. A greater concern isthe presence of a barrier such as wallpaper or a thick
boundary layer, which is also used as a means of Pbpaint abatement. While the vast majority of Pb paintwas used on trims and exterior coating, which wouldbe less likely to have these barrier layers, this con-cern would require further consideration.
D. Cosmetic Damage
The XRF measurement is nondestructive and non-marking, making it attractive for many Pb paint in-spection needs, especially for home sale surveys. Inan effort to anticipate the consumer response tothe more destructive LIBS method, we evaluatedthe laser damage spot following sampling. For nearlyall of our test cases, the damage area was no largerthan 0:5mm in diameter and typically closer to100 μm. The visual impact of the damage site in-creased when the laser penetrated deep into the sub-strates, resulting in slight darkening around thecrater perimeter. Figure 5 shows a picture of a testsample following sampling with various laser inten-sities used to penetrate a layer of paint. It is appar-ent that the LIBS method may be controlled toprevent damage beyond the many common surfaceimperfections also seen in the image. In addition,it is much less destructive than sampling for labora-tory analysis. Finally, because paints have a high de-gree of homogeneity, a LIBS-based instrument is notexpected to require a greater number of sample loca-tions than XRF despite its small sampling volume.
E. Portability
From a marketing point of view, the portability of aLIBS inspection tool is a critical factor if it is to com-pete against the existing XRF instruments. As a re-ference, the entire RMD’s LPA-1 consists of a single
hand-held unit with dimensions of 13 × 28 × 23 cmand weight of 1:4kg [19]. Although less inviting thana complete hand-held unit, until the technologyevolves, a LIBS Pb paint inspection unit would likelybe too bulky to fit into a single assembly. Therefore,themajority of the weight, including batteries, can beplaced on a belt or shoulder-mounted pack. Fiber-optic coupling also makes it possible to separate ahand-held probe from the bulk of the electronics,which can be advantageous for difficult to accesslocations.
Along with our current research, previous workhas demonstrated that the Ocean Optics spectro-meter is sufficient for use in a portable system [15].The HR2000 measures ∼15 cm × 10 cm × 5 cm andweighs 0:54kg [20]. However, while still available,the manufacturer has discontinued the currentlyemployed detector array [21]. Although further in-vestigation is required, it is hoped that similarinstrumentation performance at comparable priceand size can be realized.
Available compact lasers for portable LIBS systeminclude those from Kigre, Inc. Table 1 lists some ofthe attributes of the Kigre lasers [22]. While somemodels have repetition rates that would be too lowfor Pb paint inspection, models with 10 to 20Hz ratesare in development. Commercially available laserswith energies from 10 to 100mJ and repetition ratesaround 100kHz have also been used for LIBS butwould require further investigation for use in Pbpaint inspection.
For a LIBS system, the laser is the primary drainon the battery life. As a reference, RMDs LPA-1 usesa battery with 3300mAh at 7:2V and will operate for8 h. Based on specifications for the Kigre MK-88laser [22], we estimated that a hand-held camcorderbattery would allow operation for >15; 000 lasershots. This is a sufficient number of laser pulsesfor inspection of one complete household. Batterycosts of <50=unit make it reasonable for inspectorsto keep several on hand for multiple inspectionsthroughout the day. In addition, consumer batterytechnology is rapidly advancing, further suggestingthat this factor would not be a barrier to the market.Because the lifetime of the MK-88 has been reportedto be >40; 000; 000 shots depending on the operatingcurrent and environment, assuming an upper aver-age of 15,000 shots/inspection and 1000 inspec-tions/year (5 per day for 200 days) the laser wouldlast over 2.5 years.
Fig. 5. Picture of the paint crater created by the LIBS process.Each was formed with four laser shots. The laser energies rangefrom 28:4mJ (indicated with an arrow on the left) to 2:2mJ (indi-cated with an arrow on the right). The smallest crater has a dia-meter of ∼75 μm. The paint was applied to an Al substrate, andtypical paint imperfections can been seen. The region above thearrow on the left was formed with over 100 laser pulses andhas a dark ring from ablation of the Al.
Table 1. Specifications of Compact Lasers Available from Kigre, Inc.
The biggest barrier for a LIBS inspection tool forPb paint that can compete with existing XRF in-struments is the availability and initial cost of theappropriate components and the compact laser inparticular. To be competitive, we estimate that thetotal bill of materials must be less than ∼5000. As-suming a discount for large volume purchases, thespectrometer and the optics alone might be severalthousand dollars. Beyond that, existing laser sourcesthat can meet the necessary performance specifica-tions range in cost from 5; 000 to 10; 000. Until thesecosts decrease, it will be extremely challenging tomanufacture a competitive instrument that wouldbe an attractive replacement the existing Pb inspec-tion units.Even if it is difficult to promote LIBS as a direct
replacement for in-home Pb-in-paint inspectionmethods, some of its advantages could be exploitedin complementary applications. For example, alter-native methods for Pb paint inspection are neededfor postabatement screening where the sample sizeis too small for XRF and laboratory analysis is tooslow. The unique aspect of depth profiling also offersa useful alternative inspection tool. While it may notbe critical in the majority of inspections, the locationof the Pb paint layer can offer a homeowner addi-tional information regarding abatement decisions.Other advantages of the LIBS system include theability to reach difficult to access locations using fi-ber coupling and substrate-independent measure-ments, especially on metal structures such asboats and housing exteriors.
G. Regulatory Concerns
At present, inspectors that use XRF instrumentsmust be trained to work with radioactive sources.In addition, the units must be returned to the man-ufacturer every year or two for source replacement.The manufacturer must then store and dispose ofthese sources according to government regulations.The use of a laser is also a safety concern since eye
and skin protection must be addressed. However,most U.S. states would relieve the inspector from un-dergoing special licensing as long as appropriatemanufacturer safety precautions are incorporatedto prevent unintentional access to the full laser en-ergy. Such precautions as an optical sensor thatwould only allow laser emission when the unit isagainst a flat, opaque surface can be readily imple-mented. If an eye-safe Class I laser (e.g., 1:54 μm,<7mJ) can be utilized, it would meet all licensing re-quirements for both user and manufacturer [23].
4. Conclusions
It is apparent that a LIBS-based unit offers severaladvantages over hand-held XRF instruments for Pbpaint inspection. In particular, the ability to detectthe Pb concentration as a function of depth adds acapability not found in other inspection methods.In addition, a portable unit can have sufficient sen-
sitivity to readily determine the federal abatementlevels. By promoting these advantages the technol-ogy will be more readily accepted by both the userand federal agencies. However, until federal regula-tions change or component cost are reduced, LIBSinstrument manufacturers will likely be compelledto seek niche markets offering noncompetitive solu-tions to Pb paint inspection challenges.
The authors wish to thank from Michael J. Myersfrom Kigre, Inc., for helpful discussions. This re-search work reported in this paper was supportedby the U.S. Department of Homeland Security (con-tract #HBCHC060150).
References1. “CPSC announces final ban on lead-containing paint,” Office
of Information and Public Affairs, Washington, D.C., 2September, 1977.
2. “Residential Lead-Based Paint Hazard Reduction Act of1992,” EPA Public Law 102-550, 102nd Congress–2nd Ses-sion, enacted 28 October, 1992.
3. G. Markowitz and D. Rosner, “Cater to the children: the role ofthe lead industry in a public health tragedy, 1900–1955,” Am.J. Public Health 90, 36–46 (2000).
4. “Why lead-based paint was used,” http://www.leadlawsuits.com/index.php?s=community, retrieved on 24 March, 2008.
5. This information can be found at the National Safety Councilwebsite, http://www.nsc.org/issues/lead.aspx, retrieved on 24March, 2008.
6. D. E. Jacobs, “Lead-based paint inspection,” in Guidelines forthe Evaluation & Control of Lead-Based Paint Hazards inHousing (Daine , 1995), Chap. 7.
7. A. Mathee, H. Rollin, J. Levin, and I. Naik,“Lead in paint:three decades later and still a hazard for african children?,”Environmental Health Perspectives 115, 321–322 (2007).
8. R. Jenkins, X-Ray Fluorescence Spectrometry (Chemical Ana-lysis: a Series of Monographs on Analytical Chemistry and ItsApplications), 2nd ed. (Wiley, 1999).
9. L. J. Radziemski and D. A. Cremers, eds., Laser-Induced Plas-mas and Applications (Marcel Dekker, 1989).
10. D. A. Cremers and L. J. Radziemski, eds., Handbook of Laser-Induced Breakdown Spectroscopy (Wiley, 2006).
11. A. W. Miziolek, V. Palleschi, and I. Schechter, eds., Laser In-duced Breakdown Spectroscopy (Cambridge University Press,2006).
12. W. B. Lee, J. Y. Wu, Y. I. Lee, and J. Sneddon, “Recent applica-tions of laser-induced breakdown spectrometry: A review ofmaterial approaches,” Applied Spectroscopy Reviews 39,27–97 (2004).
13. D. W. Hahn, W. L. Flower, and K. R. Hencken, “Discrete par-ticle detection andmetal emissions using laser-induced break-down specroscopy,” Appl. Spectrosc. 51, 1836–1844 (1997).
14. B. J. Marquardt, S. R. Goode, and S. M. Angel, “In situ deter-mination of lead in paint by laser-induced breakdown spectro-scopy using a fiber-optic probe,” Anal. Chem. 68, 977 (1996).
15. B. C. Castle, A. K. Knight, K. Visser, B. W. Smith, andJ. D. Winefordner, “Battery powered laser-induced plasmaspectrometer for elemental determination,” Journal of Analy-tical Atomic Spectrometry 13, 589–595 (1998).
16. R. C. Weast, M J. Astle, and W. H. Beyer, eds., 67th ed., CRCHandbook of Chemistry and Physics (CRC Press, 1986).
17. M. Corsi, G. Cristoforetti, M. Hidalgo, D. Iriarte, S. Legnaioli,V. Palleschi, A. Salvetti, and E. Tognoni, “Effect of laser-induced crater depth in laser-induced breakdown spectro-scopy emission features,” Appl. Spectrosc. 59, 853–860 (2005).
18. J. M. Vadillo, J. M. Fernandez Romero, C. Rodriguez, andJ. J. Laserna, “Effect of plasma shielding on laser ablation rateof pure metals at reduced pressure,” Surf. Interface Anal. 27,1009–1015 (1999).
19. This information can be found at http://www.rmd‑lpa1.com/,retrieved on March 24, 2008.
20. This information can be found at http://www.oceanoptics.com/products/hr2000.asp.
