Office Phone: 301-405-1859
 Office Address: 0130
 Group Website



  •  B.S. Chemistry, 1970, Muhlenberg College, Allentown, PA
  •  Ph.D., Chemistry, 1974, University of Maryland, College Park, MD

 Professional Experience

  • Professor of Chemistry, Department of Chemistry and Biochemistry, University of Maryland, 1998-present
  • Associate Professor, Department of Chemistry and Biochemistry, University of Maryland, 1984-1998
  • Visiting Professor, EPA Central Regional Laboratory, Annapolis, MD, 1992
  • Senior Scientist, Martin Marietta Environmental Center, MD, 1982-1984
  • Senior Scientist, Lawrence Livermore National Laboratory, Livermore, CA, 1975-1982

Research Interests

Measurement of aerosol particle composition and size distributions; instrument, advanced multivariate receptor model (PDRM-I, PDRM-II); intentional tracer methods development for quantifying human exposure and understanding the sources and fate of urban aerosol particles; multivariate methods for gamma-ray spectroscopy analysis/Nuclear Forensics; and Radiocarbon measurements of individual compounds in fatty-food matrices.

Professional Societies

American Chemical Society (ACS), American Association for Aerosol Research (AAAR)

Major Recognitions and Honors

  • University of Maryland High-Frequency Aerosol Sampler, U.S. Patent #6,732,569
  • University of Maryland “Rainmaker’s Luncheon” for top 100 Grantees in 1999, 2000, 2001, 2002
  • Ph.D. Student awards: Spectroscopy Society Award (Christopher Kidwell), Spring, 1999
  • Spectroscopy Society Student Award (Peter Caffrey), 1998
  • Precious Metals institute’s award for student research (Ana Suarez)1996
  • Society of Environmental Toxicologists and Chemists “best student paper” award (Terri Quinn), 1994.

 Significant Professional Service and Activities

92 peer-reviewed publications, 16 book chapters; ; 98 invited and 202 contributed talks. Guest Editor, Special Issue, Volume 38, or Particulate Matter, J. Amer. Assoc. Aerosol Sci., 2005; Lead PI for the Baltimore “Supersite” Project, 2000 to 2005; Expert witness support to University of Maryland at Baltimore School of Law, Havre de Grace Rubble Fill. 1995/1996; NIH, NSF, DOE proposal reviewer; EPA Supersites PI Meeting organizer, Las Vegas, February 25-26, 2004; Consultant, International Truck and Engine Co. (2002-2003, 2005-2007) and NIST (2000, 2001); Expert Reviewer, USEPA Speciation Program, 1999; Program reviewer, So. Florida Water Management District workshop on deposition of phosphorous to the Everglades, 1997; Organizer, Symposium on Intentional Environ. Tracers, ACS National Meeting, Boston, MA, August 23-27, 1998; Member, Scientific and Technical Advisory Committee, Chesapeake Bay Program, since 1993; Member, U.S. EPA Air Toxins Workgroup, since 1992; Member, State of Maryland Air Quality Control Advisory Council (AQAC) since 1985; AQAC Vice chairman, 1990-1997, AQAC Acting chairman, 1995-1997; Guest Worker, NIST Reactor Physics Group, since 1982. Expert Advisor to the Philippine Nuclear Reactor Institute’s receptor modeling program, on behalf of the International Atomic Energy Agency, 1995; Reviewer, U. S. EPA 1994 Exploratory Research Program, Environ. Chem. and Physics of Air, March, 1995. Evaluator, Radioisotopes and Trace Analysis Program, MIT Reactor Laboratory, 1994.

Students Mentored

14 post-doctoral, 13 Ph.D., 11 MS; University of Maryland “Rainmaker’s Luncheon” for top 100 Grantees in 1999, 2000, 2001, 2002; Ph.D. Student awards: Spectroscopy Society Award (Christopher Kidwell), Spring, 1999; Spectroscopy Society Student Award (Peter Caffrey), 1998; Precious Metals institute’s award for student research (Ana Suarez)1996; Society of Environmental Toxicologists and Chemists “best student paper” award (Terri Quinn), 1994.


We developed the University of Maryland Semicontinous Elements in Aerosol Sampler (SEAS; Kidwell and Ondov, Aerosol Sci. Technol. 35:596-601, 2001) to collect urban air-borne particles in a form suitable for elemental, major ionic, and biologic assays (cytokines and endotoxin) using standard analytical platforms, with little sample preparation. Detection limits SEAS samples by Electro-thermal Atomic Absorption (ETAA) permitted elemental constituents to be measured at 30-min with precision and accuracy often better than standard (X-ray fluorescence) methods on 24-hr filter samples (Pancras et al., Analytica Chimica Acta 538, 303-312, 2005). At 30-min sampling intervals concentration excursions can readily be related to specific sources, e.g., power plants, incinerators, and steel mills. Contained in the concentration (yt) vs. time (t) profile for these and other pollutants is the intrinsic horizontal-plume dispersion width, a parameter that is sensitive to both source-receptor distance as well as atmospheric conditions. To exploit this information we apply a Pseudo-deterministic (PD) approach to solving the general bi-linear receptor modeling (RM) equation,

cast in units of the average source emission rate and source (j)-to-receptor, dilution rate (Vj,t, sm), wherein a Gaussian plume equation is used to constrain solutions for the latter. With high-quality, high-frequency concentration data, we are able to identify previously unrecognized sources. e.g., plumes from an animal feed supplements manufacturing plant and a construction waste burning site, both in Tampa, FL, were identified by chemical composition and wind-direction (Pancras et al. Atmos. Environ. 40:S467-S481), but neither had been previously recognized as having measurable emissions of toxic substances. With application of our PDRM to another set of measurements made in Tampa, SO2 emissions were predicted to within -6% of the emission rates calculated from continuous emission monitors at 4 coal- and oil-fired utility power plants, including two differing only by a few degrees of arc (Park et al., J. Geophys. Res., 110, No. D7, 1, 2005), an accomplishment appraised by one EPA scientist as a “quantum leap” in the field and PDRM as the most important advance in receptor models achieved in the EPA “Supersites” Program). Emission rates of metals were also predicted, but no emission measurements were available for comparison.


Ondov Figure One





Our previous work was limited to11 elements that were feasible to determine by ETAA.  Ph.D. Student Gregory Beachley is currently analyzing SEAS samples collected in Baltimore during the Baltimore “Supersite” project ( Inductively-coupled plasma-mass spectrometry (ICPMSß) using our new Thermo-systems instrument. Beachley reports sensitivity comparable or better to our ETAA method, but 25 to 27 elements are routinely determined in a few minutes per sample, as compared with 11 elements at one hour per sample. More elements should permit resolution of more sources, and needed to understand the contributions of the many sources in the industrial areas of South Baltimore. Greg has been optimizing ICPMS methods and will apply PDRM to selected data sets to determine optimum sets of tracers for resolving power plant emissions from the complex mix of other sources.

More recently we have been developing methods for the purification of sub-ppm levels of individual phthalic acid esters in fatty foods for radiocarbon 14C analysis by Accelerator Mass Spectrometry.  This work is being conducted to determine the fraction of contemporary carbon in these compounds for the US Food and Drug Administration.  Phthalates are suspected of being human endocrine disruptors.  If found to be of industrial (petrochemical) origin, the food industry may need to develop processes to further reduce or eliminate phthalate contaminants. A 7.5 million concentration factor was required to obtain sufficient quantities of the most abundant phthalate, bis(2-ethylhexyl) phthalate (DEHP).  Only 26%±3.8% of DEHP extracted from Stilton Cheese was found to contain contemporary (hence “naturally” produced) (Nelson et al. Contemporary Fraction of bis(2-ethylhexyl) Phthalate in Stilton Cheese by Accelerator Mass Spectrometry, Radiocarbon, in review).  Market butter contained 8-fold more DEHP and a preliminary measurement suggests that it too is mostly synthetic (Ph. D. student T. Tong, UMCP, unpublished data).  In the course of this work, we developed our own non-negative least-squares deconvolution method to accurately determine compound purity in our extracts, and especially to determine that no additional compounds were eluting under the DEHP peak in GC-EIMS and high-resolution GC-MS total ion current chromatograms

Nuclear Forensics.  Traditional gamma-ray spectrum analysis involves finding and fitting individual peaks and identifying component nuclides by their gamma-ray energies, most often by searching a library of photopeak energies.  Half-life determinations may be used to confirm peak identifications, however, this has generally required manual manipulation of spectra collected at appropriate time intervals.  Modern desktop workstations now provide sufficient power to apply multivariate methods to the problem of component nuclide identification and resolution of complex gamma-ray spectra.  Moreover, the advent of modern time-tagging analog-to-digital converters provide spectral data ideally suited to analysis by automated multivariate analysis methods and provide the ability to construct optimally timed series of spectra for half-life analysis. We have conducted simulations of a series of two-component mixtures of spectra of neutron irradiated concrete dust and natural Uranium nitrate were accurately resolved for mass-mixing ratios ranging from 100 to 100,000:1 using a binary form of the general model:Ondov Equation 2013

We recently applied the method to a multi-element standard reference material (SRM, 1648a, “Urban Particulate Matter) to explore the efficacy of nuclide identification and quantitative analysis by implementing the method in screening and analysis modes, respectively.  Analytical results for 8 certified elements were within ±10% of certified SRM values, using a simple binary-model and 9 of 9 nuclides for which standard spectra were constructed were identified in the SRM at or above the 95% confidence level.  Accuracy of the results was limited by large residuals caused by small deviations in the centroids of peaks in the SRM and reference spectra.  We are currently investigating counting and modeling methods to improve accuracy.

Print Friendly